We’re focusing this week on post-All-Star break career stats for pitchers and also for hitters in order to at least consider some players who historically have appreciably improved and declined as the season has worn on.
These splits of course are arbitrary end points. So we can’t invest in them too heavily. However, it’s certainly theoretically plausible that some players wear down and others do not. After all, we have sprinters and marathoners in track and field, not just “runners.”
As with all data that splits up larger data, I like it way better for pitchers than for hitters. So let’s start there, with the pitchers who have ERAs at least 0.60 better or worse after the break than before the break.
Now the hitters, where we’re looking at only .100 OPS points gained or lost in the second-half, a very small group.
Let’s make a full season out of the all-star break and see which pitchers are doing better and worse 2015-to-2016 than perceived generally based on our other arbitrary end points.
The value of doing this, albeit it still being arbitrary, is that others are not. So we can see a big consecutive sample of data that others are ignoring, making it easier to bet on at a profit.
We’re sorting by ERA from the break last year to the break this year. And it’s for pitchers with at least 162 innings since the last All-Star break. The ones we skipped are the pitchers who we would expect to be ranked about where they are or the pitchers in the vast middle that are generally replaceable in mixed leagues. I’ve also included all the starting pitching categories.
Let’s look at well-hit average allowed with the help of our friends at Inside Edge , stat provider to MLB teams. But first let’s differentiate this from the other hit-quality metrics.
We started with line drive rate, which is limited by excluding many homers and also by excluding Ks. Additionally, many line drives are hardly well hit — we’ve all seen the flare to the middle infielder, i.e, the “soft liner.” Then we advanced to exit velocity and that’s superior when you make various adjustments to the angles, etc. And it’s averaged, which I don’t like because I want to know how many times a guy hit the ball hard similar to how many times a guy hit a homer.
Inside Edge does well-hit average, which includes strikeouts as at bats where the hitter failed to hit the ball well. Okay — common sense! However, we have a human element here, where scouts have to agree that a ball qualified as well-hit.
Homers allowed are a quandary for fantasy baseball owners. They are worth about 1.4 runs allowed every time they occur yet they are relatively isolated events even at the most extreme levels. So the question becomes whether they are fact or fluke. In other words, do we blame pitchers for allowing homers and, if so, how much?
Stats like xFIP just normalize a rate based on fly balls allowed for everyone, essentially stipulating that homers allowed are completely random. But this seems like an overcorrection; if pitchers are not to be blamed at all for homers, what are we even doing here in projecting them? Hits are random. Homers are random. Are strikeouts the fault of the hitter, too? Obviously this becomes ridiculous when you proceed even halfway down the slippery slope.
First, the 2016 leaders through Wednesday in HR/9:
So how much blame do we place on Ian Kennedy or Drew Smyly et al for these dingers? Let’s look at leaders through June in prior years and how they fared the balance of the season. Last year the worst 25 had an average HR/9 of 1.44 through June and they averaged 1.33 from July 1 on. This is bad news for xFIP.
The third time is not a charm in baseball. Meaning, the third time through the order. This year, pitchers allow a .778 OPS the third-time through the order compared with .717 the first time and .744 the second. Of course, individual mileage will vary. We want guys who are capable of pitching deeper into games because each out they get on their own increases the likelihood that a tied or trailing game can turn into a win.
While Scott Pianowski put values on starting pitchers the rest of the season earlier this week, we're going to focus on the pitchers who have proven to be much better than average — and much worse — the third-time around (minimum 80 at bats vs. batters ). You could of course note career averages for broader context but pitchers are really more “of the moment” performers versus hitters, who you can usually count on to level out to the back of their baseball cards.
We’ve isolated pitchers who are either very good (OPS allowed .630 or lower) or really bad (OPS allowed .900 or greater). Stats are through Wednesday.
We have enough 2016 data to put closers under the stat microscope to form a power ranking for all of those who have at least five saves through Wednesday, regardless of whether they’re currently in the job.
We’re using two metrics to accomplish this. The first is of course (K-BB)/IP because dominance in this area is arguably more important in the one-inning role than it is overall. The second is ISO Allowed (ISO is slugging average minus batting average), which is a proxy for hit quality because you of course want your closer to have to allow three hits to give up a run.
We’ve made an index of just these two stats. That’s simply adding the ranking of each of the 34 qualifiers in (K-BB)/IP and also in ISO Allowed and then adding those rankings, so the lower the sum of these rankings, the better the pitcher.
Today we isolate the recent month-long sample for starting pitchers to examine their (strikeouts-walks)/inning pitched.
I’d never do this with hitters because I can’t really think of a reason generally why a hitter would greatly improve in a slice of a season beyond a natural peak-age progression. Of course, hitters do radically improve and sustain that improvement but I don’t view it as something you can confidently bet on at its outset. But there are so many reasons why a pitcher can improve or decline for at least the remainder of the season including refining a pitch, developing a new one, eliminating a pitch, optimal health, injury, a mechanical change that significantly boosts or reduces velocity, a sudden emphasis or fear of throwing inside, etc., etc.
Plus the best part of isolating current samples is that they are masked by the full-year where perhaps some of the issues mentioned above were conversely in play.
Finally, just drop Jaime Garcia, Marcus Stroman and Francisco Liriano.
We’re going to take a short break from pitching this week and use our Inside Edge well-hit metrics to look at select hitters in the context of the MLB-wide average rate of .138.
While that’s the main statistic, I also noted each hitter’s slugging percentage through Wednesday and also their rate of quality at-bats. Of course a quality at-bat is any at-bat resulting in a hit or walk, but also includes well-hit outs and any at bat regardless of outcome that lasts at least seven pitches. The MLB average for this is 39 percent. (It also includes sacrifices and hit by pitch.)
Ryan Zimmerman is hated due to declining production and the ever-present injury risk and is just 38 percent owned but we see that his declining hitting is a bit of a myth. He’s also on pace for about 24 homers and 80 RBI with a healthy number of runs scored. Well-hit says his .277 BABIP is way low. I’d project him to hit about .275 going forward.
Michael Salfino at Special to Yahoo Sports 4 mths ago
We’re told that batting average allowed is largely luck, making so much bottom-line pitching performance (ERA and WHIP) largely random. This is depressing. But the solution isn’t to curse that our faults lie in the stars but rather to strive to better isolate luck by cross-checking batting average against contact type.
We use well-hit data here from Inside Edge, where scouts review each batted ball for whether, to their eyes, it was well struck. Balls out of play — homers — are counted as they should be as well-hit. Grounders are assessed. Not all line drives are well hit. Strikeouts count because the stat is tethered to at bats. The MLB well-hit average this year is .134, meaning that pitchers allow batters to hit the ball well 13.4 percent of the time.
Michael Salfino at Special to Yahoo Sports 4 mths ago
We’re going to focus on relievers this week in Pitching by the Numbers, using a combination of obvious run prevention and less obvious dominance and control — irrespective of saves.
But first a statistical analysis of Matt Harvey focusing on the central question of whether the game’s most disappointing pitcher to date is a buy-low candidate.
Let’s start with FIP, where his ERA now sits at an expected 3.66. His actual ERA of 5.77 is driven by an absurd .390 BABIP. While line-drive stats say this is earned given that clocks in at 34 percent, I’m not a fan of this stat. There are many soft line drives.
Hector Rondon not having save opportunities is a fluke. He’s arguably the best closer after Aroldis Chapman, who is just the master of our surplus K stat. Craig Kimbrel is worth way more than the conventional metrics that so heavily weigh his completely non-predictive ERA. While Mychal Givens is looking very closer worthy, his path is blocked by the solid Zach Britton.