I love reading Rob Neyer’s M-Th baseball column on ESPN.com. Today he discussed an Orioles-Angels game in which Angels closer Troy Percival gave up three runs in the bottom of the ninth inning to blow the save and allow the Orioles to tie the game. The Angels came back to score in the top of the tenth, and another Angels pitcher with an unspellable name took over for the tenth inning to end the game.
Normally, thanks to baseball’s somewhat arcane rules on how to determine the winning and losing pitchers, the Angel who pitched the 10th would have been awarded a save, and Percival the win. This is, of course, utterly crazy, which is why Rule 10.19 allows the official scorer the discretion to say “Hell no! Percival pitched like my grandmother. Unpronouncable Japanese Reliever gets the win.”
Neyer attributes this decision in part to the fact that the scorer works for the Orioles, the home team, and therefore was free of any pressure to give the stats to the hometown boys. This assumes there is a hometown bias in how baseball games are scored.
If this bias is real, there should be a statistically significant difference between the errors a team is charged with at home and on the road. But I can’t find any statistics on road/home team error splits. Can any baseball fans help me out?
[sub]Meanwhile, my beloved Rangers have tied the game at 4-4. Nobody’s noticed yet, but they’ve been playing quite well for the last couple months. Enjoy your fluke season while it lasts, Mariners fans! Aaron Sele’s annual late-season breakdown begins any day now![/sub]
There is no way that someone can be 100% objective therefore it should show in the statistics, even if the occurence rate is low (say once every 3 games). Of course that could be explained away, by saying that the home team is more familiar with the field, gets home cooking, etc.
I am going to do the quick and easy version of this, just to see if more in depth examination is worth while.
The following statistics are extracted from the STATS, Inc. 2000 handbook, and reflect the 1999 MLB season. I am presenting, for each team in the AL, the errors charged to that team at home, the errors charged to that team on the road, the errors charged to all opponents at home, and the errors charged to all opponents on the road.
I have NOT analyzed these in advance - I am transcribing data, and will see if I see anything obvious
So - while opponents error rates stay basically the same, a given team seems to record fewer errors on the road then at home.
The problem now comes with interpretation. There could still be a bias in the home town scoring - but that bias is reflected in a desire to help the stats of the home town pitching staff (which looks better when there are more errors, and thus fewer earned runs).
I probably can track down other years of the same sort of data - so this could be subject to further examination.
I can’t offer any hard data, only my impressions, after watching baseball for many years.
I often disagree with official scorers, but they’re very rarely biased. I think they almost always try to render accurate, impartial decisions. If they slip up, most of the time, it’s simply because they’re human.
The only times I’ve ever seen blatant, DELIBERATE miscalls have been during games where the hometown pitcher had a potential no-hitter in progress. I mean, suppose Greg Maddux has a no-hitter going against the Mets through 6 innings, and Mike Piazza hits a hard grounder right down the third base line. Chipper Jones dives for the ball, but it bounces off his glove into left field.
In that situation, will the scorer rule “hit” or “error”? It probably depends where the game is played. If the game is at Shea Stadium, the scorer (usually a local sportswriter) will probably decide that the ball was well-hit, and Chipper Jones would have had to make an exceptional play to get Piazza out… so he’d rule “base hit.” On the other hand, if the game were in Atlanta, the scorer would rule “error” to protect Maddux’ no-hitter.
By the way, I’m NOT picking on Atlanta! Scorers in almost any city would do the same.
First of all, Shigetoshi Hasegawa has a relatively easy name to pronounce IMO. Especially if you ignore his first name.
There may be some bias in official scorers, but I doubt it ever turns up enough to be statistically significant. If you consider the number of plays made and the number of errors committed, there just aren’t that many chances to make a difference.
Right up my alley. I work for the Colorado Rockies A-Short Season affiliate, the Tri-City Dust Devils. I have only scored two games myself but work beside the official scorer.
Our official scorer is not biased, at least not intentionally. But I can’t say the same for the rest of the press box. He is constantly harped on by both radiomen, and occasionally he is even summoned to the managers or umpires office for a little “discussion.” He kind of has a reputation for leaning toward the error side, note we are near the top in NWL pitching:), but he is consistent and statistically is in line with most of the league.
But that has to be one of the most thankless jobs in all of sports anyway.
Thanks to all who have responded. Just for clarification, I’m not arguing that there’s any intentional bias, but rather that exactly the sort of pressures Cyberhwk describes could lead to
So far, I think dorkbro has come up with the best evidence, though heck if I can tell what it means. I’m actually quite mystified by those numbers. If I read your columns correctly, individual AL teams have somewhat more (~5%) errors at home than on the road, but their opponents have virtually identical home and road errors.
Huh? All I can figure is that this is a result of interleague play, and that since NL teams get thrown into the opponents’ columns, the NL teams must have had more errors away than at home last year to make up for the AL’s surplus of home errors. (Does that make sense?)
Anyway, I’m starting to suspect that the error rates (in the big picture) are essentially identical at home and away, but I think more data would be required before reaching that conclusion.
“could lead to an unconscious bias.” Sheesh, where’s my head this morning? Man, I’m lucky I’ve got a two-month vacation starting at 5:00 this afternoon!
When they’re on the road, and behind going into the bottom of the ninth, the game is over. Crudely, let’s estimate they play on average 8 1/2 innings in the field, instead of 9 (ignoring extra innings), so multiply 639 by 9/8.5 to get 677. It’s still kind of low, but not too bad. Of course, you have to do this to one of the opponents’ column also: 688*9/8.5 = 728 (Is this the right column to scale?), so the scaled total columns are:
695 (Home Team, home errors)
677 (Home Team, away errors)
728 (Opponents, away errors)
685 (opponents, home errors)
Actually, that is helpful, since I’d not considered that visiting teams play an inning less defense about half the time. Taking this into account, it looks like the individual teams’ home and away errors converge fairly well (from ~5% to ~2.5% more at home than away), while their opponents’ home-road error splits actually diverge (from almost equal to ~6% more on the road).
