Quality Starts and George Bush
Posted By Jeff Fletcher on June 29, 2009 11:17 am
It would seem to me that the Quality Start has been around long enough that people like me wouldn’t still need to be defending its value to old-time baseball curmudgeons. Yet, I found myself wasting a solid half hour of my morning on Sunday arguing with Dave Feldman. Feldman is a pretty smart guy, one of the Bay Area’s official scorers, a stat guy for A’s television broadcasts. He knows a lot about baseball.
Which is why it’s fun for me every once in a while to tell him he’s wrong.
Feldman’s argument, like that of most traditionalists, is that the QS is BS because you can get a QS by pitching six innings and allowing three earned runs, which is a 4.50 ERA. A 4.50 ERA, he says, is hardly “quality.”
That is just plain silly. He’s using the minimum standard for the statistic and applying that to the whole thing. It’s like saying “George Bush was a bad President, so we shouldn’t have a President any more.”
Using Baseball-Reference‘s invaluable Play Index, I did some research. Since the start of 2008, there have been 3,449 QS, and only 11 percent of those have been the minimum 6IP-3ER.
To me what makes a QS “quality” is that if a starting pitcher meets that minimum standard, he has given his team a pretty good chance to win the game. How good? Over that span, the winning percentage of teams whose pitcher had a QS was…
.674.
Your team is more than twice as likely to win when your pitcher has a quality start. Even at the minimum quality start, that 6IP-3ER game, the winning percentage was .486. So at worst, in that very bottom 11 percent of the group of QS, your team is just about break even. At worst.
(As an aside, I wouldn’t mind seeing the QS changed to three runs, not three earned runs. You obviously haven’t given your team a very good chance to win if you give up six runs, but only three are earned. If you adjust it to 6IP and 3R, the winning percentage goes up to .688.)
I also contend that the pitchers who get the most quality starts happen to also be the best pitchers, the guys who win the most games and have the lowest ERAs. The guy the purists hate, someone who racks up minimum quality starts, just doesn’t exist. (Well, he does, but he’s very rare. More on that later.)
Here is a list of the pitchers with the most quality starts since the start of 2008. You’ll see the leaders are Johan Santana, Tim Lincecum, Dan Haren, Zack Greinke, Cliff Lee, Chad Billingsley, Mark Buehrle, Roy Halladay, CC Sabathia and Ted Lilly. Funny, that looks like a list of pretty good pitchers to me. Also, if you look at their ERAs, the highest ERA among the top 20 is Ted Lilly’s 3.87.
ERA is still the best number, in my opinon, for telling you how effectively a pitcher pitched. But the QS percentage adds the consistency factor. Look at these two pitchers’ numbers since the start of 2008…
| ERA | Run Support | QS% | |
|---|---|---|---|
| Pitcher A | 3.49 | 4.9 | 67.4 |
| Pitcher B | 3.44 | 5.1 | 56.1 |
Which pitcher’s team won a higher percentage of his starts? Is it Pitcher B, who had a lower ERA and got more runs of support? Heck no, it’s pitcher A. Pitcher A is the Angels’ Joe Saunders (32-14, .696 in his starts) and Pitcher B is the Reds’ Edinson Volquez (26-15, .634). It means that Saunders was more consistent than Volquez, and his team won more. (Yes, I know the Angels are better than the Reds, but since we’re just comparing them on the day these two guys pitch, and we’re taking into account the run support on those days, the teams are comparable.)
On the other side of the coin, there are a few pitchers out there who have bad ERAs, but give you an unusually high percentage of QS. Jon Garland is one of them. Over the past two years he’s got a 4.94 ERA, but a 57.4 QS%. His teams have a .553 winning percentage in his starts, which is pretty darn good for a 4.94 ERA. In fact, it’s probably why he’s been allowed to keep taking the ball every fifth day with an ERA that high.
My point is that the QS tells you something. It doesn’t tell you everything, but no stat does. If you use it in conjunction with ERA, you’ll get a pretty good indication of what a team’s winning chances are when a certain pitcher starts, and winning games is what matters.
As with all stats this one doesn’t really say anything in a vacuum, but when looked together with other stats like ERA, walks to strikeouts, BA of balls put in play, total innings, etc you get a fuller picture of effectiveness. This stat is just another puzzle piece in figuring that out.
Although Feldman has some flaws in logic as you point out, it would be interesting to know what the average innings pitched per quality start. Based on the pitchers listed above, I am guessing that those guys are also the guys that are going to be at the top of the list for innings pitched and complete games which further illustrates how flawed his thinking is, especially with the group ERA being considerably lower than 4.50.
the problem with the quality start stat is that if a pitcher goes 9 innings and gives up 4 earned runs (4.00 ERA), or 8 innings and 4 earned runs (4.50 ERA), that does not get counted as a quality start.