Workhorse pitchers

Jury duty update: I got selected to serve on a case that begins tomorrow and is expected to last for three-to-five days, plus the deliberations. I'll continue to blog when I can. Today, with a one-day reprieve, I'm back to work and will go to tonight's Orioles-Yankees game.

A few days ago, before he even knew of my schedule, NaOH offered me some research he had performed on the always provocative topic of starting pitchers, pitch counts and old-school work ethics. I told him I'd love for him to write it up as an item here, if he was so inclined. He agreed, so I thank him the meat to today's post. I think it's pretty, pretty good:

First and foremost, my thanks should be extended to two people. Tom Tango was kind enough to review this for accuracy, and he also offered a couple of suggestions for refinement. The availability and cooperation of someone like Tango is a tremendous asset for others, and the amount of time, effort, and work he freely shares is remarkable.

Of course, my thanks also go to Ken. While my timing may have been fortuitous since he's busy with jury duty, his receptiveness to this unsolicited piece speaks highly of him.

Lots of people have been debating Jerry Manuel's removal of Johan Santana from a 0-0 game after 8 well-pitched innings and only 105 pitches from Johan. Since the Mets' bullpen proceeded to lose the game in the ninth, the move has fanned the flames for those in the anti-Manuel crowd. Frankly, the game outcome and commonplace nature of this move, along with Manuel's managerial efficacy, don't interest me.

But I am interested in the broader topic which emerged in the comments here. What makes a pitcher a workhorse? Who in today's game has established themselves as one? The unspoken hope, of course, is that a workhorse starter is also a premium-level performer, not just an innings-eater.

I did some research, and here's what we have: There are just 18 pitchers who have at least 6 years of service time and have accumulated 600 IP as a starter since the beginning of 2007. Considering how we envision a durable pitcher being a guy who gives a lot of innings, this is really only looking for pitchers who have averaged 180 IP over the last 3.33 years.

Note that Cliff Lee technically just missed the cut by 7.2 innings, but I decided to include him for his reputation as a workhorse and for his proximity to the 600-inning threshold, so this bumps the group total to 19 pitchers.

These 19 pitchers are basically the cream of the crop in terms of well-established, durable starters over the last three and one-third seasons. Based on how much these pitchers are paid, experienced durability does not come cheaply. The average 2010 income of these pitchers is $13.59MM.

Let's look at how those pitchers have fared since the beginning of 2007.

2007–May 2010 Workhorses

Ordered by IP/start

Halladay and Sabathia are well above everyone else in terms of innings pitched and quality. Of the rest, Santana stands out for his exceptional quality and inning per start, but he's clearly below the top two by virtue of having thrown about 105 fewer innings. Even if he hadn't had surgery last year which cost him eight or nine starts, he'd still be about 50 innings behind those two. For Santana, that's the equivalent of about 7½ starts.

Among the remaining 16 pitchers, let's eliminate one from the quality discussion: Livan Hernandez. Despite his impressive innings total, his performance is easily the worst of all these pitchers. From this group, he's the only pitcher with a below average ERA+, and it's not even close to average at 86. Presumably, teams signing him recognize they will get a large amount of low-quality innings, and this likely accounts for him being the only pitcher listed who makes less than the league average. The Nationals, entering the year in need of an innings-eater, signed him for $900K.

That leaves us with 15 pitchers. From here, it's pretty easy to see which are the better ones simply by looking at who outperforms the group averages. Take away the four outliers – Halladay, Sabathia, Santana and Livan – and the quality and value drop. Here's the average numbers for those 15 pitchers:

Presuming everyone remains healthy, the best remaining pitchers are Lackey, Buerhle, Beckett, Oswalt, Lee, and Lilly. Of those five, though, only Buerhle has avoided the disabled list since 2007.

Considering the average performance of these 15 pitchers, the idea of an established workhorse starter in today's game really means a guy who throws a decent number of innings with a performance level which is pretty good but not spectacular. For some comparison, the 5 pitchers to win the last six Cy Young Awards from 2007-2009 have had an average WAR of 7.1 and an ERA+ of 170. From our group, only Sabathia and Lee have won a Cy Young during this time, though Zito (2002), Halladay (2003) and Santana (2004) have won previously.

And for those who think that pitchers today are coddled, well, it depends on how you look at things. In terms of total innings pitched per year and the number of starts per season, yes, pitchers today do have a lighter workload. But in terms of pitches per game, not much has changed in terms of the averages. (Really, if you think the idea of limiting pitchers to around 100 pitches is new, go read that last link.)

Of the all-time, top-20 pitchers in innings pitched, 10 of them are recent enough that most readers here probably remember at least a portion of their careers. We can use Tom Tango's pitch count estimator to determine how many pitches these 10 threw per start. Tango's formula is actually quite simple and remarkably accurate:

(Batter Faced x 3.3)+(K x 1.5)+(BB x 2.2)

To demonstrate how well it works, I ran it against the 19 pitchers above, all of whom have actual pitch-count data available for 2007-2010. Fourteen of the pitchers' actual per game averages were within one or two pitches of what Tango's formula estimated. Five matched identically. Tango's method overshot Halladay's average by four pitches and Zito's by six. Considering what our eyes have told us in the past few years about these two – Halladay is uncommonly efficient and Zito has been uncommonly wild – the discrepancies make sense.

So let's see how some of the recent legends of innings pitched did during their careers.

Recent-Vintage All-Time Workhorses

Ordered by IP

Those ten pitchers averaged 109 pitches per start. The top-ten pitchers by total innings from the 2007-2010 list averaged 102 pitches per start, a difference of about two batters.

I imagine most people think the average for these legends can't be correct considering we often hear about how much more pitchers used to throw. In addition to many of them pitching on four days of rest, many of us probably remember recent pitchers like Clemens throwing more than 150 pitches in a start. But that's where another change has occurred: The extreme numbers for pitch-count totals have narrowed.

If you read the linked piece above about the mythical 100-pitch count limit, you already understand this: Instead of pitchers throwing 140 pitches in one start followed by 70 in the next – for an average of 105 pitches – pitchers today are more likely to have pitch-count extremes around 80 and 110 pitches, again for an average of 105 pitches.

On top of that, we live in an age of specialization. Teams are carrying at least seven relievers, and they're being employed for a variety of reasons. Key among the rationales are matchups – e.g., lefty vs. lefty – and the increasing amount of evidence showing that starters are less successful after each successive time through a batting order. As Mitchel Lichtman found, here's how hitters perform against a pitcher as their in-game plate appearances increase:

Even if your team has a pitcher of Halladay's or Sabathia's quality, the advantages to bringing in a fresh arm are tremendous, both in terms of a specific game's outcome, but also for the decreased likelihood of an expensive investment coming down with an injury.

More importantly, for the original question of what makes a workhorse, in today's game it appears to be a pitcher who gives his team about 6½ innings per start. But not all guys pitching this amount seem worthy of also being called an ace.

At this point, there only seems to be 2¾ established aces who are also workhorses. Unless you're following the Phillies, Yankees or Mets when one of those three pitch, it's important not to confuse your team's best pitcher with being an ace and being a workhorse. The odds are he's not both.

--Thanks so much, NaOH. It's me again. Big win for the Mets last night, saving some face after Monday night's fiasco. With Santana pitching the rubber game today, going for the Mets' first road series victory of the season, I was curious how Santana had done so far this season in rubber games. I could remember two off the top of my head, the first one terrible and the next one great.

Here's how he has done in the final games of series, in which the series victory was up for grabs:

Five starts, 29.2 IP, 19 ER, 5.77 ERA, 18 K, 10 BB, 2-2 with a no-decision.

More than anything, it shows how your overall numbers can be impacted by one horrendous outing. But I was curious.

--Luis Castillo appears headed for the disabled list. The bad news for the Mets, long term, is this increases the chance that Alex Cora will vest his $2 million option for 2011. Cora already has played in 30 games, and he needs to play in a total of 80 to earn big bucks next year for being grindy, gutty and leadershipy.

--Carlos Beltran appears to be making progress. Until Beltran plays in a minor-league game and starts his rehabilitation clock (20 days), I'd be very hesitant to even start projecting when he might return.

--Jerry Manuel addressed the recent struggles of Ryota Igarashi. In the same notebook, David Lennon mentions htat the Mets are optimistic Oliver Perez will soon accept a demotion to the minor leagues.

If he does, great for the Mets, and perhaps good for Ollie, although I'm not convinced it will make much of a difference at this point. But as this Perez debate rages on, there are two points in the yakosphere (copyright Neil Best) that I felt compelled to address:

1. "It's easy to say the Mets should cut Ollie when it's not your $20 million being wasted." Well, yeah. But it's also not my $400 million coming in. The reason I feel justified in calling for Perez's release _ although, again, I'm fine with the minor-league demotion if he accepts it _ is that, first of all, there are precious few indications that Perez can be a useful starting pitcher again; the drop in his velocity is alarming, and he'll never be a crafty pitcher like fellow lefty Jamie Moyer, or former teammate Pedro Martinez.

But just as important, as we discussed yesterday, there is precedent for such a money dump, and the Mets should be very familiar with it, as they took on Gary Matthews, Jr. only after the Angels ate $21.5 million over two years.

2. "The Mets' clubhouse is weak and can't handle the trauma of Ollie hanging around." Puh-leeze. Is the Mets' clubhouse the toughest clubhouse to ever inhabit a big-league stadium? Probably not. But what's the difference between the toughest and least tough clubhouse, ever? How can we quantify it? And how can it compare to, say, the difference between the best pitching staff and the worst pitching staff, ever?

The Mets' collapses of 2007 and 2008 occurred because they didn't have enough pitching, both years. All other explanations are ancillary.

--On Twitter, the New York Post's Mike Vaccaro has been soliciting suggestions for poetry depicting the Mets' starting rotation, in the spirit of "Spahn and Sain and pray for rain." I went with, "Johan and Mike, and...can you also pitch, Ike?"

--Out in the Bronx, Javier Vazquez pitched very well and Curtis Granderson homered off Orioles lefty starter Brian Matusz. So while it's natural to start evaluating Brian Cashman's 2009-10 offseason - with very low grades, to date - it's kind of early to do so.

--Robinson Cano could've been a Ranger, Jim Baumbach recalls, and Jim notes the irony of Alex Rodriguez (who could've cost the Yankees Cano) playing an integral role in Cano's development.

--Jorge Posada, always honest, shared his frustrations over not being activated off the disabled list yesterday.

--Joba Chamberlain pitched well, easing (for now) fans' stress over his up-and-down season.

--Mark Teixeira thinks he'll play tonight, after leaving during last night's game with a bruised left foot.

--Nice piece by Alex Speier on Adrian Beltre, who has been a very good Red Sox signing so far. My sense just a few weeks ago was that Red Sox Nation was not happy with Beltre, and I didn't quite understand why. Perhaps that vibe is changing as the Red Sox continue to climb back into the race.

--Talk to you later from Yankee Stadium.