June 8, 2016

#KillTheWin

During last night's Red Sox-Giants game, NESN's Dave O'Brien touted tonight's pitching match-up of David Price and San Francisco's Madison Bumgarner as a dandy because ... (here it comes) ... both pitchers are 7-2 this season.

I have mentioned O'Brien's eagerness to tout a pitcher's won-loss record as the most important gauge of his success and effectiveness too many times this season. But it's maddening - and it makes him sound so ignorant. (Not long ago, he noted that Joe Kelly was "undefeated" even as Kelly's ERA was climbing to 8.46.) You'd think that someone with O'Brien's vast experience would know better - would see evidence in the box scores every single day that a pitcher's record is all but useless - but perhaps he's too insulated and does not venture outside of his broadcasting bubble. O'Brien will sometimes admit that, in the case of someone like Price, yes, the ERA is a bit higher than you'd like to see, but still ... 7-2!

Note: Among qualifying MLB pitchers, Bumgarner's ERA of 1.91 is tied for third. Price's mark of 4.88 ranks 84th.

Brian Kenny, Sports On Earth, May 31, 2016:
You think pitcher wins make sense. They don't. ...

There have been 17 outings this year where a pitcher has thrown at least seven innings of shutout ball and not gotten the win. There have been 48 outings where a pitcher has gone at least seven, and given up one run, or no runs. 48. We're only two months into the season! We think these things don't happen often, but they do. ...

Let's widen the lens. I wrote this up for the #KillTheWin chapter in my book, "Ahead of the Curve," which comes out in July. In that chapter I choke out the win from every conceivable angle. For now, let's use one example.

I picked the type of game where we can all agree on it being good enough to get a win most all of the time: eight innings, two runs. That's a 2.25 ERA. That ERA would be good enough to be among the Top 10 most years throughout baseball history. Yet from 1920 to 2014, a pitcher throwing such a game, got the win -- brace yourself -- just 33.6% of the time.

Soak that in. A pitcher leaves after eight innings, giving up just two runs and receive big applause from the crowd, along with pats on the back in the dugout. Yet two-thirds of the time, throughout live-ball history, that guy doesn't get a win. That number may be a bit skewed from the old days when pitchers threw complete games and eight innings is the full distance for a road loss. But two-thirds of the time?

We cite statistics because we believe they have value. That there is correlation to the physical activity we have just witnessed. The win, after all, is early sabermetrics, an 1800's attempt to isolate performance, once teams began using multiple pitchers. But the correlation to the performance is too loose and haphazard to be bothered with.
(Bold = my emphasis.) I must check out Kenny's book - which will be published in July.

6 comments:

Maxwell Horse said...

O'Brien actually did it twice. He mentioned it casually toward the very end of the game. But the one that took the cake (the one I mentioned in the thread) was (I think) starting in the 8th inning when the Sox were coming up to bat.

He didn't just casually mention how both Price and Bumgarner had 7-2 records. There was a big windup where he was leading up to the punchline that was meant to slay the audience. "Get ready for this," the tone implied. "Are you ready?" And then he lays down the fact of their winning records like, "Can you believe it? It's true. Oh, yeah, it is so on tomorrow."

Jere said...

The weird thing about pitcher win-loss records still being thought of the way OB thinks of them is that it's not like it took some modern, super-computing epiphany to realize that they don't mean much. It's the type of thing where, when you're about 8 years old, you figure out that a pitcher can give up 10 runs and win or give up 1 run and lose. It's something we've all known all our lives. So it's baffling that OB would say it as if it's the ultimate way to judge pitchers.

Clem said...

O'Brien likes his numbers. He was reading the nonsense "sellout streak" propaganda about tonight's game, which the Giants learned from the Red Sox, in the face of cameras panning over ample empty seats in the middle of a close game. Instead of having a personality and maybe cracking a joke, he defensively said "doesn't mean those seats haven't been sold, even if the people don't show up." Numbers don't lie to Dave O'Brien!!

FenFan said...

Somewhat unrelated but Ichiro is 29 hits away from the "magic" 3,000 career MLB hits mark; including his totals with the Japan Pacific League, he has 4,249 hits, which is seven shy of Pete Rose's professional mark

FenFan said...

What can you say about guys like O'Brien and Lyons who seem hell bent on ignoring the possibility that there are other numbers out there to judge a pitcher's value, or any player for that matter?

I get it: a pitcher with seven wins and an ERA below two is probably having a great season, but a pitcher with seven wins and an ERA north of six? Isn't that telling right there?... and that's from the perspective of the numbers that these guys cling to like a rotary phone.

I don't pretend to understand how WAR, ERA+, or OPS+ is calculated, but obviously some people who (a) enjoy baseball and (b) enjoy statistics are at least giving us some different perspectives that perhaps we should consider and, hey, even adopt as some new standards.

allan said...

They also do not seem to understand the weaknesses in batting average, as they still call the guy with the highest average "the best hitter in baseball". This from a stat that ignores walks (which are great because a guy gets on base and does not make an out; that's the batter's entire job) and treats home runs and singles exactly the same.