Combat mechanics question

[robo40]

Does anyone know why expected kills goes down as the target unit gets smaller? Specifically, an enemy fighter with 4 left is expected to lose 2, but after being reduced to 2 by a different unit the original unit is only expected to kill 1. Thanks!

The combat log mechanics have the same probabilities, but for some reason the expected return is lower.

[VirgilInTheSKY]

Have you picked the Commander Trait Bad Luck, which makes your damage dealt no bigger than the expected number?

[robo40]

No, but in this case the expected number is reduced.

[VirgilInTheSKY]

Does your target have Prudent hero then? For each point of strength below full strength the unit will get +1 Defense of all kinds.

[Bee1976]

Does anyone know why expected kills goes down as the target unit gets smaller? Specifically, an enemy fighter with 4 left is expected to lose 2, but after being reduced to 2 by a different unit the original unit is only expected to kill 1. Thanks!

The combat log mechanics have the same probabilities, but for some reason the expected return is lower.

Thats a know thing/issue. I experience this all the time on 10-20% rnd. My main fighter is expected to inflict 4 kills on a 6 str unit, i attack this unit with another unit first and inflict 2 dmg, and my fighter should be able to take it out, but suddenly he only inflicts 3 dmg and the enemy unit survives with 1 str left.
as an example, so the same thing you described.

i think (!) this is a problem in the combat calculations and rounding of the numbers. In most cases my fighter is able to kill ne enemy unit even if the prediction says he wont.

[robo40]

Yeah, that's what's happening. Roundoff doesn't make sense though, it's based on #of attacks not defenders.

Must be something more to the combat calcs than is being shown in the log.

[cutydt02]

This mechanic may inherits from OoB, where the less aircrafts, the more survivability. It makes sense in air combat only where concentrating firepower of multiple squadrons is much harder.
I must admit that without lethal/veteran trait AA, it’s so hard to deal with enemy fighters. But a fast deployment + camouflage 88 can wipe out entirely enemy airpower in single turn with ease

[robo40]

Hmm, maybe. The problem is that if the aircraft attacks first it has X expected kills, if it attacks last it has X-1. Maybe a loss of Mass Attack?

[robo40]

Happens with ground units as well. And how the heck can you get a predicted result of negative suppression?

[RVallant]

Are you factoring in the mass attack?

When you've already made the first attack, the other unit no longer gets the mass attack bonus, which may push them under the damage by 1 point.

[robo40]

Mass Attack only affects Init, correct? So an equal or higher init shouldn't lose expected damage by losing Mass Attack, if I understand the mechanics correctly.

Also, what does it mean when the predicted results call for negative suppression? That doesn't make sense to me.

[VirgilInTheSKY]

Mass Attack only affects Init, correct? So an equal or higher init shouldn't lose expected damage by losing Mass Attack, if I understand the mechanics correctly.

Also, what does it mean when the predicted results call for negative suppression? That doesn't make sense to me.

If enemy is able to shoot back, losing Initiative might cause less of your strength points fire before the return fire of your enemy thus reduce the damage you can deal.
Blue predicted number means all damage are supression, no kill can be made. What do you mean by negative suppression?

[Sequester Grundleplith, MD]

It has do with accuracy being percentages I think. Because there will be situations where I need to kill eg 10 HP on a unit for an overrun, but my unit can only do 9. So then I hit it with artillery for one damage, but now my unit only does 8 damage.
BUT if I have a recon next to it, I can *sometimes* get it back up to 9. Since all recon does is provide an accuracy bonus, the mechanism at play probably involves something in the shots/hits, hits/kills percentage chains

[robo40]

If enemy is able to shoot back, losing Initiative might cause less of your strength points fire before the return fire of your enemy thus reduce the damage you can deal.
Blue predicted number means all damage are supression, no kill can be made. What do you mean by negative suppression?

But having equal or higher base Init would make the loss of Mass Attack irrelevant (if I understand mechanics correctly.

Negative suppression shows up on the expected results when you hit 'L'.

[robo40]

It has do with accuracy being percentages I think. Because there will be situations where I need to kill eg 10 HP on a unit for an overrun, but my unit can only do 9. So then I hit it with artillery for one damage, but now my unit only does 8 damage.
BUT if I have a recon next to it, I can *sometimes* get it back up to 9. Since all recon does is provide an accuracy bonus, the mechanism at play probably involves something in the shots/hits, hits/kills percentage chains

Yeah, this is exactly what I don't understand. It could be accuracy based, but having Recon next to the target could just be negating some other effect.

[dalfrede]

This was posted in [06 Feb 2020 23:29].
It is the only explanation I have seen.

This phenomenon is not entirely intentional, there is no special rule in place or anything, but it is a direct consequence of how this kind of a game works. It might be unintuitive on the first glance, but if you think about it, it will actually make sense.

Imagine that you attack a 10 strength unit and kill 5 points on average. Real results will be distrbuted equally around 5, you will get a lot of 6s and 4s, some 7s and 3s and so on. Prediction in this case will be 5.

Now imagine that you attack a 5 strength unit and also kill 5 on average. Often your attack will kill 5, slightly less often 4, occasionally 3 etc. But you will never kill more than five. It is impossible because the target does not have more than 5 points. So when you average out all possible combat results, you will get a number less than 5 as well. Note that you will not see this effect all the time, but only when expected kills are similar to unit's health.

This is how it always works in this kind of a game. The logic which I described above applies the same way to Panzer Corps, to the old Panzer Generals and I think to many other games too. The only difference between Panzer Corps 2 and previous games is that it tries to provide a more accurate prediction. I actually did lots of tests (like 100000 repeats of every combat), averaged the result, compared it to prediction, and when there was a difference, I fixed it. I did it because in Panzer Corps 2 prediction is no longer just a prediction, but an integral part of all combat results at the default 50% randomness, so I felt it needed to be as accurate as possible. But this desire to provide predictions which are true to reality suddenly backfired, by giving results which players don't expect. At this point I'm not sure what's the best solution is.

[Sequester Grundleplith, MD]

This was posted in [06 Feb 2020 23:29].
It is the only explanation I have seen.

This phenomenon is not entirely intentional, there is no special rule in place or anything, but it is a direct consequence of how this kind of a game works. It might be unintuitive on the first glance, but if you think about it, it will actually make sense.

Imagine that you attack a 10 strength unit and kill 5 points on average. Real results will be distrbuted equally around 5, you will get a lot of 6s and 4s, some 7s and 3s and so on. Prediction in this case will be 5.

Now imagine that you attack a 5 strength unit and also kill 5 on average. Often your attack will kill 5, slightly less often 4, occasionally 3 etc. But you will never kill more than five. It is impossible because the target does not have more than 5 points. So when you average out all possible combat results, you will get a number less than 5 as well. Note that you will not see this effect all the time, but only when expected kills are similar to unit's health.

This is how it always works in this kind of a game. The logic which I described above applies the same way to Panzer Corps, to the old Panzer Generals and I think to many other games too. The only difference between Panzer Corps 2 and previous games is that it tries to provide a more accurate prediction. I actually did lots of tests (like 100000 repeats of every combat), averaged the result, compared it to prediction, and when there was a difference, I fixed it. I did it because in Panzer Corps 2 prediction is no longer just a prediction, but an integral part of all combat results at the default 50% randomness, so I felt it needed to be as accurate as possible. But this desire to provide predictions which are true to reality suddenly backfired, by giving results which players don't expect. At this point I'm not sure what's the best solution is.

Yea, this makes sense. It also reinforces the power of overstrength and rapid fire. If you have a 10hp/attack unit trying to kill a unit with 8-10hp, it is much harder to get the necessary kills than if you are attacking 12 or 15 times

[robo40]

Thanks, not sure how that applies to running in chess mode, though. There is no randomness in chess mode. 10 shots doing X kills doesn't change when the target size goes down, unless there's something unusual in the algorithm.

Thank you for looking that up, dalfrede. Can you search by post times? I couldn't find the thread that was in.

[robo40]

An example of what I mean:

10 shots at 50% accuracy and 60% kill rate should produce 3 kills in chess mode. Why would reducing the number of targets reduce the expected kills? I understand it can max at 3, but it shouldn't go down to 2 kills just because there are only 3 targets.

[Nustad]

Actually Rudankort is right. To take robo40 last example I calculated the expected kills (binomial distribution):

P(0 kills) = 0.7^10 = 0.028 (need to miss 10 times)
P(1 kills) = 0.7^9 * 0.3 * 10 = 0.12 (hit 1 miss 9 times is possible in 10 combinations)
P(2 kills) = 0.7^8 * 0.3^2 * 10*9/2 = 0.23 (hit 2 times and miss 8 times is possible in 10*9/2 combinations)
P(3 kills) = 1-P(0 kills)-P(1 kills)-P(2 kills) = 0.622 (law of probability total probability is 1)

So expected kills is

0*P(0 kills) + 1*P(1 kills) + 2*P(2 kills) + 3*P(3 kills) = 2.446 !!!

So in chess mode 2 kills.

With 10 targets every hit would be a kill and the expected kills would be exactly 3. So yes, decreasing the number of targets below the number of shots reduces the expected kills.