Nationals Baseball: Expounding on "should win series"

Friday, June 04, 2010

Expounding on "should win series"

FJB brings up a nice point that I wanted to break out in ol' mathematical form, for you guys to think about. Just because you are playing a bad team there is no guarantee that you will win those games.

Forget for a moment the dozens of little things that can effect the chance of winning a single game, who are the starting pitchers, who's in and out of the lineups, who's injured or not, hot or not, is the bullpen rested, who's at home, is the team in turmoil, etc etc etc. Let's just assume none of that matters for the time being and that the Nats are an exactly average baseball team and the Astros are about as bad as the Nats were the past two years.

The latter point would put the Astros at a winning percentage of about 37%, and the former point means we can ignore the fact the Astros would win fewer games against teams better than average and vice versa.

Given the fact that the Astros would win 37% of their games against the Nats - what are the chances of the Nats outcomes in a 4-game series.

Lose All - 1.9%
Win 1 - 12.8%
Split - 32.6%
Win 3 - 37.0%
Sweep - 15.8%

Of course this is nonsense for the reasons stated above. The games aren't completely random. But it gives you an idea of what to usually expect going into a series, even with a team that you think you are far superior to. A split is far more likely than a sweep. There's a decent chance you could win only one game.

And this is against arguably the worst team in the league. Give the other team just a 41% chance of winning and the split becomes the most likely outcome and your chances of a sweep start falling below your chances of losing 3 games.

I'm probably not saying anything you don't already know from years of watching baseball. Sweeps are hard. Your team will sometimes lose and win games it "shouldn't". But it's sometimes nice to remind yourself that with an average team going into a favorable stretch against bad teams (or a worrisome one agains good teams) you shouldn't be looking for much past a few games around .500.

1 comment:

Bryan said...

It's not so much that numbers favor you heavily, as that they favor you more heavily agianst bad teams than they do against good ones.