Monday, June 25, 2012

How To Analyze a Baseball Player

Evan Grant of the Dallas Morning News chose to analyze Michael Young this afternoon. Specifically, he chose to discuss the recent calls from the sabermetric, and observant, circles of the baseball world, which have been calling for Michael Young to be removed from the Fifth spot in the lineup. He chose to do so by discussing statistics that Michael Young has produced in this year and in years past, as well as the performance of his peers in important situations. By doing so he upheld the required traditions of the sports-writing profession:

He presented statistical evidence
He interpreted it
He offered his conclusions

And, in the finest tradition of the sports-writing profession, he defiled those concepts with a thoroughness and rapacity that would amaze and disgust the most virulent trumpeter of calumnies in the Roman Senate of Antiquity.

Let us, the discerning sabermetric public, analyze the five points that he presents from the perspective of someone without a deadline to meet, an agenda to fulfill and a locker-room to gain access to.

  1)    “Young’s track record is that he hits.”

Grant’s first point is somewhat longer in actuality, but the entire argument can be rendered in this single declarative phrase. And he is, in the strictest sense, correct. Michael Young, as a player of the batting/fielding persuasion in the sport/game known as baseball swings his bat at the onrushing ball and makes contact 83% of the time (a full 3% of the time above average for his career). To be more specific, Young has, for his career “hit” exactly 5% better than the average major leaguer. When he was a decent fielding shortstop that made him a minor star. As a defensively challenged designated hitter, his career rate would make him a below average player. His production THIS SEASON = 30% BELOW average, makes him a bad player.

  2)    Whether he puts up the typical numbers of a No. 5 batter or not, Young is the Rangers’ veteran leader.

Questions of clubhouse leadership aside, suffice it to say that this quote was written about a man who publicly complained about team management and demanded a trade when he was replaced by a superior player not once but twice. That he has been feted instead of disemboweled by the national sports media speaks to their remarkable penchant for terming white infielders scrappy leaders regardless of their on or off the field performance.

  3)    No Replacements”

The title of the third section of Evan Grant’s article is perhaps the most remarkably and obviously idiotic pair of words since “Mission Accomplished.” To claim that there are no replacements for Michael Young in the Fifth spot in the lineup requires a dexterity with statistics that would flummox most congressional chart makers. Mr. Grant does his best by explaining why Nelson Cruz, Texas’ slugging right fielder, and decidedly not the best statistical choice to move up in the lineup right now (that would be Mike Napoli), would fail at the role:

“For his career, Cruz is a .253 hitter with an .819 OPS. In the sixth spot it’s .862 and in the seventh spot it’s .928. When the Rangers’ lineup works right, it’s most dangerous because Young does his thing in the fifth spot and Cruz and others do theirs lower in the order.”

Notice that Mr. Grant is not arguing that Cruz has struggled with runners in scoring position this year, he hasn’t, or that he has hit worse than Young historically, he hasn’t (14% above average for his career versus Young’s aforementioned 5%). Instead, he trots out Cruz’s statistics based on where in the lineup he is hitting. Now let me be clear, I believe that Young should bat lower in the lineup because the fifth spot comes to bat more frequently than the seventh or eighth spots, where his production suggests he belongs. I do not believe that a player’s production is adversely affected by what spot they are placed in, as Mr. Grant seems to most fervently believe. Or if so the decrease in production caused by a player being in these “high-pressure, run-producing spots” is more than outweighed by the additional at bats that the better (if supposedly stressed) batter receives.

Mr. Grant’s fourth point so depresses me as to the state of statistics in baseball reporting that I am not sure if I can go on. I shall, however, because evil must be opposed… even when it is incredibly tiresome to do so and the evil itself has no real effect on the lives and livelihoods of human beings.

Here it is, quoted in its entirety:
  4)    Grade on the curve: Young is hitting .200 with runners in scoring position since May 1. That’s very out of character and ranks in the bottom 15 among 79 AL players with at least 40 plate appearances with runners in scoring position. But, here are guys who are batting less in the same period with RISP: Alex Rodriguez (.196), Curtis Granderson (.190), Derek Jeter (.171) and Robinson Cano (.156). For the year, Young is still at .269 with RISP, a touch better than Cruz (.268) and significantly better than Ian Kinsler (.254) and Napoli (.231). Offensive numbers are down all across baseball.”

Instead of using florid prose to convince you of the idiocy of this statement, I will simply list things that are wrong with it:
A)   Arbitrary starting point of measuring period. (baseball players, unlike May Flowers and April Showers, are not affected by when months begin and end)
B)   Cherry picked statistics: Young is batting .200 with runners in scoring position… and on a 2-1 count Young is batting .227… you now know just as much about Young’s hitting this year as you did when you read Grant’s statement.
C)   Cherry picked comparisons: Rodriguez’s, Cano’s and Jeter’s batting averages have nothing to do with the discussion, and by admitting that Cruz’s batting average with runners in scoring position effectively disproves his point, as if Cruz is a better hitter overall, and a better hitter in the clutch, then… #clubhouseleader, debate over.

Number 5 makes little sense from a statistical, logical, baseball, or strategic perspective, here it is in its entirety:
  5)    The Angels aren’t doing any better: The Angels split the No. 4-5 spots between Mark Trumbo and Kendrys Morales. Morales is hitting .269 with RISP for the year, tied with young. Since May 1, his OPS of .533 with RISP is worse than Young. The Angels aren’t panicking and stay on a stalking pace with the Rangers. For the year, Morales has shown a tad more pop than young, but the numbers are otherwise similar. Why should the Rangers be the ones to break stride and make a change. They are leading this race.”

After conveniently ignoring the slugging ways of Trumbo moments after mentioning him, Mr. Grant chose to excoriate Kendrys Morales and his .533 On-base+slugging percentage with runners in scoring position (since May 1).

Putting aside the fact that over the season as a whole, with runners in scoring position Morales beats Young in terms of OPS .683 to .644, I would like us all to merely appreciate that Evan Grant actually knows what OPS is. That does not redeem him or his writings, I would merely like it to be noted that Evan Grant knows what OPS is.

I am enjoying the small sabermetric victories today.

From a strategic standpoint, if your primary opponent chooses to do something that hurts them, you do not have to match their self-destructive tendencies point for point. For the same reason that Francisco Pizarro did not give up his artillery and arquebuses when he battled stone-sword waving Incas, the Rangers should not bat a bad hitter fifth just because the Angels have been batting a mediocre hitter fifth from time to time.

Evan Grant used statistics, he interpreted, and he drew conclusions, but at no point did he do so correctly, and everything he did was colored by bias. In short, this was not how to analyze a baseball player.

Wednesday, June 20, 2012

Simply Perfect

I now accept that I just watched the best game I will ever see. Matthew Cain pitched nine perfect innings AND tied Sandy Koufax’s record for the most strikeouts (14) in a perfect game. My favorite pitcher, for my favorite team, pitched the best game I have ever seen (Cain’s game comes in a three way tie for the second best game score—ever). After the shock of the moment wore off and I stopped uncontrollably sobbing, my “thinking” brain rebooted, and I thought about baseball, statistics and perfection.

Half the fun of baseball is watching the game; the other half comes from talking about the statistics behind the game. For, me the motto of baseball becomes a variation on Leahry’s “turn on, tune in, drop out”—watch it, record it, graph it out. Through numbers, we think about the game in new ways. We can celebrate the statistical impossibility of DiMaggio’s hitting streak, or poke fun at the exact terribleness of the Mendoza line. The language of baseball becomes numbers and the stories they tell.

And we celebrate excellence with numbers. A player who exceeds a numeric benchmark surpasses merely great and becomes exceptional. 714 homeruns defines Babe Ruth; Henderson has his impossible number of steals, Ripken showed up every day, and Pete Rose kept getting hits. When a player approaches 3,000 hits or 600 homeruns, fans watch every at-bat, inhaling the excitement surrounding the achievement. When the player crosses that line, he is something new. A 30 game gap can’t explain the great divide between a 300 wins and 270.

For me, the feeling of a perfect game fell into a different category. Cain’s game packed the energy and emotion of Barry’s homerun chase into a single night. But it was also different. Like the homerun chase, you knew it was coming; every at-bat built towards the end. Unlike the chase, which seemed inevitable, the perfect game built out by out, with the unnerving potential to come crashing down.

The perfect game occurs in spite of baseball. Failure is accepted in baseball. “Even the best players fail more than half the time.” Ted Williams struck out. Tom Seaver surrendered homeruns. Like a lesson right out of my childhood— nobody is perfect.

But for one night, one beautiful night, Matt Cain pitched to 27 batters. Only 14* of those batters ran towards first base. After one historical catch, a handful of brilliants plays, and no bad pitches— Matt Cain was perfect.

Maybe that’s why I don’t talk about something so fragile; baseball is not supposed to be perfect. Streaks end, records are broken, and the best teams lose to the worst teams. We expect imperfection and celebrate players who succeed despite our expectations. But for one night in baseball, everything can be perfect. And I will never see a better game.

*There was one strikeout on a ball in the dirt.

Monday, June 11, 2012

Clutchitude and Suckitude

The Giants’ pitching staff has the third highest clutch rating, as measured by, in the Majors.* The Giants rank behind only the over-achieving Orioles and Pirates in terms of their performance in high leverage situations. Clutch ratings are subject to a great deal of variation due to luck and small sample sizes. Additionally, there is some question as to how much control pitchers have over their performance in clutch situations. Those concerns aside, it does serve as a simple way to analyze how well a particular team or player has performed when the game was on the line.

*For a detailed description of how Clutch scores are calculated, see the Fangraphs glossary. For now it is enough to know that it measures how well players perform in high leverage situations when placed in a “context neutral environment”  and measured against the entire league and each player’s own historical performance.

Unsurprisingly, the best performers on the Giants’ pitching staff include the preternaturally skilled Madison Bumgarner, the enigmatically brilliant Matt Cain and the remarkably competent Santiago Casilla.

Surprisingly, to those who have paid little attention to the Giants’ season thus far, two time Cy Young Award Winning, All-Star Pitcher Tim Lincecum ranks second to last on the team in clutch score. (In last sits the execrable Steve Edlefsen)

Lincecum’s struggles have been well documented this season. The combination of Lincecum’s past success, his quotable personality, his cultural cachet and the true depths of suckitude to which he has descended have produced a plethora of articles on the subject.

It's obvious that there are two Tim Lincecums now:
1. The pitcher out of the windup, who will never have Madduxian command, but who can still strike out hitters on three pitches, strike out the side, and look like the pitcher we remember
2. The pitcher out of the stretch, who is sprinkled with weaponized Jonathan Sanchez dust

Grant’s hyperbolic point is essentially correct, with the bases empty, Tim Lincecum strikes out 11.67 batters per nine innings, walking 3.43 and allowing less than one homerun per nine.

With men on base he strikes out 7.16 and walks 6.61.

There is no possible world in which those numbers are acceptable from a major league pitcher.

As good as Lincecum has been at keeping runners off base- allowing only a .312 OBP with the bases empty- his awfulness at keeping those runners who do reach base from scoring is truly remarkable.

The 39 batters who have thus far faced Lincecum in high leverage situations have produced a batting line of .344/.474/.679. For comparison purposes, in Albert Pujols’ spectacular 2008 season he complied a .357/.462/.653 line. Lincecum has, in effect, turned every batter he faces during a high leverage situation into vintage Pujols.

The most horrifyingly interesting statistic that Lincecum’s train wreck of a season has produced thus far is that in high leverage situations, he has stranded -18.1% of runners who have are on base. That negative sign was not a type-o. In a league which strands 72.5% of runners overall, even if the negative sign had been a type-o, his performance still would have been Edlefsen-worthy.

Now consider how clutch the Giants’ pitching staff has been this year. Whether or not their performance in high leverage situations is sustainable is not at issue. Instead, merely admire the remarkable cultchitude™ of the Giants’ pitchers, who have managed to perform so well as a group when the game is on the line that they have been able to overcome the massive negative impact of 72 innings of Bad Lincecum.

Since 2009, when the all-pitch, no-hit Giants model fully emerged to provide endless Torture as only the Giants can, their various pitchers have led the majors in clutch score, ranking a full point ahead of their closest competition, the Lost Anaheim Angels of Angeles.

As hopeful as that long history of clutchitude may make a Giants fan, it is worth noting that those years (apart from this one) all involved superb seasons from Lincecum both in terms of standard pitching success and in clutch rating. If the Lincecum who has been pitching out of the stretch remains befuddled by “weaponized Jonathan Sanchez dust,” or baseball gods forbid, his stretch suckitude seeps over into his pitches delivered from the windup, Giants fans could be in for more torture than they reckoned for.

Monday, June 4, 2012

Three High School Hitters...

The Oakland Athletics had three first round draft picks in the Amateur draft tonight. The A’s were gifted with the 11th overall pick and two supplemental picks at the end of the first round, provided as compensation for the loss of free agents this last offseason. Led presumably by Billy Beane,* the A’s chose a high school infielder with each pick.

*Have you accepted Billy Beane as your lord and savior yet?

When the first selection, that of Addison Russell, High School Shortstop from Florida was announced, the Sabermetric experts were surprised, the Baseball Network crew were shocked and appalled, Harold Reynolds remained Harold Reynolds, but with a slightly shriller voice… In short, the world as we know it was not brought to an end, but to many baseball insiders and mildly informed fans, Beane’s selection of a High School player ran contrary to many of their preconceived impressions of his drafting philosophy.

For those few of you who did not read Moneyball, the short explanation is this: Beane determined that College players were laughably, absurdly better investments than high school players. This was because college players would, of necessity, accumulate a sizeable body of statistics during their careers, and against a much higher level of competition than their high school equivalents. The physical bodies of college players would also have more time to mature, injuries would occur and, most importantly to the perpetually cash strapped A’s, their bonus demands would steadily drop as they lost negotiating leverage. Beane’s preference for college players was backed by the estimable Bill James, and has since been confirmed by a plethora of statistical studies.

So why then did Beane abandon his statistically grounded study to sweep up high school hitters by the metaphorical truckload?*

*Beane has not selected a high school player in the first round since the Jeremy Bonderman debacle of 2001, again, read Moneyball for more details.

There are 4 factors that I believe contributed to this seeming change of tactics, each of which is merely representative of a broader theme within the game.