Talk:UFO Crash Recovery
Recovery of Alien Alloys
First of all, thanks to Seb76 for providing me with the Power Source explosion damage offsets in the executable. Basically what I did for the first series of tests was to edit the power source to explode with 250 maximum power all the time (which is 180 base + 70 max randomizing factor). This is the worst case scenario. See the pics below for damage dealt to the Medium Scout.
So there wasn't much left of the UFO after this. I count 7 tiles remaining with the alien alloys flag set and since the amount you recover is divided by 10 for the mission, the total you see is 0 ( INT(7/10) = INT(0.7) = 0 ). This checks with the EOM (end of mission) summary. Obviously, no navigations were recovered either. I'll be adding pics and summaries for each of the ships as I get to them. After the max is determined, I'll do the min damage too so that we can get an average. That'll help nail down a range for some of those recovery values. --Zombie 21:14, 17 November 2008 (CST)
The damage dealt to the Medium Scout when the PS explodes with 180 HE is fairly minimal on the inside and almost unnoticeable outside. Still, some floor tiles were destroyed and the Navigations didn't make it again. Final tally: 9 AA.
Since you recover 0 Alien Alloys when the destruction is max and 9 when the damage is min, the anticipated average amount of AA recovered should be 4 units.
Nothing really new here, PS detonates and leaves 28 alien alloys to be recovered (I'm not going to count the number of tiles left as it's 280 or more). The 2 navigations are untouched, but the blast leaves a big 5×5 hole in the roof. The inner wall in the ship is obviously absorbing some of the explosive yield and curbing the upwards damage as the hole in the roof is smaller than that of the Medium Scout.
PS yield is so low that damage never propagated outside the craft. For this reason I didn't include a pic of that view. Internally, only part of the engine room walls were destroyed. This coupled with the fact that the explosion didn't make a hole in the roof increased the number of alien alloys to 31 units.
The anticipated number of Alien Alloys recovered for the average crash is calculated to be 29 units (INT(28+31)/2).
After running about 40 missions I eventually saw the two blast configurations. In the west scenario, a vat of alien food was set ablaze by the blast, while in the east scenario the food was just damaged. For the vast majority of missions you run the final tally will be 90 Alien Alloys, 13 Alien Food and 8 Alien Surgery recovered. If you are able to detonate a Smoke Grenade anywhere near the burning food vat early enough in the west scenario (even outside the UFO will work as the spread of smoke is not hindered by UFO walls), you'll snuff out the fire and recover the vat. For a $150 investment in the Smoke Grenade, you'll gain $5,000 in the sale of the extra food.
The blast produced by a PS with 180 HE is fairly small and occupies only a 4×3 area. This is just enough to open up the wall into the central lift room and destroy the second Power Source. In both the East and West configurations, you should recover 14 Alien Food, 8 Surgery, 10 Navigations and 94 Alien Alloys
The anticipated number of Alien Alloys recovered for the average crash is calculated to be 92 units (INT(90+94)/2).
Fairly straight-forward on this ship. The only issue is that the craft has a lot of tight corners which the game has trouble with during explosions (MCD problem mainly). So there are a couple holes in the floor on L1 where soldiers could fall through if they venture too close. Final tally was 59 Alien Alloys, 5 Navigations and 4 Exam Rooms. You can clearly see some odd behavior on blast propagation as there are holes into the small entranceway room on the ground floor.
If only one PS explodes with maximum power (either side, the craft is supposedly symmetrical), you recover 64 Alien Alloys or if you put it another way, you lose 6 units. Extrapolating to full destruction on both PS's you would thus expect to lose 12 when you actually lose 11. This is due to integer rounding which is expected.
Below are a couple pics from one of the three scenarios possible (both PS's explode with min damage). You can extrapolate what the other two should look like. When both explode, the amount of recoverable Alien Alloys is 65 (or 5 AA lost). When the West PS explodes, you get 68 AA (or 2 AA lost), while just the East yields 67 AA (or 3 AA lost). Obviously, the East PS scenario is causing more damage somehow. Why? Well, it's due to the orientation of the inner UFO walls. See, in the West the inner wall is located 3 tiles away from the blast (on the west side of tile 3). In the East the inner wall is located on the west side of the tile, but 2 squares away. Since the wall is closer, the blast is just able to destroy the segments. So the Abductor (as well as any UFO) is never perfectly symmetrical, but more powerful blasts may mask the asymmetry introduced by map construction limitations. If you are wondering why there is an undamaged tile and a half running to the North of each PS, it's because those combination tiles have an armor rating of 100, 20 points more than the normal UFO floor tile counterparts (another MCD issue).
Anticipated average recoverable AA for the W+E scenario is 62: INT((59+65)/2). For just the West it's 66: INT((64+68)/2) while just the East should yield 65: INT((64+67)/2).
Ran about 50 trials with this craft type (probably more) and saw all the configurations except where the South Power Source explodes. I know that PS explode with a progression so if they follow a W-E-N-S arrangement (Where did I read this? I think Zaimoni or MTR mentioned it.) that would explain the rarity of the South one exploding as there are 3 other units which have a 75% chance of detonating first... As you can see, there are slight differences when each PS explodes, but it can only damage so much. In each case the final count was 74 Alien Alloys, 8 Entertainment and 8 Navigations.
This time I saw all the possible blast configurations so I'll need to revisit the max damage scenario for this ship and try to get the South PS to blow. In any case, for the West, South and East configurations, you recover 87 AA. The North configuration yielded one less unit with 86 AA.
The anticipated number of Alien Alloys recovered for the average crash is calculated to be 80 units (INT(74+86)/2) or (INT(74+87)/2).
This one was a piece of cake. In the first mission all 4 PS's detonated leaving 191 Alien Alloys. Since the undamaged ship has 231, that comes to 40 total units lost or 10 per engine. The second had the West and South PS's destroyed which left 211 alloys ((231-211)/2 = 10/engine). In the third mission the West, South and East PS's were gone and the final tally was 201 AA ((231-201)/3 = 10/engine). It took about 5 more trials and I found that only the North PS had exploded leaving 221 AA ((231-221) = 10). So this verifies that for each Power Source detonation, you lose 10 Alien Alloys for a maximum cap of 40 units.
There are only four possible outcomes when power sources detonate on this UFO. Scenario 1: center PS detonates, no PS's remain, 124 Alien Alloys recovered. Scenario 2: West PS detonates, East PS recovered w/50 Elerium and 129 AA too. Scenario 3: East PS detonated, West PS recovered w/50 Elerium and 129 AA too. Scenario 4: West & East PS detonate, no PS remain, 120 AA recovered. In all cases you recover the stuff located on the floors above: 4 Navigations, 2 Surgery and 20 Food. In the last two pics in the series, the pattern clearly shows that the West PS explodes first destroying the central PS and leaving a "blast shadow" of untouched UFO floor tiles down the center. Then the East PS explodes damaging the blast shadow tiles and killing the other overlapping tiles. The reason why there is a damaged (but not destroyed) tile in the West-Center of the craft is because the damaged UFO PS in the middle has the same HEB as an undamaged one which ultimately curbs damage down the middle.
HE re-direction / channeling?
Very interesting! What really surprises me is the apparent upward channeling effect on the Lg Sct vs the Harvester, as shown by greater damage to the floor/ceiling tiles above. Isn't this unexpected, since walls etc may block HE, but, as far as we know, walls etc don't redirect the blocked HE elsewhere. But unless the external LgSct ceiling is weaker than the internal Harvester ceiling, and assuming only 1 PS detonated in each case, doesn't this look like evidence of redirection/channeling of HE blast? Spike 03:43, 18 November 2008 (CST)
- I figured it out. UFO floors go through a progression when damaged. Initial armor is 80 but the associated death tile has an armor of 50 which makes a total armor rating of 130. The PS explodes at 250/2 damage for terrain = 125. 125<130 so therefore, the floor above is always damaged (but not destroyed). For exterior UFO roofs, those tiles have 100 armor but no death tile, and thus, no additional armor to sop up damage. No death tile = hole. --Zombie 09:54, 18 November 2008 (CST)
So in effect, it would seem that (unexpectedly) the interior floor/ceiling of a UFO has more net blast-resistance than the external roof. Great deductive analysis there Zombie. Spike 10:56, 18 November 2008 (CST)
- Exactly. I also figured out why one roof tile and wall segment remain on the North side of the Medium Scout. The UFO chairs have 20 HEB. By the time the PS blast reaches the chair, the damage has dropped to 115. Now the chair HEB is applied: 115-20=95. The next floor tile takes 90 which damages it and the wall segment sees 85. Its armor is 100 so it doesn't get destroyed. As for that ceiling piece remaining, recall damage at that distance is 90 due to the chair. That is less than the 100 ceiling armor so it survives. For the West wall, I think the reason why it isn't destroyed is twofold. First, the display panel between the navigation consoles has 70 HEB while PS damage is 110. 110-70=40 which isn't enough power to destroy the wall. The navigations have 40 HEB. 110-40=70 which also isn't enough for destruction. The reason why the outlier (urm, diagonal) West wall segments aren't destroyed is because of chairs again. That 20 HEB (plus the distance from GZ) is just enough of a drop to keep power below 100.
- On the Harvester, the East power source has 30 HEB so it soaks up most of the damage as the tile behind it remains undamaged. So is that wall segment beyond that.
- I'll be adding the Abductor pics and summary shortly. There may be more pics for this ship as there are 3 outcomes: West PS survives + East detonates, East PS survives + West PS detonates or East + West PS detonate. In the case where only one of the PS's detonates, the number of alloys may be identical but I haven't checked yet (symmetry is hard to judge on UFOs since some exterior walls (N&W) are not constructed the same). The only trial I ran so far was where both PS's exploded. --Zombie 11:48, 18 November 2008 (CST)
Power Source Damage
Zombie, I know that testing is your specialty, but I know my personal games have jived with some of your results. The most glaring being that in the Harvester, sometimes the PS detonates with such force that the floor tiles of the hallway above are destroyed(making the hallway unusable if you lack flying suits). Also, I've seen the navigation panel chairs in a Large Scout fried by the explosion as well(And very rarely, the panels themselves, IIRC.) Arrow Quivershaft 23:42, 19 November 2008 (CST)
- Don't really know what to say. Except for the UFO PS executable offsets hacked to always dish out 250 damage, I'm using an unmodified (and archived) CE version of the game. Are you using (or have you used) XcomUtil? Editors have been known to screw with results. It's possible Scott T. Jones changed some MCD values for the custom ships. The only other thing I could think of is that we don't fully know how PS explosions function in the exe. Perhaps editing the min damage to 250 and the randomizer to 0 has an unanticipated effect? Seems doubtful as the exe code is fairly straight-forward and I've been studying explosions for a couple years now. After I finish the Supply Ship, I'll go back and dump in a fresh copy of the game files to see if it changes anything. Other than that, the only way past this is to keep an unmodified exe with the usual PS explosive power and run some tests the naked way. This would be my last resort though as I'd need to reload somewhere on the order of 80-90 times to hit max damage. Feel free to verify though. I'd suggest grabbing a copy of daishiva's MapView and looking at the tile MCD's of the Large Scout and Harvester to see if normal floor tiles have 80 armor + 0 HEB and the inner walls have 80 armor + 80 HEB (you don't really need MapView for this, but a graphical result is easier).
- That said, I don't remember ever having trouble with the Harvester having an impassible hallway on L1. Abductor, sort of, due to the corners.
I just read a little of the historical discussion below and MTR mentions that UFO Floor tiles have 80 armor and 50 HEB (not 80). I know for a fact he was using XcomUtil and not an unmodified version. Scratch that, upon rereading he said that the dead floor tile has 50 armor which is correct. Still, not using an unmodified version of the game throws up all sorts of red flags. --Zombie 00:29, 20 November 2008 (CST)
I do in fact use XComUtil(without it, the game is very unstable and is prone to resizing my resolution when it crashes), so that could be it. I'll check some more when I have time, which probably won't be until around Christmas. Arrow Quivershaft 00:34, 20 November 2008 (CST)
Alright then, thanks for clearing that up! Arrow Quivershaft 00:41, 20 November 2008 (CST)
Zombie, I have not seen such an impressive dedication to blowing everything up since Bob & Doug McKenzie in Celebrity Film Farm Blow-Up.
On a slightly serious note, a question about the test method. Given you are already forcing the randomiser to return 0, wouldn't it be easier (and a closer simulation) to hack the randomiser to return 1.0 - then you would not need to hack the base damage up to max. Probably it would make no difference - though if it was different, that would show that the PS Damage formula is subtly different somehow from what we currently think.
PS for completeness can we also have a picture of a blown-up Small Scout? Spike 03:56, 20 November 2008 (CST)
- Not sure what you mean with the randomizer. A randomizer of single point here or there probably wouldn't even be noticeable on the battlescape. 5 points maybe. 10 points probably would.
- Small Scouts cannot be blown up since they do not have a power source. Suppose I could edit one in the center to see what would happen if it did. I'm guessing L1 would be completely gone. The MCD armor levels are right at the edge though (120 tile armor) and the damage of the PS at GZ+1 (the edge of the wall) would be (250-10)/2 = 120.
- If my testing methodology is found to be sound, then eventually I'd like to hack the min value through it's range like we (being MTR and myself) did for the weapons/grenades on the battlescape. I'd also like to keep the min at 250 and hack the randomizer up to the max to see whether the blast is constrained (or capped) after reaching a certain distance from GZ. Lots of tests, so little time.
- Anyhow, I was thinking about the ordering of PS blasts on the battlescape last night after studying at the Supply Ship. If the game checks the UFO map file to see if any tiles are set to PS, then it would stand to reason that the way it checks is the same way the .MAP file is structured: top to bottom, West to East then North to South. That would completely explain the patterns we see sometimes for overlapping explosions. Makes complete sense. Still doesn't explain why I didn't see the second blast configuration for the Harvester (West power source survives it's saving roll only to be destroyed when the East one goes). Might need to run a bunch more trials. --Zombie 08:39, 20 November 2008 (CST)
- I did document somewhere the order of explosion and it works as you said, except that there is not check for a PS tile type, but for a special MCD value (so it is possible to make other tile types explodable). Seb76 13:35, 20 November 2008 (CST)
- OK I was slightly joking about the Small Scout. IIRC they are totally destroyed by any crash landing (and getting them to crash land is quite a feat, requiring one Cannon and lots of patience). I misunderstood you about the "randomiser" - I thought you meant that you were hacking (overwriting) the random number generator function code in the game so that it always returns 0. This would actually be a very useful hack for testing purposes - to force the RNG to always return 0, or 1.0, or 0.5 Even just being able to set the RNG seed would be good, as it would allow you to do controlled experiments with repeated identical conditions (maybe). Spike 10:59, 20 November 2008 (CST)
- I also wrote somewhere that the RNG is the one from the microsoft CRT so it is possible to force the seed to a fixed value. The RNG actually does not return a value between 0 and 1, but between 0 and a maximum value that you specify. Anyway, forcing to return 0 or the max value is not a big deal. Seb76 13:35, 20 November 2008 (CST)
Oh! Well I think we know where the RNG code is located. However, I wouldn't even consider hacking that in any form as it will make game-wide changes, not just with the PS. That's a bad thing for testing unless you know for sure it doesn't mess with other things. When a Small Scout crashes, it is intact on the battlescape since it doesn't have an engine which explodes. The reason why Small Scouts sometimes don't show up is because of XcomUtil (I think that map was left out for some reason). --Zombie 13:29, 20 November 2008 (CST)
Min, Max, Average
It's occurred to me that unfortunately it's not likely that the average amount of recoverables is going to be the average of the min and the max. I doubt it's a linear distribution of recoverables across the range of possible damage levels. Still, knowing the min and max in itself will be useful - just need to remind people not to draw unjustified conclusions from the min and max. Anyway, Z., your observations might suggest it actual is pretty linear so let's wait and see. I guess you could do the 'median damage' case and check if the recoverables are midpoint between min damage and max damage. If they are that's a fair test for linearity. Spike 22:31, 22 November 2008 (CST)
- The average between the min and the max probably will not be a real world value, just a an average over time (ie over many missions). It's the same idea as MTR's Elerium-115 recovery on the article page. Suppose I could run some missions with the UFO PS explosion dealing 215 HE (ave between 180 and 250) to get something close to a "true" average. Though, I suspect that may not be anywhere near anticipated average numbers (if the MCD armor rating of tiles is high, then the amount recovered will be shifted higher). We'll see though, it's still too early to tell. ;) --Zombie 22:58, 22 November 2008 (CST)
Destruction of Carried Artefacts
As you noted elsewhere, recovery of items like Mind Probes (at over $300K each) can have a very big impact on the overall recoverables value. Thinking about the impact of Power Source explosions... presumably you would only lose artefacts where there are at least 2 explosions overlapping on the same point, the first of which kills the alien, and the second has enough force (not much) to destroy the object on the ground. Or is there also an 'overkill' situation where a single explosion can destroy the alien and also destroy the objects that the alien is carrying? If the destruction of carried artefacts only occurs on overlapping tiles then it's probably easy to make a statistical judgement on the loss of artefacts, based on the percentage of total tiles that are caught in an overlapping blast. If we can make some assumptions about the location of aliens (which is probably not random but based on routes and spawn points, hmm). Spike 09:05, 23 November 2008 (CST)
- Yes, the only time you would possibly lose artifacts is overlapping PS explosions. Recall that only the Supply Ship fits this description. Breaking this down even further, there is only one possible way out of 4 where there will be an overlapping explosion. The overall chance that a PS will explode is 75% and dividing that by 4 leaves approximately 19% that the overlapping situation will happen.
- Now, there are four nodes which could spawn an engineer and only 2 of those fall into the overlap range. The spawn priorities for those nodes are 2 and 3 which is just a little higher than the usual. The probability an alien engineer being placed in a node with a priority of 2 is not double the chance of a node with a value of 1 as these numbers are weighted averages. So, if we add the spawn priority numbers for all the engineer nodes we get a value of 7 (1+1+2+3). That means an alien engineer will be placed in the 3 spawn priority node 43% of the time, 29% for the 2 and 14% for each of the other non-overlapping nodes with a value of 1.
- According to the Alien Deployment numbers in the exe, you'll only see one Engineer on Beginner-Genius skill levels and 2 on Superhuman so the vast majority of the time you will only vaporize one alien+kit. Additionally, there are 3 possible kit combinations an engineer could be equipped with and that depends on how far along into the game you are (this is just a guess as we haven't figured out what triggers one kit to be phased out, but it is a valid theory nonetheless). I can split the loss of income a little though (all values include the kit+corpse):
- Early $112,880 (PP+2PPC)
- Mid $159,080 (PR+2PRC)
- Late $225,630 (HP+2HPC+AG)
- Early $112,880 (1) - $225,760 (2)
- Mid $159,080 (1) - $318,160 (2)
- Late $225,630 (1) - $451,260 (2)
- So that's about all I can do about the total loss due to overlap. Better than nothing I suppose. --Zombie 20:13, 23 November 2008 (CST)
"Better than nothing"? How about "truly, staggeringly awesome!". Wow. Spike 20:50, 23 November 2008 (CST)
Historical Discussion on UFO Explosions moved from Talk:ExploitsF
NOTE: a lot of this may be obsolete but is preserved here for posterity pending a good hard edit from someone who knows the subject better than me! Spike 13:57, 19 November 2008 (CST)
Two related questions (have they been answered here?): 1) How long do crash sites persist?, and 2) If one Power Source (PS) blows up, yes, it will destroy E115 within a particular distance, but will it cause them to explode?
[snip] page 311 of the OSG says PSs have a 70% chance of exploding, which sounds about right to me. A table of probabilities of finding intact PSs for the various UFOs could be generated. I would have thought the answer to #2 is that they don't cause them to explode; they're not an explosive object. But in my experience, the ground floor of a Terror Ship is always the same - all four PSs are always blasted up, and the surrounding interior walls are always in the same bad shape (i.e., they're always all gone, except by the door at the south).
Anyway, a probabilistic thing would be fun to do. But the question of successive explosions has been puzzling me. Could E115 trigger something special in crashes? (Successive explosions even though it's not registered as an explosive?) E115 has always had that odd boolean checkmark at OBDATA.DAT. - MikeTheRed
... As for the explosions - I guess during the pre-detonation stages, the game may take into account nearby power units that get caught in the blast and detonate them as well. The 70% chance still holds true for isolated units. Can we add extra power units to the existing craft maps to test this I wonder? Or what about setting this boolean flag for enemy corpses, and crossing our fingers? Heh. That might not work. - NKF
Hehe, it might be fun to play with those booleans. Or not. ;)
I guess that, since PS explosions are performed by code that is "above" normal battlescape properties (i.e., PSs are not explosive per se), it's entirely possible they also programmed in, that they trigger successive detonations. It could be fairly easily tested, in theory... There are only 3 UFOs where PSs can "interact": Harvester (2 PSs), Supply (3), Terror (4). Supply UFOs might be the most interesting in terms of teasing out whether there are successive detonations. FWIW, straight odds (no interaction) of having 2 PSs intact is 9.00% (.32), 3 intact is 2.70%, and 4 intact is 0.81%.
Remind me, is there some quick way to "lift the fog of war" and see the whole map? Maybe with Mapview? (I don't have it installed ATM; I recently got a new PC.) If so, it would be fairly easy to see e.g. if terror ships ever vary their explosion pattern, etc. Also any differences between combat early or late after a crash.
Maybe we should move all these comments, maybe even this whole section, to Discussion? - MikeTheRed 17:43, 10 January 2007 (PST)
Moved - and discussions split up and line breaks inserted for your reading convenience. Coincidentally, Mike, it has been a while, but I'll let you decide what to do with your earlier comment up above.
As for the maps - I know XComutil has an option to clear the fog of war and make everyone on the map visible (and sets the lighting for the whole map to the brightest setting until you move). Use the command line utility's VIS command. Like so: xcomutil game_n vis wrt - game_n being the directory for your tactical savegame.
NKF - Cool, I just tried that XcomUtil and it worked like a charm. And guess what... I happened to try it on a Terror ship... and only ONE PS was blown up (although it did, of course, destroy the other 3). Damn, that seems like low odds... and OMG it instantly explodes (cough) the possibility of successive explosions. One other thing can be done via a screenshot of it... one can determine a PS's explosive strength (but maybe not its radius). (Has anybody determined this stuff yet?) Anyway here goes (dusting off Explosions memory cells):
Edited after I looked closely and got the equations right - MikeTheRed 17:48, 12 January 2007 (PST)
- UFO floors (their "initial" tile) are destroyed out to radius 3. They need 80 HE to kill them.
- Tiles are hit with average/2 damage; this damage decreases by 5, each tile outward from Ground Zero. So average/2 must be 80-84 at tile 3, and 95-99 at GZ.
- Twice this value (for the average) is 190-199. A PS explosion is almost as strong as a blaster bomb. (Notice how blasters scorch UFO floor tiles out to 4 tiles, 1 more than a PS, and are average damage 200.)
- Initial destroyed tiles are not blown through to raw earth anywhere, even at GZ. The slightly stronger blaster also does not blow through to raw earth at GZ on UFO floors. (Can someone look up the death tile armor strength for a damaged Power Source? The GZ PS was not raw earth, either.)
- Those walls around the PSs on a Terror ship are only destroyed up to where the floor tiles were; they also are 80 initial / 50 dead tile "armor".
- PSs at 6 straight-line radius and 3 diagonal (walking TUs means its considered 4 tiles away) are destroyed, consistent with UFO_Power_Source info that they have 50 "armor"... average/2 HE strength is 65 at radius 6.
- All the Elerium is gone, of course... it only takes 21 HE.
- A Snakeman was also killed, at a straight-line radius of 7 (just past the radius-6 PS). Here, the average/2 HE should be 60, and strength against units 60 to 180. It's a Superhuman game, so the lowest stats (Superhuman Snakeman Soldier) would be Health 45 and armor ranging from 16 Rear to 20 Front, for a minimum of 45+16=61 to kill. Easily enough to kill this weakest example; stronger aliens near to the edge of the blast (wherever it is) might have a chance of only being injured.
Does anyone ever remember seeing aliens injured by crash-landing? In my vague recollection of times when I MC everything, they are all either fine, or dead, at crash sites... never injured.
FWIW there wasn't any other vulnerable stuff around, to tell the whether the blast went farther than radius 7. In theory it could extend to radius 18, with a GZ strength of 190-199... but the largest blasts seen in the game are both clipped at radius 11 (blast and fusion tank).
- MikeTheRed 17:33, 11 January 2007 (PST)
I've seen injured aliens from pre-BattleScape power plant explosions. They're generally toast if the walk into a fire :( [This is from studying XCOMUtil reports at DIS:2.] It's more common with Large Scouts, but not unthinkable with crashed Terror Ships.
Even if it always is the same amount of time, the "in theme" way to code it would be a countdown timer (like manufacturing, research, and time to next UFO).
-- Zaimoni, 9:25 Jan 11 2007 CST
Ok I just fixed a glaring error in my math above. That's better. The soldier was practically gauranteed to be dead. FWIW a supply ship where only one "outer" PS has exploded will confirm whether 11 is the limit of the blast radius. It's probably also the best place to try to find an alien caught at the edge of the blast, to see where its edge is. Not that it's a real big deal. The probabilities of finding intact Elerium for various ships can be stated already. Probably. :)
A new wrinkle that may affect it was raised by my Terror Ship example: Only the "top most" PS exploded. And as you will recall, I had previously thought all 4 PSs exploded, and only the door and the walls near it (at the "bottom") were left. In hindsight, maybe PSs have a fixed "order" in which they explode, and what I had actually seeing is that 3 exploded (which is good for the odds of 70%) - but it was the "top 3". If the "bottom" last one had exploded, the door and walls around it would've been gone. But that's not what I remember encountering most/all the time. See? Perhaps they explode, starting from the top-most one, down. Depending on some dice roll as to how many would explode. This possible predictability in terms of the order (top to bottom) in which they are "slated to explode" could mean that e.g. seeing a particular Battleship "foot" intact tells you something about which others might be intact, at a glance (and how much E115 is left). Something worth testing with a few reloads some night.
As for effects on E115 survival probabilities in general, it looks like the Supply ship is the only ship where there is a chance for this potential "fixed order" to affect the likelihood. The central PS will always destroy the other two PS's E115. If it is more or less likely to explode due to a fixed order - if the left/top one "goes first" - this would affect the probabilities. For all other UFOs, though, the PS spacing is such that any PS explosion kills all the others' E115 - or not at all (battleship, harvester).
Thanks for the feedback on the injured guy, Zaimoni. It sounds like PS blasts are like blasters... it's possible for someone to survive it near the edge, but not likely. Combined with how I usually try to farm intact elerium (i.e., not crash UFOs), probably explains why I don't remember seeing anyone injured from crashing.
- MikeTheRed 17:48, 12 January 2007 (PST)
Did some more testing, on a Terror ship. I made a savegame with a Terror Ship crash site not yet visited, then visited it 10 times, using XcomUtil's "lights on" switch. Some observations:
- Right Zai, there can be injured folks. Three times out of 10. One of these was stunned, even. Oddly, this stunned one briefly showed in the equipment pile at initial loadout(!), but disappeared from there once I finished the loadout (and appeared in the blast area).
- Dead or injured guys are always standing right next to a PS. There were always 1 to 2, dead or wounded. The PS they stand next to is random, out of the 4 on a Terror.
- Very very oddly: Out of the 10 reloads,
- 4 had identical radius-3 UFO-floor-damage blast patterns centered on the top PS, like the first one described above. (The four includes that first one.)
- Four had an increased level of damage (floor damage radius 5), centered on the top PS, still.
- One had an even larger blast pattern of radius 8 or more; it could have been more, because it damaged the floor of the whole interior PS room, but did not extend past any of the walls (all were blown out). And it was still centered on the top PS.
- The final one had floor damage to radius 4, and was centered on the left PS.
This doesn't make any sense to me. *scratches head* Ok, that's enough for now. - MikeTheRed 19:22, 12 January 2007 (PST)
So a Power Source explodes with a power = 190-199? As a quick calculation a while back, (see this post for the specifics) I determined the PS has a strength of 200 on a desert landscape. Guess that desert tiles aren't sensetive enough to pick up on that 1-10 point difference in strength you calculated.
It's easy to create your own ship designs with daishivas mapview program: just delete everything on the ground floor of a UFO and stuff a PS smack-dab in the middle. When you visit the crash site, all you see is either a PS (if it didn't explode) or a huge crater - lol.
I wonder if BB's custom tile testing landscape could be tweaked to nail the power of the explosion down exactly? If not, I suppose I could edit the alien spawn points on a ship so that they are clustered around the PS. After properly editing alien stats of the race contained in the UFO, shoot the UFO down and visit it many times. Log the alien stats near the explosion over a few thousand reloads and we should have our answer (extrapolating damage to take the extra tile into consideration). Might need to call in a favor to BB for a "special" logger for this, but hey, it's possible. Heck, it may be easier to just edit the UFO floor tile properties to show this. ;) --Zombie 23:13, 12 January 2007 (PST)
Hiya Z! Yes well... in the follow-up testing, it suggests that PS explosion strength is variable. But this (and the finding) don't make sense relative to what you might've expected - i.e., just let some of the PSs blow up or not, at a fixed strength, with a probability of 70%. I get the feeling it may be jerry rigged in the code. Especially disconcerting is how there were bigger or smaller explosions - all centered on the top-most PS. 9 times out of 10.
If it helps any, here's a montage of the four types of damage seen in those 10 samples. The radius-3-floor-damage (seen 4x) is in upper right, radius-5 middle right (4x), radius 8+ lower right (1x), and radius 4 left PS is on left (1x). You can clearly see how all three of the pix on the right are centered on the top-most PS; notice the explosion blockage "shadow" due to the other PSs in the pic on the lower right. (Did I get my math right, when I said the strength seems to be 190-199 for the pic in upper right?)
I am so rusty on all the details of map editing. But if you can combine BB's numerical tiles with a UFO with no floor or interior walls, it should be good to go! Why not stick in more than one PS so we can see what it does relative to this head-scratcher of, is only one PS exploding, but with variable strength? A part of my head scratching is, with Battleships, it's clear that you don't always have only one PS explode. - MikeTheRed 09:13, 13 January 2007 (PST)
Whether Medium Scout Navigation consoles survive a pre-BattleScape power plant explosion is inconsistent. [It's relatively rare, but does happen.]
-- Zaimoni, 9:25 Jan 11 2007 CST
Huh, ok, I guess that means it can be variable even if there's only one PS. I had been thinking, maybe it was loading all the strength of multiple PSs blowing up, into that one PS that was blowing up, on the Terror Ship. (Did that make sense?) But if even a lone-PS ship has variability, that tells a lot about it.
Hellfire and damnation, lol - MikeTheRed 17:45, 13 January 2007 (PST)
I've done some testing with Supply ships - which has laid my confusion with Terror ships to rest. Obvious in hindsight, really... the first PS to explode on a Terror Ship destroys all the other PSs due to their proximity. This doesn't happen on a Supply ship, between the two "edge" PSs... which leads to some interesting pictures, when heavy blast damage is laid on other heavy blast damage. And yes, the explosion strength definitely seems to be variable. So, some things so far: 1) PSs appear to have a "designated order of explosion", 2) blast strength is variable but quite high (enough to melt UFO floors), 3) if a PS explosion destroys other PSs, they will not explode (blush), and 4) the radius of the blast appears to be quite large.
As for that radius, it's hard to pin down with these fairly indestructible floors and walls. It is probably less than 12, and my best guess ATM is 10. But it's hard, because it could be variable (including dependent on strength), and these UFOs don't give up secrets easily. Z, are you making a test setup? If so, I'll stop wasting my time. :-) - MikeTheRed 20:26, 13 January 2007 (PST)
Analyzing the various pix I've collected shows PSs can have a blast strength of 180 to 240+ (!) Average HE. That 240 could have been anything up to 260, but was constrained by walls containing damage, etc. Initial UFO Floor armor is 80 and Dead Tile UFO Floor armor is 50, which would take 260 average HE in one explosion, but this was never seen, even in the 240+ HEs which destroy both the initial PS (50 armor) and the dead PS (70 armor). 260 HE is over the "theoretical" limit of 255 HE... We can't set weapons to higher than this, but that doesn't necessarily mean the game can't, somehow. (Can't damage to units exceed 255? Did you ever try a maxxed shot weapon vs. units with maxxed health and high armor, Zombie?)
Another pic clearly shows that a PS blast of average HE 240 only has a radius of 10. Looks like 10 is the max radius for PS blasts. I say "max" because it may be smaller, if blast radius is linked to explosive strength, as it is for projectiles. My initial impression is that PSs don't follow the blast pattern for any other explosives, though.
If you are using numerical tiles Zombie, they should make things a lot clearer than I can. For one thing, I can't get exact numbers. And there are many restraints. But it must be a bit of work... you have to try to get a specific type of UFO to crash in a specific kind of terrain, right?
FWIW I might point out that quite a few of the crashed Terror UFOs looked the same (only 4 unique configs out of 10), but all 5 of the Supply crashes were unique. I guess it underlines how severe the effect of immediately destroying the other PSs can be. Try to space your PSs 11 tiles apart, Zombie. :) Or not... either way informs or confirms theories.
Interestingly, in one of my pix there is a Supply PS in the middle position that first got blasted from one side to a dead PS, then the death PS got vaporized from the other side.... and all that's left is the flames that come with a dead PS, laugh.
Ok, that's enough for now - MikeTheRed 22:23, 13 January 2007 (PST)
I have been rather busy the last couple days but I'll look into recreating a testing scenario with a custom UFO floorplan today. I think this will have to be a modified Battleship since that map has the largest editable area (30x30). However, BB's hacked desert landscape can only be used on scouts due to the MCD limits. A Large Scout will have to work for now (map size of 20x20). If I do not use BB's desert landscape (as may be necessary if we want to add extra Power Sources), I could probably create my own desert map (sans those nice numbers in BB's example) just to get an overview of the situation. All crashed UFO's will therefore have to be shot down over the desert landscape to see any changes to the terrain.
I haven't tried a maxed weapon strength with a subject having maxed armor and health yet. (I'll do this tonight to see what happens). However, if shot weapons follow explosive, then I'd assume weapon damage can go above 255. (Remember my tests with the Blaster Bomb? Subjects could receive more than 255 damage points with a direct GZ hit. The theoretical max was 300 [200*3/2] while I observed a max of 298 over 55 trials). --Zombie 13:26, 14 January 2007 (PST)
Mike, you may want to take a look at the post I made here.
I ended up using BB's numerical tileset and a modified Medium Scout map for testing the blast strength of the PS. It worked really well.
BTW, a Large Scout has a map size of 20x20 so by placing a PS at each corner of it the explosions shouldn't overlap. This may be handy for testing multiple explosions. 2 would work fine, but 4 PS modules could also be crammed on the map to check on overlap and/or the explosion order. --Zombie 18:51, 17 January 2007 (PST)
Thanks for that other post! It seems pretty clear that there's an explosion order. 9 times out of 10 probably isn't coincidence. But it could have been! Hehe - MikeTheRed 19:39, 17 January 2007 (PST)
- Does anyone ever remember seeing aliens injured by crash-landing? In my vague recollection of times when I MC everything, they are all either fine, or dead, at crash sites... never injured.
- I've seen injured aliens from pre-BattleScape power plant explosions. They're generally toast if the walk into a fire :( [This is from studying XCOMUtil reports at DIS:2.] It's more common with Large Scouts, but not unthinkable with crashed Terror Ships.
Probably the reason for this dichotomy is people are thinking too advanced - as if the game simulated the explosion while in the air when the UFO was intercepted. What most likely happens is the game spawns all the aliens in their starting points around the map, then fires the explosion. So only the aliens that are in the ship, near the PS, can be damaged or injured, where more realistically you might expect more to be in the vicinity (especially for medium scouts of course). Clearly this means the amount of injured aliens is going to tend to be very low compared to more realistic expectations.
Especially with the smoke of the explosion still around you could interpret this as the power systems have an emergency shut off when they are too damaged, which causes the crash, and the explosion only happens when they try to restart the engines/PS before they are fixed when the detect your team landing in the area, hence the explosion happening in the alien turn before you land. --Sfnhltb 10:46, 10 March 2007 (PST)
How long do Crashed UFOs stay on the ground?
Would love to have an authoritative answer, and it really should be on the page. I'll try to time in in the meantime.Oathbreaker 13:17, 9 February 2013 (EST)
- I'm pretty sure it's 3 days (72 hours), and it's stored at LOC.DAT offset 6-7. However, "pretty sure" is not strictly authoritative. :) I also looked for it on UFOpedia, couldn't find it, and so I started watching LOC.DAT, once ... but didn't stick with it. Maybe LOC.DAT will make it easier for you to work with? For the ultimate test, try changing it for a crash and see what happens. -MikeTheRed 17:16, 9 February 2013 (EST)
Small scout survivors
Are any of the survivors in the UFO on small scout crashes?
- If not, this at least maks it less disturbing. Presumably the survivors jumped off, or got a few seconds "grace period" before the explosion, which they used to escape. In programming terms, they're likely spawned directly outside the ship before the explosion is applied.
- If yes, then I have no answer.Medinoc (talk) 12:13, 8 May 2015 (EDT)