October 22, 2003
One of the things I've been struggling with over the last couple of weeks is the idea that Mike Mussina
isn't a big-game pitcher. As it is, I think he's tremendously underrated; while he doesn't have any awards or 20-win seasons under his belt--his best years have coincided with some pretty bad run support--he's been healthy and consistent for the better part of 13 years. He has 199 wins and is already a Hall of Fame candidate, and a good bet to be inducted once the automatics from his era are enshrined.
Mention his name in some quarters, however, and be prepared for a torrent of abuse. Mussina has seen his reputation sullied by the vagaries of the support he's received, particularly in the postseason. After last night, his career postseason ERA is 3.06 in 15 starts and one relief appearance, but he has just five postseason wins. In some circles, that's enough to diminish what has been a great career.
It shouldn't. Here are Mussina's 15 postseason starts. "RS" is run support while he was the pitcher of record. "TR" is the number of runs his team scored in the entire game.
Date Opp Dec IP H R ER BB SO GS RS TR
10/04/96 Cle N 6.0 7 4 3 2 6 48 4 4
10/11/96 NYA L 7.2 8 5 5 2 6 47 2 2
10/01/97 Sea W 7.0 5 2 2 0 9 68 9 9
10/05/97 Sea W 7.0 2 1 1 3 7 73 3 3
10/11/97 Cle N 7.0 3 1 1 2 15 80 0 1
10/15/97 Cle N 8.0 1 0 0 2 10 88 0 0
10/13/01 Oak W 7.0 4 0 0 1 4 72 1 1
10/18/01 Sea W 6.0 4 2 2 1 3 58 3 3
10/27/01 Arz L 3.0 6 5 3 1 4 34 1 1
11/01/01 Arz N 8.0 5 2 2 3 10 71 0 3
10/04/02 Ana N 4.0 6 4 4 0 2 36 6 6
09/30/03 Min L 7.0 7 3 3 3 6 54 0 1
10/08/03 Bos L 5.2 8 4 4 2 4 39 0 2
10/13/03 Bos L 6.2 6 3 3 2 10 58 1 2
10/21/03 Fla W 7.0 7 1 1 1 9 67 2 6
Totals 5-5 97.0 79 37 34 25 105
That's a tremendous record. One-hundred five strikeouts in 97 innings against playoff teams is excellent, and the composite line looks like your basic three months of Mike Mussina
. He's been essentially the same pitcher in the playoffs that he's been in the regular season, maybe even a little better when you consider the quality of competition.
Look at the individual starts. Of the 15, 10 come in at a game score of 50 or higher, and five are over 70. That's not the record of a postseason choker. Moreover, there are some amazing performances in that list, including two starts, in the 1997 ALCS and the 2001 Division Series, in which Mussina didn't allow the other team to score in a game his team needed to win to stay alive.
One of the things I noticed is that Mussina seems to struggle on long rest in the postseason. Of his five worst postseason starts, four of them came with more than the usual four days' rest. Through 2001, though (the last year for which I have data handy), there's no indication of a comparable split in his regular-season stats, so it's probably not meaningful.
The big number up there is Mussina's run support. Mussina has received no runs-he could not possibly have won the game-in a third of his postseason starts. He received one run--he had to throw a shutout to win--on three other occasions. Just three times has Mussina received more than three runs of support, and in one of those, he was taken out of the game after just four innings (October 10, 2002 against the Angels). Fascists get better support in Berkeley than Mussina has gotten from his teammates in October.
Mussina's won-lost record isn't just a poor measure of how well he's pitched, it's contraindicative. He hasn't been a .500 pitcher; he's been a .700 pitcher with .300 run support. Nevertheless, many people can't get past that 5-5, and want to use that deceptive data point to impugn Mussina's character.
Contrast Mussina's performance record with that of clutch god Andy Pettitte.
Date Opp Dec IP H R ER BB SO GS RS TR
10/04/95 Sea N 7.0 9 4 4 3 0 40 4 7
10/02/96 Tex N 6.1 4 4 4 6 3 46 2 5
10/09/96 Bal N 7.0 7 4 4 4 4 47 3 5
10/13/96 Bal W 8.0 3 2 2 1 3 70 6 6
10/20/96 Atl L 2.1 6 7 7 1 1 17 0 1
10/24/96 Atl W 8.1 5 0 0 3 4 74 1 1
10/02/97 Cle L 5.0 9 7 7 1 3 23 3 5
10/06/97 Cle L 6.2 6 4 4 0 2 48 3 3
09/30/98 Tex W 7.0 3 1 1 0 8 75 3 3
10/09/98 Cle L 4.2 8 6 6 3 1 22 1 1
10/21/98 SD W 7.1 5 0 0 3 4 69 3 3
10/07/99 Tex W 7.1 7 1 1 0 5 65 2 3
10/17/99 Bos W 7.1 8 2 2 1 5 58 3 9
10/26/99 Atl N 3.2 10 5 5 1 1 21 1 6
10/04/00 Oak W 7.2 5 0 0 1 3 66 3 4
10/08/00 Oak N 3.2 10 5 5 2 4 23 7 7
10/13/00 Sea W 6.2 9 2 2 1 2 49 4 8
10/21/00 NYM N 6.2 8 3 3 1 4 47 2 4
10/26/00 NYM N 7.0 8 2 0 3 5 59 2 4
10/11/01 Oak L 6.1 7 1 1 2 4 55 0 0
10/17/01 Sea W 8.0 3 1 1 1 7 78 4 4
10/22/01 Sea W 6.1 8 3 3 1 1 43 9 12
10/28/01 Arz L 7.0 5 4 4 0 8 59 0 0
11/03/01 Arz L 2.0 7 6 6 2 1 17 0 2
10/02/02 Ana N 3.0 8 4 4 0 1 28 1 6
10/01/03 Min W 7.0 4 1 1 3 10 72 4 4
10/09/03 Bos W 6.2 9 2 2 2 5 51 4 6
10/15/03 Bos N 5.0 8 4 4 2 5 38 6 6
10/19/03 Fla W 8.2 6 1 0 1 7 76 6 6
Totals 13-7 179.2 189 86 83 49 111
Pettitte's career postseason ERA
is 4.16, more than a full run higher than Mussina's. Despite that, he has a much higher winning percentage and has won nearly half of his starts. Mussina's mean (60.9) and median (58) Game Scores are both significantly higher than Pettitte's (49.5 and 49). Mussina has stronger peripheral stats as well.
Pettitte's better postseason record is directly tied to better run support. Mussina's run support per nine playoff innings is 2.97. Pettitte's is 4.36. The Yankees have been shut out four times with Pettitte on the mound and scored one run for him four times; that total of eight "can't win" starts matches Mussina's, but Mussina has ten fewer playoff starts.
The difference between the reputations of Mike Mussina and Andy Pettitte is that when Pettitte has done his best postseason work, his team has scored. In Mussina's five postseason starts with Game Scores above 70, he is 2-0 with three no-decisions, having received four runs of support in 37 innings. Pettitte has five postseason Game Scores above 70 as well; he's 5-0, receiving 18 runs of support in 39 innings.
That's the difference between a reputation as a clutch god and one as a poor postseason performer: being in the right place at the right time with the right lineup.
Tuesday night, Mussina followed up last Thursday's heroic relief appearance-no win, no save, just three innings that may have saved the Yankees' season-with another excellent postseason start: seven innings, nine strikeouts, one run. He might well have had a shutout, but the Yankee defense conspired to give up a run in the first inning, allowing two poorly hit balls to go for a double and an RBI single. Mussina even came back after a 39-minute rain delay, and although he clearly wasn't the same pitcher, dodged trouble long enough to get the game to Mariano Rivera, Cyborg Reliever.
The hard-luck pitcher last night was Josh Beckett, who watched his team score just one run in seven innings, and who wasn't even in the game when the run that pinned him with a loss was scored. Of course, Beckett has a halo--and not an undeserved one--after his amazing work to eliminate the Cubs last week, so the Marlins' loss won't be seen as a failure on his part.
Reputations are made and unmade not by performance, but by how that performance is described by the media. Because good stories can push facts to the background, it's critical to look past the mainstream take on a situation and find out what really happened. Mike Mussina isn't a choker any more than Andy Pettitte is clutch; they've both had good and bad starts in the postseason, but Pettitte has had more opportunities to succeed and been fortunate enough to get some runs when he's needed them.
Last night, Mike Mussina had Andy Pettitte's luck, and Josh Beckett had Mike Mussina's luck. If we have a Game Seven, things could easily reverse themselves, and it has nothing to do with the character of the players.
- Beckett, who pitched very well for the fourth straight time, was squeezed pretty badly in the fourth inning when he gave up his first run. He walked Jason Giambi and Jorge Posada in the frame; either might have been called out on strikes on the key pitches.
Like I've said, balls and strikes right now are essentially random, and the players know it. If you honestly don't think a machine can do better than this...hell, a random number generator can probably do better than this. Just give the umps a dice tower and a 20-sider...strike 1-12, ball 13-20. The results wouldn't be noticeably different, and the illusion that an ability is on display would be lost.
The same goes for check-swing calls, especially on appeal. If you can discern a pattern to those--I'm thinking here of Alex Gonzalez in last night's game, a swing called a check--you're a better man than I am.
Maybe the current technologies aren't there yet, but very soon machines are going to be able to call balls and strikes much, much better than the umpires. At that point, and we're not far from it, baseball is going to have to make the leap. The game deserves better than to have its most important space be treated with such disdain by the men charged with upholding its sanctity.
- Beckett doesn't get off completely scot-free based on the umpiring. Hitting Hideki Matsui with a 1-2 breaking ball is borderline
indictable, and contributed to that run in the fourth. With a few exceptions,
Beckett's curve seemed to give him trouble last night. He was more effective
when he was using his fastball.
- One of the developing concepts of the postseason is that idea that the
outcomes don't necessarily tell you which is the better team, but which is the
better team in those particular games. The influence of luck on the outcome is
disproportionate, because each game bears such importance in the final result.
I mention this in the context of the Marlins coming up empty in the sixth
inning last night. With one out, Ivan Rodriguez doubled off the wall in left. Miguel Cabrera followed with a line single to right field. The ball was hit too hard for Rodriguez to score, and Marlins' third-base coach Ozzie Guillen correctly held him up. But when right fielder Karim Garcia bobbled the ball, Guillen tried to send Rodriguez home. The runner had slowed down too much, however, and had to hold at third. Three batters later, the inning was over and the Marlins still had one lonely run.
You could run that sequence 100 times, and on probably 50 of them Rodriguez scores. Maybe he doesn't decelerate into third quite so much. Maybe he picks up Guillen's changed signal a split-second sooner and heads for home. Maybe he gets a slightly better jump off of second base and is sent home to begin with. These things happen in microseconds, and if you don't like the word "luck" you can go with "physics."
Last night the microseconds, the physics, didn't fall the Marlins' way. It cost them a baseball game. It happens.
- So, what do you figure, there are maybe 3,000 baseball people at the games? Execs, players, ex-players, all kinds of people associated with the game who have interesting and entertaining stories to tell.
Three minutes into last night's rain delay, Fox gave the air back to its local affiliates, so instead of baseball promotion, we here in L.A. got "Married With Children."
I honestly don't know how you make a decision like that. Maybe I don't want to know.
- The game resumed after a 39-minute delay and the teams played through intermittent showers the rest of the way. It's highly unlikely that you'll ever see an in-progress World Series game called off, and almost as unlikely that we were going to have a second rain delay last night, no matter how many outfielders ran like they were on ice. Fox didn't want to go deep into the night, and there's no way MLB wanted a rainout and the logistical headaches that would cause.
One of the least attractive trends in recent years is that baseball gets played in weather unsuited for the game. This is in part due to improved drainage systems that keep fields playable longer, but it also has to due with the capacity crowds at many ballparks. With some teams' entire schedules sold out, rainouts can't be handled with rain checks. Tickets have to be refunded or, in some cases, exchanged for tickets the following season. That costs teams money, and they don't like that. So games are played in nasty, cold, wet weather, just so refunds don't become an issue. The quality of the game, the quality of the experience for the fans, and sometimes the safety of the players takes a back seat to the money.
I think last night would have looked more unusual 20 or 30 years ago. We've become used to the sight of baseball being played in the rain, even though we know it shouldn't be that way.
- The fascination with Derek Jeter is generally unwarranted. He's a very good, very smart player who also happens to be overrated due to his athleticism, personality, and a decade of good teammates.
That said, I don't think you can say enough about his baserunning. If he's not the best baserunner in the game, I don't know who is. He's fast, smart, and seems to have an instinct about risks and rewards. By advancing to third on a medium-depth fly ball to Juan Pierre last night, he scored an easy run on Matsui's subsequent single, instead of being involved in a play at the plate.
As is the case with defense, baserunning is not a "little thing," or an intangible, although it gets lumped in with those sometimes. It's just an area of the game that we haven't gotten around to measuring yet. (I know STATS, Inc. used to do stuff in their Baseball Scoreboard book, but that product no longer exists, and I haven't seen comparable information since its demise.) Jeter puts runs on the board with his baserunning; certainly less than 10 a season, but he's a positive contributor that way, and there probably aren't 50 players who make a difference worth mentioning in this area. I suspect that we'll be able to confidently evaluate players' baserunning performance by the end of the decade, maybe sooner.
- What follows seemed really important at the time, and just as irrelevant now. I'm going to get into it, though, because I know that some of the readers are table-game players, and you might find the thought process interesting.
With two outs, the bases loaded, and a 2-1 lead in the eighth inning last night, Joe Torre sent up Ruben Sierra to bat for Karim Garcia. After Sierra was announced, Jack McKeon removed Dontrelle Willis in favor of Chad Fox.
I figured Torre had McKeon roped. Now he could hit Nick Johnson for Sierra, which would give him his best OBP guy, his most patient hitter, batting against a pitcher with control problems who had no place to put the runner. Even better, Johnson could stay in the game at first base, with David Dellucci, already pinch-running for Jason Giambi, staying in to play right field for Garcia. It seemed like the perfect plan.
Torre apparently disagreed, allowing Sierra to bat against Fox (he struck out). I thought it through, and what I came to was that Torre preferred getting Mariano Rivera into the game as far from having to bat as possible. If he'd used Johnson for Sierra and the inning had ended before the #9 spot batted, Rivera would have batted in the ninth inning.
Now, I understand that, but when I go through the decision tree, I think that Rivera in the #9 spot was not only not that undesirable, but a vanishingly small price to pay for the greater chance of scoring in the eighth.
Fox is right-handed, and both Sierra and Johnson would be batting left-handed against him, so there's no edge there. Sierra hit .284/.346/.427 against right-handers during the season, which is a reasonable approximation-probably a little generous-of his skills. The key number with a one-run lead, of course, is the OBP. The value of the second run is considerable, especially with Cyborg Reliever coming into the game.
Johnson hit .285/.429/.482 against right-handers this year, also a pretty good approximation of his skills. His .429 OBP against righties represents about a 25% better chance of scoring one run than Sierra's .346 does, and the rest of Johnson's skills create a marginally better chance of scoring two or more. Beyond the stats, sending the patient Johnson up against the closest thing we have to Nuke LaLoosh in this World Series is the move that maximizes the chance of scoring anything.
If you let Sierra bat, you get Johnson batting in the ninth inning unless both Sierra and Aaron Boone reach base in the eighth, in which case you're up at least 4-1, anyway. Rivera is safely ensconced in the #7 spot for his two innings of work. If Sierra makes an out, Johnson bats second in the inning behind a low OBP (Aaron Boone), lessening his chance to have an impact. If Sierra reaches, you have at least a two-run lead and maybe more, decreasing the value of any runs Johnson might create no matter when he bats.
However, if you let Johnson bat for Sierra, you get an immediate boost in the chance of scoring a key run. If he succeeds, the worst-case scenario is that Rivera leads off the ninth inning with a multiple-run lead (assuming, arguendo, a scoreless bottom of the eighth). If Johnson fails, you have Rivera batting second in the ninth behind a low-OBP hitter. If Boone reaches, Rivera gets to bunt. If he doesn't reach, Rivera is batting with one out, and while that's suboptimal, once you have one out, the difference between the expected runs with Johnson and Rivera batting is much smaller.
I think Torre let his desire to not have Rivera scheduled to bat in the ninth inning drive the decision when that was really not an important factor. There were few scenarios where the Yankees would have been significantly worse off with Rivera ninth than with Johnson ninth, and the difference pales next to the added value in getting Johnson up against Fox with the bases loaded. Sure, you can concoct extremely risk-averse reasons-what if Rivera gives up runs? What if he gets hit by a pitch?-but that's managing to not lose, and Torre has never done that with this team in the postseason.
Like I say, it ended up being meaningless when the Yankees got four runs in the ninth. I'm just enough of a [fan/seamhead/dork] to think that the whole process is interesting.
- Speaking of Aaron Boone, he's certainly having an interesting week. Probably the worst player in the first 26 innings of the Series, with three errors and more missed swings than an off night at Plato's Retreat, he rendered that little mental exercise above moot with a solo home run leading off the ninth, his second after-midnight home run in six days.
Maybe Boone just needs to be platooned. He'll get all the a.m. playing time, and the Yankees can find someone to take the post-meridien at-bats. Maybe the Yankees can lobby for a spot in the Red Sox' Patriot Day game. If they lead off Boone, they should be able to get him up twice before noon.
- You know what has to be nice? Being soaking wet at 11:20 on a Tuesday night, knowing you're not getting home until 1, maybe 1:30, with work the next day, and being forced to listen to a 30-years-past-relevance child star do a "Star Search" version of about the fifth-best standard about America.
David Cassidy?!? What, Leif Garrett was busy? Peter Frampton begged off? Davy Jones couldn't get past the "You Must Be This Tall To Annoy the American Public" sign?
- True story: the Yankees have hit three-run homers following Alfonso Soriano's last two walks.
OK, it's cheap, because Soriano got himself picked off and had nothing to do with the first one, but the Yanks should use this as positive reinforcement.
Maybe we can set a betting line, like ESPN The Magazine does for goofy things. What side of this prop bet on the number of times we'll hear each phrase today do you like?
"Roger Clemens' last start" 2 1/2 "Past a diving Jeter"
I tried to get inside Roger Clemens
' head before his last final start, which turned out to be a mistake. I won't do that this time; I have no idea how this being his current final start will affect him. None. I do know that, this being Game Four, it is his final final start. There can't be a next final start unless...you know, I don't even want to imagine what kind of scenarios Bud Selig and Jeffrey Loria might concoct to bring us a Game Eight.
I do know that he was up in the zone in his Division Series outing against the Red Sox, which was his seventh or eighth "final start" after his final regular-season start, his final start at Fenway Park (which was only his next to final start at Fenway Park), his final start in the All-Star Game, his final start at Yankee Stadium (also just his next-to-final), his final start in a foreign country, his final start in front of a record-low crowd and his final start with nasty heartburn.
This matchup isn't as bad for the Marlins as Mike Mussina was. Clemens works up and down with the splitter and fastball, and has shown a fairly persistent reverse platoon split since joining the Yankees. With a bunch of right-handed hitters who can drive a good fastball but who will chase once they fall behind in the count, Clemens' success will again come down to getting ahead in the count and avoiding leaving his fastball up in the zone. There's not a lot of middle ground here; look for a 3.2-7-6-6-4-2 line, or a 7-4-1-1-2-10 one.
Carl Pavano saved the Marlins' bacon in Game Six against the Cubs, and he's being asked to do much the same thing here. Pavano was much more effective at Pro Player Stadium than elsewhere in 2003, with a much lower home-run rate. The Yankee offense relies on homers and walks, and Pavano is stingy with both at home, making him a tough matchup for the Bombers.
The Marlins' story is splitting from that of last year's Angels at the wrong time. Remember that while the Halos made contact and ran the bases well and got good relief pitching, they also smacked the bejesus out of the ball for three weeks. The Marlins matched that in dispatching the Giants and especially the Cubs, but have stopped driving the ball in the World Series. They have three extra-base hits, all doubles on Tuesday night.
While the media is fixated on the idea that Juan Pierre can create runs, MacGyver-like, from a fungo bat, some leftover baseline chalk and a ball-strike indicator, the Marlins need the middle of their lineup to launch some horsehide if they're to keep following the Angels' script.
The Fish have one last gasp in them, but it's not tonight. Yankees, 5-4.
Joe Sheehan is an author of Baseball Prospectus.
Click here to see Joe's other articles.
You can contact Joe by clicking here