There are two types of stat-heads. On the one hand, you have the type that ignore the hissing from the traditional media elites who scoff at decimal points and sigmas. On the other, you have the type who want to educate. I really respect the latter type, because--hoo, boy---if that isn't tough work...
So let's dedicate today's box score to those writers, players, and fans who lead us unflinchingly into greater understanding of baseball.
Table of Contents
It's been a tough year for Mets fans. So it's with an extra degree of sympathy that I would like to acknowledge the work of Amazin' Avenue's James Kannengieser. He's been under attack from Mike Silva for his use of (particularly, though not exclusively) UZR and WAR. Instead of getting fed up, though, James is all about raising consciousness--a sort of sabermetric Paolo Freire. Mike Silva had this to say:
As for using WAR, any stat that using [sic] a fictitious player as its benchmark automatically gets thrown in the trash by me.
Instead of going the Fire Joe Morgan route (watch out for that reunion, though), Kannengieser says this:
A replacement level player's production can be easily replaced by readily available talent at minimal cost to a team. A team full of replacement players would be expected to win anywhere from 45-50 games a season, which varies based on league and some other factors. Current examples of "replacement players" are Willie Bloomquist, Cory Sullivan, Luis Ayala, Willy Taveras, and Nelson Figueroa. Basically half of the Mets roster fits this description.
I think Reds fans only wish Willy Taveras were a fictitious player.
If you don't love players who love sabermetrics, I think you might be lost. We've heard all about Banny, but I'd like to share yet one more piece about him:
He was able to analyze specific numbers and apply them to his pitching.
"Quite unsuccessfully at first," he admits. "But I think now I've really come to some good conclusions regarding the data. I'm top five in the American League in ground-ball percentage now, which is just proof that the system works."
"If Bill James had a 90-mph fastball, he'd be me," Bannister says.
I think Tom Tango speaks for all of us when he says:
I can’t think of a better spokesperson for saberists than Banny.
But it was his younger brother, Alex, a business economics major at Missouri, who got him interested in baseball's deeper statistics. Scherzer said they had a months-long debate after Alex told him a pitcher has no control once a batter hits the ball.
"It took about a year of arguing with him for me to realize that actually is the correct way to think," said Scherzer, referring to BABIP, a stat that can be used to gauge a pitcher's luck or quality of the defense behind him.
He also says he is fascinated by tERA. That stands in stark opposition to Banny, who is a self-avowed xFIP guy:
"I think the ultimate stat for a pitcher is xFIP," Bannister said before pausing and offering a wry grin. "I know that’s getting really technical. It’s fielder independent pitching adjusted for your home-run rate back to the league average.
Did he just successfully demonstrate knowledge of regression to the mean? For what it's worth, Scherzer sports a tRA of 3.33 versus Bannister's 4.01. As for xFIP, Bannister's is 4.35, while Scherzer's is 3.95. Comparatively, xFIP says there is a smaller difference between Scherzer and Banny than does tRA. Could this account for their varying preferences?
Hal McCoy might not have been the most stat-obsessed beat writer, but he wrote intelligently and embraced the blog format, despite the fact that he has been blind for the last six years. And now that he is retiring, it was fitting that he announced his decision in a blog post:
I feel like I still have my fastball at the keyboard and can deal with the curves thrown my way.
He never was a Luddite. He embraced blogging, chomping on a cigar while he hacked away on his magnified computer screen. Hal has been legally blind for the last six years and he could have gone to Florida with Nadine for good. But he kept trucking, 150 days on the road a year, because reporting on the Reds was embedded in his DNA.
There is much to lament in a loss like this, but I believe sportswriting will be alright. It's too much fun not to go on.
For all the boosting of stats and online baseball writing, it wouldn't be nearly as eventful if it weren't for the people who are down in the trenches, whether it's reporting or regressing. So let's take this moment to reflect on a few interesting bits of sabermetrics.
At Baseball Analysts, Dave Allen looks at Joel Pineiro's newfound control and ability to induce ground balls. He argues that, armed with PITCHf/x data, we might need to reevaluate how we go about weighting past performances. Perhaps, he thinks, we can weight the past less if we have better data from the current season.
Obviously we do not expect Pineiro to continue to walk under 3% of batters faced and get over 60% of his balls in play on the ground. An estimate of true talent and expectation going forward must include some weighting of past performance and regression to the mean. But I think the PITCHf/x data, just like scouting data, can be used to adjust the weighting, maybe weight this year even more heavily if we expect him to use this pitch break down going forward, or regress to different pool, one with this breakdown of pitches, to get a better estimate of his true talent going forward.
It's an interesting thought, and one that caught the eye of Tom Tango, who responded:
The question you are always asking yourself is: how much of what I see is real, and how much of what I see is a change from the past. The answer is ALWAYS greater than 0% and less than 100%. Our job is to figure out the numbers in-between.
Something like ground ball rates on contact, he argues, stabilize more quickly. Therefore, the regression to the mean and weighting of past performance do not need to be so heavy. That does not mean, however, that they do not have to be done at all.
Finally, Tango points to a very fascinating survey of current defensive statistics by three knowledgeable statisticians. I highly recommend it.
Points if you know the differences, and of course I'm more interested in your reasoning than what your verdict is.