May 1, 2015

BP Unfiltered

Everything You Could Have Learned This Week, 5/1/15

by Ian Frazer

DRA has changed everything for me. Before its release on Wednesday, I lived in a run-down house with two filthy roommates, my car was a rusty mess with a broken driver-side mirror, and I was a senior in college set to graduate into a weak job market with nothing but a journalism degree.

All those things are still true, but hey, now I have a pitching metric that properly contextualizes each plate appearance! And so do you!

Monday

Make fewer mistakes (than your opponent), win more games! Pretty simple: The 2015 Mistake Index, by Scott Lindholm, Beyond the Box Score

The third tab, Win Pct Corrrelation, gets to the heart of the issue, because it's one thing to quantify a team's mistakes, but doing so makes sense only when compared to how many mistakes their opponents make. If the Mistake Index is to have any value, teams that make fewer mistakes than their opponents should have more success. Like any other measure, there will be outliers, but in general, it's a correlation that holds up well.

Better pitches = a better pitcher, generally! (By ERA, at least,) Again, pretty simple: Pitch Arsenal Scores, by Saul Jackman, The Hardball Times

The beauty of this technique is that it allows the data to dictate how important each component of a pitcher’s arsenal is in explaining his overall performance (measured here as ERA). Having a high swinging strike rate on a particular pitch type is valuable to a pitcher only insofar as that swinging strike rate is negatively correlated with ERA. For example, swinging strike rate on four-seam fastballs is very strongly correlated with ERA (coefficient of -37.7, p<0.001), while swinging strike rate on sliders is correlated with ERA at a slightly lower magnitude (coefficient of -27.4, p=0.001), and swinging strike rate on curveballs is only weakly correlated with ERA (coefficient of -14.5, p=0.154). This indicates that while a high swinging strike rate on any of these three pitches leads to a lower ERA, it is more useful to have an elite four-seam fastball than it is to have an elite slider or curveball, all else equal.

Similarly, the model supports the intuition that swinging strike rate has a stronger effect on ERA than does groundball rate. Indeed, only the change-up and knuckle curve (the latter was a relatively uncommon pitch type in 2014) have larger negative coefficients on their groundball rates than on their swinging strike rates. While groundball rates are generally negatively correlated with ERA, all other pitches primarily derive their value from high swinging strike rates.

Tuesday

The strike zone appears to change measurably during different times of the day: The Twilight Strike Zone, by Frank Firke, The Hardball Times

I really don’t have a good guess for what’s going on here, which is fairly disconcerting. I don’t know enough about the subtleties of my possible explanations to assess how likely they are, so I’m very open to other suggestions about what’s happening here.

Still, if these effects are real and anywhere near as large as I’ve estimated here, they could have substantial implications. Even if they are only measurement issues, the magnitude of difference they imply suggests they should be included in pitch-framing and other strike zone calculations. If they aren’t just measurement issues, these findings suggest there’s an advantage to be gained by pairing pitchers with the environments that play best to their strengths, depending on where they throw most of their pitches.

Hitters are seeing relievers earlier than ever, and those relievers are now better than ever: David Ortiz on Cubs' Kris Bryant and why hitting is harder than ever, by Tom Verducci, Sports Illustrated

Look how the gap has grown. Starters lose their stuff essentially at the same rate as they did 30 years ago, but batters now hit just .227 when they see a reliever for the first time! That’s crazy.

"The game has changed, and it’s not just velocity," Nationals pitcher Max Scherzer said. "It used to be that as a starting pitcher you wanted to have three pitches. Now it’s four—with command. It used to be that a relief pitcher needed only one really good pitch and one so-so pitch. Now you see more guys coming out of the bullpen establishing three pitches. And now that guy can get out righthanders and lefthanders. The cutter is just one pitch that you see more, but no matter what the pitch is, the trend is pitchers are adding pitches, whether they’re starters or relievers."

Being an ace bunter doesn't make the defense as uncomfortable as they'd have us believe: Death of the Renaissance Man, by Russell Carleton, Baseball Prospectus

In situations with this configuration of baserunners, hitters had an overall .311 BABIP (again, 2010-2014). If we assume that our hitter was league average and facing a league average pitcher, with two outs–when we know he won’t be bunting—we estimate his BABIP would be .306. With no one out, the model predicts a BABIP of .321. Big effect, right? Fifteen points!

Not exactly. I re-ran the same regression for the group that had four or fewer bunt attempts and got roughly the same results. I took it down to a group that had zero bunt attempts for that year and got roughly the same results. In fact, if we look at situations where a runner is on first and not on third (again, second base is optional), with no outs, there was an overall BABIP of .321 and with two outs, a BABIP of .299. That’s a difference of 22 points. Even if we make the rather tenuous assumption that with our potential bunters, we are only seeing a 15 point spread, and their 7 point advantage is due to their bunting prowess, we’re not seeing much of an effect.

Taking a hitter who gets 600 PA in a season, a league-average hitter will be in a potential bunting situation (runner on first, none out, no one on third), 6.8 percent of the time (41 PA). The ball was in play in 68 percent of the time (down to 28 PA), and 7 points of BABIP would mean two-tenths of an extra hit each year over the course of a year. The actual number might vary, but that’s the level of magnitude that we’re talking about. Being a well-known bunter might score you an extra hit every five years in a sac bunt situation because the third baseman is playing in a little closer and you swung away.

Wednesday

If you have dreamy blue eyes, don't worry, because 1) You're probably getting, like, a ton of chicks and 2) It doesn't demonstrably present a disadvantage in hitting a baseball during the day: How Does Eye Color Affect Day/Night Splits?, by Gerald Schifman, The Hardball Times

We’re given results that run completely against our Bayesian prior. The lightest blue-eyed players hit better in the daytime, while grade 2s and 3s were basically equals in the day and night. The results are even more unexpected if we consider the green-to-brown continuum from buckets 5–9, as hitters’ daytime stats worsen as their eye color darkens. Hitters in bucket 5 hit +6.3 points better in the day, an effect that keeps shrinking until getting to the darkest brown-eyed players, who post the worst day split out of all groups at -4.2 points.

It’s important to point out that we’re dealing with pretty small samples throughout the bins. That may be surprising, given that the buckets are comprised of tens of thousands of plate appearances, but split effects such as these require greater amounts of PA. Here we’re taking a standard split (day/night) and breaking it up in nine ways. The corresponding uncertainties are reflected in standard deviations that are often larger than the effect sizes, culminating in confidence intervals that include 0 well within their range. Saying that the light green types hit 2.5 points worse in the day—leaving us 95 percent confident that the true estimate is between -14.3 points worse in the day and +9.2 points better—just doesn’t tell us much.

DRA is here, and it's spectacular: Introducing Deserved Run Average (DRA)—And All Its Friends, by Jonathan Judge, Harry Pavlidis and Dan Turkenkopf, Baseball Prospectus

Today, we are transitioning to a new metric for evaluating the pitcher’s responsibility for runs that crossed the plate. We call it Deserved Run Average, or DRA. Leveraging recent applications of “mixed models” to baseball statistics, DRA controls for the context in which each event of a game occurred, thereby allowing a more accurate prediction of pitcher responsibility, particularly in smaller samples. DRA goes well beyond strikeouts, walks, hit batsman, and home runs, and considers all available batting events. DRA does not explain everything by any means, but its estimates appear to be more accurate and reliable than the alternatives. As such, DRA allows us to declare how many runs a pitcher truly deserved to give up, and to say so with more confidence than ever before.

We now have graphs illustrating the influence of pitch location on batted ball type. They're not too surprising, though: Batted Balls: It's All About Location, Location, Location, by Jonah Pernstein, Fangraphs

Much like vertical location, BABIP and wOBA follow similar patterns to each other, and of course, a pitch in the middle of the zone is more likely to fall for a hit — and more likely to fall for an extra-base hit — than a pitch on the edge.

Thursday

If pitchers are consistently staying away from the strike zone against a certain hitter, it's a pretty strong sign of an impending or in-progress breakout: Fear Factor: How Pitch Location Helps Reveal Batter Breakouts, by Ben Lindbergh, Grantland

Pitchers don’t pound the zone against a player unless some combination of past performance, advance scouting, and pitcher’s intuition tells them they can get away with it, so a hitter’s average called strike probability alone tells us something about how big a threat he is. The graph below plots ISO against called strike probability for every batter-season in the PITCHf/x era with at least 1,500 pitches seen.

The trend is easy to see: The higher ISO is, the lower called strike probability tends to be, and vice versa. The correlation between ISO and CSP is -0.49, which puts them halfway between a nonexistent relationship and a perfectly intertwined one in which the two stats rise and sink like opposite ends of a seesaw. The correlation between CSP andTrue Average (TAv), BP’s all-inclusive offensive rate stat, is a slightly weaker but still significant -0.32. Of course, a hitter’s willingness to swing at pitches outside the strike zone plays some part in how far away pitchers work. The nine batter-seasons with the lowest average called strike probabilities all belong to one of three players: Pablo Sandoval, Vladimir Guerrero, and Alfonso Soriano, all of whom are known for a lack of plate discipline as well as a profusion of power. On the whole, though, the less pitchers fear a hitter, the more probable strikes he sees.

Trends, trends, trends, like, for example, that hitters are doing well against first pitches and that lefty hitters appear to be back on top: Closing the Lefty Disadvantage, by Matthew Trueblood, Baseball Prospectus

If nothing else, I would expect the dampened value of contact—again, April sees lower BABIPs and less power on contact than any other month, historically—to make taking pitches and getting deep into counts more rewarding early in the season. Instead, we’re going the other way, however narrowly. I’m willing to call this no change for the time being, but it’s strange. The difference lives mostly in the fact that players who take the first pitch are having a very difficult time using that patience to reach base at an increased rate. In that sense, it reads as a continuation of the trend we’ve seen recently, whereby pitchers have poured the ball into the zone so reliably that batters are better off simply attacking the first pitch they feel they can handle.

Friday

Being analytically-focused reflects favorable on a team's projected GDW — the value of its acquisitions and call-ups compared to the assets it got rid of: 2015 Front Office Preseason Evaluations, by Stephen Shaw, Banished to the Pen

Overall, the chart seems a little scattered, but there are a few trends that can be seen as well. A good majority of the “All-In” teams project to have a high gross domestic wins value. The “Nonbelivers” and “Skeptics” show the lowest predicted GDW numbers on the chart. These two categories also contain 3 out of the 4 lowest predicted GDW values.

Remember, GDW is trying to help us evaluate general managers by parsing out team winning percentage and only looking at each team’s controlled moves. Based on Steamer’s 2015 projected WAR numbers it would seem that up to this point analytic friendly teams are making better moves then teams who are not. Of course, the true test will be whether these projected values hold up.

The luxury tax appears to have curbed payroll spending, but revenues keep soaring, which is an unappetizing combination for the MLBPA: MLB's Evolving Luxury Tax, by Nathaniel Grow, FanGraphs

This evolving luxury tax framework has contributed to the reduction in the players’ share of overall MLB revenues in several ways. First, and perhaps most importantly, the fixed payroll thresholds have failed to keep up with MLB’s growing league revenues. While the luxury tax threshold was originally set at 90% of the average MLB team’s annual revenue in 2003, it has dropped to only 63% today:
...
In 2003, for example, the $117 million luxury tax threshold was just $13 million under the average MLB team’s annual revenue of $130 million. By 2014, that gap had increased to $111 million:
...
Even this comparison doesn’t tell the entire story, however. The primary purpose of the luxury tax was not – at least originally – to restrict the spending of the average MLB revenue team, but instead to curb the spending of MLB’s highest revenue franchises. Because the largest market teams generate revenue figures well above the league average, this means that the revenue gap between the luxury tax threshold and the teams it was primarily intended to restrict is now, in fact, much larger.