The Very-Early-Season-Win-Share All-Starsby Dave Studeman
May 18, 2006
It's time to talk Win Shares again. Bill James's creation appears to be one of the most popular parts of The Hardball Times site, based on the many e-mails I've received asking when they would be available. Win Shares are complex, however, and I like to take the time to make sure I have them right. I think I have, and the good news is that they will be updated every week for the rest of the season. They're available on this page.
With all of the recent talk about Win Probability Added, you may get the two stats mixed up, but they're different. WPA calculates the impact of a play, based on "real time" progression of the game. Win Shares doesn't care about the context of a game (except for relievers—more on that later). All events, like home runs, strikeouts and errors, have the same consistent impact. But both stats do something cool—they estimate how many wins each player contributed to the team.
I won't get into too many specifics of how Win Shares are constructed. If you'd like, you can find more detail in this article. We've changed Win Shares a little bit from Bill James's original conception, but not so much that they can't be compared to historical figures. Most fundamentally, our system permits individual Win Shares to fall below zero when warranted, because we found that certain players' totals were being diminished by having too many "zeroed out" players on their team. I first wrote about this two and a half years ago, and haven't changed my mind.
We've also equipped Win Shares with a few wrinkles to provide additional context.
- Expected Win Shares are the number of Win Shares that a player would have accrued if he were average, given his time at bat, in the field and on the mound. It's used to approximate playing time and provide a fair, comparison baseline for each player. For instance, designated hitters don't field, so you'd expect them to accrue less Win Shares. Please note that we calculate expected Win Shares for National League pitchers based on how other pitchers have batted, not on how all batters have batted.
- Win Shares Percentage (WSP) is a simple rate stat, defined as WS/(2*ExpWS). It's calculated this way so that the average player is a .500 player, just like an average team is a .500 team. The only flaw in this approach is that players can be better than 1.000 and worse than .000, but that's not a big problem once you get used to it.
- If you want, you can calculate Win Shares Above Average by simply subtracting ExpWS from WS. However, we prefer to use Win Shares Above Bench (WSAB), which is WS minus 70% of expWS (or 60% for starting pitchers). These percentages aren't perfect, but they are an estimate of the number of Win Shares an average bench player would contribute. They provide the best baseline for comparison purposes, and WSAB is our preferred Win Share metric.
Catcher League Name Team Bat Pitch Field Tot ExpWS WSP WSAB AL Hernandez BAL 5.7 0.0 1.4 7 3.2 1.104 5 AL Posada NYA 3.2 0.0 2.1 5 3.0 .907 3Catchers receive a lot of fielding credit in Win Shares, especially for their caught stealing results. Ramon Hernandez has nabbed 10 of 22 baserunners. That said, Hernandez and Jorge Posada are the two best-hitting catchers so far this year. Victor Martinez has also been batting well, but he hasn't hit well with runners in scoring position (a factor in batting Win Shares), and his caught stealing stats have been terrible; he has ZERO fielding Win Shares.
First Base League Name Team Bat Pitch Field Tot ExpWS WSP WSAB NL Pujols STL 12.1 0.0 0.8 13 4.2 1.522 10 NL Berkman HOU 9.7 0.0 0.6 10 4.4 1.179 7Albert Pujols is King, of course. At this pace, he would rack up 56 Win Shares. The record is 59, by Honus Wagner in 1908. As great as Pujols's bat has been, he's also tied for the most fielding Win Shares. The key fielding stats for first basemen are range and errors (including errors by other infielders, to a small degree). In 1999, Tino Martinez racked up 5.2 fielding Win Shares at first, which I think is the second-highest total ever. Overall, Jason Giambi is the top American League first baseman.
Second Base League Name Team Bat Pitch Field Tot ExpWS WSP WSAB NL Utley PHI 7.1 0.0 1.0 8 3.9 1.043 5 AL Lopez SEA 4.5 0.0 2.0 6 4.0 .807 4Chase Utley has been turning heads everywhere, but are you a little bit surprised that Seattle's Jose Lopez is ranked second in the majors? He's batting .299/.320/.488 and he's batting .425 with runners in scoring position. With two fielding Win Shares, he's tied with Aaron Hill for the major league lead. Hill, however, has -2.2 batting Win Shares.
Overall, second basemen have 99 Win Shares, less than any position other than designated hitter.
Third Base League Name Team Bat Pitch Field Tot ExpWS WSP WSAB AL Chavez OAK 6.7 0.0 0.5 7 3.3 1.088 5 NL Ensberg HOU 6.5 0.0 0.4 7 4.2 .835 4 NL Encarnacion CIN 6.0 0.0 0.2 6 3.7 .838 4Eric Chavez, of course, is having a fine season, with 25 Runs Created already. The Reds' Edwin Encarnacion has been inconsistent in the field, and his fielding Win Shares show it. But he's been dynamite at bat.
Miguel Cabrera actually leads all third basemen with 34 Runs Created, (second to the Astros' Morgan Ensberg) but he's hurt by his glove (only .2 fielding Win Shares) and the Marlins' poor "Pythagorean performance." The Marlins are actually four wins under their projected won/loss record (that is, projected according to runs scored and allowed) and this hurts Cabrera's relative Win Share performance. It helps Encarnacion, whose Reds are three games above their Pythagorean projection.
By the way, it appears that Alex Rodriguez is fielding well once again. He's got 1.4 fielding Win Shares, tied with David Bell and Scott Rolen for the major league lead.
Shortstop League Name Team Bat Pitch Field Tot ExpWS WSP WSAB NL Greene SD 4.9 0.0 1.8 7 3.9 .849 4 NL Lopez CIN 6.1 0.0 0.6 7 4.4 .768 4 AL Cabrera LAA 5.0 0.0 1.3 6 3.8 .828 4That left side of the Reds' infield is the best in the majors so far. Felipe Lopez has done nothing to shake his good hit/no field reputation. But Khalil Greene is the shortstop leader, despite a .227 batting average. He's got a .447 slugging percentage with eight home runs (playing in PETCO Park), and he's batting with runners in scoring position. His 26 Runs Created are fourth among all shortstops.
John Dewan's Fielding Bible compared Adam Everett and Derek Jeter in the field and declared Everett the best-fielding shortstop in the majors and Jeter the most overrated. The trend continues this year, as Everett is second among major league shortstops with 2.5 fielding Win Shares and Jeter is almost last with 0.3. I have no idea why Jhonny Peralta has 2.7. I expect that to change.
Outfield League Name Team Bat Pitch Field Tot ExpWS WSP WSAB AL Swisher OAK 8.0 0.0 0.4 8 3.6 1.182 6 NL Abreu PHI 7.6 0.0 0.5 8 4.0 1.014 5 NL Alou SF 6.5 0.0 0.4 7 2.6 1.310 5 NL Beltran NYN 4.8 0.0 1.8 7 2.9 1.161 5 AL Rios TOR 5.6 0.0 0.8 6 3.0 1.068 4 AL Dye CHA 5.2 0.0 0.7 6 2.5 1.188 4Win Shares lumps all outfielders together instead of breaking them out by left, center and right. That's because fielding stats aren't separately available for each of the three positions throughout major league history. Still, fielding doesn't matter a lot for outfielders (Carlos Beltran leads all major leaguers with 2), it's what they do at the plate that matters. Otherwise, the outfield leaders closely mirror the Runs Created leaders.
By the way, Mets fans, you read that right. Beltran is the fourth-best outfielder in the majors so far. He's earning his money.
Designated Hitter League Name Team Bat Pitch Field Tot ExpWS WSP WSAB AL Thome CHA 10.1 0.0 0.0 10 2.8 1.836 8 AL Gomes TB 8.7 0.0 0.1 8 2.8 1.570 7Isn't it kind of weird that, at .255, designated hitters have the lowest batting average of all AL positions? Or that Rondell White has -3.5 Win Shares as the designated hitter?
Starting Pitcher League Name Team Bat Pitch Field Tot ExpWS WSP WSAB NL Glavine NYN 1.4 6.7 0.0 8 2.6 1.552 7 AL Contreras CHA 0.0 7.9 0.0 8 2.4 1.615 6 AL Kazmir TB 0.0 7.4 0.0 7 2.7 1.374 6 NL Arroyo CIN -0.7 8.0 0.0 7 2.7 1.339 6Talk about contrasts. Glavine and Contreras are continuing the recoveries they started last year while Kazmir and Arroyo are the young studs finding their way. Note that Glavine is leading all major league pitchers with 1.4 batting Win Shares, based on his .500 batting average and five Runs Created in only 12 at bats. He's only eighth in pitching Win Shares.
I've always felt that starting pitchers are underappreciated by Win Shares, which is why we prefer to use WSAB (which has a lower baseline for starters).
Relief Pitcher League Name Team Bat Pitch Field Tot ExpWS WSP WSAB AL Papelbon BOS 0.0 5.6 0.0 6 2.3 1.214 4 NL Gordon PHI 0.0 4.5 0.0 4 1.8 1.258 3 NL Sanchez NYN 0.0 4.2 0.0 4 1.5 1.394 3 NL Coffey CIN 0.0 4.1 0.0 4 1.5 1.337 3One of the reasons Win Shares appears to underrate starters is that it gives relievers extra credit for pitching in "clutch" situations. In the zero-sum game of allocating wins, something has to give, and it's often the starting pitchers.
Even though Win Shares doesn't use WPA, it does a decent job of using saves and holds to calculate "clutch" pitching. Papelbon, Gordon, Sanchez and Coffey, the top four Win Share relievers, are in the top eight of Baseball Prospectus's Relief WPA tables.
That's all we have for Win Shares today, boys and girls. Raise your hands and send an e-mail if you have any questions.
References and Resources
Please note: these Win Share totals are a bit more current than the totals in our Win Shares pages.
Pete Simpson first turned me onto Win Shares and built the spreadsheets that we still use today. I don't know where you are now, Pete, but thanks again.
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.
<< Return to Article