are you sure there's no way to cheat?

[gggol]

Or that odds calculations are correct, or that battle outcomes really follow the odds? Sure there isn't some subtle bug in the battle code, maybe something that unintentionally favors units produced in the northern and western sections of maps over the southern and eastern sections, or favors units produced by players whose turns are sooner in the turn order, because they are listed first in the data structures, and get to roll and apply hits first, something like that?

In this ladder game I'm playing against Solo, I'm now up to somewhere around 15 losses out of 15 fights that I had between 50% and 80% odds of winning. Have only 1 win where the odds were against me. I did about average in unimportant fights out in the wilderness, but lost all the ones that really mattered, over cities and powerful units. The only important wins for me happened when my odds were better than 90%. The luck is now so lopsided that I think it rather likely my opponent has found and been using some kind of cheat. It's possible I have just had awful luck, but I would have to be very unlucky to have that much.

I can think of one way to cheat that might work. Start a fight, and disconnect from the Internet. If it goes well, reconnect, and take your win. If it doesn't go well, let the browser timeout, then reconnect after it has reset the game to before you made the attack. Can try again and again even against low odds, until you win. Don't know if that works, but it might. Doesn't seem that could work against me when it's my turn, and about half of those fights happened on my turn.

[KGB]

gggol,

The battles happen on the server, not in your browser. So disconnecting won't help.

As far as I know, there are no bugs in the battle calculator. I wrote a separate one using the same rules to test Piranha's and to verify his odds. I haven't tested mine against his recently but I haven't seen any need to because I have not seen anything strange (like losing a lot of battles where I am favored).

If you can give me some of the battles (units on both sides and bonus's) then I can work out whether everything looks correct.

KGB

[gggol]

Unfortunately, there isn't enough battle history for me to review those fights. I'll remember as best I can.

I calculated some of the odds myself, and found agreement with the numbers the game reports. But the way the battles actually went has me wondering if the battle code really follows those odds. Most particularly, lots of weak units can't seem to win against a few powerful units, even when the odds are as good as 80% to 85%. One fight like that was elf, lt cav, 2 pikemen attacking a blessed valkyrie in a forest space. The game reported odds of over 80% in my favor, but I lost. Another 80% fight I lost was kraken, 2 scouts, 1 or 2 pikeman vs 1 heavy cav in city with 10 defense. I don't expect to win every such fight, of course. 80% is not 100%. But losing 4 out of 4, and another that was 70% odds of winning, and a bunch of fights with odds ranging from 50% to 65%, has me suspicious that it's not bad luck, it's some kind of problem.

[KGB]

gggol,

The examples you stated (1 Valkrie, 1 Hv Calv) are interestingly enough both battles against a single enemy unit. In those cases, regardless of the odds, you won't be able to do any partial damage when you lose because there is only 1 enemy unit so you either kill it or you don't.

When those kinds of battle happen it always feels worse than losing a bigger battle (say 8 vs 4) where you do some partial damage by killing some of the enemy. I wonder if those are what's sticking out in your mind.

KGB

[gggol]

They were all battles of many weak units vs few strong units. Which has me suspicious that there is something wrong. Weak units don't do well, but should do a little better than that.

I went through several battles where I had 20 heavy inf and pikemen in a city, and my opponent attacked with a super stack, containing a high level paladin and dread knight, a siege engine, and heavy cav. I lose all 20 units, and he loses 3. This is not a surprising outcome, given how overmatched the weak units were. And BTW, that's why I think you ought to raise the minimum str to 10, or change how fear and chaos work. It's a huge swing to take heavy infantry from 25 (thanks to city walls of 10) down to 5 str. That last drop from 10 to 5 str cuts their expected damage down to nothing.

But that's not the problem I'm talking about here. The problem is that the weak units are still losing even when they have odds upwards of 80% in their favor.

[piranha]

I think its quite safe to say that there is no cheating going on.

About the results from battle. I've had my share of outcomes that made me think there must be something wrong, but when I test the results everything is right. KGB also checked and got the same results and even if there are new bonuses since we did those checks the calculation is the same. If something is wrong the numbers shown in battle should also show that.

Its a bit unlucky to lose with 80% but its nothing extreme. Being 65% isn't the same as "should win" its more like pray to god that I win this battle that I shouldn't have gone into unless I'm totally desperate. You can lose a lot of 65% in a row and there is nothing strange about it. Grab a dice and start throwing and you will see .

[Solo]

What I could say.
As for me battle results' delta was in normal limits in this party. There was some battles where I was lucky, but often during all the game I was surprised that my units was killed when they had to win. I think it could be usual for most of us way to notice bad evens for us and forget good events for us taking good events as a matter of course.

There is another thing also. Sometimes I noticed that success percents are not right (IMHO). So I look atunits' strength and calculate chances and my results are different and battle go like for my results. Could percentage calculation be a little not right?

Also I often see that strength of units is shown not right. For example, the stack of elves with a red dragon fight against stronger units without any unit that could change strength and remove morale. The red dragon must give +12 in morale but shown strength of elves is usual. Meanwhile battle result is right (the red dragon add strength to elves).

[LichKing]

I have to say.. I don't think that there's cheating or errors, but to lose those >80% hero battles is really annoying. 90% (and less, as piranha said in another thread) is a blessing, but maybe isn't enough. I'd really like there were a hero skill that could affect battle behaviour. For example:

Normal battles (without hero): like now (90%, 85% if total STR >80 and <400, 80% if total STR >400, if I'm correct)
With hero: 1st level 80%, then a "Field Vision" hero skill to improve automatic winning chance (every 15 skill points, down 5%, up to 65%)

[KGB]

LichKing,

I'm not sure a 'field vision' skill that increases hero stack odds is such a good thing. Heroes already have a huge advantage with their own bonus's and the fact most players place their best units/stack bonus units with their heroes. So killing such stacks (past about turn 10) is already very hard for non-hero stacks and making it harder is only going to tilt the game even more toward hero stacks.

I know I wouldn't like to see such a hero skill unless it was SEVERELY limited to say 1% per hero level capped at 5 (so a hero would cap out at 85%). Then there is no need to spend skill points on this feature.

KGB