Seidnotes: WARP Speed

Last week, I wrote an article titled “The Rodrigo Lopez All-Stars” to uncover how many pitchers have thrown 200 or more innings at or below the replacement level after pitching as poorly over the prior three seasons. Lopez pitched exactly 200 innings and finished the 2010 season with a 0.0 WARP; the figure should have been expected given his track record, but the innings total proved surprising. While researching for that article I found the worst WARP totals for pitchers who similarly logged 200 or more frames. Most of the pitchers came from past eras, as it is now far less common for a bad pitcher to record that high of an innings total. Actually, it is far less common for any pitcher to record that high of an innings total these days, which made Lopez’s stats so perplexing.

One of the names on the list of pitchers with the worst WARPs in 200 or more innings was Denny McLain, who produced 1.5 wins below replacement in 1967, a year before famously winning 31 games for the Tigers and adding 8.9 wins above replacement. In the comments of my article, the question was raised as to where that spike in WARP would rank on an all-time scale. After all, it cannot be too common for a pitcher to fluctuate that wildly. Improving from one or two wins to five or six might make sense in the right context, but to go from costing a team over a win and a half to adding approximately nine? That would be like being Pauly Shore’s sidekick in a movie and then landing a starring role in a Scorsese film.  

Going strictly from one year to the next—I’ll explain more after the table—here are the biggest positive improvements in WARP for pitchers dating back to when teams were named the Beanaters and every fourth player was named Red or Lefty. Note that I removed pitchers whose jumps were directly attributable to vast playing time increases:

These were the only five pitchers that improved by more than nine wins from one year to the next while accruing similarly high amounts of playing time in both seasons. Not so surprisingly, McLain’s ascension tops the list. Examining the list, Perry’s improvement looks strange on the surface given that he had a 2.76 ERA in 1971, but league ERAs were also much lower in his heyday. In that 1972 season, Perry posted a 1.92 ERA and struck out 234 batters in 342 2/3 innings, allowing just 17 home runs. Gibson’s 1967 season was decent, but in 1968 he set the record for the lowest single-season ERA for a starting pitcher, thereby explaining the reason for his sharp uptick.

McNally pitched less in 1967 than in 1968, but threw so poorly in that first year that the improvement still merits inclusion. He improved from a 4.54 ERA to a 1.95 mark and saw his peripherals drastically improve: 70 K and 38 BB in 1967, 202 K, and 54 BB in 1968. Carlton, much like McLain and Gibson, had an historic second season in the span. He spent 1971 with the Cardinals, producing a 3.56 ERA with mediocre peripherals. The next season, he won 27 games for the Phillies with a 310/97 K/BB ratio and a 1.97 ERA. That these seasons were historically good shouldn’t be surprising, as the WARP figures in those years would need to be notably high to garner such a big improvement. Carlton’s 1972 ranks as the sixth-highest WARP for a pitcher; Gibson’s 1968 season ranks 10^th; and Perry’s 1972 ranks 18^th. McLain and McNally experienced such drastic improvements not only because they pitched well in the second year but because they pitched so poorly in that first season.

How about the opposite query? Here are the biggest declines in WARP from one year to the next:

Interesting that Carlton shows up on both lists, as he went from a mediocre season to one of the 10 best of all-time, and then back down to slightly above average. Of course, everyone would see their numbers decrease as compared to a 13 WARP season, but it stood to reason that Carlton may have been able to remain elite in 1973 after such a tremendous 1972 campaign. Turnbow came out of nowhere to take over the Brewers closer role in 2005 after toiling in obscurity with the Angels for a few seasons, adding 5.8 wins above replacement, which led all senior circuit relievers and ranked second only to Mariano Rivera. The next season, he posted the worst WARP in the entire league. Oh, how extremes are fun!

But going from one year to the next probably is not the most optimal way to answer a question like this. Fluctuations across two seasons are one animal, but it is less likely that a pitcher will improve or decline substantially in a season compared to his production over the level established during the prior three seasons. Framing the question in that manner, where does McLain’s WARP spike rank? I turned to the database, added the WARP totals over a three year span, computed the average wins above replacement during that span, and then compared that average to the fourth year of each span. Here are the five biggest improvements: