Jump to content


Photo

How Does Explosion Damage Work -


  • Please log in to reply
123 replies to this topic

#1 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 14 October 2005 - 11:27 PM

The wiki says that damage drops off by 10 for every tile. But, attached find the overhead view for a blaster bomb (mag x2 for ease of viewing; game cursor is on center of explosion).

A BB makes an 11-square radius of damage, if the center is 0. Twelve, if you count the center. A BB is rated at 200 High Explosive.
  200 damage / 12 tiles = 16.7 damage per tile
   200 damage / 11 tiles = 18.2 damage per tile
Damage to the scenery was all or none... there was no point where e.g. only grass was affected, but not stone garden walls. So there seems to be no doubt about how far it extends.

16.7 per tile is not anywhere near 10 per tile. What am I missing? :hmmm:

Also Alien Grenades (HE 90) extend by offset 6, or 7 if center is included.
  90 damage / 6 tiles = 15 damage per tile
   90 damage / 7 tiles = 12.9 damage per tile
Help?

edit by j'ordos : gif image replaced with jpeg image

Attached Thumbnails

  • AfterBlasterBombX2.jpg

Edited by j'ordos, 15 October 2005 - 05:52 AM.


#2 Blehm 98

Blehm 98

    Colonel

  • [Global Moderators]
  • PipPipPipPipPip
  • 2,626 posts

Posted 14 October 2005 - 11:38 PM

that might be direction and armor stuff thats changing it up though, but i could be mistaken
Top Secret Xenocide Status report

BF2 Hackers =5SF= have busted
]sD[ Engageo <-- couldn't get him banned though, no screenshot of him ingame
]sD[Nomisser
an AK guy
The anti-logarithm of the logarithm of X plus the logarithm of Y equals X times Y, or 10^(logX + logY) = XY
Posted Image
I hate spider solitaire...

#3 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 14 October 2005 - 11:45 PM

Thanks for reply, Blehm!

As far as I can tell, the wiki only meant scenery, and that's all I mean, too, as the pic shows.

Effect on units will be the next question, after this one is squared away ;)

Edited by MikeTheRed, 14 October 2005 - 11:46 PM.


#4 Blehm 98

Blehm 98

    Colonel

  • [Global Moderators]
  • PipPipPipPipPip
  • 2,626 posts

Posted 14 October 2005 - 11:59 PM

Now, i am either about to embarass myself, or make myself look good, one or the other, and i kind of let myself ramble, so don't be too worried about offtopic stuff ;)


you said damage to scenery was all or none, i believe that NKF and a bunch of us had a conversation in the TFTD forums about grenades. The reason why the damage to scenery is all or none is i believe because scenery can have lots of armor, but usually not much health (in its own way). So thats why small explosives may not do anything, but bigger explosives will go right through it. I really know very little about scenery though, but i assume it has armor on different sides of it (actually, two sides, north and west, but there is also grid and content) and each has its own armor determined in its .pck or something like that. I figure that all types of terrain (like all basic ground tiles of one type, like cabbage, etc) have very similar amounts of armor when compared to each other, so that you won't have (as NKF described it) "swiss cheese scenery." Now explosives will do a set amount of damage only at their base, and for either the 8 or the 16 tiles surrounding the base (forgot which) take all damage to their bottom armor. This probably applies to tiles too. After you get that far away though, the explosion will do damage to the armor facing the explosion. Since there are multiple tiles in each single space, i would figure that a mean or some sort of average is figured out, and then applies to the damage as armor normally would, and then the individual tiles take damage according to how tough they are. That might explain some of the wierdness of the way that damage came out.

I have no idea if i just answered your question or not, but i tried explaining how damage works and such, so i may have helped.. if i didn't, then i just made myself look like a fool :P
Top Secret Xenocide Status report

BF2 Hackers =5SF= have busted
]sD[ Engageo <-- couldn't get him banned though, no screenshot of him ingame
]sD[Nomisser
an AK guy
The anti-logarithm of the logarithm of X plus the logarithm of Y equals X times Y, or 10^(logX + logY) = XY
Posted Image
I hate spider solitaire...

#5 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 15 October 2005 - 12:15 AM

Thanks greatly for the reply Colonel!

Testing shows that there's not any damage on aliens past the circumference described above. I think there is some misunderstanding with what was posted, or my understanding of it.

Where'd you get your name from, soldier? It's an intriguing one.

#6 NKF

NKF

    Commander

  • [Global Moderators]
  • PipPipPipPipPipPip
  • 4,798 posts

Posted 15 October 2005 - 12:51 AM

As Blehm said, scenery has armour as well, but it doesn't technically have any health. This means scenery has a sort of 1 hitpoint system, although the more damage done, the worse the replacement tile/wall will look when the armour is breached. Different tiles have different armour levels, so how much damage done to the ground will vary depending on the terrain. Even if it looks like no damage is done, that's not to say damage wasn't done. Most of the time your armour will have soaked up the last remnants of the damage so you wouldn't notice it at all.

If you really want to test damage drop-off along the edges, I suggest editing your unitref soldiers to have 0 all-round armour. Use soldiers in their standard jumpsuits, as they practically have no damage modifiers, making them most sensitive units to damage.

- NKF
Lord High Generalissimo Ruler Supreme of Norm's Anti Pedant Society (NAPS).

Number of members: 1

#7 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 15 October 2005 - 01:04 AM

I'll work on that, NKF - and thanks for the reply. Are there any old posts that go into deep detail? That's what I'm after. The wiki page didn't mention armor vs. units at all, or the explode-through effect (HP=1). So you can see my confusion. Reminds me of the brain ship sci fi series, where a grapefruit sized mass controls vast shielding resources. I'm ultimately after an equation for damage roll as a function of distance. So, armor causes a simple minus on damage potential? A roll from 0 to max - armor? I'm getting the feeling the wiki glossed over a few things that you and/or others know well :)

Edited by MikeTheRed, 15 October 2005 - 01:11 AM.


#8 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 15 October 2005 - 09:16 PM

I did some tests on the Blaster Bomb's damage radius a while back and noticed the same thing. Instead of the anticipated 19 or 20 tile radius, damage was restricted to only the first 11 or 12 tiles.

A lower damage for the Blaster Bomb would be one theory, but this does not agree with experimental results. A different damage drop-off per tile would be another theory, but the OSG (both for EU and TFTD) are fairly clear that the 10 drop-off is a given. The only other explanation is that the ground itself "sops up" (or absorbs) a portion of the damage. Each tile from ground-zero absorbs part of the explosion until the damage left is < tile armor.

Assuming the larger radius for both explosions (BB and Alien grenade), each tile should be absorbing roughly 3-4% of the total damage. That works out to around 6-8 points of damage. By default, this should be the armor rating of each tile.

Like NKF mentioned, different tiles have different armor ratings. I did some tests on blast radius vs. terrain type and concocted a couple of crazy tables listing the apparent damage from each explosive in the game:

          Dirt/    Grass   WhtTile   Asphalt    Tin      White     Concrete
           Weeds    /Lawn    Roofs     Roads    Roofs    Shingles   Walkway
Grenade     7         7       3         3        5         5          3
Proxy      11        11       7         7        9         9          7
Alien      13        13      11        11       13        13         11
Hi - X     13        13      13        13       13        13         13
Small R     9         9       7         7        9         9          7
Large R    13        13      13        13       13        13         13
Blaster B  23        23      25        25       25        25         23
AC - HE     5         5       1         1        3         3          1 
HC - HE     7         7       3         3        5         5          3

             Corn     Wheat    Desert   Mars     Polar    Mountain     Alien
            Fields    Fields    Sand    Sand      Snow     Tundra      Floors
Grenade       7         7        9       0         1         0           0
Proxy        11        11       13       0         5         0           0       
Alien        13        13       13       3         9         3           0
Hi - X       13        13       13       7        13         7           3
Small R       9         9        9       0         5         0           0
Large R      13        13       13       5        11         5           1
Blaster B    23        23       23      23        23        23          21
AC - HE       5         7        7       0         0         0           0
HC - HE       7         7        7       0         1         0           0

"White Shingles" refers to the roof on the gas station complex.
"Mars Sand" refers to the soil on the Martian surface in Cydonia.
"Alien Floors" refers to the floor of an alien base. Alien ships were not checked.
"Concrete Walkway" refers to sidewalks near buildings or roads.
The numbers are not radius, but rather diameter. To find radius, subtract one and divide by two. Look at the Blaster B rows on each table. Depending on terrain, BB's can damage anywhere from a diameter of 21 to 25 (radius of 10 to 12). ;)

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#9 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 16 October 2005 - 11:14 PM

Great table, Z. Wow, that must have been a lot of work! I tidied it up a little (collapsed duplicate columns etc.) and posted it to the wiki, together with HE ratings and cost. Perhaps you or NKF can take a look and make sure I worded the "tile absorption" idea correctly.

When compared to HE rating, there seem to be some oddities. For example, the Prox Grenade is 70 HE and the Small Rocket is 75, but the Prox G has a wider blast diameter. Could its "height" when exploding have made a difference? Also the humble grenade (50) did more damage to grass than HC/HE (52).

Any ideas why these differences were seen?

My original reason for looking at explosions, was to see if I can find an alternate way to get experience than the firing squad with standard pistol. IOW, can I find a way to do less damage with each hit. But most explosives have the big drawback that either they can't be reaction fired at all, or else you only get 1 reaction point and can't control just where the center of the explosion is very well. (But you can still get a lot of experience points for firing accuracy!) Hmm... still working on the concept. The Small Launcer is another interesting one to look at. Will also take a look at incendiary hits... I don't know much about them yet.

#10 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 16 October 2005 - 11:50 PM

Great table, Z. Wow, that must have been a lot of work! I tidied it up a little (collapsed duplicate columns etc.) and posted it to the wiki, together with HE ratings and cost. Perhaps you or NKF can take a look and make sure I worded the "tile absorption" idea correctly.

<{POST_SNAPBACK}>

Indeed, it took some time to put together. The main problem was finding all the landscapes and/or scenarios necessary to test the explosive blasts out. The wiki section looks good, and I'm glad you used the word "theorize" because that's all it is right now - a theory. But, until some more experiments are done, it is the best hypothesis so far. I have a couple tricks up my sleeve yet, so don't count me out. :)

When compared to HE rating, there seem to be some oddities. For example, the Prox Grenade is 70 HE and the Small Rocket is 75, but the Prox G has a wider blast diameter. Could its "height" when exploding have made a difference? Also the humble grenade (50) did more damage to grass than HC/HE (52).

Any ideas why these differences were seen?

<{POST_SNAPBACK}>

The only thing that comes to mind at the moment is that the grenade-type weapons are closer to the ground and thus the blast is more widespread. Since the fired explosives are shot straight down, some part of the blast may be unevenly absorbed by the tiles. Give me some time and I'll recheck those weapons to make sure the numbers are correct. :wink1:

Small Launcer is another interesting one to look at. Will also take a look at incendiary hits... I don't know much about them yet.

<{POST_SNAPBACK}>

I also did a test for incendiary rounds:
Ammunition            Affected Area
   AC - I                1 3 5 3 1 
   HC - I              1 5 5 7 5 5 1
   RL - I            1 7 7 7 9 7 7 7 1

So the Auto Cannon - Incendiary round will look like this:

    X          1
   XXX        333
  XXXXX      13531
   XXX        333
    X          1
The radius/diameter for the Stun Bomb is unknown at the moment, but I would assume it follows roughly the same results as the Alien Grenade. (Just a speculation). ;)

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#11 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 17 October 2005 - 12:00 PM

That sounds great re: theories on blasts, and such. If you come up with new stuff, let me know (and I'll stick it in the wiki). Or you're welcome to wiki it yourself of course...

I am thinking that explosions might not be worth it, relative to experience points. Too variable and little or no reaction training. But I will try to get some harder numbers. For one thing, I'm realizing I haven't collected good data even on how much Muties are hurt by pistol shots. Much less blast damage, to compare against. I see the Damage Formula in the wiki and want to make sure I'm following it ok...

In your example of incendiary blasts, do the numbers mean, how many turns the tile will be on fire? I'll be snarfing that to the wiki too, shortly. =b

JFG and NKF have posted stuff on incendiary damage here and here. But I'm not entirely clear on how many damage points are dished out. There's damage from the initial explosion, but I can't seem to find any hard numbers on that... and then, if on fire or in fire, it's approx. 1 point per turn? And otherwise no damage except for the weirdness about, fire or smoke impact being repeated any time an IN round goes off anywhere on the map? NKF, Z, or anyone else who already knows? Again, I'm doing this with a mind toward experience points, so I'll be looking at that. I assume "on fire" damage isn't attributed to the person who fired the IN round...

I'll be doing some testing, but try to avoid re-discovering, as much as I can. :P

#12 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 17 October 2005 - 05:07 PM

If you come up with new stuff, let me know (and I'll stick it in the wiki). Or you're welcome to wiki it yourself of course...

<{POST_SNAPBACK}>

But, but, but... you are doing such a nice job. :) Seriously though, I'm having some trouble learning how to use the darn formats in the Wiki. I suppose it just takes some time to learn the software like posting in forums such as this, but man. Tables are my lifeblood and it looks pretty bad that I am unable to figure out how to make them work in the Wiki. And when I get frustrated, I fall back doing things that make me happy - like testing! :wink1:

I am thinking that explosions might not be worth it, relative to experience points. Too variable and little or no reaction training. But I will try to get some harder numbers. For one thing, I'm realizing I haven't collected good data even on how much Muties are hurt by pistol shots. Much less blast damage, to compare against. I see the Damage Formula in the wiki and want to make sure I'm following it ok...

<{POST_SNAPBACK}>

Explosions from what? Thrown grenades, HE ammo, or both? Just by throwing the grenade you improve TA, and because it is a primary stat, it might increase a secondary also. As for firing HE shells, I'm not sure exactly how experience is calculated with a non-lethal shot. But surely a lethal kill should increase experience. In addition, since the shell was fired from a gun, there is always the chance that the soldier might get a reaction shot off.

By my calculations, a beginner level Muton should be getting a 0 to 31 damage range (with a normal pistol, dam=26) before armor is taken into consideration. Factor in armor (10 to 14) and the damage dealt should be between 0 and 21 or 17 - depending on rank. A HE blast of the same magnitude should do between (Zombie crunches some numbers)... 13 and 39 points of damage. Again, factoring in armor the health loss range would be:
  • Muton Soldier: 3-29

  • Muton Navigator: 1-27

  • Muton Engineer: 0-2
I suppose it's only fair to mention that these are theoretical since I have not got around to testing Mutons out in the on-going Damage Modifier Tests in the StrategyCore forums. But one thing is for sure: If the Official Strategy Guides tables are correct, then the numbers I presented above would also be correct.

In your example of incendiary blasts, do the numbers mean, how many turns the tile will be on fire? I'll be snarfing that to the wiki too, shortly.  =b

<{POST_SNAPBACK}>

Nope, it only shows the number of tiles affected:
     13531
      ↓↓↓↓↓
   1→   X
   3→  XXX
   5→ XXXXX
   3→  XXX
   1→   X


JFG and NKF have posted stuff on incendiary damage here and here. But I'm not entirely clear on how many damage points are dished out. There's damage from the initial explosion, but I can't seem to find any hard numbers on that... and then, if on fire or in fire, it's approx. 1 point per turn? And otherwise no damage except for the weirdness about, fire or smoke impact being repeated any time an IN round goes off anywhere on the map? NKF, Z, or anyone else who already knows? Again, I'm doing this with a mind toward experience points, so I'll be looking at that. I assume "on fire" damage isn't attributed to the person who fired the IN round...

<{POST_SNAPBACK}>

Again, from the Damage Modifier Tests I proved that impact damage (I use the term loosely here since it is not really the impact which causes the damage) from ANY incendiary round/rocket is always between 5 and 10 points of damage - if the unit catches on fire. If the unit fails to catch fire, it gets no damage.
This is therefore a discontinuous range: 0 or (5-10).

"Units that catch fire receive from 5 to 10 damage points per turn until the fire is out. Units standing in a fire receive from 1 to 12 damage points per turn."

I haven't tested the second part of the quote, but it should be correct - assuming a 100% susceptibility to fire. If a unit is more/less susceptible, it may affect these numbers - but again, this hasn't been proven yet. (Man, we still don't know 1% of how fire behaves). :)

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#13 NKF

NKF

    Commander

  • [Global Moderators]
  • PipPipPipPipPipPip
  • 4,798 posts

Posted 17 October 2005 - 10:27 PM

Wiki tables?

http://meta.wikimedi...wiki/Help:Table

You can also use html tables, but I think it would be a good idea to use the wiki version instead, just to keep things standard. I loathe using html in wiki coding, but sometimes it's the only way some things can be done.

For more information on editing wiki pages:

http://meta.wikimedi...nts#For_editors



- NKF
Lord High Generalissimo Ruler Supreme of Norm's Anti Pedant Society (NAPS).

Number of members: 1

#14 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 17 October 2005 - 10:38 PM

Thanks NKF! Much appreciated. I think that helps. :idea:

Now to practice... ;)

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#15 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 17 October 2005 - 10:55 PM

But, but, but... you are doing such a nice job.  Seriously though, I'm having some trouble learning how to use the darn formats in the Wiki.

Heh! I don't mind doing it... I just hope I can catch all your nuggets of wisdom. ;)

Eh, I tried wiki tables per se at first, and could not get them to look good. So now I just use "code" insertions. Judging from all the other pages, most everybody else feels the same way. I use a combination of Excel and Textpad to work around the hassles of formatting with spaces. Textpad has columnar (or box) cut and pasting, which makes it easy to e.g. space out columns or tighten them up. But thanks for that link, NKF!... I'll give wiki tables another try.

Unfortately, throwing only increases throwing accuracy, unlike any other primary stat. Also, the number of kills per se doesn't matter for Firing Accuracy, only the number of hits. If a hit also results in a kill, so be it. :) Have you seen the Experience part of the wiki lately?

I just did some tests on exposives of all kinds. In desert sand, everything you wrote in your table was correct, even the stuff that seemed contradictory relative to HE power. So that's that - there's something else going on. Oddly, some explosive types seemed to extend smoke farther than tile damage, and some didn't, consistently.

I found a few more things that I can't recall seeing anywhere. Stop me if you already know them :)

1) Explosion damage counts as a hit (toward Firing Accuracy and secondary skill increases) even if the alien wasn't hurt, just as long as they were in the blast radius. So it must've taken a run at the Damage Formula and come up empty. Of course, you only see undamaged guys at the edge of a weak blast, and then only occasionally.

2) Prox Grenades DON'T GIVE ANY EXPERIENCE POINTS for Firing Accuracy, unlike any other explosives. (This is with throwing it; you do still get a Throw experience.) I guess that their unusual nature precludes tracking hit attribution? Anyway, something to keep in mind if you want to get credit for hits (and kills)! :P

3) At the edge of weak blasts (low HE weapons), mutons always have little or no damage. But at the edge of the mighty Blaster Bomb, they will often be killed outright, even mutons. (Actually it can be fairly random - you can have aliens not be killed 4 tiles from the center, and everybody die 8 tiles away, or vice versa. But on average of course they all die, hehe.) Anyway it seems clear that BBs do more damage right up to the very edge of their radius. Also seen with e.g. Large Rocket. So this is another way in which there is something more going on than that quote about damage evenly falling off by 10... if this were true, you would expect pretty much the same damage at the edge of every type of blast, just before it fell off to zero.

All in all, they must've done something fairly funky in order to model the "blow around" effects you see any time there are (greatly) varying terrain types in one blast. Somewhere in their funky calculations must be a decrease by 10, since the OSG stated it so clearly. But exactly how, I have yet to understand.

I wrote down a bunch of figures for damage done by different explosives and will put them in something and get some stats and compare to what you and the Damage Formula say. Earlier tonight, before testing, I had mainly used Blasters, and assumed that explosions were too nasty any way you sliced it, even at the edges. But now that I see that weaker HE can leave edge guys barely touched (but still give Hit experience points), maybe it's back in the running :) But explosions still give little or no reaction experience though :(

Thanks for the clarification on Incendiary - I know I saw something like that before, but couldn't seem to find it again. I'll be working on that too... only 5-10 damage points? Could be sweet for building experience...

- MTR

#16 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 18 October 2005 - 12:51 AM

Have you seen the Experience part of the wiki lately?

<{POST_SNAPBACK}>

Yes, I have. Unfortunately, most of that info hasn't sunk into my noggin yet. :wink1:

I just did some tests on [explosives] of all kinds. In desert sand, everything you wrote in your table was correct, even the stuff that seemed contradictory relative to HE power. So that's that - there's something else going on.

<{POST_SNAPBACK}>

Whew! Well, at least it checks out. I did have a chance to recheck some "problem" areas on the table, but everything was fine as you mentioned.

1) Explosion damage counts as a hit (toward Firing Accuracy and secondary skill increases) even if the alien wasn't hurt, just as long as they were in the blast radius. So it must've taken a run at the Damage Formula and come up empty. Of course, you only see undamaged guys at the edge of a weak blast, and then only occasionally.

2) Prox Grenades DON'T GIVE ANY EXPERIENCE POINTS for Firing Accuracy, unlike any other explosives. (This is with throwing it; you do still get a Throw experience.) I guess that their unusual nature precludes tracking hit attribution? Anyway, something to keep in mind if you want to get credit for hits (and kills)! :P

<{POST_SNAPBACK}>

Now that you mention it, I do remember #1. Makes sense though, because a hit is a hit is a hit. But the Proxy, well, that's news to me. Interesting! :)

3) At the edge of weak blasts (low HE weapons), mutons always have little or no damage. But at the edge of the mighty Blaster Bomb, they will often be killed outright, even mutons. (Actually it can be fairly random - you can have aliens not be killed 4 tiles from the center, and everybody die 8 tiles away, or vice versa. But on average of course they all die, hehe.) Anyway it seems clear that BBs do more damage right up to the very edge of their radius. Also seen with e.g. Large Rocket. So this is another way in which there is something more going on than that quote about damage evenly falling off by 10... if this were true, you would expect pretty much the same damage at the edge of every type of blast, just before it fell off to zero.

All in all, they must've done something fairly funky in order to model the "blow around" effects you see any time there are (greatly) varying terrain types in one blast. Somewhere in their funky calculations must be a decrease by 10, since the OSG stated it so clearly. But exactly how, I have yet to understand.

<{POST_SNAPBACK}>

Hmmm... Perhaps the game does use the 10 drop-off in damage up till 1 or so tiles inside the blast. Then it "dumps" the remainder of the damage on the final tile or so. *Zombie thinks for a bit* Nah, that would only be 100 or so points of damage for a BB explosion, and it wouldn't account for the damage potential with Large Rockets either.

I think that in this case, we need to call in a favor from our good friend NKF:

Do you think you could rustle me up one of your legendary saved games? Nothing too fancy or elaborate: just a mission with 1 soldier carrying a Blaster Launcher (and a BB or two), and around 13 soldiers. Those 13 should have no armor (0) and have health bumped up to 200 (maybe more if possible). I'd really like to do some testing to see exactly what is happening inside the blast radius. Only then can I figure out exactly what is going on. (Barring that, just tell me how to MS-edit soldier.dat to remove the standard Kevlar armor, and bump health up to a manageable level so I can create my own mission).

Thanks! B)

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#17 NKF

NKF

    Commander

  • [Global Moderators]
  • PipPipPipPipPipPip
  • 4,798 posts

Posted 18 October 2005 - 02:47 AM

I can easily come up with such a scenario, but I need a bit of time - I have a number of things on my plate at the moment. Procrastination tends to revisit you and bite hard!

If you're in a hurry, here's what you can do:

Your savegame's unitref.dat file needs to be opened in binary mode with 124 columns,

Cols 19 to 23 and cols 30 to 34 - for current and max armour. You'll want to set them all to 0 (use ctrl+p+2 to insert a 0). another way of identifying the armour is if you're wearing a flying suit. Look for the following sequence of letters, twice, and note the case: nZZPF

Health can be found at column 14 and 27. To enter a value of 100, enter small caps "d". To enter 200, alt+2+0+0.

For the aliens, I highly suggest you give all but 1 critical wounds. Fill, say, column 66 with 100 wounds to kill it off. Then set the remaining unit's energy recharge to 0 - that's at column 36.

Let me know if you still need my assistance.

---

Speaking of the proximity mine not giving any experience: What happens to the experience of the unit at index 0? It could be an unattributed kill, like death by roasting, but it's worth finding out.

- NKF
Lord High Generalissimo Ruler Supreme of Norm's Anti Pedant Society (NAPS).

Number of members: 1

#18 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 18 October 2005 - 12:20 PM

Thanks for pointing out what to do, NKF - Otherwise I was unsure I would've gotten it right. I can fix it up for Z (attached). 250 HP and 0 armor. This is from my explo. test so there are tons of diff. explosives in the piles. I didn't kill the critters cuz you can easily do it (just in case you want to test them too), but I did set the lone Navigator to 0 recharge. There's also an FBL tank in the Craft, for a little extra fun ;)

NKF, good point... now I remember reading about unattributed kills. But unfortunately I had a Hovertank in position 0. It's counters did not change but its a tank (should they?). Nobody else's "erroneously" changed either. If anybody has a savegame with prox mines ready to kill somebody (and no tank!) I can quickly test it. I don't use Prox Gs very much so I only had the one explosion test setup. Still have yet to hack stuff into units' hands.

Death by roasting is unattributed? :( I still want to play with fire some vs. experience counters... ultimately we can make sure the wiki says these kinds of things in relevant Experience text. I'm still ISO better stat training... no luck just yet.

I've just started analyzing my blast test results and noticed something that I can't recall seeing: Health damage has an extremely specific correlation with armor damage. Specifically, multiply armor damage by 10, then subtract from 1 to 10. Example: A muton with 7 armor damage will lose 60 to 69 HPs. Entirely consistent for all 230 samples with diff. kinds of explosives and up to 11 AC damage. There are two logical exceptions: 1) If armor is totally gone, the sky's the limit. 2) If there is exactly 1 armor damage, there will be at least 1 HP damage (instead of 0-9 per usual rule). Guess they wanted to make sure somebody got hurt.

The wiki "Damage Formula" page talks about armor damage, but I don't recall seeing anything this specific. Also, this is for explosions; dunno about shot damage. Anyway... continuing to analyze explo results...

BTW Z the Hi-Ex also kills most guys, even near edge of blast. Just like Large Rocket, and very unlike humble grenade.

Attached Files



#19 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 18 October 2005 - 08:37 PM

First of all, many thanks to both NKF and MikeTheRed for taking the time to provide me with a decent saved game to work with. I really appreciate it! :)

I've just started analyzing my blast test results and noticed something that I can't recall seeing: Health damage has an extremely specific correlation with armor damage. Specifically, multiply armor damage by 10, then subtract from 1 to 10. Example: A muton with 7 armor damage will lose 60 to 69 HPs. Entirely consistent for all 230 samples with diff. kinds of explosives and up to 11 AC damage. There are two logical exceptions: 1) If armor is totally gone, the sky's the limit. 2) If there is exactly 1 armor damage, there will be at least 1 HP damage (instead of 0-9 per usual rule). Guess they wanted to make sure somebody got hurt.

<{POST_SNAPBACK}>

So what you are saying is that: Health Drop = Armor Drop * 10, right? That's nice to know.

Okay, onwards to Blaster Launcher testing. After fooling around with the afore-mentioned saved game to get it the way I wanted it, I started to gather some numbers. Some preconditions need to be mentioned to explain my rationale though:
  • I only tested three tiles out from ground-zero, but not ground-zero itself due to the fact that a unit standing at the center might sop up some of the blast strength.
  • The tests were carried out at the very right (East) of the map for two reasons. First, only half the blast would be on the map. Secondly, if the BB would somehow miss the target, it would harmlessly fly off the map and not strike unanticipated targets.
  • The Blaster Bomb was fired from Level 1 to Level 0 (the ground) to mirror a grenade blast as close as possible. In addition, this eliminated the need for a ground-zero target which could skew results.
From past experience, I know that 200 or more gathered values in X-COM are needed for a basic understanding of the mechanics. I picked 250 values just to be safe. A full 1000 values is undoubtedly needed for a better insight. I subtracted 250 from the current health numbers to show damage inflicted.

Dist from GZ:    1       2       3
Minimum:         96      85      85
Maximum:        250     250     250
Range:          155     166     166
Median:         198     172     172
Mode:           250     250     250
Ant. Ave¹:      173     168     168
Ant. Ave²:      190     180     170
Act. Ave:       190     175     169
Deaths:          50      21      16
% Deaths:       20%      8%      6%
¹ Found by taking the average of the min and the max.
² Average predicted by the OSG.

As you can see, the 10 point damage drop-off does indeed seem to be in effect (at least for the first three tiles from GZ). Also, the farther a unit is from the blast center, the lower the probability that the unit will die (like we didn't already know this). One tile away from ground-zero and the unit had a 20% chance of getting killed, while it went down to 8% for 2 tiles out and 6% for three tiles out.

A haven't compiled a frequency table yet, but it is pretty obvious to me that the probability of getting killed happens far more often than anticipated. I calculate that death should occur around 1.5-1.6 times out of 251, while actual results show 50, 21 and 16 times out of 251. Pretty big difference, eh?

Anyhow, I'm off to gather the remaining 750 values necessary for a complete understanding of what happens near ground-zero. After that, I'm going to check (with a lower sample size of 100) to see if a soldier standing on GZ plays a role in damage dealt to units further out. If it doesn't matter, then I can safely place all 12 or 13 soldiers inside the blast radius and collect another 1000 values/soldier to check all the tiles out. Wish me luck! :)

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#20 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 18 October 2005 - 09:35 PM

As you can see, the 10 point damage drop-off does indeed seem to be in effect (at least for the first three tiles from GZ). Also, the farther a unit is from the blast center, the lower the probability that the unit will die (like we didn't already know this). One tile away from ground-zero and the unit had a 20% chance of getting killed, while it went down to 8% for 2 tiles out and 6% for three tiles out.

A haven't compiled a frequency table yet, but it is pretty obvious to me that the probability of getting killed happens far more often than anticipated. I calculate that death should occur around 1.5-1.6 times out of 251, while actual results show 50, 21 and 16 times out of 251. Pretty big difference, eh?

<{POST_SNAPBACK}>

Never mind. I think I stumbled upon the answer myself. :)

From the damage modifier tests, explosive ammo is calculated a bit differently from other ammunition. So according to the equation I calculated months and months ago, the Blaster Launcher should be doing anywhere from 100 to 300 points of damage on ground-zero. Assuming that the 10 point drop-off affects both min and max damage, the damage per tile away from the center should be the following:

Dist from GZ          Damage Range
       0                100 - 300
       1                 90 - 290
       2                 80 - 280
       3                 70 - 270
       4                 60 - 260
       5                 50 - 250
       6                 40 - 240
       7                 30 - 230
       8                 20 - 220
       9                 10 - 210
      10                  0 - 200
It sort of follows the results I gathered so far: 90-290 ≈ 96-250. But, the health of the soldiers is only 250! Therefore the ones closest to GZ are sometimes getting dealt more damage than their health. That increases the chances for death:

Dist from GZ        Max Damage       % to Die
       0                300             25%
       1                290             20%
       2                280             15%
       3                270             10%
       4                260              5%
       5                250              0%
       6                240              0%
       7                230              0%
       8                220              0%
       9                210              0%
      10                200              0%
Ah ha! That mirrors my results exactly! :)
Still, I do not know what exactly happens beyond 10 tiles out. But one thing is a fact: assuming a maxed-out soldier of 61 health, death should be guaranteed at GZ with no armor, and there should be a 30% chance of living at GZ+10. If the max drops by 10 for the remainder 2 tiles, your naked soldier should keep on kicking 32% and 34% of the time.

Unfortunately, it seems that the "ground sopping up damage theory" is incorrect for the time being. Perhaps the BB is hard-wired to detonate a certain radius? That's my best guess at the moment. B)

[Edit] Sorry, that equation should be: Health Drop = (Armor Drop * 10) - INT(Rnd# * 11). Right? :doh:

- Zombie

Edited by Zombie, 18 October 2005 - 10:29 PM.

The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#21 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 18 October 2005 - 10:35 PM

So what you are saying is that: Health Drop = Armor Drop * 10, right?

Very close to that. 10 * Armor Damage, then subtract 1-10. Or, if you will, range of 10(AD-1) to 10*AD-1. With exception that it can't be less than 1 for AD=1. And excluding cases where armor was totally blown out. Here's the data:

Armor  Min   Ave.  Max
 Damage HPs   HPs   HPs    N
    0    0    0.00    0   12   <- top line was in blast radius
    1    1    5.13    9   39      and did get "hit" experience
    2   10   15.14   19   44
    3   20   24.51   29   37
    4   30   34.25   39   32
    5   40   44.27   49   22
    6   50   54.39   59   23
    7   60   62.43   66    7   <- starts breaking down here
    8   70   72.14   79    7      due to lack of data... I was
    9   83   85.50   89    4      trying to sample edge of 
   10   91   91.33   92    3      blasts, not center
   11  107  107.00  107    1
It's based on damage from a single explosion, but some preliminary testing vs. shots shows that it holds true for shots, too. It was for mutons and silacoids, in case that matters.

Note that you cannot get this precise estimate if you compare armor damage to health damage after several hits or shots. E.g. at the edge of std. grenade blasts you can take several blasts that only do 1 or 2 health damage each and 1 armor damage each time, and wind up with 3 armor damage but only 5 HPs gone. But if I look at each hit individually (before and after), I see it every time. So far, anyway (fingers crossed). Anyway, it holds for all explosive damage, and light weapons. Don't see why it wouldn't for heavy weapons.

Here is something I put together that ought to be of interest one way or another. A secondary check, if you will:
 DFE   HE   MinHPs AveHPs  MaxHPs   N
 
   1   Gren     0     1.27     5    15
       AC-HE    0     2.50     8    10
       HC-HE    2     7.90    22    10
       SR       8    23.33    42    15
       LR      10    24.80    48    15
       BB      37    79.44   125    27
 
   2   Gren     1    11.60    20    15
       AC-HE    4    12.10    21    10
       HC-HE    7    21.10    37    10
       SR      12    33.60    57    15
       LR      15    36.60    60    15
 
   3   Gren     5    18.47    34    15
       SR      17    43.40    70    15
       LR      20    47.12    73    17
 
   4   LR      30    60.50    91    2
 
   5   LR      30    56.50    83    2
DFE = "Distance from edge of blast", where just outside the blast radius is 0, and just inside is 1.

Sorted on ascending distance from edge, then ascending average damage by bomb type. So damage goes up quickly as you move inward (higher DFE). This test data only applies to Muton Soldiers hit on a side (not close enough for underarmor), in Desert terrain. In my game, Muton Soldiers have stats of 10 for side armor and 125 HPs. (Armor differs by game difficulty, doesn't it?) Armor breakthrough and damage that killed is included in this table (but not the first one).

I only tested Blasters at their edge because I figured they weren't what I was after... also note they were the only one to kill Mutons outright. And that's at their edge.

My Ns are very small, but they showed me all I needed to see: the edges of std. grenade blasts are probably the gentlest hit possible in the game (for which you can get experience - correct me if wrong!). It is entirely possible to have every man on a team get lots of hits from Mutons, probably the 11+ needed for max stat increase, for every soldier. If combined with the fact that Mutons can easily give you all the reaction shots you need (with std. pistols)... well, you can see where I'm going with this. But it may be too much of a hassle for anyone but the most dedicated to bother with. (Not hardly as much fun as pure firing squads, and they do take some time to set up just by themself, already.) Anyway, I'll be mulling it over. At this point it's looking like there's not gonna be a free lunch hidden away.

NKF and Z, I played with incendiary some... and, yep, never got an experience point for 'nothing. Not unless my tank (slot 0) is eating up unattributed hits.

Z, I just saw your new post... that sounds about concerning Health Drop, but don't make me try to remember how to get Int(rnd*11) to work just right... will that have a min of 1? Anyway, I posted up the actual data...

Speaking of which, look what you just pasted up! It'll take me a little while to think what all it implies, but basically... it looks SWEET! :Cucumber: Looks like you found the "missing link" as to how it "decreases by 10", but still does substantial damage. Gotta love it when things fall together. Thanks!

Got a url for your old equation? I wonder how readily the results might be extended to other HE types.

As for "hard wired to detonate", you can sort of see a couple things through the mist. If the AVERAGE blast strength is 200 (as it is for you at GZ) and it decreases by 10 per tile... since there's no reason for the game to give everything away on its Ufopedia screens, I'd say they did a pretty good job of saying how it works, without going into all the details, eh? There's a lot of variability in other damage equations and e.g. the double damage and weapon susceptibility things, shrug.

But wait... why are blast radii different sizes? Hmm...

Getting late here... great stuff!

#22 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 18 October 2005 - 10:58 PM

Z, I just saw your new post... that sounds about concerning Health Drop, but don't make me try to remember how to get Int(rnd*11) to work just right... will that have a min of 1? Anyway, I posted up the actual data...

<{POST_SNAPBACK}>

Son of a gun. Sorry again! Final equation: Health Drop = (Armor Drop * 10) - INT((Rnd# * 10) + 1). Yep, that's the one. :doh: :doh: :doh: (It's late here too, and I messed up the order of operations).

Got a url for your old equation? I wonder how readily the results might be extended to other HE types.

<{POST_SNAPBACK}>

I found the equation with HC-HE shells so it should work with all HE types. Basically, it's a "one-size-fits-all" equation in X-COM. Quite a complex little bugger to figure out, so if you have any trouble understanding, let me know. Here is a link to that particular post (in the StrategyCore forums): Clicky

Please note that the equations are only valid for 1x1 units. I'm still working on the 2x2 units after a very long hiatus (apologies to NKF). In addition, there is still that nagging question of how stun is calculated according to health drop. I'm close, but it still needs some work. :)

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#23 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 19 October 2005 - 08:55 PM

Thanks for the link! Yes, it all makes sense now. I'll try plugging in the various HEs and see how it compares to my "DFE" data and the armor modifier. (I personally am still getting a handle on how armor stops damage.) Perhaps I'll post to the wiki, damage dropoff over distance for all HEs like you just did for BBs. Will be somewhat busy the next few days, though...

Earlier I had meant to add, I entirely understand why you're concerned about how units in a blast are arranged. Given how the blast alters its pattern based on terrain, obviously any units "behind" (or maybe next to) others are liable to get different damage. In fact if you line up a solid wall of units, esp. 2-3 deep, they are liable to "block" the blast "downwind" from them, and leave undisturbed ground.

Also is damage to tiles stored anywhere? Would this be the "4th" bit in MAP.DAT? With the reg. grenade, at first Desert just gets a dimpled look. With more grenades, dimpled desert starts turning into "raw earth" from GZ outward. (I'm just wondering if that also might be some indicator of damage needed to produce effects, to wit, can one choose a particular HE setting and predict just what tiles will be affected how, and where? If effects on tiles can be understood this well, then also the effects of "blast around" etc. can be visualized as a picture painted in the sand as it were. And thus get a better handle on blast propagation in uneven terrain. Just something I'm toying with.... don't know if I'll do anything with it. But it'd be hard to do without understanding tile effects. Might help to read them directly from game files, esp. if "tile damage level" per se is stored somewhere.)

Also it ties into, blast radius (or lack thereof). Why is it so short in alien bases? BTW, want me to hack an alien base savegame for you, according to NKF's instructions? If so, how many aliens would you like left alive (11 Mutons, 2 Silacoids, and 1 Celatid)? It currently only has Blasters and Alien Grenades...

Have you ever had any thoughts concerning how some explosions consistently produce smoke that is outside the blast radius (as determined by terrain damage) whereas others don't? The grenade has one of the oddest patterns (in desert). Shortly I'll give a look at whether any actual hits occur there (where there is only smoke but no tile impact).

Whenever you're able to post stun bomb stuff, that's great. I intend to look at it sometime soon. It may be the one last trick up the sleeve relative to getting experience.

Sorry for all the questions. Too much to discuss at once, even if you know answers for everything I've asked! Anyway, more food for thought...

Mike

P.S. If you want it (or anyone else), I have a MS Access mdb that can directly read Unitref data from a savegame at the punch of a button. This can make e.g. looking at or capturing health damage, stun damage, etc. a lot easier in some situations. Especially when collecting a lot of data; you can stick each trial directly into a growing dataset. But it's very homegrown and would be real hard to work with if you don't know Access (and possibly VBA).

#24 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 19 October 2005 - 10:16 PM

NKF - Although I don't have a savegame that does have a prox grenade but doesn't have a tank, something else occurred to me: I primed a regular grenade and then just dropped it. Then, the tank in slot 0 DID get credit for the "unattributed hits". So tanks will take unattributed credit. (I did not see it get any with Incendiary, though... is that supposed to give unattributed Hits credit?)

Then I screwed around some more... accidentally set off a prox grenade... lo and behold... The unit that SETS OFF the prox g gets the Hits credit. NOT the person that primes it.

That's why I wasn't seeing any points previously - I was tossing the prox, and aliens were setting it off. But aliens don't get experience - only XCOM units. When I set off my own (ouch!), I get the XP.

Gives a whole new meaning to "learning from experience", eh? LOL

#25 NKF

NKF

    Commander

  • [Global Moderators]
  • PipPipPipPipPipPip
  • 4,798 posts

Posted 19 October 2005 - 10:53 PM

That's... well, good to know. So it's a good idea to not treat proximity mines as a religion, just as a weekend hobby, or somesuch.

The unattributed kills for grenades isn't actually unattributed. See, every object in the battlescape gets assigned a 'last owner' flag so that it can award experience to the proper unit for an area-effect weapon that isn't as instant as, say, a rocket. Unfortunately, all items in the game default to '0' until thrown. Just dropping it will have all credit go to the unit in slot 0.

So, priming your grenade and dropping it will simply result in the soldier in slot 0 (or the tank), getting credited for the kill. Now if you want someone other than the soldier/tank to get credit for dropping the bomb, you must have the person you want to take the credit thrown the grenade at least once. Basically, you could have soldier A prime and throw the grenade, then get suicidal soldier B to pick it up and barge into a Medium scout door. After the fireworks, Soldier A will be credited for any aliens hit in the detonation.

----

Re: Smoke beyond blast radius: Perhaps the smoke is there because damage was done, but the terrain 'armour' was able to absorb it all, and not get affected by the blast. Light smoke particles were created anyway - it doesn't know that damage wasn't done.

Or perhaps smoke does get created a bit beyond the blast area on purpose.

Here's a potential spanner I'd like to toss into the works: Mountain terrain. Not sure what that signifies, but I thought I'd like to mention that anyway because Mountain Terrain weakens all previously indestructable structures - such as UFO walls and Skyranger landing gear and ramp - on the lowest level. ;)

- NKF

Edited by NKF, 19 October 2005 - 10:57 PM.

Lord High Generalissimo Ruler Supreme of Norm's Anti Pedant Society (NAPS).

Number of members: 1

#26 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 19 October 2005 - 11:10 PM

Earlier I had meant to add, I entirely understand why you're concerned about how units in a blast are arranged. Given how the blast alters its pattern based on terrain, obviously any units "behind" (or maybe next to) others are liable to get different damage. In fact if you line up a solid wall of units, esp. 2-3 deep, they are liable to "block" the blast "downwind" from them, and leave undisturbed ground.

<{POST_SNAPBACK}>

Well, just for giggles I set up this exact scenario today to see what happens. Going strictly by ground damage, the soldiers in a row didn't "protect" terrain in back of them. In short, the same radius appeared with the same damage profile. Still, that doesn't mean that the soldier’s health has changed. More testing needs to be done to confirm or deny anything yet.

Also is damage to tiles stored anywhere? Would this be the "4th" bit in MAP.DAT? With the reg. grenade, at first Desert just gets a dimpled look. With more grenades, dimpled desert starts turning into "raw earth" from GZ outward. (I'm just wondering if that also might be some indicator of damage needed to produce effects, to wit, can one choose a particular HE setting and predict just what tiles will be affected how, and where? If effects on tiles can be understood this well, then also the effects of "blast around" etc. can be visualized as a picture painted in the sand as it were. And thus get a better handle on blast propagation in uneven terrain. Just something I'm toying with.... don't know if I'll do anything with it. But it'd be hard to do without understanding tile effects. Might help to read them directly from game files, esp. if "tile damage level" per se is stored somewhere.)

<{POST_SNAPBACK}>

I am familiar with the scorched earth/cratered earth scenario you mention. Need I say it? I made a table of that info too a while back too. But I only tested it on desert sand:

Weapon Type     Scorched     Total tiles damaged
Grenade:            0          (9 normal)
Proxy:              3         (13 normal)
Alien:              7         (13 normal)
Hi - X:            11         (13 normal)
Small R:            3          (9 normal)
Large R:            9         (13 normal)
Blaster B:         23         (23 normal)
AC - HE:            0          (7 normal)
HC - HE:            0          (7 normal)

For the most part, the scorched tiles are always smaller then total tiles damaged (last column). As for where the actual damage numbers are stored in the game files, I couldn't tell you. Daishiva (who wrote a map view program for X-COM) would be the man to ask. Unfortunately, I tried to get on his website this week to download the program, and it seems to be down for the count. I know Bomb Bloke wrote a map program too, but I am unsure whether it displays armor/health/constitution numbers for tiles/objects.

Also it ties into, blast radius (or lack thereof). Why is it so short in alien bases? BTW, want me to hack an alien base savegame for you, according to NKF's instructions? If so, how many aliens would you like left alive (11 Mutons, 2 Silacoids, and 1 Celatid)? It currently only has Blasters and Alien Grenades...

<{POST_SNAPBACK}>

Hold your horses there, Mike! I'm not even finished determining what happens on desert sand, let alone alien base tiles. :) If you want, go ahead and create it. But I can not pin down an exact time when I will get done with the current BB tests.

Why is the blast radius shorter in an alien base? Alien Alloy floors are incredibly tough, for one. But, I'm not totally convinced that terrain doesn't play a role. That's why we need to do a good job figuring out how Blaster Bombs do damage on one tileset first. When that is done, we can then move on to the alien base and mountainous terrain like NKF mentions. (This is how I test things, so please humor me: eliminate unnecessary variables, collect values, analyze numbers, and finally try to base this against other things and possibly formulate an equation to describe all possible scenarios. Following a strict set of rules like this reduces mistakes and disinformation – both which I hate more than anything). :D

Have you ever had any thoughts concerning how some explosions consistently produce smoke that is outside the blast radius (as determined by terrain damage) whereas others don't? The grenade has one of the oddest patterns (in desert). Shortly I'll give a look at whether any actual hits occur there (where there is only smoke but no tile impact).

<{POST_SNAPBACK}>

Just from observation in the current BB tests, smoke only occurs on damaged tiles. The smoke will last 5 rounds after the impact. But I haven't really seen smoke propagate outside the blast radius. Obviously something is going on with the grenade. What exactly, I don't know, but I'll look at this a little closer when I get to the grenades. =b

Whenever you're able to post stun bomb stuff, that's great. I intend to look at it sometime soon. It may be the one last trick up the sleeve relative to getting experience.

<{POST_SNAPBACK}>

I suspect the stun bomb works in the same manner as HE blasts do. Both are area effect weapons, so it stands to reason that if an alien is caught in the blast the soldier will get experience points. Hmmm... I recall one of my soldiers had his stats bumped up after catching an alien in a stun bomb explosion. Because he ran out of Stun Bombs, another soldier had to kill the alien. The guy who fired that one stun bomb saw a stat increase after the mission.

Sorry for all the questions. Too much to discuss at once, even if you know answers for everything I've asked!

<{POST_SNAPBACK}>

As you can see, I don't know the answers to all your questions. But I am more than happy to help you solve them. :)

P.S. If you want it (or anyone else), I have a MS Access mdb that can directly read Unitref data from a savegame at the punch of a button. This can make e.g. looking at or capturing health damage, stun damage, etc. a lot easier in some situations. Especially when collecting a lot of data; you can stick each trial directly into a growing dataset. But it's very homegrown and would be real hard to work with if you don't know Access (and possibly VBA).

<{POST_SNAPBACK}>

Actually, Bomb Bloke was kind enough to write me a unitref logging program to make gathering those numbers easier for the damage modifier tests. Sure beats writing out all those values out by hand and then entering them into a spreadsheet like I did for 5000 or so trials. If anything, I should be offering that logging program to you. For now, I'm set Mike. But thanks for asking! B)

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#27 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 20 October 2005 - 06:30 PM

NKF, thanks for that very clear explanation of hit attribution with grenades!

Grenades are pretty weak. Someone in an FS can handle it, just a couple of squares away. (I haven't done the math yet to see if they can even take GZ unscratched.) Combined with the info on attribution, it means that if somebody wants to pump up Hits experience by having aliens stand at the edge of a grenade blast... if one of your soldiers throws a grenade but misses the center of the circle (of which aliens are arranged along its edge), another soldier can be close to GZ, pick up the grenade, and drop it in the center. Then walk a couple tiles away. The bad thrower will still hit all the aliens.

Sometime soon I hope to get a mathematical model of just how well this works for experience points vs. e.g. various aliens. But I'm still collecting data and coming to grips with damage, armor levels, etc. Zombie's stuff is great but I have yet to extend it to edge of grenade blasts, vs. different terrains, etc. etc.

I guess the fun is in the journey, too :)

Zombie - agreed re: coming to know one experimental setup 100% before moving on to others. Can we get you anything? Tea, coffee? ;)

Most HE blast zones have one extra tile damaged out at the edge, "directly lateral" (north, south, east, west) of GZ. But not the grenade... it has 3 along its lateral edges.

I tested whether there were any hits on aliens beyond the tile-damage edge of grenades. Thrower's Hit counts seem to me to be the most sensitive indicator of blast effect, considering how grenades can do zero actual damage to a Muton, but still cause the grenadier to get Hit XP (inside the edge of a grenade blast).

Result: No XP. So there must be one of two things happening: either the tile has totally absorbed (zeroed out) the blast without being damaged (as you suggest, NKF), or there's just something odd going on with the smoke (and it's not really the blast radius). Maybe the blast gets a wee bit of rounding that the smoke calculations don't? (Remember that grenades don't get that one-extra-lateral tile that most patterns do.)

Either way, for my purposes of trying to find the least-damaging ways to get XP, I now know that nothing's happening where there's only smoke.

the soldiers in a row didn't "protect" terrain in back of them.

In double checking, it was two rows of four soldiers bunched together. But the place where the blast damage was different was the bottom of ramp for Avenger (directly behind them), which otherwise would've been the edge of blast pattern IIRC. Dunno if the ramp affects it. I'll see if I can reproduce or see any other such potential "protection". Unless/until I do, don't worry about it :)

Would anyone happen to know whether damage, per se, will be exactly the same for an alien in a given situation, whether they are mind controlled or not? Some of my repetitive testing of amount of damage will be easier if I can MC the aliens. I realize that 1) I won't get XP then, and 2) aliens can then get Fatal Wounds. But otherwise, should the damage be the same? My impression is that it is, but it's pretty variable so it's hard to tell.

Arg NKF... I don't want to think about Mountain terrain just yet ;) Maybe later after more common terrain is well understood. Can a Hvy Plasma shoot through outer UFO hulls in mtn. terrain?

I'm going to test extent of stun bomb radius etc. now...

BTW has anyone noticed that Silacoids damage tiles that they move over? I can't recall noticing it before (or reading about it), but in the testing I'm doing in the Desert, if I let him move on his own, more often than not he leaves bare earth as he passes over the sand.

#28 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 20 October 2005 - 07:42 PM

Notes on stun bombs... apologies for what's already known...

Stun bombs have a diameter of 11 tiles as shown in the image. This is for Desert.
StunBombPattern.gif

IF it's handled as an explosion, note that it's not like any others; no HEs have diameter=11 for desert.

Soldiers do get Hit experience for any alien in the "blast" diameter, even if the alien got 0 stun.

Stun amount is definitely higher at GZ and decreases toward edge. But it can still be quite high at the edge; it is quite variable all over the diameter.

Finally, higher armor means greater chance of no stun at all (but hits are stilled recorded).

My FS guys at center of blast got no stun damage about half the time; the other half, it varied up to approx. 40 (but was usually pretty low). This is on only a half dozen tests.

Even with direct hits, none of my guys were stunned enough to go down - FSs are very protective of stun bomb damage.

But it was a small sample... sometimes it will be enough, esp. if their health is low. My troops are experienced, with high natural HPs.

All in all, it seems to follow explosion dynamics quite closely.

Edit: Conclusion re: using Small Launcher for Firing Accuracy experience:

The grenade is much better than SL.

Even at its edge, Mutons can get up to 60 stun with SL. So within only a few stun bombs (which can be a hassle to aim properly), aliens can go down so deeply that it'll be a big hassle, and run you through many medkits, if you're trying to keep them all awake through a cycle of letting every one of your soldiers make a shot with Small Launcher at grouped aliens.

In comparison, each grenade only does 0-5 hits on Mutons at edge of blast, a much more "sustainable harvest", and grenades can always be accurately placed. (Even if you miss the throw to the center of some aliens set around the edge of blast radius, someone can pick up the frag and place it right. Frags barely touch FSs, and picking up and dropping does not change "ownership" of hit XP.) Finally, stun bombs are expensive to manufacture (need 1 Elerium; cost 7k to make and sell for 15k), which makes them an awful economic choice for intensive skill training over the long run. Grenades cost $300 and can be delivered anywhere in the world overnight.

Edited by MikeTheRed, 20 October 2005 - 07:56 PM.


#29 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 20 October 2005 - 07:56 PM

Zombie - agreed re: coming to know one experimental setup 100% before moving on to others. Can we get you anything? Tea, coffee?

<{POST_SNAPBACK}>

Caffeine? Nope, trying to cut down. Just water, thanks! Though tomorrow it might be a beer. The one thing I would really like is more hours in a day. Can you get me that? :wink1:

I'll see if I can reproduce or see any other such potential "protection". Unless/until I do, don't worry about it :)

<{POST_SNAPBACK}>

Well, I'm still worried about it. For my tests, I moved the entire compliment of soldiers away from the Avenger and lined them up right to left from the edge. I did not see a change in blast radius, but that’s not to say that all those soldiers are influencing explosive damage, especially near the edge. As I've said before: "eliminate variables". =b

Speaking of variables - I decided to run a test with 4 soldiers grouped around GZ. My Blaster Launcher guy would fire the BB to a corner above the soldiers, and then send it on an angle down to GZ. Problem was, the blast radius was not centered on the intended GZ. Nope, it was shifted one tile in the last direction which the BB followed. So if it was headed down on an angle to you, the blast was shifted closer to you by one tile. I'm glad I recognized this fact, because otherwise all those trials would be for nothing.

Right now, I started a brand new trial with 4 soldiers clustered around GZ, but my Blaster Launcher guy will only be firing left-to-right and down instead of straight ahead and then diagonally down. Stupid CE Blaster Bomb bug. If it wasn't for that, my tests would be done. :(


Would anyone happen to know whether damage, per se, will be exactly the same for an alien in a given situation, whether they are mind controlled or not? Some of my repetitive testing of amount of damage will be easier if I can MC the aliens. I realize that 1) I won't get XP then, and 2) aliens can then get Fatal Wounds. But otherwise, should the damage be the same? My impression is that it is, but it's pretty variable so it's hard to tell.

<{POST_SNAPBACK}>

Fatal wounds do nothing to affect health lost. As far as I could tell from some testing this, it’s a random roll. What the fatal wound range is has yet to be determined. I suspect it is based on the total health of the unit in question.

BTW has anyone noticed that Silacoids damage tiles that they move over? I can't recall noticing it before (or reading about it), but in the testing I'm doing in the Desert, if I let him move on his own, more often than not he leaves bare earth as he passes over the sand.

<{POST_SNAPBACK}>

Sure, alien-controlled Silacoids can burn through practically anything - including alien alloy floors! But, MC them and they loose this ability for your turn. It's a good way to search for these guys - look for the burnt tiles. Like breadcrumbs to grandma’s house, it leads you right to them. :)

As for your Stun Bomb trials, I figured as much. It acts like an explosive in terms of area. But damage? It is definately different. Compare and contrast the alien grenade (HE=90) against a Stun Bomb (Stun=90). The alien grenade will have a min damage of 45 and a max of 135 at GZ. But the Stun Bomb is different: it has a min of 0 and a max of 180. (This assumes a 100% susceptibility rating to both types.

Using a flying suit? They have 20% resistance to Stun (or 80% susceptibility). That comes out to 144 max damage. This may or may not account for the apparent difference in blast radius.

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#30 Danial

Danial

    Captain

  • Moderators
  • PipPipPipPip
  • 778 posts

Posted 20 October 2005 - 08:00 PM

According to Daishiva's MapView program, all desert ground tiles have an "Armour" rating of 5. The "Death Tile" then has 25! So in my eyes, that means 5 to burn and 30 to scorch. And just so you can compare, a UFO floor tile has 50 armour and its death tile has 50 on top of that.

Back to the desert though, the hills have 30, and plants have 15-30 depending on their size.

There's also a "HE Block" value of 5 on the hills and 4-12 on the various foliage. I'm assuming that causes a damage drop-off behind those items. Absorbing that blast.

Not sure if any of that helps your calculations in any way, but I thought I'd let you know anyway :)

I've attached Daishiva's MapView if anyone wants it. The Armour and HE Block ratings are kept in the terrain's MCD file. Make sure you change the 'Paths' to the games or the program won't work:

Attached Files


Edited by Danial, 20 October 2005 - 08:21 PM.


#31 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 20 October 2005 - 08:32 PM

According to Daishiva's MapView program, all desert ground tiles have an "Armour" rating of 5. The "Death Tile" then has 25! So in my eyes, that means 5 to burn and 30 to scorch.

Also, hills have 30, and plants have 15-30 depending on their size.

There's also a "HE Block" value of 5 on the hills and 4-12 on the various foliage. I'm assuming that causes a damage drop-off behind those items. Absorbing that blast.

Not sure if any of that helps your calculations in any way, but I thought I'd let you know anyway :)

I've attached Daishiva's MapView if anyone wants it. The Armour and HE Block ratings are kept in the terrain's MCD file. Make sure you change the 'Paths' to the games or the program won't work:

<{POST_SNAPBACK}>

Great stuff Danial! Thanks for attaching Daishivas MapView program. Much appreciated! :)

Hmmm... I'm trying to mathematically determine the blast diameter for the Stun Bomb. Assuming a simple ratio of 13/90 for the alien grenade, the Stun Bomb should have a diameter of 26 tiles with 180 damage, or 20 tiles with 144. Neither works. Something else is going on. Let's see... 144/11 = 13 points of damage per tile, or 180/11 = 16 points of damage per tile away from GZ. I wonder if terrain tiles are resistant to stun somehow? :(

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#32 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 20 October 2005 - 10:15 PM

OMFG Danial, MapView is incredible. To think so much goes into just the ground we tread upon :) Thanks so much for sharing that! I wonder how much time Daishiva spent on that... or if he was some superwiz that did it in a weekend, laugh.

Daishiva's interactive website seems to be down (except for that last link). I wonder if there's a replica somewhere?

That rings a big bell re: an HE block value for hills. Hills certainly don't simply disappear to nothing if caught in a blast, leaving a nice big circular blaster pattern.

It's also consistent how the top armor is 5 but the death armor is 25... no wonder grenades at first dimple everything, then more are needed to start slowly hurting the ground more, outward from GZ. Is it remembering past damage to current armor, even? Hmm, will play with that a little.

If "HE Block" is in the code (for terrain), it certainly makes it more reasonable it might be in there for units. It's getting late here, but I'm thinking a simple test is to pack some soldiers together, then blow off a grenade on the far side of the group. Grenade gives weak damage; desert "top armor" is weak... could be a sensitive indicator of whether units do HE Block.

We can supply water and beer... even bourbon, Z. But not time LOL Wish we could work on XCOM time... 24 hours/day.

For my stun bomb tests, I moved my soldier up to the ceiling, then fired straight down. For blaster, I steered it to a point at ceiling 1 tile to the side of ground target, then went down. Yep, aim is a hassle, esp. for the BB. It's also another reason the stun bomb isn't so good for earning experience... the hassle of proper aim (and possibility of not hitting the middle of the bullseye, if aliens are arranged around edge of blast radius).

My impression is that you are liable to get up to N Fatal Wounds, when you get 10N health damage. Only an incredibly small sample though, and I was deliberately trying to produce small health damage. So this wouldn't have let me see any potential "max allowable" caps. I might take a look at Fatal Wounds some day.

If anybody can figure out stun damage, Z, I'm thinking it's you. =b However my guys in FSs never got entirely stunned, and they only had 40-60 HPs. So, if the max is 144 (w/Resistance), it must have been subtracting armor as per the Damage Formula. But it was a tiny sample, ~16 datapoints in all.

Danial, let me take this opportunity to thank you for all the work you do on the wiki. Research Time is something I'd always wondered about, and the Damage Modifiers are much easier to read now. Plus all the many other things. Did you make those armor pix yourself? They're cool!

I'm packing it up for the night... will tip a brew for you tomorrow night, Z ;)

#33 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 20 October 2005 - 10:42 PM

Daishiva's interactive website seems to be down (except for that last link). I wonder if there's a replica somewhere?

<{POST_SNAPBACK}>

All his links at the StrategyCore forums are definitely down. But his website is not. Click here for the main page and here for the download page. I totally forgot you need microsoft's .NET framework to run MapView though. Better get that too, even though it's huge.

We can supply water and beer... even bourbon, Z. But not time LOL Wish we could work on XCOM time... 24 hours/day.

<{POST_SNAPBACK}>

I hear ya, Mike. :)

If anybody can figure out stun damage, Z, I'm thinking it's you.  =b However my guys in FSs never got entirely stunned, and they only had 40-60 HPs. So, if the max is 144 (w/Resistance), it must have been subtracting armor as per the Damage Formula. But it was a tiny sample, ~16 datapoints in all.

<{POST_SNAPBACK}>

You are correct, armor does subtract from total damage (hence why I always test with no armor because that makes determining damage that much easier). So a Stun Bomb does 0 to 180 damage max. A Flying Suit (I finally figured out what FS stood for) is 80% susceptible to stun. 80% of 180 is 144 damage max. Assuming a hit to the front plates of the Flying Suit (110 stopping power) damage is then reduced to 144-110 = 34 stun damage max. That's assuming you didn't muck with the Flying Suit's properties. If you didn't, then I have no idea why damage was only between 40 and 60.

It's possible that the OSG's table is wrong concerning damage modifiers for X-COM units. If that is the case then assuming 100% susceptibility, the range should be 0 to 180-110 = 70. Still doesn't explain the min of 40 though. The only other thing to consider would be the sample size. 16 points is hardly representative. So I think that is what the problem is. :wink1:

I'm packing it up for the night... will tip a brew for you tomorrow night, Z ;)

<{POST_SNAPBACK}>

I'll do the same for you. Thanks for your help and tests! :)

- Zombie

Edited by Zombie, 20 October 2005 - 11:06 PM.

The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#34 Danial

Danial

    Captain

  • Moderators
  • PipPipPipPip
  • 778 posts

Posted 21 October 2005 - 05:07 AM

Danial, let me take this opportunity to thank you for all the work you do on the wiki. Research Time is something I'd always wondered about, and the Damage Modifiers are much easier to read now. Plus all the many other things. Did you make those armor pix yourself? They're cool!

Sure did make them myself. Just like I made the floor plans for the Battleship myself B)

#35 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 21 October 2005 - 05:38 PM

Sorry Z, I wasn't clear - I meant that the soldiers had 40-60 Health. And that they were never stunned all the way up to that (so that they went unconc). So the 34 stun max is right on target.

I saw the Online Help for MapView, but the link wasn't working. I posted a message on Daishiva's forum... and he replied within 5 minutes that he overlooked it and would get right on it! 'Zat service or what? Cool site too... is that guy too good or what. B)

I'm setting up to gather a lot of stats on hits from the pistol vs. the grenade. Should also teach me/confirm about damage etc. and perhaps elucidate fatal wounds and other things a little.

Danial, those floor plans are very cool too! You've got some real skillz. It really helps the wiki to have the many things big and small that you do. Whereas I usually only make a useful contribution to one small area... but hopefully it's an important one :)

Drinking a brew for ya Z. Tastes good! Just thought you'd like to know =b

#36 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 21 October 2005 - 09:38 PM

Drinking a brew for ya Z. Tastes good! Just thought you'd like to know  =b

<{POST_SNAPBACK}>

Same here! What are you drinking? A cheap domestic or an import? I pulled out the big-guns for you Mike: Paulaner Oktoberfest. Prost! :cheers:

Sorry Z, I wasn't clear - I meant that the soldiers had 40-60 Health. And that they were never stunned all the way up to that (so that they went unconc). So the 34 stun max is right on target.

<{POST_SNAPBACK}>

You had me worried there for a second, real worried. I thought I understood weapons, but you never know. The game always seems to pull a quick one to prevent us from understanding stuff completely.

I'm setting up to gather a lot of stats on hits from the pistol vs. the grenade. Should also teach me/confirm about damage etc. and perhaps elucidate fatal wounds and other things a little.

<{POST_SNAPBACK}>

Well, I gathered some values today trying to formulate a link between helath loss and fatal wounds. Preconditions: Soldier Health: 200 (thanks NKF), soldier Armor: 0, weapon: Heavy Laser (don't ask, yet).

Okay, so the purpose was to shoot at a soldier with the Heavy Laser and write down the health and fatal wounds. Simple enough. The reason why I used the Heavy Laser instead of something a little more, um, advanced, was because the game claims it does 85 damage while the OSG claims 80. Assuming the 100% susceptibility rating to Laser, my troops would therefore see a range of 0 to 160, or a range of 0 to 170 depending on who is right. Seemed like a good time to kill two birds with one stone, if you catch my drift. :)

This is just a rough approximation for now, but I think 205 values are sufficient to start a discussion. Without further ado, here are the numbers:

            3 Fatal      2 Fatal      1 Fatal      0 Fatal      Total
Minimum:        7           11            4            1          1
Maximum:       168         169           168           8         169
Range:         162         159           165           8         169
Median:       59.5        102.5           87          3.5         81
Mode:           34         106            59           3          59
Ant. Ave:     87.5          90            86          4.5         85
Act. Ave:     74.3         94.4          90.0         4.3        82.6

First of all, take a look at the Total column, specifically the Min and the Max: 1 to 169. Problem solved. The game is correct when it says the Heavy Laser does 85 (average) damage. The OSG is therefore wrong. :NyaNyaNya: (Sorry David Ellis, I just had to do that).

Just like I mentioned in a previous post, the number of fatal wounds is a random roll... sort of. How fatal wounds are calculated though, is still unknown: Is health loss used? Probably. The problem comes about when looking at the values above. There is a 96.1% chance that a soldier will receive 1, 2 or 3 fatal wounds (with a max damage of 170). There is a very slim chance (3.9% overall) that a soldier will receive no fatal wounds. To look at it another way, from the range there is roughly a 4.7% chance your soldier will see no fatal wounds. This may be as high as 5%.

Look at the max for 0 Fatal wounds: a value of 8 maximum damage points. Now look at the mins for the other columns: 7, 11, and 4. See an overlap? I do. I suspect that given more values, the ranges will increase - mostly to the lower side. But there is also a little room to improve upwards too. Another problem is the low count for 0 Fatal Wounds. Eight values is not sufficient to gain a complete understanding of what is going on.

MikeTheRed: If you gather some numbers with the normal pistol, please make sure you are not shooting at a Muton, a Cyberdisc or a Zombie (not me stupid, the alien variety). Those aliens have susceptibility ratings which do not equal 100%. That could mess up your results. If you read this before you gathered too many numbers, make sure the health of the unit in question is also 200, and all-around armor is 0 (just like my guys). That way we can compare apples to apples. Perhaps by looking at both data sets, we can put our heads together and solve this little mystery. :wink1:

Anyhow, I have a date with a computer game known as X-COM. Can Zombie do this 800 times in a night? Tune in tomorrow to find out! LOL

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#37 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 21 October 2005 - 10:46 PM

Good stuff, Z.

I've been distracted tonight. Spent 25% of my time attempting to design an applet that compares savegame results across every round of, guys shooting at each other and/or getting explosion damage. Spent the other 75% working on other things I need to do, and/or mind wandering. Am about a quarter of the way toward a serious number-crunching applet. It's gonna be a b$tch to set up. This is why I kept remembering other things I needed to work on. I'm no Daishiva, laugh. Just learning programming as I go.

Correct me if wrong: You never saw more than 3 Fatal Wounds (FWs)? I suspected there was an upper limit.

My brain is not working 100% right now - thanks to all the beers I drank for you - but I think that something missing here might be a lower limit on # of FWs possible. Which is to say there's something going on such as, Health Point damage (-HPs) equals 10-19 for 1 FW; -HPs 20-29 for 2 FW, etc. Like I say, at lower levels, I was seeing a correlation. Our intuitive experience says it can't go through the roof. Instead of hitting them with low damage, you took the other end and found there's 3 FWs max with high HP damage, eh? See where I'm going with this?

Those numbers I just posted are pure thinking out loud. But I suspect there is a maximum sticking its foot in the door. Then after that is the issue of body facing and damage to specifc body parts... does it just keep filling up all over, or does XCOM really use an aimed-at-parts model?

Don't be too rude toward David... writing game guides is heck on the writer. We'll provide him plenty of beer after your jeer, eh?

Alas I think I have to use Mutons. Two reasons: 1) I don't have any other savegames close at hand. 2) Mutons are what I'm really after, since they're best for building experience.

We come at this from different angles (and that's a good thing). I believe in getting simple things I want to know, known first, and moving on from there, too. Just like you. The thing I most want to know is re: gaining experience. Mutons are best for that. :) Please speak up if you see me going astray. But Muton XP is what I'm focussing on, more so than general models.

Actually, do I only need to change Unitref[00] to change AP resistance? Hmm

One other thing I might've tried would've been to have soldiers shoot each other. But the hits experience counter (Unitref[81]) is key to knowing if someone was actually hit, and it only works for hits on enemies. Yes, I am tracking shooters vs. targets. I want to know how often the pistol misses and/or how often armor totally deflects a hit. Thus the agony over applet design. I am thinking of multiple health levels ands multiple armor levels in my design. Low health, high armor, high health, zero armor, etc. The applet will track individuals across each round of shooting (approx. 5 times per round for pistol; can be much more for grenade) and the battlescape won't need to be reloaded until one or more targets are dead. Just to save that bit of time. Have I missed any variables?

I thank you for the Paulaner Oktoberfest. Never met Prost, but he sounds like a good guy. I go for mid-level stuff... Bass, Heineken. There is no such thing as a high volume USA beer that's any good, excepting Sam Adams and perhaps Michelob as a last resort. Sort of like when you run out of every option in XCOM and your last resort is rookie Michael Blake trying a long shot at an alien menace holding a Blaster.

#38 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 22 October 2005 - 10:36 PM

Correct me if wrong: You never saw more than 3 Fatal Wounds (FWs)? I suspected there was an upper limit.

My brain is not working 100% right now - thanks to all the beers I drank for you - but I think that something missing here might be a lower limit on # of FWs possible. Which is to say there's something going on such as, Health Point damage (-HPs) equals 10-19 for 1 FW; -HPs 20-29 for 2 FW, etc. Like I say, at lower levels, I was seeing a correlation. Our intuitive experience says it can't go through the roof. Instead of hitting them with low damage, you took the other end and found there's 3 FWs max with high HP damage, eh? See where I'm going with this?

Those numbers I just posted are pure thinking out loud. But I suspect there is a maximum sticking its foot in the door. Then after that is the issue of body facing and damage to specifc body parts... does it just keep filling up all over, or does XCOM really use an aimed-at-parts model?

<{POST_SNAPBACK}>

You are correct, no more than 3 fatal wounds were ever inflicted (I would have listed that otherwise). And indeed, I totally understand your point about low and max damage. Hmmm... that brings up some other interesting ideas too. *Zombie scribbles something to himself* What would happen with low weapon damage and high health? Or what would happen with low weapon damage and low health (just below weapon strength)? Or even a max of say 240 weapon damage and a higher health of 250 (or even a lower health so possible more FW are inflicted). My gut tells me the beginning health of a soldier plays a role in determinations such as these.

I recall seeing more than 3 fatal wounds for soldiers getting shot at with a pistol and low health (at least I think so). In any event, my tests pushed normal weapon power / health to the limit. But editing obdata.dat to give a weapon power just under 250 may yield some interesting results. I'll leave this for a later test.

FW's do not "fill" certain areas first and then "spill over" to different parts. If you keep shooting (aiming) at a particular area of the body, that area will continue to take on fatal wounds for as long as 2 conditions are met:
  • The unit remains conscious.

  • The unit does not die.
But there may be an upper limit on the number of FW's a certain body part can handle - this should become apparent with a normal pistol and a soldier with very high armor. Some reloading may be necessary (ie if a unit takes on a fatal wound, save the game, if it doesn't, reload until you get another FW). An emergency edit may be necessary to "up" the targets health if it is near death.

To minimize hitting other areas of a soldiers body, I have my shooters pop the target directly in the back (read: no turning). Turning to shoot a target will reveal a different orientation and therefore a different area of the body will be hit. After a few quick checks with my medics, all shots were hitting the torso. =b (As I always say, "minimize variables").

We come at this from different angles (and that's a good thing). I believe in getting simple things I want to know, known first, and moving on from there, too. Just like you. The thing I most want to know is re: gaining experience. Mutons are best for that. :) Please speak up if you see me going astray. But Muton XP is what I'm focussing on, more so than general models.

Actually, do I only need to change Unitref[00] to change AP resistance? Hmm

<{POST_SNAPBACK}>

I understand now. We are on two different pages. You are still on experience point determinations, while I on the other hand am after a model to explain FW's. Carry on. BTW, from BB's unitref notes, [00] is just a flag for the type of unit. I'm doubtful if this changes damage modifiers. AFAIK, that value is still hidden in the game files somewhere. Correct me If I'm wrong anyone.

I want to know how often the pistol misses and/or how often armor totally deflects a hit.

<{POST_SNAPBACK}>

Here may be a dumb question, "but why"? If a pistol misses, the XP counter stays the same. If the shot connects, then the XP counter goes up. Simple as that. Testing this out would only be good for Firing Accuracy determinations.

Second part, same question: "why"? That info can be determined from the damage equation given in the Wiki. It's as easy as plugging numbers into that equation. (I did calculations like this everyday for 6 months so humor me here):

Random scenario:
Beginner level Muton mission. Your soldier is carrying a normal Pistol and you want to know what is the chance that a Muton Soldier's front armor will absorb a direct hit without lowering health.

Vitals:
Muton Soldier front armor = 10
Muton Soldier health = 125
Muton susceptibility to Armor Piercing ammo = 60%
Pistol listed power = 26 (average)

Calculation:
Max damage for a Pistol against a Muton is lowered due to susceptibility. Modified max damage = INT(26 * 2 * 60 / 100)
2 is the max modifier and 60/100 is the damage modifier. Therefore, Modified max damage = INT(52 * 60 / 100) = INT(3120 / 100) = INT(31.2) = 31
The range of values the Pistol can deal is 0 to 31 points of damage (or a range of 32 different values).

The Muton Soldier has a front armor rating of 10. It will absorb up to this much damage without changing health. In other words, it will negate the first 0 to 10 points of damage a Pistol can dish out (11 range). Since the Pistol's range is 32, the probability that a direct hit will be blocked by Muton armor is simply 11 / 32 * 100 = 34.375 = 34%.

Okay, say you want to know the probability of doing damage to a Muton's health. Weapon range of the pistol is still 32, but instead of absorption we want penetration. 10 points of damage is absorbed, so we need at least 11 to penetrate. 32 - 11 = 21 values which will breach. 21 / 32 = 65.625 = 66%. 21 is also the maximum number of damage points which will get applied to the Muton's health. Assuming the lucky Muton was initially uninjured, it should have an ending health of 125 - 21 = 104.

I don't know. Is this kinda what you were asking? There is no need to run through countless trials just to arrive at the same number you could calculate in about 1 minute. :wink1:

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#39 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 23 October 2005 - 07:24 AM

There is no need to run through countless trials just to arrive at the same number you could calculate in about 1 minute.

Hmm. Well. (cough) Now that you've laid it out entirely clearly for me... point taken. :) I'll copy this clear example to the wiki. *sheepish grin*

I'll do some calculations concerning how many shots one can make until a Mutie will die (including armor decreasing). This will show the expected "harvest" of experience points per Muton. I suppose I only need to calculate two endpoints: Average number of shots til dead (using average shot damage), and minimum number (using max shot damage).

Fortunately we don't have to worry about Fatal Wounds for non-MC'd enemies.

I guess we can assume that armor damage will follow in line with health damage, based on the relationship I posted.

FWIW, the Muton's resistance to AP is yet another reason (in addition to high health and relatively high armor) that they are good for reaction training.

Which reminds me... maybe I'll add to the shots per Muton calculations, a situation where the player is constantly turning the Muton so as to shoot his highest remaining Armor side... see how much longer he lives then.

What about grenade damage? As soon as you get to a point where you have a handle on its damage at the edge of the blast, can you post that (or do you know it already)? Could that use some testing?

Another thing I might do is use MapView to pull out Armor values and HE Block (if any) for common ground tiles on the various terrains. It should show a good relationship to the blast diameters you found. Are those the only values of interest? For common ground tiles, and their death tiles, I suppose.

#40 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 23 October 2005 - 09:29 AM

I just made a Small Launcher entry for the wiki.

Question: Does anybody know if stun bombs ever go through the Under Armor? Will they always be hitting side armor (front, left, right, rear) because it's a shot weapon, or will they go through Under armor (at GZ and GZ+1) because it acts like an explosion??

#41 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 23 October 2005 - 10:00 AM

I just made a Small Launcher entry for the wiki.

Question: Does anybody know if stun bombs ever go through the Under Armor? Will they always be hitting side armor (front, left, right, rear) because it's a shot weapon, or will they go through Under armor (at GZ and GZ+1) because it acts like an explosion??

<{POST_SNAPBACK}>

Nice job on the Small Launcher entry, Mike. :)

Can Stun Bombs penetrate Under Armor? Sure. Why not? Armor is armor. There is nothing special about Under Armor except maybe the lower blockage values.

Think about orientation when looking at things like this. It is true that a Stun Bomb cannot "splash over" to affect under armor from a shot fired at the same level on the battlescape. But raise the unit in the air and put the shooter on the ground. If the shot doesn't miss the target, it should hit under armor. The most likely spots where you can see ground-based soldiers in the air is on Grav Lifts. ;)

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#42 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 23 October 2005 - 10:17 AM

Thanks Z, good point/reminder about shooting from below...

Think about orientation when looking at things like this. It is true that a Stun Bomb cannot "splash over" to affect under armor from a shot fired at the same level on the battlescape.

What about how explosions hit Under Armor at GZ and GZ+1, even with shots from the side like HC-HE. (Right?) Do you think the stun bomb is the same? If so, then Under Armor is probably the single most relevant measure, since most of the time a stun bomb is shot at a single unit (i.e., hits usually occur at GZ). (And I should update the wiki!)

#43 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 23 October 2005 - 10:48 AM

Fired high explosive shells act exactly the same as Stun Bombs. Say you fire a Small Rocket at a soldier standing on the same level of the battlescape. Depending on on orientation, the target will only suffer damage damage on F, L/R, or Rear armor. Under armor should remain unaffected.

Aiming the shot at the ground instead of the soldier might create "reflection" damage like grenades do to under armor. Let me level with you though: I never tested this out so this is still just theory.

Wait. Do you mean to tell me that a target at GZ or GZ+1 might get damage to under armor, even though the shot connected with the torso? This migh have some serious rammifications to the damage equation for HE blasts... though maybe not. If a shot "drifts" off course from the intended target and detonates on the ground instead, then I could see how damage would affect under armor - even though the shot was fired from the side.

I think you mean grenades though, right? They affect under armor. :)

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#44 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 23 October 2005 - 11:18 AM

Right, I was thinking about grenades. Actually I hadn't gotten all that clear in my mind. I thank you kindly for your assistance. ;)

#45 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 23 October 2005 - 12:14 PM

I just did some testing with Rockets (large and small), HC-HE, and AC-HE. Firing from above, directly over target, from the ceiling (height=4). Every one of my shots hit the "ground" a.k.a. Under Armor damage in GZ and GZ+1. This was with 2-3 shots of each weapon type. Nobody within GZ+1 ever got hit anywhere but Under Armor.

Then I tried shooting directly into the "side" of a group of mutons and soldiers, with my shooters on the same level as the targets, thus shooting into their "side" aspect. I included the Blaster this time.

Every single time, only Under Armor was hit within GZ+1. Outside GZ+1, hits were to one of the sides, as expected.

I watched carefully and only counted when it hit the intended target, i.e. no chance of a miss that hit the ground behind. However, with the graphics and instant explosion, it's hard to say whether it looked like it was hitting the ground of the target, or the target itself. Still, all my shots looked level. And the results were always consistent.

This is actually what I thought I remembered from my previous testing (and why I was confused in my prevous message). However, the concept that only Under Armor is hit within GZ1 was not an "issue" at that time, so I didn't pay close attention. I still have my data though and on review see that also with them, in every case, Under Armor was hit in GZ1. (Also, I only used hits on intended target, so, no hitting of ground behind.) N=58 within GZ+1, 220 for >GZ+1.

Either Under Armor is ALWAYS hit at GZ1 of explosions, or else there's something funny going on where you hit the ground almost every time (instead of target's side), and I did not do enough testing to see it on the rare time it did not hit the ground.

I suspect the former is true, knowing XCOM... that you always hit Under Armor within GZ+1.

This is not how you thought of it? If not, do you mind doing a little testing to make sure I wasn't doing something wrong?

Edited by MikeTheRed, 23 October 2005 - 12:15 PM.


#46 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 23 October 2005 - 02:58 PM

Wow, I didn't actually anticipate the programmers getting this creative with explosive damage. :)

Exactly true, Mike. All your tests were spot on. I did the same tests as you but switched things around after I confirmed everything. Grenades, Small Rockets, actually all HE blasts which detonate on GZ or GZ+1 only affect under armor. Even soldiers on the diagonal from GZ still got under armor damaged.

Now I progressively moved the center of the blast away from the soldier. Units standing on GZ+2 only received damage to the area of armor where the blast went off. For instance, if the blast happened in back of the soldier, the explosion damaged rear armor. It didn't matter what the landscape was either.

It makes sense now that I think about it. Shrapnel from an exploding mine usually does severe damage to the lower extremities (legs). Move farther away from the blast and damage mainly happens to the upper area of the body (torso, head). That's why sergeants tell their troops to hit the deck if a grenade goes off nearby: a lower profile reduces injury. Only the few pieces of shrapnel traveling horizontally will actually reach the soldier in the prone position. But, have the soldier standing and pieces come flying from different trajectories increasing the chances for wounds.

Like I said before, I really didn't think that the programmers used this accurate of a model to describe explosives. That's impressive. Now that I know exactly how explosives behave, I am starting to gather some numbers with the normal grenade. Since we now know that fired or thrown explosives react in the same manner, it's safe to use just grenades instead of fired weapons (no chance for "drift"). Zombie thinks that starting at the edge and working my way in may be more beneficial. :D

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#47 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 23 October 2005 - 03:57 PM

Cool. I wondered if I was losing my mind :P

Well, actually to me it doesn't make as much sense. If an AC-HE hits you in the head, it's gonna get your underwear? :laugh: But, "whatever"... we all know they did some simplifications. Actually it makes testing easier. (Now that we know how it works!)

I just did a little calc on the grenade, using your equation. For the grenade, using your method, I get:

For HE Rating 50: Min 25, Max 75, Range 51, Average 50.5

And a table of damage over distance as follows:
Dist Min  Max   Ave
   0   25   75   50.0
   1   15   65   40.0
   2    5   55   30.0
   3    -   45   22.5
   4    -   35   17.5
   5    -   25   12.5
   6    -   15    7.5
   7    -    5    2.5
Unfortunately, this doesn't agree with my numbers in post #21 in Desert terrain, where the blast diameter was 9, radius 4, so DFE 1 = 4 in the table above. Also keep in mind that my test results were for Muton Soldiers ("side" armor=10; all data is for side armor with fresh armor).

Should it decrease by 10? Looking at my table, esp. the mins and maxxes, it looks to me like it's suggesting it drops by 15 (and is already going to 0 at the minimum by DFE=2). Also it suggests that the range is 30, not 50 or 51. You're testing it more now?

If the decrement per tile is 15, a calculated table looks like:
Dist Min  Max   Ave
   0   25   75   50.0
   1   10   60   35.0
   2    -   45   22.5
   3    -   30   15.0
   4    -   15    7.5
MUCH better. Spot on with my results, if you keep in mind that Mutons have 10 armor. AND spot on with the observed blast radius. Remember how I saw smoke beyond the edges of grenade blast sometimes? Maybe it's due to rounding being different for smoke vs. blast, a.k.a. the decrement of 15 means it ends right at the edge... maybe the smoke doesn't round as well as HE damage. Also remember how the grenade has an unusual blast pattern relative to other ones AFAIK (no lone tile of damage sticking out due laterally)... maybe it's because it's the only one to decrement by 15?

Hmm... actually that minimum at DFE=3 is too high. Especially the one for DFE=2 should be 0, too. Maybe it's just too small a sample, but I doubt it. Also more numbers for a Muton Navigator and Silacoid were gathered but not shown in post #21 (to keep that table simple). The numbers are entirely consistent with a drop of 15 (inc. Silacoids 130% vulnerability to explosion) except again the Navigator did not have the minimum expected... but for these other two critters I only have five datapoints, which isn't much at all.

It's cool that you're testing grenades... they're what I'm most interested in. :)

I just fleshed out the Explosions part of the wiki. Broke it away from Damage Formula, and stuck a lot of your stuff (and a little of mine!) in there. Comments or edits welcomed!

Hey I hate to break the news but we're all out of beer. :crying1: Oh well, it's Sunday anyway, eh?

#48 Zombie

Zombie

    Colonel

  • Xenocide Creative-Text Departmen
  • 2,947 posts

Posted 23 October 2005 - 04:24 PM

Well, I have a couple cold ones here yet (I can't get much testing done while drinking). Unfortunately, I'm all out of Paulaner. But I do have a couple cold bottles of Spaten Optimator. *Zombie opens both bottles and hands one to Mike* Prost! Now be careful with this stuff. The alcohol content is like 9% or so. :wink1:

Nice job on the Wiki by the way. Looks good. One of these days I'm gonna have to gather all my equations and present a "formal" conclusion on how HE is calculated. I have a couple theories which I am still testing, so that's why I held off adding it.

So you are saying a 15 damage point drop-off per tile? Hmmm.... It all makes sense now. Remember Danial's post:

According to Daishiva's MapView program, all desert ground tiles have an "Armour" rating of 5.

<{POST_SNAPBACK}>


Ah ha! That's where the extra 5 points come from! The ground actually is interfering with (or sopping up) blast damage!!!! *Bells and whistles go off, and streamers fall.* No longer is this a theory. So to summarize, the OSG is still correct when it mentions a 10-point drop-off per tile. It's just that the tiles themselves are taking off a bit too when they are damaged. =b

- Zombie
The Mr. Grognard of X-COM

Zombie: Empirical data's your only man, when formulating a research plan.
A soldier's death is never in vain if it makes the formula more plain.
A few dozen make a better case for refining that third decimal place.
They call me Zombie because I don't sleep, as I slowly struggle to climb this heap,
of corpses, data points, and trials, but from the top - I'll see for miles!


#49 NKF

NKF

    Commander

  • [Global Moderators]
  • PipPipPipPipPipPip
  • 4,798 posts

Posted 23 October 2005 - 04:46 PM

I just want to interject briefly about area effect damage:

Damage is always done to under armour at ground zero, and a 1-tile radius from it. All other damage is directional. Basically, a 3x3 area in the centre of the explosion.

This information is tucked away in some forgotten corner on the wiki. Perhaps it needs to be brought forward into a more obvious article?

- NKF

Edited by NKF, 23 October 2005 - 04:47 PM.

Lord High Generalissimo Ruler Supreme of Norm's Anti Pedant Society (NAPS).

Number of members: 1

#50 MikeTheRed

MikeTheRed

    Captain

  • Forum Members
  • PipPipPipPip
  • 179 posts

Posted 23 October 2005 - 04:52 PM

Good idea... here we are!