I think a key concept to include in the analysis is the idea of standard deviation, or variance. Sure, you win 63% of your trades, but that doesn't mean you will win 63 out of your next 100 trades.

Have you considered the possibility of what losing 50 out of your next 100 trades will do to your equity? Assuming that you will always win 63 out of your next 100 trades will lead you to increasing your risk % per trade higher than what it actually should be (for an optimal result).

In poker, winning players also suffer from inevitable drawdowns. A noted mathematician and poker pro said that a mandatory bankroll to survive these drawdowns would be equal to:

(9s^2) / (4*WR)

Where s is your standard deviation per x hands, and WR is your win rate per x hands. This assumed that your "bad luck" would never exceed 3 standard deviations from your mean win rate, which I think was about 99.7% of the time those x hands occured.

Basically what I'm getting at is to determine your optimal risk level (so that your drawdown never exceeds a certain % of your total equity, analogous to the bankroll for poker), you not only need to know your win rate, but also the variance you experience.

I'm not sure how much variance varies (haha) between different trading systems. Maybe it's normally distributed and we can calculate some sort of baseline average that can be used for most trading systems as a start.

Taking that standard deviation, you could then say I want to never experience a drawdown larger than X%, Y% of the time (Y could never be 100 in theory, but you can get to be 99.7% or 3 standard deviations from the mean as a reasonable figure). X could be what you wanted it to be, most likely your account equity minus the margin required to keep your trades open.

I hope that helps get this started.