[V3.2 Ground Combat] Soldiers/Sebillians counted as dead in mission (but still standing) and not counted as dead post mission

[wulf 21]

As requested in linked thread, I open a new one to discuss the behaviour where a soldier or sebillian would be counted as dead in mission (so other soldier can move onto same tile) but is displayed upright and still can regenerate (sebillian)/bleed(human) and will not be counted as dead after mission.

I have startet seeing this behaviour just recently (most likely starting with 3.0) and it seems to happen more frequently than the pure graphical bug (This was maybe one or 2 times per whole closed beta game time playthrough). I did not see it happen on psyon, yet.

I had actually one more occurence in Build V3.1 with a sebillian, this time he was out of UFO (did not take any screenshots). It was regenerating after being killed. However the sebillian stayed dead still and the mission ended properly when killing last alien (so not through capture and hold of UFO) but was not counted as dead post-mission.

I have one new occurence in Build V3.2 with a human soldier:

• one of my soldiers was hit on alien turn
• on being hit, there was displayed message that soldier is not bleeding now from 1 wound (which only makes sense if the game thinks the soldier survived the hit)
• on my turn however, I no longer had control over the soldier and it was possible to move other soldier onto same tile (First screenshot)
• I did not receive any further "The soldier is bleeding and looses 5 HP per turn" messages for this soldier
• after mission however (which was pretty much desaster, only one soldier really survived the mission) the dead but still standing soldier (Jane wilson as it turned out) was counted as alive in the end (Second screenshot)

So there seems to be some issue in the code for removing HP per turn (human bleeding) or adding (sebillian regenerating) where some part of game counts the human/sebillian as dead, another part as alive. The alive part only seems to affect GC mission ending triggers during one turn but then carry over to overall mission result seen in Geoscape.

Edited April 1, 2019 by wulf 21

[wulf 21]

Have one more occurence with a sebillian here. The one with the green blood below in the screenshot was killed by grenade explosion but then regenerated at the same time. However it was still dead (possible to go on same tile, not possible to target). Again mission completed properly right on last alien kill, but one alien was counted as not killed post mission. So again, it seems the not couning as dead part only affects game ending trigger just during round. (i.e. it is not triggered that the game should immediately recheck winning conditions because an alien just died, but when the next check is done anyway, either on new round or last kill, it counts the sebillian as dead and completes the mission)

[wulf 21]

Yet another occurence where human soldier is affected.

It was same pattern again ie. soldier was hit and I got the message that this soldier is bleeding right after hit very shortly (but not on new turn). During mission actually only 2 Xenonauts survived, yet, 3 were displayed as alive after mission.

This time I noticed the affected soldier was actually loosing only a little more than 20 HP on alien hit, so not nearly enough to kill her. I am pretty sure the health bar of Michelle Davis actually depicts how much max HP she had left after "killed". (I know that Samantha Ross and Zhao Mengjie were really not killed as one is visible in screenshot after it happened and I know my sniper survived).

After it happened I loaded the savefile from before the mission and saw she was one of the soldiers that actually started the battle wounded already (last 2 screenshots).

Reported that instance with F11 additionally.

So I think what's happening is that some calculation that decides if a soldier will be dead after the hit is wrong in some cases where the "current max hp" (or whatever it's called, I mean the HP value to which a soldier can be healed back up during mission) is significantly lower than the absolute max hp. It may be that the same thing might happen on sebillian. Now that I think about it, it could be that it was only happening after a sebillian already did regenerate for one turn (so it already had its "current max hp" lowered before the "kill"). It might be the reason why it does not happen for psyons - as they arent regenerating and do not heal each other, such value is probably not implemented for them in the first place (or it is there, but never lowered).

During mission (soldier with blood is "killed")

After Mission:

Screenshots from savefile before mission:

Edited April 1, 2019 by wulf 21

[wulf 21]

Just encountered it on a sebillian again. I guess I'll stop adding screenshots as they really don't tell much more at this point than they already told.

What was noticalbe about this one is that I killed it in one turn (which kind of invalidates my assumption that the sebillians needed to regeneratre first). However this kill was in multiple small hits, each of which did little damage, which lowers the max HP which potentially can regenerate. So my assumption that some "will the target die on hit?"-calculation can be wrong under these circumstances might still be right.

Edit: Actually the bug does not require the "current max hp" to be lowered before the hit. I just tried what happens if I let a grenade explode in the skyranger right on first turn of first mission. Result was the 6 Xenonauts died, 2 stayed standing and after aborting the mission it said that 4 survived. So while its random, it does not seem too hard to reproduce actually. It might be as simple as that the part of the code that checks if a hit will kill, assumes a little more damage than is actually substracted after the hit, so it has wrong result in some border cases.

Edited April 2, 2019 by wulf 21

[Chris]

Yeah, your thoughts here are rather perceptive. We're still not completely sure what's happening with this bug because I'm concentrating on features rather than bugfixes right now, but I did have the "death animation not played" issue happen to me in the editor a couple of times in the last week and I was able to check what was happening. Turns out in each case the "dead" unit actually had positive HP values; e.g. the unit would be tagged as dead but actually still have 3HP out of 50HP or something.

I can't be certain until the coders look into it but it seems likely to me that that the "death" check is being based off the new (lowered) modified max HP rather than the old value, which means if a unit has 3HP left but lost -15 Max HP the death check will think they have -12HP and mark them as dead. But I've got to reproduce it in the editor with some added logging before we know for certain.

[Chris]

Yup, I spent a few minutes testing it now I know what to look for and I've found reliable reproduction steps - it's much easier to do in the dev version of the game where we can disable the weapon damage variation. The bug does indeed occur if a unit is left with only a few HP and I'll get the coders to look into why and hopefully fix it up before the next update.

For now I'm going to assume that fixing that issue will actually also fix the Sebillians not dying bug, but please report it again if it continues in the next build.