The 90% rule

[grg121]

Taking two of my recent battles as examples I would like to ask about the 90% rule: Which number does it refer to? To the 'probability to win' or to 'attacker chances'? (These battles occured in single games, so the result doesn't really matter, but similar battles happen in many games)

1st battle: Paladin level 1 +scout against eagle at ruins. The scout hits first.
The strenghts are: paladin=37 +scout=14 against eagle=18.
The statistics were: Probability to win: 90%, Attacker: 89,4%, Defender: 10,6%, Outcome: 10,6% (I lost!)
So the 90% rule doesn't refer to the 'probability to win', or the probability should be 90,1% to ensure victory?

2nd battle: I had the same battle 4 times (at gold mines called "villages" in the map "Abandoned kingdom", at the center of the map). The occasion was: Barbarian with strength=58 and 3 hit points against 2 pikemen +1 light infantry (with strengths: 25+25+10). I won all 4 battles, but the statistics were slightly different:
2a (first battle): probability to win: 88%, attacker: 90,2%, defender: 9,8%.
2b (second battle): probability to win: 88%, attacker: 87,6%, defender: 12,4%
2c (third battle): probability to win: 88%, attacker: 93,3%, defender: 6,7%
2d (fourth battle): probability to win: 88%, attacker: 94,3%, defender: 5,7%
So, what do these numbers say? If the 90% rule refers to the 'probability' I could have lost any of these battles.
If it refers to 'attacker chances' I would never lose the three of the battles (2a, 2c, 2d) but there is a chance I could have lost the other one (2b). (Since the battles were identical this may also mean ''You never know the result of the battle 'til you hit")

Which of those are correct?

[KGB]

Grg121,

Welcome to the world of "Too Much Information" In an effort to make everything transparent the game has tended to confuse new players like yourself.

So the 90% rule doesn't refer to the 'probability to win', or the probability should be 90,1% to ensure victory?

The latter: It refers to the percentage *required* to guarantee victory for the attacker.

Probability to win: - This is what is required to guarantee victory (this number changes from 90 down to 81 depending on the total strength of all units involved so that large scale battles require lower probability to guarantee success).
Attacker: This is the attackers probability to win the battle
Defender: This is the defenders probability to win the battle

If it refers to 'attacker chances' I would never lose the three of the battles (2a, 2c, 2d) but there is a chance I could have lost the other one (2b).

This is the correct interpretation of those 4 battles.

The differences in attacker chances (90.2, 87.6, 93.3, 94.3) in 4 identical battles come from the fact the game does 1000 simulations of a battle to generate the attacker/defender win %. As you can see, 1000 simulations (which isn't that many) still allows for about 7% in variation (87.6 - 94.3) which in your case means that in 1 of your 4 battles you could have lost. This is exactly what happened in your first battle (Paladin+Scout) where you were 89.4% chance to win and were just under the 90%. If that battle was simulated again it could easily show you at 91% next time and guarantee you victory.

KGB

P.S. Note that the 90% rule has an opposite but unstated rule called the 10% rule. That is, if you were to make an attack where you are a real long shot that is less than a 10% chance to win, you are guaranteed to lose (the defender in that case is 90% likely to win so they are guaranteed not to lose).

[gil]

grg i have no idea how this works i have given up trying soulman seems to have fiured it out hes got a calc you can dawnload

you can also use my tumb X3 rule

eagle 18*3=54 power

youll need 54 power to kill that for sure

if your not using a barb

[grg121]

Thanks!