Home team record in extra innings
Posted by Andy on August 6, 2007
In my gut, when my favorite team plays in extra innings, I've always felt like they have a better chance of winning when playing at home then on the road. I decided to find out if that's true or not. I'm sure this data is available somewhere, butI just used the PI Team Pitching Game finder to figure it out. This is what I did: Going to the Play Index Team Pitching Game Finder page, I reasoned that in an extra innings game, the home team (win or lose) always pitches more than 9 innings. (In fact, they always pitch at least 10 innings.) So, in the center column, I checked the bubble for "Home", and in the right column, I set IP to be greater than or equal to 9.1 innings. Next, in the left column, I clicked the bubble for Win, and then I summed those results by year. Then, I went back, changed that selection to Loss, and again summed by year. First, here are the results since 2000, through games of Saturday.
Year Home wins Home losses W% ----------------------------------------- 2007 78 74 0.513 2006 105 80 0.568 2005 97 85 0.533 2004 115 103 0.528 2003 108 89 0.548 2002 115 85 0.575 2001 94 101 0.482 2000 101 101 0.500 TOTAL 813 718 0.531
So the home team usually has a slight edge, although as you can see the visiting team actually won extra-inning games more often in 2001. Since 2000, the home team has a winning percentage of 0.531 in these games.
I also went back and summed it all the way back to 1957. For all extra inning games 1957 through Saturday, the home team has won 4846 and lost 4449, for a winning percentage of 0.521. Looking back through the decades, there have been some ups and downs, but for every decade the home team has had a winning percentage of at least 0.510.
Just as I wrote this, on Sunday afternoon, the Phillies scored two in the top of the 11th against the Brewers...so much for the home team being favored 🙂
I wonder what the rationale is for the home team winning more often in extra innings. Mind you, with 400 more wins than losses over 50 years, that is an average of just 4 games per year swinging in favor of the home team. My best guess is that it is psychological. The home team always knows what is required of it. If the visiting team scored in the top of the inning, then the home team must score to avoid losing the game. I think this provides a little bit of extra focus, which makes only a very small difference in the long run. At first I had a theory about the use of closers, but since the W-L record in extra innings has been pretty constant since 1957, I don't see how closers could factor in.
Another interesting tidbit would be to look at W-L records for 9-inning games that were tied after 8 innings. I am guessing that the home team has just about the same winning percentage. (Perhaps Tangotiger will post the win expectancies for these situations.)
August 6th, 2007 at 10:11 am
What is the winning percentage of home teams in games that are 9 innings or less? This would be the number to compare with the .521 winning percentage.
August 6th, 2007 at 10:33 am
Good thought, frankman.
Using the PI, I searched 2000 to 2006, just set it to home game wins, then home game losses, games of 9 innings or less, summed by year.
for home team wins:
2000 1211 Ind. Games
2001 1179 Ind. Games
2002 1199 Ind. Games
2003 1227 Ind. Games
2004 1184 Ind. Games
2005 1209 Ind. Games
2006 1222 Ind. Games
And home team losses
2000 1015 Ind. Games
2001 1054 Ind. Games
2002 1026 Ind. Games
2003 1005 Ind. Games
2004 1026 Ind. Games
2005 1039 Ind. Games
2006 1022 Ind. Games
Total home team wins = 8431
Total home team losses = 7187
That's a winning percentage of 0.540 for home teams in games of 9 innings or less. So, in actuality, home teams win extra inning games slightly less than games of regular length. I suppose that is basically a regression to the mean, so to speak. In other words, over the course of 9 innings, the home team tends to outscore the visiting team. But in any one inning (such as a decisive 10th inning), those larger trends are less apparent and the record drifts slightly more towards .500.