RLakin
All Star
In baseball, numbers don't lie; or do they?
By JOE POSNANSKI
Columnist
Batting average is a stupid statistic, if you think about it. Have you ever walked into a movie halfway through? Years ago, I walked in on "Raiders of the Lost Ark" about an hour in. I would tell people, "Oh yeah, I saw that. It was OK."
And they would say "Man, how about that scene when Indiana Jones ends up in the pit with the snakes ..."
And I would say, "Uh, missed that."
That's batting average. You may learn something about a player, but you miss the best scenes. Batting average tells you how often a player gets a hit. It doesn't tell you how many times a player gets on base. It doesn't tell you what kind of hits a player gets. It doesn't tell you much, really.
Still, when people talk about winning a "batting crown," they mean best batting average. When team statistics are listed, the highest batting average goes first. When people talk hitting, they talk batting average.
Well, we'll try to give you a little different look today. Nothing too complicated. But we'll offer up a few statistics to consider. With Michael Lewis' fine book, Moneyball, on the best-seller lists, everybody in baseball is talking about statistics.
See, baseball statistics are different. They actually affect the way the game is played. This happens sometimes in other sports -- counting sacks has had an impact on football, driving distance has had an effect on golf, the assist has had an impact on basketball (especially in Utah, where every time a bell rings, John Stockton gets an assist).
But it is fair to say statistics have changed baseball more than any other sport, partly because baseball has more statistics than any other sport.
"What do players get paid for?" one baseball executive asks. "Batting average, home runs and saves. And those aren't team statistics. Those are selfish statistics. They don't tell you how much you help the team."
Well, let's take a quick look. There are a billion baseball stats out there, and I don't understand 98 percent of them. But here are a few that might open your eyes a little bit.
Batting average vs. OPS
Let's look at two players. One is Seattle's Ichiro Suzuki. He is hitting .346 and is second in the batting race.
The other is New York's Jorge Posada. He's hitting a lousy .250.
Obviously, Suzuki is much better at getting on base, right? Well, actually, no. See, Posada walks more than twice as often as Suzuki. Batting average, based on some decision made 100-plus years ago, does not incorporate walks, even though every Little League coach knows that a walk is as good as a hit. Posada also gets hit by pitches fairly often.
Posada's on-base percentage is actually higher than Suzuki's.
This is just one of the problems with batting average. It does not credit the men who earn their walks and get on base. Barry Bonds, for instance, has only won one "batting title." But he has actually led the league in on-base percentage six times, a far more impressive thing.
Then, there's the issue of extra-base hits. A guy can hit .300 with nothing but singles (Rey Sanchez almost did this with the Royals). That doesn't help very much. That's why it's key to look at slugging percentage, which gives players more credit for hitting doubles, more for triples, and even more for home runs.
Slugging percentage is a great statistic, but it has never caught on, and I have a theory: I don't think any statistic with the word "percentage" in it ever catches on. If ERA was actually ERP -- earned run percentage -- nobody would care about it.
Anyway, statisticians have tried to combine on-base percentage and slugging percentage into one reasonable statistic. There are lots of variations, some very complicated, but the easiest is called OPS -- on-base plus slugging -- and it simply adds the two. It's not perfect either. But it's a whole lot better than batting average.
Here then, through Friday, were the Royals' batting-average and OPS leaders (the average OPS in the American League is .763):
Batting average
1. Mike Sweeney: .321
2. Desi Relaford: .291
3. Carlos Beltran: .290
4. Raul Ibanez: .284
5. Mike DiFelice: .283
On-base plus slugging
1. Mike Sweeney: .980.
2. Carlos Beltran: .855
3. Michael Tucker: .832.
4. Raul Ibanez: .787.
5. Desi Relaford: .784.
Wins vs. game scores
A behind-the-scenes argument raging at Kauffman Stadium involves pitcher Chris George. He has, by most standards, pitched all kinds of lousy. His ERA of 7.00 is among the worst in baseball. He has walked more batters than he has struck out. He has given up a team-leading 19 home runs, including one Friday night to Detroit's Dmitri Young that is still going.
Still, his record is a spectacular 9-5. Had he managed to get just one more out on Friday, his record would probably be 10-5.
"He's a winner," says the pro-George crowd among the Royals. These people believe some pitchers have a mystical quality -- the ability to win. And, after all, isn't it a pitcher's job to win games?
Well, Bill James invented a fun way to judge a pitcher's performance. It's called "game scores," and it fairly simply calculates a pitcher's strikeouts, walks, runs and innings pitched to give you a score between zero and 100. Only a few pitchers have ever reached 100. (Kerry Wood's 20-strikeout game achieved the highest-ever score: 105.) And you have to be really bad to score less than 10 (George managed to score a two against the Cardinals last week).
A score of 50 is an average start.
Well, according to the game scores, only four of Chris George's 17 starts have been above average. And none was much above average. His highest score, a respectable 63, came in his first start. He has managed just one above-average start -- and it was just barely above -- in the last two months.
In other words, George really has not pitched well enough to win. He has not shown the sort of subtle ability to win games. He has simply been lucky.
Chris George's last 10 starts:
• May 16: Game score: 37. Decision: loss.
• May 21: Game score: 39. Decision: win.
• May 27: Game score: 7. Decision: loss.
• June 1: Game score: 46. Decision: none.
• June 7: Game score: 24. Decision: win.
• June 13: Game score: 49. Decision: win.
• June 18: Game score: 27. Decision: win.
• June 24: Game score: 51. Decision: win.
• June 29: Game score: 2. Decision: loss.
• July 4: Game score: 31. Decision: none.
Saves vs. inherited runners allowed
The save changed the way major-league teams played baseball. The rule was invented by longtime baseball writer Jerome Holtzman, and it states that if a pitcher pitches the final inning, and his team is ahead by three runs or fewer, he gets a save.
There are other ways to get a save, but that's the one everybody uses. Every team now has a closer -- usually highly paid -- whose job is to pitch the ninth inning of relatively close games. If the rule had stated you needed to pitch two innings to get a save, those closers would start the eighth.
The rule makes the game.
Now, I do believe it is hard to close out a game. It's a different mind-set. The ninth can be a high-pressure inning, and teams pull out all their tricks to score runs. But this is a ridiculous statistic. Let's take a look at John Smoltz, who has 31 saves and an 0.81 ERA. Many think he is the MVP of the National League.
But let's take a closer look. In 23 of his saves, he came in with at least a two-run lead. How hard is it to save games with a two- or three-run lead? Well, let's take a look at a starter having a fairly disastrous season. Baltimore's Rick Helling is 5-6 with a 5.81 ERA.
Thing is, Rick Helling has only twice, in his 17 starts, allowed two runs in the first inning. And he has never allowed more than two. Mostly (13 out of 17), he gets through the first inning without giving up any runs. If he was the Braves closer, he might have 15 or 16 saves in 17 appearances.
Here's another way to judge relievers: Look at inherited runners and how often they score. After all, you bring in a reliever with men on and hope he can get out of the jam. On average, relievers allow inherited runners to score about one-third of the time. The Royals allow them to score more than 50 percent of the time, the highest percentage in baseball and a reason why Royals general manager Allard Baird is still looking for bullpen help.
Royals' inherited-runners percentage:
• Mike MacDougal: three runners, one scored; 33 percent.
• Kris Wilson: 12 runners, five scored; 42 percent.
• Jason Gilfillan: seven runners, three scored; 43 percent.
• Jason Grimsley: 38 runners, 19 scored; 50 percent.
• D.J. Carrasco: 11 runners, six scored; 55 percent.
• Sean Lowe: 12 runners, eight scored; 67 percent.
Errors, pitch counts and other stuff
Millions of calculators have burned out trying to measure fielding. It's hard to do. Baseball people will tell you that one great defensive player can save hundreds of runs a year, statisticians will tell you that's absolutely absurd, and everybody is trying to come up with an easy-to-use fielding statistic that people can understand.
In the meantime, we count errors. This, of course, is absolutely ridiculous. Errors only provide two things.
1. They tell you when a player made a blatant mistake.
2. Uh, no, they don't really tell you that either.
Earlier this year, Royals third baseman Joe Randa had a long errorless streak. I don't know precisely how long because I stopped counting on the back-to-back days at Kauffman Stadium when he botched routine ground balls. Neither, for reasons unknown, was scored an error. That's not to say Randa isn't a good third baseman -- I think he's very good, in fact. It just shows that errors are simply judgment calls, no more scientific than pass-interference penalties. That's no way to judge a defensive player.
• These days everybody's counting pitches. This is a good statistic to watch. Bill James tells a story about Sandy Koufax coming off a serious arm injury in 1964. In his first start at spring training, he threw something like 165 pitches. They didn't count in those days, and Koufax's arm swelled like the egos of this year's Minnesota Twins.
Now, everybody watches those pitch counts very closely. The Royals have been careful not to let any of their pitchers throw more than, say, 110 pitches in a game. How has it worked so far? Pretty much every pitcher has been hurt.
• My favorite baseball statistic involves Joe DiMaggio's 56-game hitting streak in 1941. DiMaggio hit .357 that year, meaning that if he got four at-bats, he had about an 83 percent chance of getting at least one hit that day.
So, what are the odds of a .357 hitter getting a hit 56 consecutive days. If he gets four at-bats per day, the odds are about 36,000 to 1. These are the same odds, incidentally, of turning the radio dial in Kansas City and not hearing at least one song by Lynyrd Skynyrd.
In case you are wondering, the odds of a .357 hitter hitting in 57 consecutive days goes up to 43,000 to 1. I can keep going, if you want. A calculator is a dangerous thing.
Baseball Stats 101
How to figure a few key baseball statistics:
• On-base percentage: Take hits plus walks plus hit-by-pitch and divide by at bats plus walks plus hit by pitch plus sacrifice flies.
• Slugging percentage: Take total bases and divide by at-bats.
• OPS: Add on-base percentage and slugging percentage.
• Game scores: Start with 50 points. Give a pitcher three points for every out he gets, two points for every complete inning he pitches after four, and one point for every strikeout. Now, subtract two points for every hit allowed, four points for every earned run and one point for every walk.
By JOE POSNANSKI
Columnist
Batting average is a stupid statistic, if you think about it. Have you ever walked into a movie halfway through? Years ago, I walked in on "Raiders of the Lost Ark" about an hour in. I would tell people, "Oh yeah, I saw that. It was OK."
And they would say "Man, how about that scene when Indiana Jones ends up in the pit with the snakes ..."
And I would say, "Uh, missed that."
That's batting average. You may learn something about a player, but you miss the best scenes. Batting average tells you how often a player gets a hit. It doesn't tell you how many times a player gets on base. It doesn't tell you what kind of hits a player gets. It doesn't tell you much, really.
Still, when people talk about winning a "batting crown," they mean best batting average. When team statistics are listed, the highest batting average goes first. When people talk hitting, they talk batting average.
Well, we'll try to give you a little different look today. Nothing too complicated. But we'll offer up a few statistics to consider. With Michael Lewis' fine book, Moneyball, on the best-seller lists, everybody in baseball is talking about statistics.
See, baseball statistics are different. They actually affect the way the game is played. This happens sometimes in other sports -- counting sacks has had an impact on football, driving distance has had an effect on golf, the assist has had an impact on basketball (especially in Utah, where every time a bell rings, John Stockton gets an assist).
But it is fair to say statistics have changed baseball more than any other sport, partly because baseball has more statistics than any other sport.
"What do players get paid for?" one baseball executive asks. "Batting average, home runs and saves. And those aren't team statistics. Those are selfish statistics. They don't tell you how much you help the team."
Well, let's take a quick look. There are a billion baseball stats out there, and I don't understand 98 percent of them. But here are a few that might open your eyes a little bit.
Batting average vs. OPS
Let's look at two players. One is Seattle's Ichiro Suzuki. He is hitting .346 and is second in the batting race.
The other is New York's Jorge Posada. He's hitting a lousy .250.
Obviously, Suzuki is much better at getting on base, right? Well, actually, no. See, Posada walks more than twice as often as Suzuki. Batting average, based on some decision made 100-plus years ago, does not incorporate walks, even though every Little League coach knows that a walk is as good as a hit. Posada also gets hit by pitches fairly often.
Posada's on-base percentage is actually higher than Suzuki's.
This is just one of the problems with batting average. It does not credit the men who earn their walks and get on base. Barry Bonds, for instance, has only won one "batting title." But he has actually led the league in on-base percentage six times, a far more impressive thing.
Then, there's the issue of extra-base hits. A guy can hit .300 with nothing but singles (Rey Sanchez almost did this with the Royals). That doesn't help very much. That's why it's key to look at slugging percentage, which gives players more credit for hitting doubles, more for triples, and even more for home runs.
Slugging percentage is a great statistic, but it has never caught on, and I have a theory: I don't think any statistic with the word "percentage" in it ever catches on. If ERA was actually ERP -- earned run percentage -- nobody would care about it.
Anyway, statisticians have tried to combine on-base percentage and slugging percentage into one reasonable statistic. There are lots of variations, some very complicated, but the easiest is called OPS -- on-base plus slugging -- and it simply adds the two. It's not perfect either. But it's a whole lot better than batting average.
Here then, through Friday, were the Royals' batting-average and OPS leaders (the average OPS in the American League is .763):
Batting average
1. Mike Sweeney: .321
2. Desi Relaford: .291
3. Carlos Beltran: .290
4. Raul Ibanez: .284
5. Mike DiFelice: .283
On-base plus slugging
1. Mike Sweeney: .980.
2. Carlos Beltran: .855
3. Michael Tucker: .832.
4. Raul Ibanez: .787.
5. Desi Relaford: .784.
Wins vs. game scores
A behind-the-scenes argument raging at Kauffman Stadium involves pitcher Chris George. He has, by most standards, pitched all kinds of lousy. His ERA of 7.00 is among the worst in baseball. He has walked more batters than he has struck out. He has given up a team-leading 19 home runs, including one Friday night to Detroit's Dmitri Young that is still going.
Still, his record is a spectacular 9-5. Had he managed to get just one more out on Friday, his record would probably be 10-5.
"He's a winner," says the pro-George crowd among the Royals. These people believe some pitchers have a mystical quality -- the ability to win. And, after all, isn't it a pitcher's job to win games?
Well, Bill James invented a fun way to judge a pitcher's performance. It's called "game scores," and it fairly simply calculates a pitcher's strikeouts, walks, runs and innings pitched to give you a score between zero and 100. Only a few pitchers have ever reached 100. (Kerry Wood's 20-strikeout game achieved the highest-ever score: 105.) And you have to be really bad to score less than 10 (George managed to score a two against the Cardinals last week).
A score of 50 is an average start.
Well, according to the game scores, only four of Chris George's 17 starts have been above average. And none was much above average. His highest score, a respectable 63, came in his first start. He has managed just one above-average start -- and it was just barely above -- in the last two months.
In other words, George really has not pitched well enough to win. He has not shown the sort of subtle ability to win games. He has simply been lucky.
Chris George's last 10 starts:
• May 16: Game score: 37. Decision: loss.
• May 21: Game score: 39. Decision: win.
• May 27: Game score: 7. Decision: loss.
• June 1: Game score: 46. Decision: none.
• June 7: Game score: 24. Decision: win.
• June 13: Game score: 49. Decision: win.
• June 18: Game score: 27. Decision: win.
• June 24: Game score: 51. Decision: win.
• June 29: Game score: 2. Decision: loss.
• July 4: Game score: 31. Decision: none.
Saves vs. inherited runners allowed
The save changed the way major-league teams played baseball. The rule was invented by longtime baseball writer Jerome Holtzman, and it states that if a pitcher pitches the final inning, and his team is ahead by three runs or fewer, he gets a save.
There are other ways to get a save, but that's the one everybody uses. Every team now has a closer -- usually highly paid -- whose job is to pitch the ninth inning of relatively close games. If the rule had stated you needed to pitch two innings to get a save, those closers would start the eighth.
The rule makes the game.
Now, I do believe it is hard to close out a game. It's a different mind-set. The ninth can be a high-pressure inning, and teams pull out all their tricks to score runs. But this is a ridiculous statistic. Let's take a look at John Smoltz, who has 31 saves and an 0.81 ERA. Many think he is the MVP of the National League.
But let's take a closer look. In 23 of his saves, he came in with at least a two-run lead. How hard is it to save games with a two- or three-run lead? Well, let's take a look at a starter having a fairly disastrous season. Baltimore's Rick Helling is 5-6 with a 5.81 ERA.
Thing is, Rick Helling has only twice, in his 17 starts, allowed two runs in the first inning. And he has never allowed more than two. Mostly (13 out of 17), he gets through the first inning without giving up any runs. If he was the Braves closer, he might have 15 or 16 saves in 17 appearances.
Here's another way to judge relievers: Look at inherited runners and how often they score. After all, you bring in a reliever with men on and hope he can get out of the jam. On average, relievers allow inherited runners to score about one-third of the time. The Royals allow them to score more than 50 percent of the time, the highest percentage in baseball and a reason why Royals general manager Allard Baird is still looking for bullpen help.
Royals' inherited-runners percentage:
• Mike MacDougal: three runners, one scored; 33 percent.
• Kris Wilson: 12 runners, five scored; 42 percent.
• Jason Gilfillan: seven runners, three scored; 43 percent.
• Jason Grimsley: 38 runners, 19 scored; 50 percent.
• D.J. Carrasco: 11 runners, six scored; 55 percent.
• Sean Lowe: 12 runners, eight scored; 67 percent.
Errors, pitch counts and other stuff
Millions of calculators have burned out trying to measure fielding. It's hard to do. Baseball people will tell you that one great defensive player can save hundreds of runs a year, statisticians will tell you that's absolutely absurd, and everybody is trying to come up with an easy-to-use fielding statistic that people can understand.
In the meantime, we count errors. This, of course, is absolutely ridiculous. Errors only provide two things.
1. They tell you when a player made a blatant mistake.
2. Uh, no, they don't really tell you that either.
Earlier this year, Royals third baseman Joe Randa had a long errorless streak. I don't know precisely how long because I stopped counting on the back-to-back days at Kauffman Stadium when he botched routine ground balls. Neither, for reasons unknown, was scored an error. That's not to say Randa isn't a good third baseman -- I think he's very good, in fact. It just shows that errors are simply judgment calls, no more scientific than pass-interference penalties. That's no way to judge a defensive player.
• These days everybody's counting pitches. This is a good statistic to watch. Bill James tells a story about Sandy Koufax coming off a serious arm injury in 1964. In his first start at spring training, he threw something like 165 pitches. They didn't count in those days, and Koufax's arm swelled like the egos of this year's Minnesota Twins.
Now, everybody watches those pitch counts very closely. The Royals have been careful not to let any of their pitchers throw more than, say, 110 pitches in a game. How has it worked so far? Pretty much every pitcher has been hurt.
• My favorite baseball statistic involves Joe DiMaggio's 56-game hitting streak in 1941. DiMaggio hit .357 that year, meaning that if he got four at-bats, he had about an 83 percent chance of getting at least one hit that day.
So, what are the odds of a .357 hitter getting a hit 56 consecutive days. If he gets four at-bats per day, the odds are about 36,000 to 1. These are the same odds, incidentally, of turning the radio dial in Kansas City and not hearing at least one song by Lynyrd Skynyrd.
In case you are wondering, the odds of a .357 hitter hitting in 57 consecutive days goes up to 43,000 to 1. I can keep going, if you want. A calculator is a dangerous thing.
Baseball Stats 101
How to figure a few key baseball statistics:
• On-base percentage: Take hits plus walks plus hit-by-pitch and divide by at bats plus walks plus hit by pitch plus sacrifice flies.
• Slugging percentage: Take total bases and divide by at-bats.
• OPS: Add on-base percentage and slugging percentage.
• Game scores: Start with 50 points. Give a pitcher three points for every out he gets, two points for every complete inning he pitches after four, and one point for every strikeout. Now, subtract two points for every hit allowed, four points for every earned run and one point for every walk.
