Larian Studios Forums Dragon Commander Dragon Commander - Beta Lone Trooper suspiciously good in autoresolve.

No, I don't agree. In RTS the success depends on my skills and decisions. When I let the AI autocalculate the whole stuff this element is gone. Therefore it shoulnd't be the same as an RTS match.

Just because a lone weak unit can sometimes do larger than expected damage, doesn't mean that that would be an effective long term strategy.

Even if that did prove to be effective (has anyone actually recommended spamming cheap units and spreading out as thin as possible?), it could easily be countered by using cheap units along the border of someone doing that, or by advancing your own cheap units.

The problem is if you remember an earlier beta they had it like many people are suggesting and had autocalc a bit firmer (you tended to win with very little casualties as soon as the win chance crept over around 60%). Not only did most people hate it as it punished you for going into RTS (you took higher casualties so many people were complaining that you were penalized for going into RTS and they were right). That would also really suck if you were the person with the 40%, I know i would be very angry if I had a 40% chance of victory and 9 times out of 10 I lost (or worse I auto-lost because the rules were too firm). If I have a 40% chance of winning I want to win 4/10 times.

If you made the rules not match the RTS then you WILL be penalized for wanting to RTS. Maybe it's a really important big battle that you want to fight (say attacking their capitol) and you have a 60% chance of victory. Under firmer autocalc rules you would never in your life take control of that battle! 60% chance to win means that you will almost always win with lower casualties and if you took control personally you would probably have a 40% chance of losing. Therefore noone would ever play battles that they have a decent autocalc chance if we reverted to how it used to be.

Not only that but I seem to recall that the campaign usually featured larger enemy armies to counter your dragon ability and you can only RTS once a turn there. Do you really want to lose almost every battle that isn't the one you RTS because the autocalc rules were too firm.

The auto-resolve results should be comparable to what the AI (or an average player not using the dragon) would do with the RTS battle.

That can't be comparable by no means. How could the AI calculate my RTS behaviour? Perhaps I would only build air units? Or only ground units? Perhaps I would kill everything with my dragon?

I think that battles with balanced armies should lead to various results based on some kind of random principle. But a 20%-20% battle shouldn't lead to a victory of the enemy by autoresovling it. Perhaps once every 20 or 30 campaigns but not in almost every game I play. This mechanics forces the player to engage in battle which isn't my take on the game. You should consider engaging in the field with a 60-40 or 50-50 or even a 30-70 balance against you in order to make a difference. But forcing the player to engage in EVERY battle is wrong.

Well, the AI isn't calculating your RTS behavior, autocalc is just calculating the results of the fight base on what an average RTS player would get in terms of losses. If you want special strategy, play the fight yourself. At least that would make sense.

As for the percentages, as I said earlier, right now they seem to be indeed your chances to win or lose, so you should see your troops losing an 80-20 fight 1 times in 5.
But people seem to have issues with that, so it might be adjusted e.g. as I suggested in the other thread by giving you practically 100% win chance once you outnumber the enemy substantially and replacing the percentage to lose a fight with a percentage to suffer losses.

Anyway the latest patch has adjusted something about the autoresolve, I think, so we might first want to see how it plays now.

If I have a huge army I shouldn't lose a "small fight" agaist a weak enemy. I should lose some units of course but not the whole battle, especially not if something like an RTS is simulated in the background. With a huge army I could overrun the enemy in the first minutes.....

The game doesn't have to calculate your RTS behaviour, merely a reasonable RTS behaviour. If you want it to use your behaviours and choices, enter RTS yourself.

If a 20% chance to win a battle means that you can not win, then it is not a 20% chance to win. One out of every 20 or 30 is a 5% or 3% chance. An 80% chance to win is not a guaranteed win, nor is a 90% chance to win, nor should it be.

You are not forced to engage in every battle. Giving results that ignore the chance to win would do that much more than that chance being accurate. Remember in an earlier beta when the auto-resolve was not accurate, and people with a low chance to win would always enter RTS because they could win against the AI that way, or even the odds with another player?

I found that if the results are significantly in favor of one person the fight still usually ends with little casualties on the larger side. How about this as a compromise. If you outnumber the enemy by a significant amount (it would have to huge, say your force is 5x or 10x the amount of gold) then each unit counts as twice as strong in autoresolve. This would keep the current system but allow massively larger forces to steamroll. This is not ideal (I really like how the system works now. I find them very realistic although I haven't played the new beta that just came out today) but could possibly be a good compromise for players who feel that they can game the system with small armies.

Well, the way it is now is a more "systemtic" approach, very Risk-like. You roll a dice and the outcome is calculated.

Another approach would be to take the balance of armies more into account which would be a more realistic approach.

In the end I understand that you can find arguments for both approaches so any solution will be legit.

Agreed.



Disagree with both points. It's an issue with how casualties are calculated in both the rts mode and autoresolve. At present it's based on a relatively direct proportion of how many units die in combat which doesn't respond well to starting with a few or even an overwhelming amount of extras. Instead of this you could match the units against each other more directly and apply far less of a weighting to what happened in the RTS portion. So even if you lost that battle between 1 Trooper vs 3 Hunters an Armour and a Shaman you might only expect to lose one of the Hunters, the other units would be pushed back.

I'd also say while every game will have some slightly overpowered strategies good games never have completely overpowered strategies. In particular though sending streams of cheap units against armies completely kills any depth in the campaign map.



This is not acknowledging the scale of the issue, it's not that cheap units sometimes do higher than expected damage, it's that singular units almost always do massively disproportionate amounts of damage against large numbers of entrenched foes. Not only is it an effective long term strategy it's the only effective long term strategy. Keeping your army behind the lines and using singular units to counter singular units is just wasting resources by never bringing them to bear, and risking them being brought into combat through the use of strategy cards.

singular units almost always do massively disproportionate amounts of damage against large numbers of entrenched foes.

That has not been my experience. I've taken higher than expected damage with a good advantage or not, gotten off lightly with even odds or barely won, and done better than expected with poor odds or simply gotten wiped out.

I'll try a campaign tomorrow against the insane AI and keep track of the battles.


Not only is it an effective long term strategy it's the only effective long term strategy.

Then why have people complained that the AI doesn't expand fast enough on Center Mass, letting you take over most of the neutral countries?


Keeping your army behind the lines and using singular units to counter singular units is just wasting resources by never bringing them to bear

So when an AI sends out singular opponents it is the only effective long term strategy. If you send out single units you are wasting resources. Are you assuming that you will never win a country with a 50/50 chance, so you can never get close to anywhere the enemy has units massed?

There is one thing naturally in favour of the lone trooper: Under no circumstances will the larger force take out more than a lone trooper.

If autoresolve corresponds somewhat to the visualisation, with units going against eachother individually (one squad on your side versus one squad on the other side), a single trooper versus 5 troopers with no other advantages or disadvantages has:

- A 50% risk of losing without causing any damage
- A 50% chance of taking out at least 1 unit
- A 25% chance of taking out at least 2 units
- A 1/8 chance of taking out at least 3 units
- A 1/16 chance of taking out at least 4 units
- A 1/32 chance of anihilating the enemy (without casualties)

Perfect casualty-distribution across 32 battles
x1: Army 5, Loner 0
x1: Army 4, Loner 1
x2: Army 3, Loner 1
x4: Army 2, Loner 1
x8: Army 1, Loner 1
x16: Army 0, Loner 1

Total loss over 32 battles:
Army 31, Loner 31

Given such a system, it seems that the effect should even out over time. For the sake of realism, the loner should be under fire more frequently than it should be able to fire when pitted against an army.

Given that units on the same side are not equal in purpose, cost or power, this is obviously a lot more complex, but I'd care to know whether or not the basic principle or the results are anything like what I describe. (Does 1 against 5 even units have a 1/6 win chance <16%>, a 1/32 (3%) win chance or something else; which is closer?)

If you're getting 16%-like numbers on that, isn't the autoresolver being awfully nice to the weaker side? (Getting to face your enemies 1 by 1, you'd have 50% five times, which is what I find to be slightly in excess of 3% chance overall. Having to face your enemies all at once should not be easier.)

Certainly it's been my experience, while occasionally a single troops will fail to kill anything it seems like better than half the time it will wipe out 3 to 4 times it's own value. Hard figures would certainly help pin things down though, especially since the matrix has apparently been tweaked in the last patch.



I might be missing something but that seems like the same problem to me, fast expansion and spreading your forces thinly is the only effective long term strategy and the AI isn't doing it.



No, maybe I could have explained that better but sending out single units is by no means a waste of resources, some die to no effect but in general they pay for themselves many, many times over. The waste is in having a large number of troops in one location, either you deploy them and suffer tremendous casualties to little benefit, or you don't deploy them and they achieve nothing. By spreading your forces evenly across the entire map you can easily and quickly wear down any force the AI brings to bear.

I fear your calculation is wrong because your assumptions are wrong.

You assume that this is like a mathematial experiment in which the lone trooper fights against only one trooper of the enemy army at the same time, so one after the other. If he survived against trooper one, he will fight against trooper two, and if he survived that fight again, he will fight against trooper three, and so on.....and then you calculated the probability of the lone trooper to win this game.

But the "reality of war" is different. In reality the lone trooper has to fight against all five enemy trooper AT THE SAME TIME. The moment he can fire one single shot at one of the enemy troopers he is hit by five shots of the enemy. His chance to win this fight is about 0%. His chance to take out only ONE enemy trooper is also about 0%. Let's assume that a trooper dies when he is hit by five shots. Then the lone trooper wouldn't survive even the first "round of fire". And even if he dies after 10 shots he would die before he could even endanger one of the enemy units.....

Therefore a "reality-based" or even "RTS-based" autocalculation of a "1 trooper vs 5 troopers" battle had to lead to a 100% chance to lose the battle and a very, very, very tiny chance to even kill one single enemy unit.

While I'm not so sure about Sinister's calculations either, I think you are potentially wrong with that quote as well (sorry).
At least given the internal logic of the game.

If we would play the battle in RTS mode instead of using autoresolve, we wouldn't see our troops directly facing of against the lone trooper.
We would have our troops start at their base. Likewise would the enemy trooper.
While we would directly march towards the enemy base, the enemy would instead start building more troops. Of course it wouldn't be a lot of troops, but let's assume that by the time we reach the base, he has produced ~5 more troopers. Suddenly the odds are totally different.
While the enemy still has no realistic way of winning the battle, he will still inflict higher losses than a single trooper could.

Now, we are talking about autoresolve here, not the RTS mode, but autoresolve should still be balanced somewhat to the RTS part (otherwise autoresolve would be vastly superior to RTS mode for such battles, which is strange considering that the RTS part should probably give better results due to it being controlled by yourself, in form of a friggin dragon).

My personal suggestion would be to first adjust the RTS part in a way that what I described is less likely to occur (e.g. increase building time/costs, put in a timer before you could start building new units in RTS, etc.). That would probably result in lower losses in a RTS match, in turn allowing for more sensible losses in autoresolve as well.

I only wanted to demonstrate that his calculations were wrong from the beginning.

But I still don't see the point why the outcome should be different. If a match between two armies is simulated like an RTS AI vs AI battle there is no point that the result would be different because the AI would be identical. If not either the AI is implemented poorly or the element of randomness is way too weighty which is again the result of an AI poorly implemented.....

I think (and that's my honest opinion) that there shouldn't be an RTS simulation in the background because it's completely pointless. You could roll some dice in the background, it would be the same. At least, rolling some dice would reflect the Risk-origins much better....


And I don't agree that autoresolving would be superior to personally engaging in battle:

Simple example: you have only one trooper, the enemy has five troopers (like in the example above):
a) You autoresolve the combat. Result: you lose
b) You personally engage in combat: Result: It would be very hard to win but there is a chance.

Another example: you have five troopers, the enemy has one trooper (opposite scenario):
a) Autoresolving. Result: you win
b) Engaging. Result: You most likely win but you could also lose if you make bad decisions or play poorly.

Third example: you have three troopers and the enemy has three troopers (balanced scenario):
a) Autoresolving. Result: 50/50, pure random principle
b) Engaging. Result: Depends on your skill but you have a fair chance to win.


So what's the result of these three examples? In two of three basic scenarios you have an advantage to personally engage in battle. Only in the scenario in which you have a far bigger army there is just no point to personally engage in battle. And that's how the game should work imo. The 20% probability to lose a battle if you fight with one trooper against four enemy troopers is just wrong imo. You should have way worse chances to lose this battle and same is true for the other way round. If you attack an enemy with a very small army your chances to win by autoresolving should be minimal (REALLY minimal). So if you want to do that, you should be literally forced to engage in combat personally.
Everything else is just "betting on the system": produce one trooper in each round and send him in a country with a big army and autoresolve the battle. After X rounds you will win this country and the enemy army by pure random principle, because the chances of winning the battle are just way too high.....

You know what I mean now?

Not to engage personally in battle shouldn't be rewarded.....

This in a nutshell. If anything we should be penalized for not fighting as a dragon rather than the reverse. That way you feel your presence actually matters. Otherwise why even bother playing RTS matches?

I don't think there is. I just said that it is (or the results are) balanced accordingly.




It used to be in older builds, for a simple reason:
When you go into RTS, you have to build a base and produce more troops.
This will almost inevitably lead to higher losses (at least for me, and I assume the average player as well).
In contrast, autoresolve produced the results one would expect from directly putting the units against each other, without any additional units produced during the fight, i.e. a single trooper did not cause such high losses as it does now.
Consequently, in such cases using RTS was a bad idea, because autoresolve used to give more consistent and lower losses.



Actually I think our oppinion on the matter is not so different, I merely argue that it is that way for balancing reasons.
Bringing it more in line with expectations would however potentially favor autoresolve over RTS again, unless RTS is rebalanced a bit as well



Before you used to be punished for doing so, which isn't so great either.

I agree with what you say if there weren't the possibility of unit production during RTS battle... The point is that 4 troopers against 1 as starting units is not such a huge advantage - especially, if you play 2x2 (even if the other two factions do not have starting units at the beginning of the battle). I think I am quite an average player and I have made the experience that if I start the battle with 4 troopers versus 1 (or vice versa) I can either win or loose very easily (depending on the map). If I rush on the enemy base with my 4 troopers, all of them will be likely killed - either due to produced units or turrets. And even if I succeed to take down this base there is still another opponent in 2x2 who has already produced Giant Blob of Death during the time while I was engaged with the first enemy. As already stated, at the end I can win or not (depending on the map), but I will very certainly suffer heavy losses.

So, my point is: 4 troopers are just a very small starting force compared to all the units which can be produced during RTS. If I engage in battle I am likely to loose all the units - and it is now pretty much what the outcome of the autoresolve is.

The problem is not the autoresolve which promotes the thin-spreading of units but rather the possibility of unit production during RTS. In earlier betas the player was punished for going personally into battle, because the autoresolve gave a result with no losses in the 4 versus 1 trooper situation. Now, it just mirrors what happens if you enter the battle as a dragon.

In the reverse situation, in which I attack with only 1 trooper against 4, I also have a pretty decent chance to win the RTS battle (because I can produce lots of units during RTS). That is also the result which I get with autoresolve.