Got to assume you're talking about points in which the game
was actually won, not all 'game points'.
First, in runouts, the server has obviously already won. In 18.75%
of the games.
In the rest (non-runouts), I'm assuming the team who just won
the point (i.e., the current server) has more points (the majority
of the time) than the team coming off the bench. Not positively
sure, I suppose, but a guess, since they have just won 2 points.
So, give them just a 50% closer rate, and you still get a majority
of server wins.
Interesting, but I agree, not a handicapping consideration.
