clock menu more-arrow no yes

Filed under:

Saber 101: Why pitcher wins and saves are obsolete statistics

New, 108 comments

HookSlide takes a semi-serious look at three traditional pitching stats that should probably be put out to pasture

Rick Osentoski-USA TODAY Sports

I don't consider myself to be an "old-school baseball fan." Sure, I grew up in an age when there were zero permanent on-screen graphics during games, and yeah, I remember a time before pitch counts were invented, and fine, ok, I go to ballgames in a suit and tie with a fedora on my head and a pipe in my mouth, but dagnabbit, who stole my newspaper?

Oh well. Let's face it, fellow veteran baseball fans: in terms of baseball statistics and the way we measure player performance, the times, they are a-changin'. Honestly, that's probably a good thing. It may have been easier to gauge a batter by his batting average, home runs, and RBI, and it may have been far simpler to judge a pitcher based on win-loss record and ERA, but it wasn't really a very accurate way of looking at things. We might as well grade a pitcher based on his height, or size up a batter based on whether he "looks like a ballplayer" or "has a certain look in his eye" because he "has the will to win," but this is a sure path to getting a job as the play-by-play announcer for the Chicago White Sox.

It's time to improve our understanding of pitcher performance, so here are three traditional pitching statistics that need to be packed up and left in the clubhouse before the next game.

Pitcher wins

This is the easiest stat to discredit, mostly because I have photos of it on Spring Break in Key West from three years ago. We've all watched members of the Tigers starting rotation toss absolute gems, only to take an "L" on their stat sheet because the offense went MIA and provided no run support. A more useful stat, if you're in a hurry because of a "fifth inning, out of beer" situation, is the quality start statistic. A pitcher earns a quality start if he throws at least six innings, and gives up no more than three earned runs. If he can do that, he should be in line for a win, and if he doesn't get the win after a quality start, there's a good chance the problem wasn't with his performance.

Here's an interesting experiment: get a package of Mentos and a bottle of Diet Coke, vow never to mix them together because you are not an idiot, and compare a pitcher's win/loss percentage to his quality start/non-quality start percentage. The difference can sometimes be enlightening, especially with the help of recreational drugs. Justin Verlander's 2013 looked awful from a win/loss standpoint (he went 13-12), but his quality start performance was a solid 13 percent better. If the baseball gods were fair and not total jerks, his win/loss record would have been 22-12 that year.

By contrast, Max Scherzer looked good in 2014, posting an 18-5 win/loss record, but his quality start statistics were a full 12 percent worse. By that measurement, he went 22-11, which is why the bum is playing in Washington now. (You probably want to point out that 22-11 is actually better than Verlander's 22-12 quality start numbers in 2013, but clearly, you have a lot to learn about letting emotions cloud your judgment.)

Tigers Top Ten Starters by Quality Starts

Top 10 Tigers Starters by QS

Saves

This particular statistic, long held to be the true litmus test by which closers are measured, has recently been exposed by scientists and historians as — to use their technical language — "a really crappy stat." The rules governing the save are long, complicated, and arbitrary. To qualify for a save, a pitcher must a) finish the game, b)not also be in line for the win, c) enter the game with a three-run lead or less or, d) enter the game with the tying run at least on-deck, e) know at least 85 percent of the words to R.E.M.'s "It's the End of the World" (excluding the words "Leonard Bernstein," because, come on).

The save, of course, in no way reflects how effectively a closer does his job. What you want out of a true ninth-inning specialist is that elusive "lights out" quality, a guy who can shut the door on the opposition quickly, painlessly, and with a minimum amount of stress. For example, Joe Nathan amassed 35 saves in 2014, good for fifth-best in the American League, whereas Dellin Betances only picked up one save for the Yankees. Who would you rather have closing games for the Tigers? Correct, the answer is "Wade Davis," who pitched 72 innings and only allowed eight earned runs, despite only being credited with three saves.

A far better statistic for a closer is WHIP, or "walks and hits per innings pitched," or "for the love of Shiva, stop putting the tying and go-ahead runs on base." Allowing baserunners is precisely what you do not want a closer to do, because this makes us stressed out and eat more pretzels in the ninth inning, and we said we were, by god, not going to eat after 8:00 p.m. anymore.

Who was one of the most troublesome relief pitchers for the Tigers in the past few years? Our old pal, Phil Coke, and his WHIP reflects the reason why: in 2014, he allowed 1.5 baserunners per inning, which also happens to be the exact same WHIP that Joe Nathan posted that same year.

Compare this to a true Tigers legend like Willie Hernandez, who in 1984 posted a WHIP of 0.90. Red-hot relievers in 2014 like Betances, Davis, and Andrew Miller posted WHIP stats of 0.77, 0.84, and 0.80, respectively. Incidentally, the best Tigers relievers since they started locking down division championships in 2011? Joaquin Benoit in 2013 (WHIP of 1.03) and Drew Smyly in the same year (also 1.03).

Tigers Top Ten Relievers by WHIP

Top 10 Tigers Relievers by WHIP

ERA

The earned-run average is not a terrible statistic, per se, but it does make for an awful business model. It's like some kind of reverse pyramid scheme, where the responsibility for all of the talents or failures of the nine-man collective filter down and land in one man's lap. A pitcher may have an ERA over 4.00, but that could be because his catcher calls a terrible game, or because his outfield is staffed by a church softball league (and is that ... yep, the left fielder actually brought a cooler of beer with him onto the field).

A better statistic here is FIP, short for "fielding-independent pitching." It attempts to take everything out of the equation except what the pitcher himself can control, and nowhere was the value of FIP made more evident than in the case of Rick Porcello.

Porcello, you'll recall, is a ground-ball pitcher. In 2013, his infield was patrolled by Cabrera at third, Peralta at shortstop, Infante at second, and Fielder at first. The only way this crackerjack squad was going to stop ground balls from getting through the infield was if those balls were hit directly at them. Even the Comerica Park statues in centerfield were watching those games going, "Wow, those guys don't move much." And so, Rick Porcello's ERA for the year was an unflattering 4.32.

But, AHA! (Literally, "voila!") His FIP for 2013 was 3.53, which suggested that, with an infield behind him not made up of stationary deck chairs with baseball gloves sloppily taped to the seats, his ERA might actually be pretty decent.

The next year, the Tigers improved their defense a bit by putting Ian Kinsler at second, Nick Castellanos at third, and Andrew Romine/Eugenio Suarez/Danny Worth/Alex Gonzalez/Jerry from Accounts Receivable at shortstop, and magically, Porcello's ERA for 2014 dropped to 3.43 — which is almost exactly where his 2013 FIP suggested he ought to be.

(Yes, this same principle works in reverse as well. The Kansas City Royals had one of the best defenses in the American League in 2014, which made their starting pitchers look a bit better than they actually were. James Shields, for instance, posted an ERA of 3.21, but with a more league-average defense behind him, his FIP suggested that his ERA should have been closer to 3.59.)

Tigers Top Ten Starters by FIP

Top 10 Tigers Starters by FIP

In conclusion, I would say to my fellow old-school fans: fear not the winds of change. Some of these newer metrics are actually an improvement over the traditional stats — except for RE24. That stat can go straight to hell. You can safely abandon pitcher wins, reliever saves, and ERA stats, but never under any circumstance should you surrender suits and pipe tobacco at ballgames. Some traditions are timeless.

For discussion:
  1. With Jose Iglesias back at shortstop to complement Ian Kinser at second, does losing Rick Porcello to the Red Sox make you want to absolutely sh*t a chicken? Explain using Haiku.
  2. Write a mathematical formula expressing how much you miss Phil Coke in proportion to the amount of acid you ingested in the late 60's
  3. Have you ever actually smoked a pipe at a baseball game? If so, can we hang out sometime?
  4. Do you remember Jerry Don Gleaton? Prove it.