confidence of being better?

[tommo]

Lets say two very similar players sit down and play heads up limit poker against each other. Lets say one of them should on average make .1BB/100 hands off the other. How long would you expect it to take for this true win rate to become apparent? This is Heads Up Limit Hold'em.

[AaronBrown]

The key variable is the standard deviation. It's hard to guess without knowing the styles of the players, but say it's 10 BB/100 hands. In that case, after a million hands, the better player is expected to be up by 1,000 BB, with a standard deviation of 1,000 BB. There's about 5 chances in 6 that he's ahead.

After 4 million hands, the better player is expected to be up 4,000 BB, with a standard deviation of 2,000. Two standard deviations is conventionally considered to be significant; so I guess that's a reasonable answer.

[tommo]

right, so I guess my actually question was what is an average standard deviation given solid players.

[tommo]

[ QUOTE ]
The key variable is the standard deviation. It's hard to guess without knowing the styles of the players, but say it's 10 BB/100 hands. In that case, after a million hands, the better player is expected to be up by 1,000 BB, with a standard deviation of 1,000 BB. There's about 5 chances in 6 that he's ahead.

After 4 million hands, the better player is expected to be up 4,000 BB, with a standard deviation of 2,000. Two standard deviations is conventionally considered to be significant; so I guess that's a reasonable answer.

[/ QUOTE ]

can you explain to me real quick how you used a standard deviation of 10BB/100 to acheive a standard deviation of 1,000 over 1 million hands?