Statistical Curiositiesby John Walsh
May 31, 2006
A look at a few strange and unusual things in the baseball world that I came across over the past week or so.
Can't Buy A Win
Last week Jake Peavy struck out 16 batters in a game, but he took the loss as the Braves dropped the Padres 3-1. Peavy surrendered just two runs on three hits and no walks. I was wondering how often a pitcher strikes out that many batters and loses.
Well, since 2000, pitchers have struck out 16 batters or more in a game 14 times (not counting Peavy's game). Not surprisingly, Randy Johnson has done it most frequently, seven times. Overall, pitchers went 9-2 with three no-decisions in those games. The toughest outings: a 20-strikeout effort by Randy Johnson against the Reds in 2001. The Big Unit left with the game tied after nine innings and the D'Backs went on to win in 11 innings. Pedro Martinez suffered a 1-0 loss to the Devil Rays back in 2000, a game in which he struck out 17.
The most amazing thing I found out while looking at these high-K games: on September 29, 2000, the Reds beat the Cardinals 8-1, with the Reds starter striking out 16 Cardinals. His name? Ron Villone. Who'd've ever thunk it?
Doing More Than His Share
As I write, Albert Pujols has hit 46% of his team's home runs. Since WWII that figure has been exceeded only once, by a guy named Sam Chapman who in 1946 hit 20 homers for the lowly Philadelphia A's. All of Chapman's teammates combined hit the same number as he did. More recently, in 1981 Mike Schmidt hit 45% of his team's homers (31 of 69). The all-time leader in this silly stat (considering only teams that had at least 20 home runs) is, of course, the Babe himself. In 1919 he hit 29 home runs while his Red Sox teammates hit just four.
Everybody knows that Pujols is hitting a lot of home runs this year. In previous years, Albert hit a home run for about every 13 at-bats. So far this year, he's rounded the bases once every 6.3 at-bats. How the heck is he doing it?
Two things: he's hitting more fly balls this year and a greater percentage of those fly balls are going out of the park. Before this year, Pujols' fly ball percentage was a pretty constant 40%, but this year he has raised that to over 50%. Also, his pre-2006 fly balls had roughly a 20% chance of going over the fence, while this year's rate is over 30%. If you put the two stats together, you find that 15% of the balls that Pujols hits fairly go for home runs. Amazing. Before this year that number was about half as big.
One last note: the increase in fly balls comes at the expense of both ground balls and line drives. This should lead to a lower batting average, which might end up costing Albert the Triple Crown. In fact, his batting average on balls in play (BABIP) stands at only .220 this year, a very low number. Still, Pujols has managed to keep his average high (.314 as I write), in large part due to his .150 home run average on balls in play. I wouldn't be surprised, though, if Pujols hit 15-20 points below his career average of .331 this year.
Chasing Chief Wilson
Those of you who read my recent piece called Endangered Species: The Three-Base Hit know I have an affection for the triple. So I noted with interest when the Giants' Steve Finley hit his eighth triple of the season the other day, putting him on pace for 27 for the season. Even the 20-triple plateau is a rarity these days: the last player to accomplish it was Cristian Guzman in 2000, and since 1980, the mark of 20 triples in a season has been reached only three times.
Finley is a good triples hitter; in fact he's the active career leader with 120. In the article I linked above, I made a simple (some would call it crude) model to estimate the number of triples a player should hit, based on his speed and which side of the plate he hits from. Through 2005, Finley had exceeded his prediction 112 to 79 (numbers through 2005 season). Based on a quick look at some splits at Retrosheet, it appears AT&T Park is a good place for hitting triples, so we should expect Finley to get his share of triples this year (if he gets to play).
Don't Even Think About Going
Brewers pitcher Chris Capuano has yet to allow a stolen base this year while he is on the mound; two would-be stealers have been caught, and he has taken care of three others on his own by picking them off. Capuano has a terrific pickoff move, the best in the majors. He's averaged 12 runners picked off per 200 innings pitched, almost double that of the next-best pickoff artist, Mike Maroth (considering the 2003-2005 seasons).
Even when not getting picked off, runners have a very hard time stealing on Capuano. Pro-rating his career numbers to one season (200 IP), this is his line in the running game: 3 SB, 2 CS, 12 pickoffs. How many runs is that worth? Using run values for stolen bases and caught stealing supplied by Tom Tango, I estimate that Capuano saves about five to six runs per season just by shutting down the running game. Or, looking at it another way, he shaves about 0.25 runs off his ERA.
You might wonder which pitcher is the worst in holding runners in recent years. The answer is Kevin Millwood, who surrenders about four extra runs per season.
Who is Pythagoras?
Here are the teams who currently have winning records (.500 or better), despite scoring fewer runs than they've allowed:
Team W - L RS RA -----+--------+------+-------+ HOU 26-26 239 263 MIL 26-25 257 282 COL 27-24 220 233So, what would you expect their records to be, given those runs scored and allowed numbers? Well, you could use the Pythagorean Formula, but that's got nasty powers of two and fractions and stuff. An easier way is this: start with a .500 record and add a win for each 10 runs of (positive) run differential. All of these teams should have records a game or two below .500, based on their run differential. On the other hand, Cleveland has won fewer games than you might expect:
Team W - L RS RA -----+--------+------+-------+ CLE 24-26 284 256With a +30 run differential, you'd expect the Tribe's record to be 28-22. I expect them to turn it around somewhat, but maybe not enough to overtake the Tigers and White Sox.
Goodbye, Mr. Spalding!
On Sunday the Braves hit eight home runs en route to beating the Cubs 13-12. It was a good day for hitting homers at Wrigley: a 13-mph wind blowing out to center field and a temperature of 87 degrees. It didn't help the Cubs any, though, who failed to hit the ball out of the park. The news services say that eight home runs in a game is a Braves record. But what is the all-time record?
The answer is 10, achieved by the Toronto Blue Jays back in 1987. They beat the Orioles in that game 18-3. A good place to look up stuff like this is Retrosheet; for example top one-game team performances are listed on this page. If you have a look, you might notice an interesting game between the Red Sox and Blue Jays on July 4, 1977. The Sox clubbed eight round-trippers but somehow managed to score only nine runs. Luckily, that was good enough to win the game, 9-6.
The Tigers weren't so lucky in 2004, when they lost to the Red Sox 11-9, despite hitting seven home runs, six of them coming off knuckleballer Tim Wakefield. Wakefield picked up the win; I wonder if any other pitcher has won a game while giving up so many homers? I'll have to get back to you on that one.
John Walsh dabbles in baseball analysis in his spare time. He welcomes questions and comments via e-mail.