July 3, 2007

Juan Pierre's bunt double and a new baseball statistic

I went to the Dodgers baseball game tonight and saw something extremely rare in baseball: a bunt double. Speedy Dodger centerfielder Juan Pierre tried to dribble a bunt down the third baseline for a single, but popped it up over the head of charging Braves third baseman Chipper Jones. The mishit bunt landed a little past third base and trickled off into foul ground. Pierre reached second easily.

Here's a YouTube of Pierre getting a bunt double in an earlier season, although in the video he appears to be intentionally bunting the ball hard into left field. What I saw was a pure bunt executed poorly, but Pierre appeared to be the least surprised person in the stadium and exploited his luck nicely.

By the way, young Dodger catcher Russell Martin, who will start the All-Star Game a week from now, looked great with four hits and non-stop hustle on the basepaths.

That reminds me of a baseball statistic that doesn't exist, but should. Hitters' statistics are biased by the home ballpark they play half of their games in, so honors like Most Valuable Player and the All-Star Game often go to guys who just happened to be at the race place at the right time to drive in 130 runs. Everybody knows that Colorado Rockies hitters aren't as good as their eye-popping statistics suggest because of the mile-high elevation means fly balls travel farther and faster due to less wind drag, but it's hard to keep track of how the other 29 parks bias statistics. This is especially true because park factor can change from year to year. For example, Wrigley Field in Chicago is normally a terrific hitter's park because on hot days the wind from the south blows out to left field, turning outs into homers. But, in the summer of 1992, it was almost never hot in Chicago due to the eruption of Mt. Pinatubo in the Philippines the previous year spewing ash into the atmosphere.

Even statistically adjusting batters' performance for park factor (runs scored at home / runs scored away) has its problems because it's not all that stable: e.g., Dodger Stadium jumped from 25th best hitter's park to 10th best from 2005 to 2006. Was this due to some objective change in hitting conditions in Dodger Stadium or just because of luck (e.g., pitchers having worse outings at Dodger Stadium than in away games)?

On the other hand, dedicated fans can typically tell you with a lot of accuracy who is the most valuable player on their own team, not from statistics, but from the number of times he was the best or almost the best player in the game. For example, Russell Martin was clearly the best player on the field last night. If I attended, watched or listened to Vin Scully (now in his 58th year announcing Dodger games!) every night, I'd have pretty accurate impressions of who the best player on the Dodgers was, and by how big a margin.

This approach is independent of park factors. For example, back when the Houston Astros played in the old Astrodome, one of the worst hitters parks in history, they came up with a long string of terrific players like Jimmy Wynn, Rusty Staub, Cesar Cedeno, Bob Watson, and Jose Cruz. But only Joe Morgan made the Hall of Fame ... because he got traded to Cincinnati, a reasonable hitter's park. But Astro fans knew how good these guys were because they saw them holding their own night after night against competition who came to town with flashier statistics..

This kind of relative game analysis could be institutionalized. Right after the end of hockey games, the announcer traditionally proclaims the top three players in the night's games. I presume a few sportswriters make up the list. This could become an entertaining tradition in baseball as well. You'd then sum the game rankings over the entire season. (It would be best to rank all the players in the game, to prevent low-average sluggers from getting an unfair boost by just including games when they were in the top 3 and leaving out games when they didn't do anything.)

Or, you could do calculate the best players in each game statistically, giving a point for every total base and on-base, as in the OPS average.

Still, a subjective ranking for each game could be useful. For example, I followed the Dodgers closely in the 1971 pennant race, and it was clear that 38-year-old shortstop Maury Wills was their most valuable player, even though his statistics were merely average (.281-3-44 and 15 stolen bases). Night after night down the stretch he made the big plays that made him among the best players in many crucial games. That year he was well-recognized for his contributions, finishing 6th in the league in the MVP voting, but he was lucky because he was a famous old player with the spotlight of a pennant race on him, so the national media heard about his contributions. Less well known players could benefit from some kind of game-by-game ranking system.

It's a little bit like how you rate movie character actors. Paul Giamatti, Philip Seymour Hoffman, Frances McDormand, and Gary Oldman don't often get the big role that every star in Hollywood wants, or the best screenplays or best directors, but in whatever movie they are in, they more than hold their own relative to the other actors. For example, in 2005, Catherine Keener out-acted Sean Penn in The Interpreter, and held her own with Daniel Day-Lewis in the Ballad of Jack and Rose, Hoffman in Capote, and Steve Carrell in The 40-Year-Old Virgin. That's a pretty good year.

My published articles are archived at iSteve.com -- Steve Sailer


max said...

Don't they keep road stats? That should be the best way to neutralize park inequities. (It doesn't neutralize factors that apply to every park but...)

The best road hitter should be the best hitter; a bad road hitter should be a bad hitter in general.

m, should work for fielding and pitching as well

Ben Kennedy said...

Sounds like you want to see Bill James' win shares

Anonymous said...

Even though this idea has not been implemented league-wide yet, it has been around for a while. I can remember people talking about Stan Musial relative to Ted Williams back in the day. And one of the things that Cardinal fans would bring up was the "Player of the Game" sign that was hung on the boards in centerfield.

The joke was that they could have fired the guy who was supposed to change the sign each day (to the appropriate Cardinal) because it was (almost) always Stan the Man.

Anonymous said...

The Cardinals used to post the name of the hero from the previous day's game on the outfield fence. This was back when they played in Sportman's Park, and the name posted was usually Stan Musial's.

Anonymous said...

I should've read Ian's comment before I posted.

Anonymous said...

Everybody knows that Colorado Rockies hitters aren't as good as their eye-popping statistics suggest because of the mile-high elevation means fly balls travel farther and faster due to less wind drag

Reduced wind drag might help a few balls make it out instead of being caught, but it's not the most important factor. What really helps hitters in Colorado is that the thin air reduces the amount pitches break. There's a technical term for a breaking ball which doesn't break: "batting practice pitch."

Anonymous said...

There's another statistic that's closer to this than win shares; I don't know what it's called but it's something like adjusted win probability.

Since baseball has such a large sample set you can check the probability of winning before and after each at bat. If a batter hits a sac fly with a runner on third and no out, what's the expected runs scored with a runner on third and no out versus one out, one run scored and no one on. Runs scored and saved can then be translated into how much they increase the team's chance of winning. After the game you can calculate who influenced the % chance to win by the most.

Do that for a season and you've got a park neutral stat that measured actual contribution to winning by each hitter (and pitcher).

- Steve Johnson

Kai Carver said...

I assume this is a typo (don't know much about baseball):

"guys who just happened to be at the race place at the right time"

If so it's an amusing slip.

Anonymous said...

Also, about the Rockies, their stadium is no longer a home run stadium. The humidor has made the balls too heavy to hit out easily. Runs are still somewhat up because it's so much harder to get movement on a pitch.

Anonymous said...

The problem with relying on road statistics is the unbalanced schedule. Some teams will play a disproportionate number of road games in bad hitters' parks.

The case of Alfonso Soriano is one example of the danger in relying too much on road statistics. When the Nationals traded for Soriano, several DC-area baseball blogs (and national analysts) slammed the move because Soriano had had very poor road stats when playing with the Texas Rangers. This analysis failed to take into account that, because the Rangers played in the AL West, over 1/3 of the Rangers' road games were played in three pitchers' parks -- Anaheim, Oakland, and Seattle. Seattle in particular is hardest on right-handed power hitters.

As a result, a lot of people were surprised by Soriano's excellent 2006 season with Washington. RFK is a very tough hitters' park. But as Frank Howard demonstrated in the 60s and Soriano demonstrated last year, if a righty is a dead-pull hitter he can still have a big year there.

As far as the keeping track of the best player in each game, the blog Capitol Punishment does that for the Nationals. Every week he chooses the "Majority Whip" for each game (the last 2 years he picked a Majority Whip for each win and a "Lame Duck" for each loss -- with the prospect of a terrible season for the Nats he decided to accenuate the positive this year.)

Anonymous said...

one of the greater sports media injustices of recent years was the almost total neglect of WhiteSox shortstop Juan Uribe by Fox (Tim McCarver, Joe Buck, and their handlers) during the Sox world series win. Typically an enviable series batting average graphic would flash over Uribe and he would then stroke another hit at the plate, all drawing a no comment and an "isn't Joe Crede fabulous" from the announcers. In the field, Uribe made an enormous contribution, producing a number of dazzling defensive plays while gunning down baserunners in routine putouts that felt more like executions (his one error in the series resulted when the ball stuck in his glove during a difficult turn and throw to first from the outfield grass, not a collosal blunder and no runners advanced). The santeria gods were apparently listening when, with two outs left in the series, I said: "I can't believe fox is so skewed and biased they're going to go the entire series without saying anything about Juan Uribe". The series ended with Uribe diving into the left field bleachers to record a flyout, then flashing power and grace in the series ender on a close play slow chopper that might have gone for a base hit. This display actually produced a verbal reaction from the booth. Hey. Hey. Fittingly, though, it was left to Carl Everett (who had an outstanding hitting series but was intentionally neglected by fox) to describe Uribe as the best shortstop in baseball, and then only in the papers.