World of Tanks Win-Rate Simulator

Simulate the effect of your own win-rate in theoretical battles where the other 29 players have random win-rates between 35% and 65%. This JavaScript demonstration is intended to demonstrate the statistical fact that one constant element will have far more effect on outcome than fourteen random elements over time, because the random elements have a “cancelling out” net effect, given a sufficiently large sample size.

To test this, enter your “nominal” win-rate, ie- your estimate of the likelihood that you will personally perform well in any given game. Then, hit the “Battle” button. As the games add up, the win-rate will slowly converge toward the number you entered. It will not match precisely, but it will show that you are the largest single factor influencing your team outcomes over a large number of games, despite the fact that you have fourteen other players on your team.

Enter your “nominal” win-rate:

	Allies	Enemies
Player 1	-	-
Player 2	-	-
Player 3	-	-
Player 4	-	-
Player 5	-	-
Player 6	-	-	Wins:	0
Player 7	-	-	Losses:	0
Player 8	-	-	Draws:	0
Player 9	-	-	Games:	0
Player 10	-	-	Win-rate:	-
Player 11	-	-
Player 12	-	-
Player 13	-	-	Max Win Streak:
Player 14	-	-	Max Loss Streak:
Player 15	-	-
Outcome

Notes:

Each time you run a battle, it will select a new team.
If you run many battles, it will compute them all for the purpose of updating your totals, but it will only show the most recent one.
The page is written in JavaScript so the code can be easily inspected, in case you are curious how it was written or you want to make sure I am not doing something dishonest.
If you enter an extremely low or extremely high nominal win-rate (like 1% or 99%), the resulting outcome will be farther away from your nominal figure than it would be if your nominal figure is closer to 50%.