John Bonnes, the Twins Geek, takes a look at some comparisons for Kielty based on Baseball Prospectus' new PECOTA projection system. PECOTA is designed to predict the chances of whether a player will "breakout" or "collapse" based on age and physical comparisons. I've got only passing familiarity with that new system, but it stands to receive a lot of attention as BP rolls out its new book and upgraded, subscription-based website in the coming weeks (about which I'll write soon). Anyway, Bonnes runs some comparisons and notes that a player such as Kielty who has a high number of walks and strikeouts doesn't project as well as a traditional stathead would expect. "It's not clear that low-contact, average-power players will succeed when pitchers start throwing more strikes until they have a lot more at-bats than Kielty has," writes Bonnes. "Kielty had a monster year last year. But to claim with any kind of confidence that he'll repeat that effort, it appears he would need at least 300, and maybe 500, more at-bats." Worth a read.
ESPN's Jayson Stark penned a column this week in which he suggested twenty-five rules changes ranging from the good (adding instant replay, toughening up the save rule) to the bad (adding a designated fielder to the roster, ditching the phantom DP tag -- you want to see a generation of middle infielders ruined by torn ACLs? I didn't think so) to the ugly (penalizing the intentional walk and limiting pickoff throws). In his inimitable style, Mike C. of Mike's Baseball Rants takes apart Stark's list with some handy research regarding intentional walk rates as well as some good old common sense. Check it out.
Not to be out-ranted (did somebody say stark, raving mad?), our friends at Elephants in Oakland have a more bilious take, as well as a few suggestions for Stark.
As for what I'm working on, DIPS 2002 will be up this weekend, and hopefully I'll return to the land of the occasionally original thought soon...
Lee Sinins, who runs the essential Around the Majors mailing list and produces an annual CD-ROM called the Sabermetric Baseball Encyclopedia, has a new offering this spring. It's called the ATM Reports Player Comments book, and it contains the kinds of handy statistical analysis on the order of Lee's other products: short blurbs about how productive a player is compared to the league average, numerous charts and statistical studies, lists of where the player fits in on a team's history or on an all-time list, and other interesting facts. Lee has kindly provided me with a review copy of the book (which is available both in a print version and as an electronic PDF), and I'll be giving it a more in-depth look in the near future.
With slow news days in the baseball world lately (the Yanks are spending money; an idiot ump is on the loose) and plenty of chaos at my j-o-b, I've been retreating to the serenity of my spreadsheets lately. It looks as though I'll be spending a bit more time there, as I've taken it upon myself to run all of the 2002 Major League pitching statistics through my DIPS 2.0 spreadsheet.
I don't take credit for the DIPS (Defense Independent Pitching Statistic) system. It was invented by a man named Voros McCracken, and he's presented DIPS numbers for the 1999-2001 seasons via his web site while explaining the system on Baseball Primer and Baseball Prospectus. The gist of it is that McCracken did some studies on pitching statistics involving balls in play and concluded that major-league pitchers do not differ greatly in their ability to prevent hits on those balls hit into play (that is, anything that's not a home run, a strikeout, a walk or a hit-by-pitch). The rate at which a pitcher allows hits on balls in play is due more to the defense playing behind him than to his own skill, and can vary greatly from year to year.
This is somewhat counterintuitive, but it's also a very helpful way of looking at pitching stats. DIPS takes the elements of a pitcher's record that are not affected by the defense -- walks, strikeouts, hit-by-pitches, homers -- and places them in a neutral context for park, league and defense. The result is a translated line of Defense Independent Pitching Statistics, including a DIPS ERA; that is, an ERA based on defense-independent pitching performance. An important thing about this DIPS ERA, McCracken found, is that it correlates better with the following season's ERA than the pitcher's actual ERA does.
For one reason or another, Voros decided not to publish DIPS numbers this year, leaving a sizeable void in the sabermetric universe. But he's already published fairly coherent instructions on how to calculate DIPS (and he encouragingly answered questions about some of the less coherent aspects of it), so I built a spreadsheet that would do the job. I used it for a few pieces about the Yankee pitchers and this year's crop of relievers figuring the sheet would give me a jump in the analysis department, but that it was only a matter of time before somebody published complete DIPS for 2002, and more power to them.
Insert sound of crickets chirping.
Nobody's done so, including myself -- mainly because I was never able to get my hands on the raw data in a spreadsheet. But via a rather mundane Primer thread, I managed to find somebody ("mathteamcoach" is his handle) who had most of what I needed. We've joined forces to share the tedium of entering Intentional Base on Balls and Batters Faced Pitching data for EVERY SINGLE PITCHER in the service of this project. It's a dirty job but somebody's got to do it, and between the two of us we're about 2/3 done. The results should be finished later this week.
