Down the Rabbit Hole of Stat Geekery

[Jeff]

I know there are a couple of baseball fans around here. And I know one or two (Kurt springs to mind) have a grasp on modern sabermetric thinking, but I was curious how much y'all follow this stuff...

So here is a list of current ideas prominent among statheads. I wanted to see where you guys are on these. Some of these are rudimentary to serious fans of today's games, others are stat geekisms.

• Pitchers have virtually no control over batting average on balls in play (BABIP). That is, they control how many home runs and walks they give up and how many strikeouts they notch, but the rest is luck, defense, and more luck.
• Given the previous idea, ERA is a shaky indicator of performance. FIP and xERA are better.
• RBIs and Wins are stupid, stupid, useless statistics.
• Batting average is substantially less useful than on-base percentage.
• Batters also experience variance on their BABIP. This explains many, many hot streaks and slumps, not to mention fluke seasons.
• "Clutch hitting" cannot be reliably demonstrated to exist.
• The most reliable indication of a team's performance in a season is not necessarily its win-loss record, but rather the difference between the runs it has scored and the runs it has allowed.
• The sacrifice bunt is almost always stupid.
• Despite what managers, announcers, and small sample sizes might seem to show, there generally isn't such thing as a hitter who is disproportionately skilled at hitting lefties vs. how he hits righties (or the inverse). Platoon advantage over large samples is remarkably stable.
• Lineup order is generally a pointless thing to worry about.
• Pitchers face a rapidly increasing risk of injury the more pitches they throw in an appearance. I'd have to expand on this, but look up "Pitcher Abuse Points" if you're interested.
• Young pitchers should generally not be allowed to exceed the total number of innings they threw in the previous year plus 35.
• The defensive spectrum is as follows: DH - 1B - LF - RF - 3B - CF - 2B - SS - C. While a player can generally move to the left on the spectrum (with a few caveats based on range and speed when moving left from C or from CF to 3B), it is considerably tougher (and usually not possible) to move from left to right.
• The position of a pitcher's arms during his motion at the point of maximum exertion can be a key indicator of risk of future injury.
• Over time, stolen bases are only really beneficial if a player is successful at avoiding being thrown out at least 70% of the time.

[Jeff]

I'll post more of these later if it turns out I'm not talking to myself.

[Poop Fresh-Herbed Pickles]

You forgot my favorite stat OPS.

[Jeff]

OPS (OBA or on-base average + SLG or slugging percentage for the uninitiated) is a good, easy to calculate stat that gives a quick glimpse at a player's combined ability to do the two most important things a hitter can do: get on base and hit for extra bases.

That said, I hardly ever use it anymore. I'm more a fan of OPS+ (as seen on http://www.baseball-reference.com ) which is calculated like this:

100*(OBA/league OBA + SLG/league SLG - 1)=OPS+

This gives you a better idea of how a player ranks against his peers, accounting somewhat for how different eras are.

Even better are the linear-weighted stuff you find on http://www.fangraphs.com

Those guys will give you the Park Adjusted Runs Above Average a guy is worth, based on his linear weighted OBA. To that they'll add the fielding runs the guy saved, measure that against replacement level, then adjust based on difficulty of position, then give you the salary the guy was worth in any given year since 2002 (when they started doing this).

Albert Pujols, in terms of the marginal wins he provides his team with his glove and bat, has been worth a staggering $36.7 million so far in 2009.

[Jeff]

Where are we on fielding stats? That's where a lot of the fun stuff is being uncovered these days.

It should be said first that fielding percentage and its brother the error are awful. The error is perhaps the most useless, arbitrary stat in all of sports, representing as it does, one man's opinion of what should have happened on a play a fielder attempted to make.

For instance, let's say, for some damned reason, the Yankees made me their new left fielder. And let's say I spent 162 games sitting on the grass in the outfield, picking dandelions. Then, one shining day in August, someone hits a loopy fly right to me and I lazily lift my glove and snag it. I'd finish the year with a 1.000 fielding percentage. Would you rather have me or Carl Crawford and his .983?

For a less humorous example, try this on for size:

In 2007, Jason Bartlett made an AL-leading 26 errors at shortstop, for a .966 fielding percentage. Edgar Renteria made just 11 errors for the Braves and posted a .977 F%. Announcers and casual fans cheerfully assume Renteria to be the better shortstop.

However, Bartlett had a Range Factor of 4.5. That means, in an average game, Bartlett successfully makes 4.5 plays. Renteria's is just 3.9. So over the course of a full season, Bartlett would make nearly 100 extra successful plays than Renteria. Let me say that again: Bartlett would make nearly 100 extra outs.. How is that possible if he makes more errors? Bartlett just flat-out gets to more balls than Renteria does. Many, many more.

Wouldn't you rather have the guy who could make an extra 100 outs, even if, in the process of getting to those 100, he bobbled, say a dozen other balls? Of course you would. And that's why fielding percentage and errors are stupid.

Another fun stat is UZR/150 which shows, over the course of 150 games, how many runs a fielder saves above the average at his position. The center fielder for the Mariners, Franklin Gutierrez has saved an astounding 22.6 runs above the average guy at CF. Even with his kinda meh hitting, this has made him worth an astounding $21.4 million this year.

[Poop Fresh-Herbed Pickles]

How do you account for Derek Jeter's statistical awfulness in the field?  Legit or an illusion?

[Jeff]

I think Jeter has good hands and terrific instincts but seriously subpar range, a problem which has been exacerbated by his stubbornly awful positioning. After finally taking some of this criticism to heart last year, Jeter has begun to play a bit further back, giving him more time to read the ball and get into place. He was actually only barely below average on UZR last year and has been 5.1 runs above average so far this year. He still may lack for range but he's at least learning to substitute with veteran experience.

As people like Mitchel Lichtman continue to improve defensive stats, and as it gets more accepted, I think you'll see more of these kinds of stories.

Though to be fair, Jeter probably doesn't know from UZR and likely did it due more to vague "aging" issues he felt. Truth is, age worsened his defense around seven or eight years ago, but his soft hands and smart playing still helped him make enough highlight reel plays that he and his apologists could easily deny the problem.

[Jeff]

Here's a good example of why errors and fielding percentage are stupid:

The 2009 Seattle Mariners have the fourth most errors in the AL this year, seventh most in MLB overall.

But you know, they're actually the best defensive team of the last decade. Here's why:

The M’s currently lead baseball in UZR by a huge margin. They are +84.5 runs above average, more than 20 runs ahead of second place Tampa Bay. Their team UZR total is the highest of any club in our data set, which goes back to 2002. In fact, the Mariners are only the fourth team in the last eight years to post a UZR over +70. The M’s passed that mark a month ago.

Amazingly, the team has achieved a record setting UZR despite making 102 errors. Their .982 fielding percentage ranks 23rd in baseball, and they have lost 13 runs from their UZR total thanks to the miscues – only the Nationals have cost themselves more runs via the error.

In fact, if we just focus on range, which is what most people associate UZR with, the Mariners are a staggering +91.2. Last year, the Rays put a track squad on the field and led baseball in Range Runs at +69.0, and they’re basically matched that with +62 range runs this year.

The 2008 Rays got to the World Series in large part because of how good their defenders were at covering ground. The 2009 Mariners have been 30% rangier than the 2008 Rays. There’s a reason the team leads the American League in ERA despite ranking 9th in the league in FIP. Franklin Gutierrez, Ichiro Suzuki, and Adrian Beltre have formed the nucleus of an historically great defensive team.

For comparison's sake, the White Sox were way ahead of the Mariners on fielding percentage this year at .981. So they don't make many miscues. Good for them. Nevertheless, the Mariners have such superior range as a team and make so many more plays, that they have saved 125.4 more runs than the Sox.

125.4!

UZR "runs" convert roughly to a win per ten. So that's 12.5 more games the Sox would have won had they manned the field as well as Seattle.

They would have run away with the AL Central crown instead of finishing third.