How Did Pedro Martinez Get Bombed?
Clayton Kershaw is coming off what was legitimately one of the best starting-pitcher seasons of all time. Though he would miss a few turns due to injury, that problem was quickly forgotten, as Kershaw still approached 200 innings and finished with both an ERA and an FIP that were half the league average. There was one stretch where Kershaw didn’t allow a single run over four consecutive starts, and that stretch was bookended by a pair of one-run outings. Yet as amazing as Kershaw was, there was one game where he allowed seven runs in under two innings to the Diamondbacks. Those seven runs were 17% of Kershaw’s regular-season total. In July, I tried to investigate what went wrong.
Kershaw, in 2014, had one of the better pitcher seasons ever. Pedro Martinez, in 1999, had maybe the best pitcher season ever. Pedro posted the lowest FIP- ever by a starter, at 28. The next-best mark is 36, also posted by Pedro. The best non-Pedro mark is 45. Over the course of baseball history, that 1999 FIP- is a full five standard deviations better than the mean. Pedro’s strikeout rate that season was 5.4 standard deviations better than the mean. I should note that this doesn’t include what Pedro did in the All-Star Game, or in the playoffs. In the All-Star Game, he struck out five of six batters, with one reaching on an error. In the playoffs, Pedro spun 17 shutout innings, allowing five hits and a .267 OPS. Pedro Martinez, that season, was probably the best that any starting pitcher has been. The statistics are unreal even before you remember to adjust them for the era.
Yet on July 18, Pedro faced the Florida Marlins and couldn’t get out of the fourth. His final line shows nine runs on a dozen hits, with no other season run total exceeding four. That year’s Marlins had one of the worst offenses in either league, and they’d lose 98 games. Just as Kershaw’s disaster was fascinating, so, too, was Pedro’s, particularly in retrospect. How did one of the best pitchers ever, in probably the best pitcher season ever, get killed at home against a bad team on the wrong side of a fire sale?
Here’s an image that you might find helpful. Here are 1999 Pedro’s regular-season appearances, including one extended appearance in relief (and excluding a brief one). This proceeds game by game, and the July 18 start is indicated by the arrows:
In that game against the Marlins, Pedro allowed 84% contact. His next-highest mark was 75%, and his season average was 64%. He allowed a slugging percentage of .727. The next-highest mark was .478. Pedro also posted his season-low in strikeout rate, at 13%. Before July 18, Pedro struck out 35% of batters. After July 18, he struck out 44% of batters. After July 18, he struck out 44% of batters. After July 18, he struck out 44% of batters. After July 18, he struck out 44% of batters. 44% of batters were struck out, by Pedro Martinez, after July 18. Sorry, I got stuck there for a moment.
That game stands as just a staggering kind of anomaly. With it, Pedro allowed an average of 2.36 runs per nine innings. Without it, he allowed an average of 2.02 runs per nine innings. More curiously, this was the outing that directly followed Pedro’s memorable and historical dominance in the All-Star Game. What that looked like, on July 13:
To this day, Pedro’s two-inning appearance is remembered for its impossibility. Pedro was taking a test with the answer book. In the first inning, Pedro struck out Barry Larkin, Larry Walker, and Sammy Sosa. In the second inning, he struck out Mark McGwire and Jeff Bagwell, and Matt Williams simply hit a groundball. In the start before the All-Star Game, on July 7, Pedro struck out 14 Devil Rays in eight innings. Going into July 18, Pedro looked like the best pitcher ever. Then half the batters he faced got hits.
So what happened? With Kershaw, it seemed to be a matter of location and bad luck. With Pedro, there’s — well there’s actually very little mystery. Here’s a clue, showing a specific selection of dates when Pedro Martinez pitched:
- July 7
- July 13
- July 18
- August 3
- August 8
Come August 3, Martinez pitched on 15 days’ rest. Because he’d been placed on the disabled list. The worst game by far in the best season ever was presumably caused by shoulder inflammation, such that Pedro just couldn’t pitch like his normal self.
The move is retroactive to last Monday. On Friday, Martinez underwent an MRI, which showed muscle swelling in the shoulder, the team said.
Martinez, 15-3 with a 2.51 ERA, missed his scheduled start Friday at Detroit after complaining that he could not loosen up during his worst start of the season last Sunday against Florida.
In a separate excerpt, it’s shown that Martinez thought he knew how he developed inflammation in the first place:
Martinez said his problem might have stemmed from overthrowing in the All-Star Game, when he started at Fenway Park and struck out five of the six batters he faced to earn MVP honors.
Every so often you hear about a player reluctant to participate in the midsummer exhibition because of the injury risk. The risk in any individual game is low, especially given the brief appearances and the reduced level of competition and intensity, but pitching is pitching and the pitchers in the All-Star Game tend to throw more or less their usual stuff. Pedro thought he might’ve hurt himself trying to play it up for the Fenway crowd. One result was one of the most memorable All-Star Game appearances ever. Another might’ve been a miserable start and two weeks of rest.
Since Pedro was having an all-time season, and since the Red Sox were in the playoff race, it could’ve been absolutely devastating to have Martinez develop a hurt shoulder. But good news would come quick, and Martinez felt good enough again before the end of the month:
I know my body real well, and I’m telling you right now, I’m ready to pitch. I’ll give you my paycheck, give me my outing.
I feel like I’ve skipped past the bad game itself. Pedro had the lowest strikeout rate of his season, and the highest contact rate of his season. It became known within a few days that Pedro just couldn’t get his arm loose, so he wasn’t pitching at anything close to 100%. In the immediate aftermath, however, Pedro didn’t blame any soreness. He said he “didn’t have anything,” which is a vaguely troubling quote. Another line:
My pitches weren’t working. I was off, probably, with everything.
This from the other side:
“He left a few balls over the plate and we were able to put a good swing on them,” Marlins center fielder Preston Wilson said.
None of this is very descriptive, but it’s easy to interpret them now as clues of a problem. Wilson indicates that Pedro had unusually bad command; he never had bad command. Pedro indicates that he had unusually bad stuff; he never had bad stuff. Everyone’s human, so everyone fails, but the magnitude by which Pedro failed that day would’ve been reason enough to suspect that something bad was going on, beyond it just being a below-average start.
In the first inning, Pedro allowed two runs, on two hits and an error. He managed a scoreless second inning, but the third brought three more runs, and four more hits, plus another error. Pedro was knocked out in the fourth, and here’s how that inning began:
- line drive
- line drive
- line drive
- line drive
- line drive
Three batters later, it was time for Rich Garces. Several innings later, the Red Sox still actually won. Pedro wouldn’t pitch again until August. On August 3, he threw five strong innings against an Indians team that scored 1,009 runs.
The story ends well, aside from the fact that the Red Sox didn’t win the 1999 World Series. Most relevant today, Pedro was able to continue his career and he just got himself voted into the Hall of Fame. But even just as far as 1999 was concerned, when Pedro came off the DL it was as if nothing had ever happened. If anything, it was as if the two weeks of rest allowed him to get even better. Over Pedro’s 77 remaining regular-season innings, he posted a 26 ERA- and a 26 FIP-. He collected 126 strikeouts and 13 walks. On September 10, in Yankee Stadium, Pedro spun a 17-strikeout one-hitter considered one of the best individual games in history. And, oh, by the way:
Martinez admitted after last Friday’s game that his shoulder was bothering him. “I still feel a little tingly, but it’s not anything I haven’t felt before,” he said.
In the best stretch of probably the best season ever, Pedro Martinez said that his shoulder still felt a little tingly. Which maybe shouldn’t be the biggest surprise — after all, that shoulder had front-row seats to watch Pedro Martinez.
Jeff made Lookout Landing a thing, but he does not still write there about the Mariners. He does write here, sometimes about the Mariners, but usually not.

Why don’t you look into why Kershaw totally collapsed in the playoffs?
http://www.fangraphs.com/blogs/clayton-kershaws-big-miss-matt-adams-big-hit/
http://www.fangraphs.com/blogs/pointcounterpoint-clayton-kershaws-7th-inning/
“After July 18, he struck out 44% of batters.”
Just…wow.
Pedro’s Bugs Bunny change up doesn’t seem like it should have been possible given the known laws of the universe.
Funny you should mention that: http://www.banknotesindustries.com/expert-like-analysis/the-physics-of-the-bugs-bunny-changeup/
I believe that is the only article on the internet to formally adapt the term “Cistullian prose”.
Buddy of mine was a minor league catcher for awhile. Pedro visited the team once while he was in AAA and at some point and ended up shaking hands with him. My buddy – who, at 6’3 220, did not exactly have small hands himself – said Pedro’s handshake felt really weird, owing to a combination of large hands (especially for for a 5’10 guy) and freakishly long fingers.
Not sure how much there is to that anecdote, but always found it interesting. It would seem to help explain some of the other-worldly movement on his pitches.
After his recent Hall induction, Pedro was on MLB Network showing off his freakishly double-jointed fingers. I got a much better understanding of how he got that unusual movement on his pitches.
How did you find Contact% for those games?
http://www.baseball-reference.com/players/gl.cgi?id=martipe02&t=p&year=1999
I would’ve guessed gin and root beer
1) I see what you did there.
2) I can’t decide if that combination sounds awful or brilliant.
it’s awesome with the right gin
As an Expos fan this just makes me sad.
was hoping this would be about Pedro Martinez’s drunken escapades…
Nah. Great pitcher but a real lightweight.
You could have posted ten paragraphs of gibberish and stuck that video in the middle and I still would have been happy.
That video, and the last paragraph.
Wait, this isn’t about a playoff appearance vs the Yankees?
Who’s your Daddy?
We know they were All-Stars and great hitters, but in today’s environment it’s sometimes difficult to remember how much better hitting was back then. Here are the ’99 batting lines for the batters faced in that video:
Barry Larkin: .293/.390/.420 (8% K-rate)
Larry Walker: .379/.458/.710 (with a 10% K-rate)
Sammy Sosa–: .288/.367/.635 (63 HR season)
Mark Mcgwire: .278/.424/.697 (65 HR season)
Jeff Bagwell: .304/.454/.591 (42 HR season)
Matt Williams 112 wRC+ season in 1999: .303/.344/.536
Adam Dunn’s 112 wRC+ season in 2014: .219/.337/.415
One thing for sure that I remember about Pedro was that after every mediocre game, he’d almost always say his stuff was great and after every awesome game, he’d say he wasn’t feeling good that day.
Just for kicks, can someone smarter than me rerun Pedro’s 1999 stats minus this July 18th start?
Without that outing, Pedro’s ERA drops from 2.07 to 1.80 and his WHIP from 0.92 to 0.88.
That’s a 1.80 ERA in a league with an average ERA of 4.86.
I do not recall if it was a formal or informal poll of AL hitters in 1999, but by consensus Pedro had the best fastball, change, and command. I don’t remember where they put the curve, which buckled my knees, and I was ninety feet behind home plate.
Pedro was an artist with a wicked brush, a Maddox on steroids, most probably literally. Good for the imp. Them boys he was cutting down were no strangers to the needle.
Chris Sale had almost the same season as kershaw.
No he didn’t.
Slightly off topic, but do people consider Dwight Gooden’s 1985 season in a similar light?
Similarly dominant results, but not nearly as dominant in the process of getting them. 6.5% BB-rate vs. Pedro’s 4.4%; 25% K-rate vs. Pedro’s 37.5%. Fewer HRs, too, in a era of 40-60+ HR seasons and a significantly stronger overall offensive environment.