What Are the Odds of Winning a Series If....Let's take the major league division series going on right now, and analyze what has happened in the past to teams in similar situations after two games (looking at the National League contests for now).
It should be obvious that there are three possible scenarios for any team after two ballgames; they can be 0-2, 1-1 or 2-0. And there's another difference, two--the team can have this record after two games at home, or two games on the road. It should be obvious that a team that loses its first two games at home is in greater trouble than a team that loses its first two games on the road, and a team that is 1-1 has done far better if they've done it on the road; they've gotten the split they're looking for and now they have the home field advantage in their favor.
So I put together a little grid to see how often teams had won a series given each of the six different situations. I only had complete data for the 2001-2004 Seasons, to which I added the Red Sox/White Sox series this year which is already in the history books.
0-2 Home: 4 Teams, 1 Series Won, 25%
0-2 Road: 3 Teams, 1 Series Won, 33%
1-1 Home: 10 Teams, 5 Series Won, 50%
1-1 Road: 10 Teams, 5 Series Won, 50%
2-0 Home: 3 Teams, 2 Series Won, 67%
2-0 Road: 4 Teams, 3 Series Won, 75%
As you can see, the data are already sorting themselves out reasonably well. We'd sort of expect 1-1 teams to show something of a difference based on location played, but there's another, subtle effect that may be coming into play; teams that play the first two games at home are slightly better than teams that play those games on the road; that's how they get to play the home games in the first place. Plus of course there is no real reason to expect the data to make too much sense over so few series in total.
