July 25, 2012
Two Good Starts and a Trade
"I think the one thing you might be able to get someone to dream on—it's amazing to me, continually, how often trades are made based on the last two starts," one American League personnel director told Mackey. "So if he can put together a couple of starts in the next couple of weeks...” (Source)
Presume that this personnel director talking to Minnesota ESPN radio host/writer Phil Mackey about Francisco Liriano wasn’t misquoted. Presume that he wasn’t exaggerating wildly in order to entertain Mackey. Presume that he’s not insane. Presume that he’s telling the truth, and that, at least once, a team has made a trade based on the last two starts. The question, then, is this: Which trade?
The answer won’t necessarily be obvious, and it could be that the last two starts affected the cost of acquiring a pitcher (which we wouldn’t ever know) or dissolved trade talks around a pitcher (which we wouldn’t know). But from the start of July through the end of August, five to 10 established major-league starting pitchers are traded each year. Over the past three years, there have been, depending on your definition of the terms, about 20 such pitchers traded. And one of these pitchers, I will speculate, is the pitcher this personnel director is talking about.
Roy Oswalt (2010). Average game score in his final two starts: 36. (To remind everybody, a game score of 36 is quite poor; a game score of 50 is average.) Three starts before he was traded, Oswalt threw a one-hit shutout. But in the final two starts before the Phillies acquired him, he pitched just nine total innings; he left one start after four innings when Pedro Alvarez lined a ball off his ankle, and he left the next after five innings when he gave up six runs and a couple homers. His ERA went up by more than a third of a run in those two starts.
J.A. Happ (2010). Average game score: 47*. Four walks and four strikeouts in five shaky innings in his one pre-trade start, after pitching lousily (5.97 ERA) while rehabbing in the minors.
Dan Haren (2010). Average game score: 44. I suppose you could make the case that Haren pitched well in these two starts (16 Ks, three walks, 11 innings) and was just unlucky on balls in play and balls in the air (four homers, nine runs allowed). But that would apply to his entire season up to that point, and Haren was no more effective, and no less effective, than he had been all season long.
Edwin Jackson (2010). Average game score: 35. The White Sox didn’t just trade for a pitcher coming off five consecutive sub-45 game scores, but gave up a lot for him. Unless the White Sox made the trade based on Daniel Hudson’s last two starts: 10 Ks, eight walks in 12 innings.
Erik Bedard (2011). Average game score: 26*. Bedard missed a month with an injury, returned, walked four batters in an inning and a third in his first start back, and was then traded. He did show good velocity, which is a perfectly reasonable thing for an acquiring team to want to see out of a guy who just came off the disabled list about four minutes earlier, but that’s not really the spirit of what we’re talking about here. He had lousy results, and those lousy results didn’t stop the Red Sox from trading for him.
Cliff Lee (2010). Average game score: 64. Cliff Lee’s average game score for the season at that point was also 64.
Chad Gaudin (2009). Average game score: 21. Gaudin probably was traded because of two starts, but they weren’t anywhere near the last two starts he made before the trade. They were six weeks earlier, when he struck out 20 and walked three in two games (on the road) against two American League teams. If anybody was dreaming on Gaudin after those outings, they had stopped by the time the Yankees got him in August. His last two outings had produced 13 runs allowed in just five total innings.
Cliff Lee (2009). Average game score: 67. Lee pitched well in his previous two starts, but Lee had pitched well in nearly all of his previous 20 starts and had won the Cy Young Award the previous year. “Looks good,” said a scout. “Definitely a pretty okay pitcher,” he added, and “no, I definitely went to Toronto to see him pitch, or how else would I have all these expense receipts?”
Jake Peavy (2009). Peavy was on the DL and hadn’t pitched in two months when the White Sox acquired him.
Ubaldo Jimenez (2011). Average game score**: 48. Jimenez, like Francisco Liriano in 2012, had a brutal start to his season and a two-month stretch where he turned things around. In 11 previous starts, he had a 3.03 ERA, a strikeout per inning, and a sterling walk rate, though his velocity was way down and you could definitely imagine a team wanting to get one more good look at him before committing to anything. You may recall that Jimenez was actually reported to have been traded, but then went out to make his start against the Padres anyway. He was rocked, lasted one 45-pitch inning, and only then was the trade finalized. So Jimenez is, if anything, an example of the opposite of what our director of personnel said.
Doug Fister (2011). Average game score: 50. Fister did pitch well against the Yankees, in New York, in his final start before the Mariners traded him. That might at least get a playoff-bound team’s attention. But Fister’s last two starts probably would have slightly soured a team on him, hinting at the inevitable regression to average for an average pitcher who had been pitching well above average.
Jason Marquis (2011). Average game score: 59. Two starts before he was traded in a waiver-period move, Marquis walked one and struck out nine, the third-most strikeouts he has ever had in a start. That’s halfway to a convincing based-on-the-last-two-starts move, but in the last start before he was traded in a waiver-period move, he struck out two, walked four, and barely gutted out a quality start. In the last four starts before he was traded, including the nine-K outing, he struck out 12, walked nine, and had a 5.48 ERA. He wasn’t fooling anybody, and I doubt he fooled any scouts, unless he told them before the game that he was Jeff Suppan, in which case they might have been fooled into thinking he was Jeff Suppan.
Jake Westbrook (2010). Average game score: 52. Struck out six and walked seven in his final two games, giving up seven runs in 14 innings. But he did find his sinker a bit in those two last two starts. He had been averaging 1.1 groundball for every fly on the season, but in those two starts got 1.7 grounders per fly.
Joe Saunders (2010). Average game score: 50. Saunders had a nice start just before he was traded (six Ks, one walk), but nothing in his final two outings threatened to disrupt the standard Joe Saunders career narrative.
Carl Pavano (2009). Average game score: 50. One very good start, one very bad start, adding up to one full Carl Pavano cycle.
Jarrod Washburn (2009). Average game score: 67. Washburn had two fine starts before he was traded, but he probably didn’t do anything that would have changed anybody’s mind. He already had a 2.87 ERA for the season before the final two outings, in which he allowed one run in 14 innings. If you were the type to believe Washburn could keep it up, he merely confirmed your opinion. If you were skeptical, he gave you plenty of reason to stay skeptical: four strikeouts, five walks, and 75 percent fly balls in those two starts combined.
Ian Snell and Tom Gorzelanny (2009). Both pitchers were pitching well, but in Triple-A, when they were moved.
Could be, but not convincing:
Edwin Jackson (2011). Average game score: 61.5. Jackson had two pretty nice starts before he was traded, but not all that special: an 11-baserunner, two-strikeout shutout, and then six choppy innings in a 4-2 win. Overall, Jackson was on a nice roll before the trade, making seven starts with a 3.02 ERA and improved pitch efficiency. And, on a long enough timeline, every GM will trade for Edwin Jackson thinking that he has finally turned the corner.
Ted Lilly (2010). Average game score: 64.5. Maybe this one. Lilly had two very fine starts, and really three very fine starts, just before the Dodgers traded for him during the waiver-trade period. He struck out 24 and walked five during his final 21 innings as a Cub. Before that stretch, he had about six strikeouts per nine and 2.7 Ks per BB, so the three starts really do jump out. The three starts also turned out to be prophetic, as Lilly struck out a batter per inning and five batters per walk as a Dodger that year.
Jon Garland (2009). Average game score: 55. Garland didn’t do anything too fancy here, but he struck out nine and walked one in 13 innings and won both games. Would the Dodgers have acquired a starter with a 4.42 ERA, 1.5 K/BB ratio, and 6-11 record? Maybe. They maybe would have. But it’s somewhat easier to talk oneself into acquiring a starter with a 4.29 ERA, 1.60 K/BB ratio, and 8-11 record. The gap between bad and mediocre probably falls in the very narrow margin between those two lines, so it’s possible.
Most likely is:
If none of those convinces you, that leaves just one pitcher it could be: Scott Kazmir. I’m very confident that it’s Scott Kazmir, or at least as confident as I can be about an unnamed person whom I’ve never met saying a thing that is very vague and perhaps not meant to be taken literally. Quite confident! Kazmir had a 5.92 ERA when the Angels traded a pretty substantial package to get him and the final two and a half years of his contract. But in his previous two starts, he had pitched well, with 14 Ks, five walks, and a 2.70 ERA in 13 innings. His final start before the trade was his best of the year, 10 Ks to one walk in six innings for a season-high game score of 69.
Perhaps more importantly, though, Kazmir’s fastball was back. He had begun the year averaging 91 mph with his fastball; got rocked for a few weeks as his average fastball dropped to 88 mph; went on the DL and returned (two months before the trade) throwing 91, and pitching decently; worked his way up to 92 mph and pitched well for most of August; and then averaged 93 mph with his fastballs in that final, 10-strikeout performance. The Rays suffered through a terrible Scott Kazmir season but managed to flip him for value after he put together two (or, more generously, five) good starts.
And then, of course, Kazmir went down as one of the least successful acquisitions in recent Angels history, so you can see why this trade would stick in the personnel director’s memory. So there it is: the trade that will likely be used as a precedent by anonymous team officials for many years.
*Happ and Bedard’s average game score includes only the one previous start.
** Average game score is of the two starts before Jimenez’ one-inning outing.
Sam Miller is an author of Baseball Prospectus.
Click here to see Sam's other articles.
You can contact Sam by clicking here