What Should We Expect from Beckett? (Part 1)

John Tomase plays around with the Baseball-Reference Play Index to come to the following conclusion about Josh Beckett:

Virtually every pitcher to struggle like Beckett did last season not only was never the same, but almost to a man failed to produce a single, solitary above-average season thereafter.

Tomase makes a number of mistakes.

First, he uses Beckett's ERA as the benchmark. This, of course, would eliminate large amounts of the historical comparables. Such an ERA was all but impossible in the pitcher-friendly 1960s and 1980s, for example. In fact, Tomase's list contains 91 pitchers who threw at least 115 innings, of which 53 came in 1993 or later. Another 18 came between 1929 and 1940, the other extremely high-offense era in baseball history. But in the 30 years between 1950 and 1980, there were exactly eight such seasons, and just four more between 1980 and 1992. That's 12 seasons of 5.75 ERA or worse in a 42-year stretch.

Then there's the innings limit. Tomase chooses 115 because Beckett threw just 127 innings. Tomase discounts Beckett's season as "a disaster that can only partially be chalked up to injuries." This means that Beckett is being compared to many pitchers who put up relatively healthy seasons, while better starters who perhaps pitched 100 innings or fewer are out of the comparison. Was Beckett at 127 innings more comparable to a 180-inning healthy but crappy starter or a 90-inning injured starter?

Finally, he picks all pitchers in their 30s. Except Beckett was exactly 30, which means many of the comparisons had their horrible seasons five or six (or seven, or eight, or nine) years away from where Beckett is now.

In short, a 29-year-old starter from the 1960s who was limited by injuries to 110 innings and posted an identical 75 ERA+ would be off this list. But a 39-year-old starter who was uninjured, pitched 180 innings and posted an ERA over 6.00 would be comparable. That's silly.

So let's rework this comparison, seeking out pitchers with an ERA+ between 70 and 80, innings between 90 and 150, and an age between 28 and 32.

This actually gets us a much larger list: 205 pitchers, and the names are much different.

For example, there's Dennis Martinez, who posted an ERA+ of 80 or below for four consecutive seasons from 1983-86, before going on an incredible late-career run (130 ERA+ over his age 32-40 seasons). There's a guy named Cliff Lee, who posted a 72 ERA+ in 2007 at age 28. He's done OK since. There's Juan Guzman, who had a terrible age 27-28 before exploding in 1996 and having a couple more solid seasons beyond that. There's John Smiley, who had a 72 ERA+ over 105 innings in 1993, then rebounded for three more above average seasons.

Perhaps the best example of why Tomase's study is problematic is David Wells.

In 1992, Wells had a 76 ERA+, very close to Beckett's 75. He threw 120 innings, just shy of Beckett's 127. But Wells was 29, not 30, and his ERA in that more pitcher-friendly season was 5.40. Wells is arguably a more comparable pitcher to Beckett, based on these numbers anyway (they are obviously very different in other ways), than Jeff Weaver or Livan Hernandez or the others Tomase cites.

And Wells went on to post a 113 ERA+ over the next eight seasons and a 110 ERA+ over the next 13 seasons. Not to say Beckett will do that, but it's a significant data point that Tomase's flawed study misses.

Wells and Beckett share something else in common that many of these other pitchers don't: They were actually really good before their awful year. Wells posted seasons with an ERA+ of 153, 131 and 114 in the three preceding seasons. Beckett posted a 145, 115 and 122 (never mind the 138, 108, 118, 95 in the four years before that.

Without going through all 91 pitchers on Tomase's list or all 205 on mine, I can't say how many had a sustained run of excellence (or at least very-goodness) before turning in their stinkers. A few clicks of the more recent names on Tomase's list shows guys like this:

• Nate Robertson, who had one above-average season at age 28 sandwiched among three average-ish seasons before collapsing at age 30. 
• Livan Hernandez, who had more seasons below average than above before falling into a pit at age 33.
• Jeff Weaver, whose season meeting Tomase's criteria at age 30 was actually his third such season to do so.
• Rodrigo Lopez, whose 77 ERA+ at age 30 had been preceded by an 88 at age 29.
• Russ Ortiz, who'd had some good seasons before utterly failing at baseball at age 31, but whose best season would have tied Beckett's third-best season. Beckett had five seasons better than Ortiz's second-best season. And on top of that, Ortiz's collapse was to an ERA+ below 70 (an ERA near 7.00), not nearly as bad as Beckett's.
• Tomase notes that Darryl Kyle is arguably Beckett's best hope (minus the untimely end, of course). Like Beckett, Kyle had an awesome age-28 season (156 ERA+), but dropped to 100 at age 29 and 88 at age 30. Then he rebounded to post a 125 ERA+ over the next two-plus seasons before dying. But even Kyle was not as good a pitcher as Beckett. His career ERA+ entering his breakout season was 91. Beckett's entering 2007 was 111.
• Greg Harris, who never posted an above-average ERA as a starter.
• Darren Oliver, whom Tomase cites, was only league average in his three seasons as a full-time starter before falling off a cliff at 30 and resurrecting his career as a reliever.
• Jack Morris, whom Tomase cites, was 38 at the time. Is that relevant? In three of the previous five seasons, Morris had been below average.
• Ramon Ramirez, whom Tomase cites, had indeed had a number of great years and was just 32, but he also was in his fourth consecutive year of missing significant time because of injuries. He managed fewer than 400 innings in those four years and had thrown just 20 the previous year and 100 the year before that. Ramirez's career wasn't over because he posted a 6.13 ERA at age 32. It was over because he couldn't stay healthy. The same may happen to Beckett, but that's an entirely different argument.
• Hall of Famer Robin Roberts was 34 at the time and had posted an ERA+ of 97 over the previous five years, but even so, he garnered a 119 ERA+ over the next four seasons.

I could go on. But I can't find a single pitcher on Tomase's list that we could even remotely consider a comparable pitcher. A huge number were so old during their dropoff seasons as to be irrelevant. The rest appear to never have been anywhere near as good as Beckett has been. When Darryl Kyle is the closest comp, you don't have much of a comp at all — and Kyle is actually a positive story (at least, in this context).

Which makes this statement pretty much as untrue as untrue can be:

They’re paying him to anchor their rotation.The same was once said of lots of good pitchers, like Morris, Cone and Stewart. Or Al Leiter, Pat Dobson, Ramon Martinez and Rick Sutcliffe.

Then they had seasons like Beckett’s 2010 and their careers were over. It’s hard to fathom that Beckett could experience a similar fate, but for whatever it’s worth, the game’s entire history is working against him.

Morris (age 38), Cone (37), Stewart (37), Leiter (39), Dobson (35), Martinez (see above) and Sutcliffe (37) are not remotely comparable. None of them.

Which means the game's entire history does not appear to have very much to say about how Beckett will perform in 2011. Maybe we can do better in another post.