The Athletic has hired a solid stable of fantasy and analytical baseball writers. The excellent Eno Sarris published a story today about how advanced stats work. He makes drudgery fun, plus it’s baseball.
The early days of the internet were extra exciting because only “advanced” people were on it, talking. There were few civilians back in that day, and there were no structures, so opportunity to talk and engage was really open.
I found my voice in alt.rec.baseball and more often alt.rec.baseball.fantasy, which were Usenets groups that attracted a lot of people with interests and passions and ideas, at a time when public internet discourse wasn’t routinely ruined by vandals.
One of the people who was encountered on alt.rec.baseball was Sherri Nichols, who is the topic of a very nice Ben Lindbergh piece on The Ringer. You can read it here.
I have no problem with Ben’s recitation of the history, and I fully embrace the idea that baseball nerdom would be better if it became more gender balanced. Or at least recognized that women were there, as Sherri was, when important ideas were being developed.
But his story is also an intriguing look at how the internet went from being the domain of academics and people with ideas to a teapot tempested with opinions.
I don’t want to tie too big a bow on it, but those of us who like real ideas miss the old days.
Eno Sarris does a good job explaining how the Astros and Jon Singleton came to a deal that pays him $10M if he busts, and $30 if he’s a success.
How good a job? Don’t take my word for it.
I read somewhere that since the earliest days of baseball pitchers, on average, through on average 100 pitches per start. Seems that pitchers would mix up very long and very short starts, and end up at 100 pitches generally.
Today, I found at the Sabernomics website, actual data on pitch counts over the years from 1988 to 2009, which shows much the same thing. Pitchers used to throw many more pitches in a game back then than they ever do today, but they also sometimes threw far fewer. And on average, the results are pretty similar.
Of course, this is all the era of the five-man rotation, long after the days of pitchers throwing both games of a doubleheader. That’s when pitchers were men.
From ZiPS inventor Dan Symborski, over at The Hardball Times, his 10 Lessons about making baseball projections. All very interesting and at the same time they feel quite familiar to me, echoing many of my experiences.
Except No. 8, which talks about the regression of inseason stats, which is not as pronounced as from season to season. This means that a player who gets off to a hot start for half a season, is more likely to continue that hot start in the second half than he is to have it continue the next year.
That doesn’t surprise me for any particular individual, but in aggregate it does.
This 10 Lessons story is part of 10 Lessons week at THT. Mitchel Lichtman’s piece about defensive statistics is full of great stuff, too.
As the editor of a fantasy football magazine, I’m aware that schedule strength is a big thing in the make believe pigskin racket. But we don’t talk about it much in baseball because it is a little thing.
But that doesn’t mean it doesn’t matter.
Jeff Sullivan, at FanGraphs, took a look at the relative strength of the divisions today. That’s a little interesting, especially the historical chart, but really, didn’t we all know that already?
Commenters also pointed out that division strength is meaningful but unevenly distributed. The best team in the division plays the four weaker teams, it doesn’t have to play itself. And the worst teams plays the four stronger teams, it doesn’t get to play itself.
So Jeff did some math stuff, I’m not sure what, and came up with a strength of schedule for each team. Assuming he got the math stuff right, we get a little more granular on the way strength of schedule affects baseball teams.
Some folks in the comments created and posted images that add info, like hitting/pitching splits to the graph.
This summary of team WAR in the site’s Depth Charts pages could be a good way to find systemic imbalances that might affect fantasy value and is worth checking out, too.
I’m pretty sure we’ve got most of these things priced in, but it can’t hurt to have more information.
Aaron Gleeman liked this story today on Facebook, and it’s a delightful romp showing how baseball’s statistical revolution is earworming Terry Ryan and Ron Gardenhire.
The thing I didn’t know and wonder about is the assertion that Michael Lewis’s first choice for a Moneyball team was the Twins. That seems, um, sick, in the, um, sick way. Whatever.
The reason to read this story is because some of the players talk about their process in a less than canned way, which is bracing. Okay, a little bracing.
Tout Wars, you may have heard, is moving it’s Mixed Leagues to On Base Percentage this year, rather than that old standby category Batting Average. The reason, as described here, is because OBP measures a player’s ability to draw walks, which is a valuable baseball skill that the traditional fantasy stats undervalue.
Some Tout Warriors are arguing that OBP is a baseball metric that measures better baseball players, but that using it in the fantasy game will break the delicate balance of fantasy baseball’s categorical imperatives. Steve Gardner, in USA Today, summarized: “Dissenters pointed out that eliminating batting average gives far too much weight to sluggers, many of whom have higher than normal walk rates, when those power hitters already get additional credit in runs and RBI for every home run they hit.”
My first response was fear that this was true. That the guys whose value would jump most were already valuable guys. That wasn’t why we’d changed the rule. But the fact is that some players in every strata of the game, from homer hitters to speed merchants, show an ability to walk, while others with those same talents don’t show that ability. The adoption of the OBP rule was intended to value home run hitters who walked more than home run hitters who didn’t walk. It was intended to value stolen base guys who walked more than those who didn’t walk. It was intended to value guys who hit for a high average who walked more than those who hit for high average who didn’t. The bottom line was, walks are a valuable skill that fantasy baseball has valued only peripherally, and as I noted here the other day in the Derek Carty is Absolutely Right post: By giving up an at-bat when taking a walk, a player hurts his fantasy value overall while often improving his real baseball team. Guys who walk get fewer chances to homer, fewer changes to drive in runs, and can even end up with a low batting average while their high on base percentage helps their team win games.
My second response was to see if this claim that the guys who would be helped most actually were the supposedly already-overvalued home run hitters. Here are the top 25 hitters with 250 or more AB with the highest walk rate in 2012. These are the guys whose value would be most improved by using OBP rather than BA (in parens 2012 PA/HR):
- Joey Votto (475/14)
Adam Dunn (649/41)
John Jaso (361/10)
Chris Carter (260/16)
Dan Uggla (630/19)
Carlos Santana (609/18)
Jose Bautista (399/27)
David Ortiz (383/23)
Ben Zobrist (668/20)
Carlos Pena (600/19)
Bobby Abreu (257/3)
Alex Avila (434/9)
Joe Mauer (641/10)
Todd Helton (283/7)
Mark Reynolds (538/23)
Mike Napoli (417/24)
Jonny Gomes (333/18)
Edwin Encarnacion (644/42)
AJ Ellis (505/13)
Dexter Fowler (530/13)
Chris Snyder (258/7)
Miguel Montero (573/15)
Chipper Jones (448/14)
Josh Willingham (615/35)
Chase Headley (699/31)
There are a few sluggers on that list, many guys who hit home runs, but certainly not only the best home run hitters. Many of these are guys whose baseball creds are mocked by fantasy players, because they don’t hit for big power and have bad batting averages. Why do they even have jobs, the neophyte wonders? Because getting on base is a valued skill. It has real value that fantasy leagues that don’t use OBP aren’t capturing. It’s also a skill that a player like Joey Votto has when his power deserts him because of injury.
So, just for giggles, who are the guys with the lowest walk rate? Who will get hurt most by the change? Let’s go 10 deep:
- Miguel Olivo (323/12)
Alexei Ramirez (621/9)
Pedro Ciriaco (272/2)
Luis Cruz (296/6)
Josh Rutledge (291/8)
Delmon Young (608/18)
Ichiro Suzuki (663/9)
Josh Harrison (276/3)
Willie Bloomquist (338/0)
Omar Infante (588/12)
It’s true, not as many sluggers here. And a lot of marginal offensive talents, or special talents (Ichiro) whose ability to hit for BA while not taking bases on balls should be noted, not applauded, by fantasy players. Welcome OBP!
(illustration adopted from bluejayhunter.com)
First off, a link to Joe Posnanski making some strong points in favor of Mike Trout as AL MVP over Miguel Cabrera this year. My favorite is his suggestion that you vote for whoever Brandon McCarthy thinks should be MVP.
Since the season ended, I eventually came to the idea that Mike Trout was most deserving of the award. The preponderance of the evidence weighs in his favor, even if I don’t think it’s quite so clear a case as some. By that I mean that despite Trout trouncing Cabrera in WAR, the award isn’t solely given to the best hitter or the best player in the league. The MVP is supposed to go the player who was most valuable to his team.
This has led some people to suggest that Cabrera was most valuable to his team because he led it to the playoffs, while Trout was only able to lead his team to third place. These people should note that Trout’s team won more games than Cabrera’s and step away.
But I think a case can be made, sort of, that Cabrera was the more important player on his team. If you use as your measure WAR, and if we’re having this discussion why not, Cabrera contributed 6.9 WAR of Detroit’s hitters’ total of 13.7 WAR, or more than 50 percent. Trout, on the other hand, was worth 10.7 WAR, which was 28 percent of the Anaheim team’s 37.9 batting WAR.
But that’s the best case, and it isn’t that persuasive, since Detroit’s total WAR (they had great pitching, with Justin Verlander worth more WAR than Cabrera at 7.6) was 36.9, while the Angels’ total was 40.5. Trout’s contribution of 26 percent of his team’s total versus Cabrera’s 19 percent of his team’s total is a decisive edge.
Which leaves one final mode of attack: dWAR, defensive Wins Above Replacement, is far from established as a reliable measure of defensive value. Even those who champion it point out that it really takes two years of defensive play to start to establish a fielder’s performance baseline in fielding WAR. In 2012, Cabrera did a decent job playing third base, exceeding expectations but probably not adding to his own value with his defensive contributions (but not hurting it either–some argue that his agreement to play third also helped the Tigers because it meant they didn’t have to play Ryan Raburn), while Trout was simply amazing. Still, if you discount his defense because the measure isn’t reliable (and don’t believe your own eyes), Trout’s contribution in WAR drops to 8.6, or 21 percent of the Angels total, which at least makes it a horse race.
I’ve enjoyed the argument about this MVP race because in discussion new ideas come up. Nate Silver, championing Trout but expecting Cabrera to win, pointed out that Trout was superior to Cabrera while leading off an inning, a not inconsiderable skill that compares nicely with Cabrera’s better stats in the clutch this year.
The bottom line, however, is that the MVP awards are given by voters or judges, and they reflect the values of that constituency. If the BBWA says these 28 voters are the judges, we have to look at who they are to see what values are reflected. They’re the bosses. There was a time when the fans’ access to the records of the game was limited, and some favored Maris while others favored Mantle, for example. Some of that argument was based on numbers, of course, but it was also personality and some ineffable human streak that drew fans to one or the other. And the judges then were Olympian.
We’re now our own best judges, as the ballots of the BBWA so ably demonstrate every time they vote, and this discussion among fans with a much broader understanding of how the game works ideally serves the purpose of helping us better understand baseball, baseball players, baseball teams, winning baseball, and the stats and numbers and opinions that help us describe them. The awards themselves are wan, the judges are suspect, but the discussion is lively, which is just great.
This collection of listserv posts from 2000 is the clearest expression of how baseball stats don’t tell the whole story about a ballplayer’s talents. Nor are statistics the last word about teams. Must reading for anyone interested in baseball statistics and sabremetrics.