September 9, 2003
From The Mailbag
Halladay, Gagne, Ichiro, and GWHBPRBITHE DOC IS IN
With Roy Halladay's impressive 10-inning shutout against Detroit Saturday, I just got to thinking: How do your excellent support-neutral stats deal with a 10-inning start like that? Or is the difference between nine and 10 innings pitched insignificant enough to be safely ignored? What about 11 innings, hypothetically?
Extra innings are dealt with by computing the odds of winning in nine, the odds of winning in the 10th, the odds of winning in the 11th, etc. separately. For example, SNW for Halladay's start is the chance that his first nine shutout innings would be good enough to win, combined with the chance that his scoreless 10th inning would get the job done:
SNW = Prob(Avg team scores >0 runs in 9 innings) + [Prob(Avg team scores 0 runs in 9 innings) * Prob(Avg team scores >0 runs in 1 inning)]
SNL is zero.
So he does get more credit than those lightweights who are satisfied with a mere nine scoreless innings.
I was looking at the reliever stats, and surprised that Eric Gagne isn't ranked higher given how unhittable he's been. Without doing any heavy math, I'm guessing he's hurt in the calculation by the Park Effects. A couple of questions: Do Park Effects affect all pitchers equally, or does a strikeout specialist like Gagne get less benefit than a nibbler like Kirk Rueter (who may put more balls in play that get caught in an expansive outfield or wide foul ground)? Since the Park Effects calculation is based on a large population (all pitchers appearing at the stadium), does "luck" or "randomness" come into play more for a reliever who has less than 100 innings pitched?
Park effects play a small role in Gagne's rating, but the real reason he ranks where he does is more basic than that: the reliever stats don't care about how unhittable a pitcher is, they just care about runs. Gagne has a great ERA, but so do several other NL relievers--Rheal Cormier, Billy Wagner, John Smoltz, etc. And in Cormier's case, he's also been terrific at preventing inherited runners from scoring, something the closers aren't often called on to do.
As to your question about the park's effect on different types of pitchers, I know there are some studies out there that isolate parks' impact on individual components of performance--strikeouts, home runs, etc. It certainly is possible for a pitcher to be ill-suited to the park he plays in. I doubt that's the case with Gagne in particular (or strikeout pitchers in general) in Dodger Stadium, though. Career, Gagne has a 3.31 ERA in Chavez Ravine, and 3.91 away.
BEST SEASONS WITH FEWER THAN 10 WINS
I was just watching the Arizona/Kansas City make-up game when something hit me. Both of these teams have winning percentages about .500 and neither of them have a 10-game winner. This brings up some questions, (including) what was the best season by a starter who didn't win 10 games? Is Brandon Webb up there?
Candidates for best season < 10 wins, excluding closers:
YEAR NAME W L SV IP BB SO ERA ERA+ ---- --------------- --- --- --- ------ ---- ---- ------ ----- 1886 BOYLE,HENRY 9 15 0 210.0 46 101 1.76 1.90 1997 APPIER,KEVIN 9 13 0 235.7 74 196 3.40 1.32 1994 CLEMENS,ROGER 9 7 0 170.7 71 168 2.85 1.65 1902 SIEVER,ED 8 11 1 188.3 32 36 1.91 1.85 1900 WADDELL,RUBE 8 13 0 208.7 55 130 2.37 1.58 1987 RYAN,NOLAN 8 16 0 211.7 87 270 2.76 1.51 1992 ABBOTT,JIM 7 15 0 211.0 68 130 2.77 1.42 1958 MILLER,STU 6 9 0 182.0 49 119 2.47 1.62 1978 SWAN,CRAIG 9 6 0 207.3 58 125 2.43 1.49 1942 NEWHOUSER,HAL 8 14 5 183.7 114 103 2.45 1.44 1987 REUSCHEL,RICK 8 6 0 177.0 35 80 2.75 1.48 1961 HOEFT,BILLY 7 4 3 138.0 55 100 2.02 2.02 1998 DAAL,OMAR 8 12 0 162.7 51 132 2.88 1.46 1916 SCHUPP,FERDIE 9 3 1 140.3 37 86 0.90 3.01
I read this in (a recent) PTP:
"Let's get the case out of the way: by any measure, Rodriguez is the top player in the league. Advanced metrics such as VORP and RARP have him a win or more above his closest pursuers (and one so far behind that the media halo is laughable)."
The laughable player being Ichiro Suzuki and his RARP of 29.5 (at the time)...well behind other MVP candidates. I thought the same thing, that the media attention was laughable. But then I thought about his defense and noticed on your own Web site that Ichiro's defense in his MVP season registers as very high in FRAR and FRAA for the last five years.
No doubt Alex Rodriguez is the MVP in the AL. And when I looked at the issue on 8/22, Ichiro looked better than he does today. On Aug. 22, Ichiro had an EqA of around .300 and was 14th in the AL in RARP. In Ichiro's MVP season, his defense garnered him 42 FRAR while his bat was worth 51 RARP. Assuming those are on the same scale, then it seems like Ichiro contributes a ton with the glove. By traditional measures (range, fld%, zone rating, Ast, DPs), Ichiro is as good if not better than in 2001. As best as I can tell, his 25 FRAA (2001) is only matched or exceeded in the OF since '97 by Andruw Jones in '98, Darin Erstad in '99 and '02, and Torii Hunter in '01. CFs all of them.
It stands to reason that this season Ichiro is worth around his 42 FRAR/25 FRAA from 2001. Assuming a fielding run is worth as much as an EqR, those are big numbers compared to Jason Giambi's 7 FRAR/1 FRAA from '02. I know that fielding runs are far from perfect, but it seems that Ichiro is as good in the OF as anyone in recent seasons and is easily the best RF in baseball today. And it is quite possible that his astounding glove makes up for his bat when compared to Giambi or Carlos Delgado. What are your thoughts? I know statheads get caught up in hitting stats, but using BP's own numbers it seems like Ichiro may very well be (or at least was) a legit MVP candidate.
Before getting to the point, let me start by saying that I am basing my estimates here on the 2.0 version of the player cards, which have not yet gone onto the Net. The most significant changes are in the fielding side; in particular, I am a LOT happier with the way the 2003 version of fielding handles outfield defense. A "casualty" of that is Ichiro; his fielding in 2001 is now "only" 33 RAR/18 RAA, instead of 42/25, which cuts into your argument substantially.
However, Ichiro does, in fact, continue to excel with the glove, getting a 27/14 from me through Aug. 31 (virtually identical to his 2001 pace, so that was a pretty fair assumption you made). The combination of his hitting (43 BRAR) gets him a WARP-3 score of 8.5 for the 2003 season, which puts him in a three-way tie for 10th in the AL. That is a pretty dramatic improvement over his 28th-place showing in BRARP.
The biggest trouble is that two of the top three players in BRARP are Alex Rodriguez and Bret Boone, who both happen to play middle infield in a way that BP's stats and the Gold Glove voters both find pretty special. A-Rod gets a 36/12 for his fielding, so he actually gains on Suzuki, netting a 12.1 WARP3. Boone gets a similar 36/13, and has a WARP3 of 11.0. These two are far and away the statistical front-runners in the AL. The entire MVP ballot, if done strictly by WARP3, would look like this (through 9/2):
1. Alex Rodriguez (12.1)
Jason Giambi is right behind this group, at 8.4.
Note that there is a 1.1-win gap between 1 and 2, a 1.4-win gap between 2 and 3, and then just 1.1 between 3 and 10. At this point, I'd say that you have to rank ARod and Boone 1-2, and then everybody else can come in any order you want. Suzuki is a legitimate MVP candidate, in the sense that he has a legitimate claim to be named on the ballot, and anyone on the ballot is a candidate. If, by "legitimate MVP candidate," you mean that there is a strong argument to actually have him win the award, then I have to say that he falls short.
(A recent) Everett AquaSox game ended when the AquaSox pitcher hit a batter with the bases loaded. I guess that makes it a game-winning, hit-by-pitch, run batted in? A GWHBPRBI? My friends and I have been thinking about that play today and have this question for you: Who holds the major league record for game-winning RBI by getting hit by a pitch in the bottom of the ninth?
I don't have a complete answer for you, but I can say that no player has had a "walk-off" HBP more than once since 1972, and that it's been done only 28 times since 1972 (through 2002).
GAMEDATE BAT INN NAME PIT V H OUTS --------- --- --- --------------- --- --- --- ---- 16-JUL-72 PIT 10 Robertson,Bob HOU 2 2 0 26-AUG-72 CHN 10 Pepitone,Joe SFN 9 9 1 22-SEP-72 LAN 11 Parker,Wes SFN 0 0 2 25-JUL-74 SLN 10 Davanon,Jerry NYN 3 3 1 22-JUN-75 CLE 10 Carty,Rico MIL 2 2 2 20-MAY-78 LAN 9 Russell,Bill SFN 2 2 1 23-MAY-79 DET 9 Kemp,Steve NYA 3 3 2 17-AUG-79 MON 9 Scott,Rodney ATL 0 0 1 19-MAY-86 BOS 9 Sullivan,Marc MIN 7 7 2 23-JUL-89 CAL 9 Joyner,Wally DET 4 4 1 11-JUN-91 CLE 12 Cole,Alex TOR 1 1 1 09-AUG-91 SFN 13 Mitchell,Kevin LAN 0 0 1 23-APR-92 NYN 13 Boston,Daryl SLN 0 0 1 08-JUL-92 CHN 10 Sanchez,Rey CIN 2 2 1 04-AUG-93 OAK 9 Bordick,Mike SEA 4 4 1 27-SEP-93 KCA 9 Gaetti,Gary CLE 5 5 2 30-AUG-95 SDN 9 Cianfrocco,Archi MON 2 2 1 01-MAY-96 MIN 10 Molitor,Paul KCA 5 5 1 15-APR-97 PIT 9 Womack,Tony SDN 2 2 1 16-JUN-97 BOS 10 O'Leary,Troy PHI 4 4 2 03-AUG-97 HOU 9 Gonzalez,Luis NYN 2 2 2 08-AUG-98 HOU 9 Hidalgo,Richard PHI 6 6 0 20-AUG-99 TBA 9 Flaherty,John KCA 4 4 0 03-SEP-99 MIL 11 Loretta,Mark SLN 4 4 0 20-JUN-01 LAN 9 Karros,Eric ARI 3 3 1 20-JUL-02 OAK 9 Saenz,Olmedo TEX 5 5 2 21-JUL-02 KCA 10 Ibanez,Raul CLE 12 12 2 23-AUG-02 LAN 9 Lo Duca,Paul ATL 3 3 1
Also, Bob Molinaro of the White Sox was hit by a pitch as the last batter in a game tied at 1-1, against Texas on July 26th, 1980, but the game was called due to rain in the 6th inning, and there was only a runner on second base, so that doesn't really fit your criteria.
LONG TIME COMING
Carlos Pulido was recently promoted to the majors by the the Twins. He last pitched in the big leagues in 1994. I wondered if you had access to any information as far as the longest time between major league appearances.
Minnie Minoso's 12-year gap (1964-76) immediately came to mind when I read your question, and in fact I think Pulido's eight-year gap would be the longest since Minoso.
As for all time, the winner is: Paul Schreiber, who appeared in 1922 and 1923 with the Brooklyn Dodgers (total of 16 IP), then not again until 1945, a gap of 21 years, when he pitched two games and 4.3 innings for the WWII-era Yankees at age 42.
Charley O'Leary also had a 20+ year gap, skipping 20 seasons from 1913 (Cardinals) to 1934 (Browns, one AB, age 51).
I was flipping through the EqA reports, and I noticed that 3B is currently a weaker offensive position than CF and (almost dead even) with 2B. Is this just a fluke, or is it a trend in recent years? Do we need to start re-arranging the "defensive spectrum" to account for this (i.e. have the standards required of an adequate 2B or CF been changed), or do you think that it is true that the current crop of 3Bs are just simply not as talented as the current crop of CFs (and on par with 2B)?
I think that this is a fluke, at least as far as second and third are concerned. The .264 EqA (registered last month) from second base is the highest value for that position in the last 10 years, with no general upward trend that I can see (since 1994: 262-258-254-261-263-264-261-261-258-264). Third base, on the other hand, has its lowest EqA in the past 10 years, with some indication of a downtrend (266-271-270-269-268-267-264-268-264-262). I don't know that it's a bad crop of current third basemen per se, but I think there are an unusual number of really bad-hitting third basemen this year (led by Beltre, Cirillo, and Tatis, players who have been substantially more productive in the recent past).
Center fielders, though, have been better than third basemen over the last decade (2001 and 1995 are the only two years in the last 10 where they did not meet or exceed the 3B EqA), and they've averaged about two points better over that time. This is nothing new; center fielders have usually been better hitters than third basemen throughout baseball history.