FinnMcCool said:
1) there are hit ppint bonuses in for ancient cavalry and the war elephant.
2) percentage of winning a round=percentage of winning an entire battle
#1 is irrelevant. We're talking about the combat system for all units.
#2 is just incorrect. It's mathematically correct, but it isn't a useful equation for calculating the chance of winning a battle. The complaints are coming from players who object to a VERY HIGH Combat Strength losing 1/3 of it's total viability due to a single round losing against a very weak unit.
It's not even mathematically correct, it's only correct if bith units have only 1HP.
Here is a good Combat Calculator:
http://forums.civfanatics.com/showthread.php?t=75765
In the above example:
Tank vs. Tank on Plains(not fortified) both 3HP
Attack 16
Defense 8.8 (10%defense bonus)
Chance of winning one round for attacker : 16/24.8 (64.52%)
Chance of winning battle for attcker: 75.7 %
Tank vs.Tank (not fortified) on forest with battle across river
Attacker 3 HP, Defender 4 HP
Attack 16
Defense 12
Chance of winning one round for attacker: 57.14 %
Chance of winning battle for attacker : 48.5%
So, dont forget about defense bonuses and Hitpoints, they can make the difference
(OTOH even in the first example winning only 10 of 20 fights is not that unlikely)
And after playing a lot C3C I can say: The RNG is perfect, it gives the results you expect from an RNG, because some streaks are part of Statistics, a RNG without streaks would be very flawed!