Why I use pitches/out and not pitches/PA

I got into a friendly Twitter debate a few weeks ago about whether pitches/out offers merits relative to pitches/plate apperance as a measure of pitcher efficiency. I strongly believe that it does, and I’d like to offer a couple of thought playful thought experiments as evidence.

Imagine two pitchers of opposite skill levels in the extreme — one really good and one really bad.

Let’s say the first pitcher is so amazing at his craft that no hitter is ever able to actually hit a ball he swings at, and he has such great control that he can throw the ball for a strike whenever he wants to. Naturally, every pitch this pitcher throws will be a strike.

And now let’s say the second pitcher is me. Now, in my defense, my last competitive appearance as a pitcher was quite successful. I pitched one relief inning, with one strikeout, one ground ball fielded cleanly with a throw to first, and one popup caught. All in that inning. Also, I was 12.

Today, my fastball probably tops out around 65 mph. From 60 feet, 6 inches away, I could probably hit the strike zone in one out of every ten tries. I would probably hit a number of batters, walk a great deal of others, and surrender long hits to the rest. The only way I could get outs is by sheer luck of BABIP or over-aggressive baserunning.

The first pitcher’s p/PA will always be 3. Mine will probably range between 2 and 4. In other words, there is not much difference, despite the monumental difference in talent level between the pitchers.

Meanwhile, the first pitcher’s p/out will also always be 3. Mine will probably range between 30 and 40. In other words, that gives you an idea of how much worse I am than that first pitcher. And that’s why I think p/out will give you a better idea of pitcher efficiency than p/PA.

Changes in Strasburg’s pitching grips since college

I stumbled upon an ESPN video of Stephen Strasburg from 2009 showing the grips on his four pitches. He still throws four pitches, but two of them different grips now. Here’s a quick look at those two:

Breaking ball:

In 2009, Strasburg called this pitch a slider. It’s now commonly referred to as a curveball, and the newer grip is certainly closer to a standard curveball grip. It has big sweeping action like a slider, but it drops a good amount, too, and is thrown significantly slower than his fastball.

“Sinker”

The other change is on what Strasburg called his sinker. Previously he used a non-standard “one-seam” grip. Now it appears he uses a standard two-seam.

David Phelps’s new (improved?) cutter

While digging around in Yankee pitcher David Phelps’s data, I noticed something a little unusual about his cut fastball: it looks like there’s two distinct “clusters” of pitches. Check it out for yourself on the right. (This graph includes data from the 2011 Arizona Fall League and 2012 MLB seasons. MLB Advanced Media calls his pitch a slider; I’m using Brooks Baseball’s designation as a cutter, which is supported by photographic evidence.)

I found that the division in spin signatures is something that can probably be explained by a mid-2012 stint Phelps had in AAA. (Phelps started the year in the big leagues, was demoted on June 15 to make roster room for David Robertson, then re-promoted when Andy Pettitte was injured two weeks later.)

Apparently in the short time he spent in the minors, he made some sort of adjustment to his cutter that is clearly visible by looking at the “before” and “after” spin deflections.

What adjustment he made, exactly, is unclear. In included a trendline to show that the pitch not only got more “rise” after his demotion, but its distribution across the chart is slightly different. That, coupled with the fact that the PITCHf/x data show no change in release point, invite the idea that he has slightly modified his grip or release action.

Due to an insufficient sample size, it’s hard to say if his cutter got markedly better because of this change. (He only threw 74 of them post-minors in 2012.) I’ll be watching this season to see how the adjustment turns out. In any case, Phelps’s cutter will be an important tool for him as he becomes a likely candidate for the back end of the Yankees’ starting rotation in 2013 and beyond.

Great thanks to MLB Advanced Media and Sportvision for the PITCHf/x data used in this post, as well as Harry Pavlidis for the Pitch Info/Brooks Baseball pitch classifications.

A scouting report on Rafael Soriano

Rafael Soriano has been a productive relief pitcher since 2003. He figures to close for the Nationals in 2013. However, he has experienced a number of arm injuries and is 33 years old. Soriano’s ERA is consistently quite low (career 153 ERA+), although there are risk factors for a drop-off in performance during his time in Washington.

Reportoire

• Four-seam fastball: 91–94 mph. In many respects, his best pitch. Opponents have hit .180 against it since 2007, ranking 5^th out of 173 relief pitchers who’ve thrown more than 1000 four-seamers since 2007. Lateral movement is minimal, often giving it the appearance of a cutter. Has good life, as shown by high whiff/swing rate (has never been below 23%). Velocity has dropped slightly in recent years, but he figures to stay pretty close to his 2012 numbers. Thrown fairly high in the zone, which has made him a fly-ball pitcher (0.63 GB/FB since 2007, per Brooks Baseball). Typically gets above-average frequency of pop-ups. He is not afraid to pitch up and in to righties, but he avoids throwing in to lefties.
• Slider: 80–86 mph. Used nearly as often as his fastball (> 40%). Thrown to hitters from both sides of the plate, typically low and away. Especially common against lefties with 2 strikes. Even with frequent use, its whiff rate is average (≈ 30%).
• Two-seam fastball: 91–94 mph. Used exclusively on lefties, especially early in the count. He runs it off the outside corner. Has decent tailing action relative to four-seamer, but little sink. Primarily a show-me pitch.

Defense-independent pitching

Soriano’s strikeout totals have been good-to-great throughout his career (26% K, 9.4 K/9). His walks fluctuate, winding up near the league average overall (8% BB, 2.9 BB/9). Home run rate is also about average (0.9 HR/9). That typically creates a good pitcher but not someone whose career ERA+ is 153. In fact his career ERA (2.78) is tiny compared to his career FIP (3.30) and xFIP (3.69). He also outpaces his SIERA (3.11) and tERA (4.06).

As a fly-ball pitcher, Soriano’s xFIP is somewhat “artificially” high because his home-run rate as a percentage of fly-balls is consistently better-than-average. Still, his ability in five of the last seven years (including his four best ERAs) to significantly outperform his FIP is significant and has a probable explanation.

It turns out that Soriano has benefitted hugely from an impressive career BABIP of .249 (FanGraphs) or .251 (Baseball Reference). In his masterful 2010 season, it was just .199. (In his previous, merely respectable season, it was .280.)

The last couple of years have seen a rebound of his BABIP to .279 and then .274. If this represents a trend, Soriano’s performance in the future could suffer. Last season, he had an anomalously high LOB% of 88% (he’s typically closer to 75–80%). It’s not hard to see how a middling walk percentage suddenly becomes more dangerous if hitters are starting to make better contact. In fact, Soriano’s ERA from 2009–2012 correlates strongly with his BB% (.94) and BABIP on his fastball (.63). Strikeouts had almost no correlation.

Potential red flags

• As discussed above, Soriano relies almost entirely on two pitches, and a key to his success has been making his fastball hard to hit well. He’s very effective against right-handed hitters partly because he can pitch them inside without fear. If Soriano’s velocity continues to decline, it could spell trouble for that strategy. Being a fly-ball pitcher with a hittable fastball and occasional tendency to walk hitters would be a bad recipe for a 9^th-inning guy. Soriano has thrown a changeup and cutter in the past, and if his fastball loses potency he might need to re-expand his repertoire.

• He has a long history of arm troubles, with a Tommy John operation in 2004, several DL stints for elbow inflammation, and various other maladies. His mechanics put strain on his elbow during the cocking phase, which explains the history of elbow issues. He also throws an unusually large percentage of his pitches as sliders, which is thought to be harmful to the elbow. As a 33-year-old, his body won’t heal faster than he did when he was 25 and had Tommy John surgery.

Other tidbits

• He’s not much of a fielder, either in range or graceful handling.

Outlook

For most of his career, Soriano has been a very good, albeit often overpaid relief pitcher. He has the mental fortitude to be a closer, and when he’s good he has the stuff for it. He will need to be very careful about his walk total and the effectiveness of his fastball. He is also at an above-average risk for arm injuries.

For the sake of preserving his elbow and to compensate for potential fastball velocity loss, perhaps the Nationals will encourage Soriano to rediscover his cutter or changeup.

This report was made possible due to the data at Baseball Prospectus, Baseball Reference, Brooks Baseball, and FanGraphs.

A scouting report on Dan Haren

From 2005 through May 2012, Dan Haren was about as reliable a starter as there was in the big leagues. In mid-2012, however, a back injury threw him off rhythm and led to his first DL stint. In five starts beginning June 9, Haren’s ERA was 8.67 with a .695 opSLG. Outside of those five starts, his season ERA was 3.55.

Repertoire

• Sinker: 88–91 mph. Used against lefties and righties alike. Thrown most often early in the count or when behind. Keeps it away from hitters, and tends to throw it fairly high in the zone. This could explain a below-average ground ball rate on the pitch. Velocity has dropped somewhat the last several years. Does not have a great amount of lateral or vertical movement.
• Cutter: 84–87 mph. Most common pitch, especially to righties. Pounds the low outside corner to righties and down and in to lefties. Has also lost velocity in recent years. Above-average lateral movement. Good whiff rate.
• Splitter: 82–86 mph. Third most-used pitch. Saw increased usage in 2012. Typically used when ahead and with two strikes. Has lost effectiveness as a swing-and-miss pitch since 2009 (from 40% to about 25%). Still has by far the lowest BAA and ISO of any of his pitches.
• Curveball: 76–79 mph. Thrown with a “spike” grip. Has gone from being a main weapon to a little-used “show me” pitch. Often used to “steal” strike one against lefties. Almost never used against righties at all. Not a great amount of depth, although it has some lateral movement.
• Four-seam fastball: 88–90 mph. Infrequent offering. Sometimes used as a two-strike pitch against lefties.

Defense-independent pitching

Haren’s most distinguishable asset is his control; he typically walks less than 2 batters per 9 innings. His strikeout totals are average, but his SO/BB ratio is still outstanding. He gives up a fair amount of home runs, however, so his FIP is good but not amazing. His career BABIP is .291, indicating no unusual trends in that department.

He has historically been a very durable pitcher; he started at least 33 games and logged at least 215 innings for seven straight seasons (2005–2011).

Potential red flags

• His solid mechanics should protect him against serious arm injuries, but the back stiffness issue was enough of a concern to lead the Chicago Cubs to scrap a deal for Haren early this offseason.
• His velocity has declined consistently for several seasons and could continue to do so in 2013. Haren has never been an overpowering pitcher, and pitchers can be effective without great speed. Still, all things being equal, faster pitches get hit less often. It will be essential for Haren to command his pitches, or else he will start getting hit hard.

Other tidbits

• For a pitcher, Haren is an excellent hitter. He is a career .223 hitter with 21 doubles in 264 at-bats.
• His fielding is decent. His range is limited, but he makes few errors (.972 fielding percentage).

Outlook

Haren is probably in line for a resurgent 2013 season. He will have the benefit of playing the National League again, an offense that should provide good run support, and a good bullpen behind him. If his back stays healthy, he will be a valuable addition to the Nationals rotation.

Baseball Reference, Baseball Prospectus, Brooks Baseball, and FanGraphs were eminently helpful in the making of this post.

Don’t vote for Todd Frazier over Bryce Harper as NL RoY

With the Washington Nationals having been eliminated from the playoffs, the next thing for their fans to anticipate is awards season. Gio Gonzalez is a strong contender for the Cy Young Award, and young Bryce Harper is the clear choice for National League Rookie of the Year. The voters atThe Sporting Newspicked Harper third, however, behind Arizona’s Wade Miley and Cincinnati’s Todd Frazier.

The choice of Miley is, I guess, defensible. He plays a premium position (starting pitcher) and performed objectively well in that role. Frazier, however, is not the National League Rookie of the Year. He had a solid offensive year but did not finish in the top 10 in any offensive category. His defense was unspectacular, finishing with -0.2 defensive wins above replacement, according to Baseball Reference. He finished at 1.7 WAR, or 2.2 if you extrapolate to 162 games.

Harper was an electric player in 2012. His raw stats — .270/.340/.477, 22 home runs, 9 triples, 18 steals, 98 runs, 8 outfield assists — are impressive enough. He also walks all over Frazier in the WAR department, at 5.0 overall, 1.4 on defense, and 5.8 on a 162-game basis. His baserunning ought to be marked even higher, but there is no real way to quantify all of the singles he turned into doubles, doubles he turned into triples, and hits he created by hustling down the line.

What makes Harper so compelling to me is that he was able to demonstrate his offensive prowess despite being treated by pitchers as though he was Mark McGwire in his prime. According to FanGraphs, using MLB Advanced Media’s PITCHf/x classifications, Harper saw fewer four-seam fastballs as a percentage of his pitches seen (26%) than any hitter in baseball. And the notoriously aggressive Josh Hamilton was also the only hitter who saw fewer pitches in the strike zone than Harper did.

The fact that, at 19, Harper was being treated as an elite power hitter speaks to a tremendous amount of respect from his peers. The fact that he could perform despite that obstacle speaks to his tremendous ability. And the unquantifiable energy he brought to a slumping Nationals offense in April is exactly the kind of “x factor” that can’t be captured by sabermetrics. In any case, Harper had the most impressive season by a teenager in decades.

Phil Hughes is pitching well: why?

Last summer, I highlighted the major hype surrounding the cut fastball, or “cutter,” that swept up the baseball world. It seems to have died down somewhat this year, and for good reason — it’s not a miracle pitch. In particular, I pointed to pitchers developing arm problems from throwing too many cutters — the prime example being Yankee starter Phil Hughes.

Hughes started throwing the cutter in 2009 and has relied on it heavily both as a starter and in his short tenure in the bullpen. His reliability as a pitcher has varied, including this year. But his last four starts have gone really well — three of them qualified as quality starts, and three of them netted him wins. The major difference is crystal clear: Hughes has dropped the cutter.

According to Texas Leaguers, Hughes threw the cut fastball about 12.4% of the time in April. Since the beginning of May, it’s below 2%. And tonight, Hughes threw zero cutters for the first time in a start since September 24, 2008.

I think the death knell for his cutter came on May 17, his last start against Toronto. His first 26 pitches were fastballs, and he only threw five cutters all game. Two of them resulted in base hits, and the last one Jose Bautista hit 353 feet to left field for a home run.

He hasn’t thrown one since.

Hughes has experimented over the years with numerous pitches, including a two-seam fastball and a slider. But perhaps he can succeed with his current repertoire of fastball, 12-6 curveball, and changeup. It’s worked so far.

Dwight Freeney suckage update

He had zero tackles yet again today, the fifth time this season. Wide receiver Reggie Wayne, with one tackle on an interception return, was more productive.

The amazing Freeney

Colts defensive end Dwight Freeney is such an impact player, he’s tied for 86th in the NFL among defensive linemen in tackles (and tied for 387th overall)!

Just a little observation

Vince Wilfork has more interceptions this season than does Darrelle Revis.