Worrying About the Recent Performance of Hamels and Cueto

Between the All-Star break and the trade deadline, most starting pitchers make only a couple starts. For starting pitchers who could be on the move, the small timeframe places those starts under a microscope for those anticipating a trade. Teams wanting to trade for the starter want to ensure that they are getting a pitcher at the peak of his abilities to help with the last few months of the regular season and potentially the playoffs. For the best two starting pitchers on the market, Johnny Cueto and Cole Hamels, recent performance has begun to raise questions about their trade value. Whether recent performances have hurt their trade value is debatable, but we can look over the past few seasons and determine whether other pitchers have gone through a dip in performance prior to a trade and compare that performance after the trade.

Some have asserted that Hamels’ last two starts hurt his trade value. While he has given up 14 runs in less than seven combined innings in his last two starts, his underlying stuff (which is great) has not been affected, there was an extended layoff between the two starts due to the All-Star break, and in the outing prior to those starts, Hamels threw seven shutout innings with six strikeouts against no walks. Hamels does have a 3.91 ERA this season, but his 3.37 FIP is still in line with his stellar career numbers.

Many of the same questions about Cole Hamels have also been asked about Johnny Cueto after lasting just four innings and giving up six walks in his last start, although the start of the game was delayed by rain. Like Hamels, this start followed another un-Cueto like start when he pitched just five innings and managed to strike out only two hitters against the Miami Marlins. Also like Hamels, there was a layoff in between the last two starts and the outing before the last two starts was fantastic, as Cueto pitched a shutout against the Washington Nationals, striking out 11 batters against just one walk. Despite the last start, Cueto is still having a very good year with his 2.79 ERA and 3.16 FIP both near his career bests.

Now compare those recent results to the most recent three appearances by Scott Kazmir, another potential trade acquisition without anywhere near the track record of Hamels and Cueto.

	IP	ERA	FIP	xFIP	SO	BB
Scott Kazmir	19.1	0.47	1.87	2.95	16	2

Has Kazmir enhanced his trade value relative to what it was a month ago? He isn’t more valuable than Hamels or Cueto, but sometimes it makes sense to weight recent performance more heavily when predicting future performance. How much a few starts should outweigh a considerably longer track record is up for debate. That said, we can look at pitchers who have been in this position before and see how they have performed prior to a trade and whether that performance carried over to the teams that have acquired them.

Using various trade deadline recaps since 2010, I identified 22 starting pitchers who were traded at or near the trade deadline. Those players varied in talent, going from David Price and Jon Lester last season to Jake Westbrook in 2010, but every pitcher was a starter for a bulk of the first half, had at least three starts near the deadline before being traded, and then subsequently started for the acquiring team. I removed those, like J.A. Happ, who after being traded to the Blue Jays, was put in the bullpen for a few weeks before starting, and Jason Marquis, who in 2011, broke a bone in his leg on a batted ball in his third start with the Arizona Diamondbacks.

I took a look at the group as a whole using FIP, measuring all of their starts prior to the three starts before the trade, those three starts before the trade, and all of the starts after the trade. The fairly simplistic graph below shows the average performance.

The numbers do go up in the final three starts before the trade and then come back down with the new team. I should note that it is possible a few outlier performances from Jake Peavy and Bud Norris in 2013 slightly skew the results — the median is actually 3.63 — although removing the top- and bottom-three still results in a 3.88 FIP.

As to why pitchers have produced a higher FIP in those final three starts, there probably isn’t a real satisfying explanation. Perhaps it’s the prospect of being traded. Perhaps it’s the unusual usage patterns that occur around the All-Star Game. Perhaps (or probably) it’s random.

In any case, pitchers were generally pretty consistent throughout the season. Only four of the 22 pitchers were more than one run different with their new team than they had been with their old team (save the last three starts), and 10 of those changed less than 0.4 runs with their new team. Focusing in on the the three-start period, where we might expect to see wide variability due to the very small sample, most pitchers were not too far from their FIP up to that point in the year. Twelve of the 22 pitchers posted a FIP within one run of their early season numbers during those three starts. Of those pitchers, all but two continued to pitch roughly the same for their new teams. Only Jake Westbrook, who had a poor 4.66 FIP with Cleveland in 2010 before his trade to the Cardinals, and Jake Peavy, who had an even worse 4.90 FIP with Boston last year, had big differences from their new team to their old team after posting similar numbers to their seasonal numbers leading up to the trade.

At the edges of the spectrum, we have six pitchers who were considerably worse leading up to their trade and four pitchers who were considerably better. Below is a chart with the players that were significantly worse. (Note: FIP is used in every case.)

	SEASON	BEG	LAST 3	LAST 3 MINUS BEG	NEW TEAM	NEW TEAM MINUS BEG
Bud Norris	2013	3.48	6.40	2.92	3.82	0.34
Jake Peavy	2013	3.73	5.97	2.24	3.79	0.06
Ryan Dempster	2012	3.18	4.65	1.47	4.08	0.90
John Lackey	2014	3.35	4.78	1.43	4.27	0.92
Cliff Lee	2010	1.84	3.12	1.28	2.99	1.15
Zack Greinke	2012	2.40	3.49	1.09	3.89	1.49

Lee and Greinke had pitched so phenomenally, they were unlikely to sustain that success over the course of a season anyway. Jake Peavy and Bud Norris went back to pitching just like they had after a rough streak, while only Ryan Dempster and John Lackey could not get back on track.

At the other end, we have pitchers who “got hot” right before their trade:

	SEASON	BEG	LAST 3	LAST 3 MINUS BEG	NEW TEAM	NEW TEAM MINUS BEG
Jon Lester	2014	2.77	1.75	-1.02	3.13	0.36
Paul Maholm	2012	4.36	3.18	-1.18	3.76	-0.60
Ted Lilly	2010	4.69	3.38	-1.31	3.98	-0.71
David Price	2014	3.11	1.78	-1.33	2.44	-0.67

While a sample of four is hardly definitive, three of the four pitchers who pitched significantly better leading up to the trade continued to pitch better than they had earlier in the season, although none of them could replicate their three-game run of success. While there is always going to be cause for concern with pitchers, and the performances of Cueto and Hamels are under the microscope, if recent history is any indication, absent injury, the small downturns in performance is not a lot to worry about. Besides, over their last three starts, the two aces have not even pitched that poorly. Here are the numbers for Cueto, Hamels, and a handful of other starters who could be on the move. (FIP is used in each case here, once again.)

	1st Half	Last 3 GS	Difference
Mike Leake	4.13	2.36	-1.77
Scott Kazmir	3.16	1.87	-1.29
David Price	2.83	2.47	-0.36
Cole Hamels	3.39	3.06	-0.33
Johnny Cueto	3.04	2.73	-0.31
Jeff Samardzija	3.40	3.48	0.08
Dan Haren	4.12	4.56	0.44
Mat Latos	3.48	4.36	0.88
James Shields	4.14	5.43	1.29
Yovani Gallardo	3.53	5.21	1.68

Perhaps another poor start from Hamels and Cueto could render them less attractive to contending clubs, but right now, their recent history is not even that discouraging. Kazmir and Leake have pitched well recently, and perhaps a contender might be more willing to take a chance on those pitchers while shying away from Shields and Gallardo. Jeff Sullivan just wrote about the success of the team that trades for ace. From a player-performance perspective based on past trade deadline deals, trying to predict two months of performance at the end of the season based on three starts in July instead of the overall track record is about as useful as one might expect, which is to say, not very.