Ten Things I Didn’t Know Last Weekby Dave Studeman
August 02, 2007
A head scratcher
Before the trading deadline came around, the Nationals signed a couple of veterans to long-term deals. Ronnie Belliard signed a deal for two years and $3.5 million, and Dmitri Young penned a two-year $10 million deal, with a third option year for $6 million if he meets plate appearance targets.
When I first heard about the Belliard deal, I didn't think much about it. Under $2 million a year seems like a pretty good price for a free agent second baseman with a decent skill set. But when I read about the Young deal, I did a double take. Yes, Young is having a very good half-season, batting .327 with a .876 OPS. But come on, people.
- He's 33 years old.
- His batting average on balls in play is .368, which will only go down.
- He may be the worst fielding first baseman in the league. His Revised Zone Rating is .634!
- You can get him out; he's batting .394 against finesse pitchers and .262 against power pitchers.
- Weird split of the day: he's batting .383 at home and .279 on the road.
Joe Sheehan expressed much the same at Baseball Prospectus, and Dan Fox drove home the point with a special Dmitri Young baserunning table (check out the article title: The Young and the Motionless). And speaking of trade deadlines, the Teixeira trade inspired a great post by long-time Ranger fan Jamey Newberg.
How much the Astros miss Adam Everett
Sean Smith recently used the latest defensive metrics, such as MGL's UZR and BIS's Relative Zone Rating, to calculate how much Adam Everett's loss has meant to the Astros. This is important, because most fans underestimate the importance of fielding, primarily because it's hard to find good fielding statistics. Luckily, the newest stats are making more in-depth analysis possible.
Sean's bottom line is that losing Everett and replacing him with Mark Loretta will likely result in a loss of 50 runs over a full season (or, five wins) in the field. When you give Loretta credit for his better bat, there's still a loss of 25 runs (2.5 wins) over 162 games.
In a somewhat more sophisticated analysis, Bagsandbidge found a smaller impact, primarily because he only counted 40 games and he also included the impact of having Mike Lamb play more games at third. Even so, he calculates that the difference between Everett and Loretta over a full year is about 20 runs. This is the sort of fielding analysis that should open some eyes.
Speaking of combining hitting and fielding stats, Vegas Watch recently combined WPA and UZR to determine the American League and National League MVPs of the first half of the year. In the second piece, Jacob is a bit skeptical that Troy Tulowitzki is really the league's MVP. It's a good point, but it's also good to remind people that Tulowitzki has had a fine year in the field, and that he's delivered some real clutch hits for the Rockies.
You can now download the Hit Tracker spreadsheet
I've gotten a big kick out of Hit Tracker since I first started reading it over a year ago. Hit Tracker calculates the length of every home run in the majors, using a complicated set of algorithms that only a physicist would truly understand.
The good news is that you can now download an Excel spreadsheet that allows you to calculate the length of home runs while you're watching a game (as long as you have a stopwatch handy). It's easy to use, and you can run the spreadsheet for any batted ball, not just home runs.
Highly recommended as a way to waste lots of time watching your next ballgame. I'm waiting for the day STATS and BIS start to use this in their batted ball stats.
I came across something called bivariate baseball score plots (phew!) the other day. It's a mouthful, but there are some really neat baseball graphics included. Specifically, the site includes a plot of the number of runs scored and allowed per game for every team, each year, in major league history. There is also a blog that talks about the plots.
You'll find lots of interesting things on these graphs, things you might not notice when combing through the numbers. For instance, the Yankees lost nine games by a 6-5 score last year. You can also modify the graphs using a number of filters (like first half vs. second half, or starting pitcher). Using it, I found that the 1965 Dodgers were 3-1 in 1-0 games when Koufax started. And, to top it off, you can download each plot as a PDF file.
My only very minor beef is that I would have put the team on the "y-axis" instead of the "x-axis," so that wins would be on top (the more natural place for them to be). But, as I said, that's really a minor criticism. So dazzle your friends and show them the bivariate baseball score plots!
How many No. 2 hitters have been MVPs
Last week, I opined that Nellie Fox really did deserve the 1959 MVP award, and also wondered how many No. 2 hitters have won MVP awards. Well, thanks to super-reader Brandon Islelib, here is the complete list for the Retrosheet years (1957 to the present):
1959 - Nellie Fox
1960 - Dick Groat
1982 - Robin Yount
1984 - Ryne Sandberg
1985 - Willie McGee
1999 - Ivan Rodriguez
In other words, the answer is six. To qualify, a batter had to spend the majority of the season in the No. 2 batting position.
Bert Blyleven's neutral won/loss record
Joe Posnanski recently gave Lee Sinins' neutral won/loss record a shout-out, using the stat to review the careers of several players, including Bert Blyleven and Jack Morris. Neutral won/loss records basically assume that each pitcher is given an average level of run support (average for the league in that year, that is). Morris has long been identified as a pitcher who received a lot of benefit from his team's offense. Blyleven, not.
The difference can be striking. Jack Morris had a 254-186 won/loss record, primarily for some mighty Tigers teams. According to Posnanski, his neutral won/loss record is 232-208. Similarly, Blyleven's career won/loss record is 287-250, but his neutral won/loss record is 313-224. So the difference between the two rises from 33 wins to 81 when you give them equivalent run support.
Personally, I wish that everyone would just stop referring to pitcher won/loss records. Teams win games, not individuals. But I know it's a hopeless case, and many fans (as well as Hall of Fame voters) persist in believing that starting pitchers have some magical hold over whether their team wins or loses, beyond their straightforward ability to keep runs from being scored.
In particular, I've seen more than one Hall voter claim that Morris "knew how to win," and Blyleven didn't. And you know what? Maybe those writers have a point. Maybe some pitchers are indeed better at "pitching to the score" than other pitchers. What do I know?
So I went back and looked at every single play of every single game that Blyleven and Morris pitched. I calculated the impact that each play had on the team's chances of winning, given the score at the time. If the bullpen blew a lead for a pitcher, the pitcher wasn't discredited for it. If he gave up runs in a close game, it counted against him more than runs allowed in a runaway. In other words, this analysis calculated exactly how well each pitcher "pitched to the score."
Okay, I didn't do all that myself. Jeff Sagarin did. As you may have guessed, the stat is Win Probability Added. It's not the most popular stat in the world, but it does seem extremely well suited to identify which pitchers might pitch to the score better than others.
The bottom line: Blyleven had 33 WPA in his career, and Morris had 18 WPA. Since every win is .5 WPA, that means that Blyleven's "record" would have had 66 more wins than losses and Morris' would have had 36. Applying these figures to each pitcher's won/loss record isn't exactly the most legitimate approach, but I'll do it anyway: Blyleven would be 302-236 and Morris would be 238-202.
Neutral won/loss says it, and so does WPA. Even Rich Lederer says it. Blyleven belongs in the Hall.
I attended the SABR conference last week and learned a heck of a lot. Most of it won't make it into this column, but I was intrigued by Mike Marshall's talk. I remember when Marshall blew away the baseball world when he appeared in 106 games in 1974. He's been controversial ever since, taking baseball management to task for blowing out pitchers' arms unnecessarily and teaching an entirely different way to pitch.
I'm not going to get into that sort of thing here, because I know nothing about it. But I was interested in Marshall's 1974, when he won the Cy Young award and finished third in MVP voting. As you can imagine, I wondered what his Win Probability Added was. According to Jeff Sagarin, it wasn't that much. In fact, Mike Marshall had 0.3 WPA in 1974.
How can this be? The guy appears in 106 ballgames and posts a 2.42 ERA in 208 innings and has 0.3 WPA? (Imagine that: 208 innings pitched, all in relief.)
Well, I don't really know. But WPA is built on a play-by-play analysis of how much Marshall helped his team win or lose games. And here's look at the batting average that Marshall allowed by different "game states" in 1974:
Tie game: .288
One-run game: .278
Two-run game: .260
Three-run game: .251
Four-run game: .250
Five or more: .227
For contrast, here is Goose Gossage's career game state breakout:
Tie game: .211
One-run game: .222
Two-run game: .226
Three-run game: .227
Four-run game: .227
Five or more: .241
In contrast to Gossage, Marshall pitched his best when the game wasn't close, and his worst when it was. That's the kind of pitching that will lead to a low WPA. Willie Hernandez may have deserved his MVP in 1984, but it doesn't appear that Marshall should have even been in the running.
Maybe it's time to lower the qualifications for the ERA title
The other day, I was thinking that we haven't seen any fluky ERA leaders for a while; no Buzz Capras or what have you. Scott Garrelts was perhaps the last fluky ERA leader we've seen, 18 years ago. And he was no Buzz Capra.
I think I know why this is happening. The qualifications for ERA leaders hasn't changed in, like, forever: A pitcher must pitch at least one inning for each team game played. And though 162 innings in a season doesn't sound like a lot, there are less pitchers qualifying each year. Here is a graph of the number of qualifiers per team since 1962:
So that's why we don't see many Buzz Capra's anymore. With relief usage so prevalent, it's harder to qualify for the ERA title, which means that it's harder for outlier years to qualify. You might say that we're better for it, but you also might say we've lost something along the way.
Bill Giles got it wrong
Last week, I recounted a story from Bill Giles' recent biography, Pouring Six Beers at a Time: And Other Stories from a Lifetime in Baseball, in which he claimed that Paul Richards once proposed trading the Houston Astros' roster for the Detroit Tigers'.
The only problem with the story, according to Warren Corbett (who's writing a biography of Richards) is that it couldn't have happened.
I'm working on a biography of Paul Richards, so accuracy compels me to point out that he was not the GM of Houston in 1967. He was in Atlanta. That's one of many flagrant errors in Giles's entertaining book. Richards did make the "I'll trade my roster for yours" offer on at least two occasions, but he would never have done so with the bounty of young talent he had in Houston.I don't understand why publishers don't fact-check memoirs such as Giles'.
Much of what Giles says about Richards' character is verified by other sources. However, Giles was loyal to Judge Roy Hofheinz, the Houston owner who fired Richards and engaged in a dirty slandering match over the money owed on his contract. The rumor about taking kickbacks from players is one of many that Hofheinz's men spread during that battle. I have a two-inch stack of papers from Richards' personal files relating to that. Very interesting reading.
The writer Mickey Herskowitz once said to Richards, "The judge is his own worst enemy." Richards replied, "Not while I'm alive."
Hank Aaron's parents hugged him to protect him.
At the SABR conference, I attended David Vincent's presentation on home run records over the years. David is a noted expert on home runs, having built a database of every major league home run in history. It was a great presentation, and Vincent showed pictures of Hank Aaron's parents hugging him enthusiastically when he broke Babe Ruth's home run record. They pretty much covered him and wouldn't let go.
According to Vincent, there was more than just pure joy at play. Aaron had received so many death threats that his parents were worried that someone might try to shoot him. So they hugged to not only congratulate him, but to protect him from potential snipers.
Just in case you needed some perspective on baseball and life...
References and Resources
By the way, Posnanski was recently interviewed by Gelf Magazine. Good read.
If you're looking for a PC-based stopwatch to time those home runs, I personally use Multitrack Stopwatch.
Dave was called a "national treasure" by Rob Neyer. Seriously. Comments about this article can be sent to him through the miracle of e-mail.