[acidca]

I don't think there is as much variance heads up as people seem to think. If I am better than my opponent, I will win most of the time. If he is better than I am, he will usually win.
Some styles have more variance than others. Let's take two hypothetical examples. I'm a significantly better player than both John and Charlie:
John is very loose and ultra-aggro. He overplays hands and never backs down.
Charlie is too tight preflop, and decent post-flop but too predictable. He doesn't veer away from ABC poker very often.
I beleive I'll win somewhere in the neighborhood of 65% of my matches against John but about 80% of my matches against Charlie.

Also, people think variance is very high in heads up matches because they look at preflop all-in odds. But this is silly. How many of your matches are decided on preflop all-ins? I know not very many of mine are.

I think that, if you were really better than ALL of your opposition at a particular level, your winrate would be in the neighborhood of 75%*

*This conjecture does not apply to turbos, and also is without any scientific basis or merit.


This being said, I think the biggest factor affecting a heads-up winrate is how often you get paired up against someone who outplays you.