The End of the ERA Era? Predicting Pitching Performance

A research study that I performed in my applied statistical modeling class this semester. Enjoy!

Advertisements

Abstract:

Athletes in today’s world garner huge sums of money. For this reason, teams need to make informed decisions on how to spend their money effectively. They need accurate measures for assessing the true talent of their players. A lot of the variation in year-to-year performance is due to luck, but there is a significant element of skill as well. I attempted to isolate the proportion of pitching performance that was due to luck and skill, respectively, through simple and multiple linear regression analysis on ERA and other performance statistics. I did this using several models that utilized various combinations of luck-based and skill-based predictors.

I found that a slightly higher proportion of pitching performance is due to skill than luck. I then created my own model for predicting ERA using significant skill indicators, and compared it to existing ERA estimators, where it fared decently. The most significant skill indicator was a pitcher’s strikeout minus walk rate. The two most significant luck indicators proved to be batting average on balls in play (BABIP) and home run to fly ball ratio. To my surprise, left on base percentage, which measures the rate at which pitchers leave runners on base at the end of an inning without them scoring, was not entirely luck based. This result indicates that some pitchers step up when there are runners on base, and others choke under the pressure of an imminent score.

Background:

There is a lot of money in sports, for both the players and the management. The management wants their team to win, so that their games get sold out and they can make lucrative television deals. Thus, they go out looking to recruit the right players to win, and they recruit them mostly by offering them large sums of money. With all of this money involved, there arose a need for accurate evaluation of talent. Scouts, who observed players, were often biased and inaccurate. Statistics, on the other hand, eventually proved to be an unbiased and accurate method of player evaluation. “The Moneyball Era” of the early 2000’s saw the rise of statistical analysis in baseball (Sabermetrics).

Many statistics had existed for decades, but as it turned out, most were luck-based, unimportant in terms of run scoring or run prevention, and/or fluctuated wildly from year to year. Sabermetricians focused on finding and creating statistics that were relatively stable across seasons, and identifying those that were not. Statistics that were stable for each player (although they differed between players) across seasons indicated a degree of reliability, and could tell analysts something about the true talent of a player.

The most important statistic for pitchers in terms of run prevention, their main goal, is earned run average, or ERA. ERA reports the average number of earned runs (i.e., runs that did not score with the help of a fielding error) a pitcher allows per nine innings (the typical length of a baseball game). ERA is an unbiased statistic; it measures an objective outcome, an outcome which is very important. However, ERA does not have a high correlation from year to year; there is a certain degree of luck involved. How well can we predict ERA from batted ball and pitch data? Or is too much of the variation in ERA due to luck for us to even come up with useful predictions?

Methods:

Data was gathered from the customizable leaderboards on the FanGraphs website. Because the leaderboards were customizable, I did not have to code for any new variables except for one, NL, which indicated which league the individual pitched in that year. The rest of the variables are as follows: ERA, HR9, K, BB, KminusBB, WHIP, BABIP, LOB, Oswing, Zswing, Zone, SwStr, IFFB, HRFB, LD, Hard, Soft, Medium, GB, FB, GBFB, Team, and Name (of the player). Each of these variables (except for Name) appears twice in the dataset, once for 2017 and once for 2018. The dataset contained data from the 140 pitchers who threw at least 100 innings in 2018 and the 134 pitchers who threw at least 100 innings in 2017. Because there was some overlap between years, there were a total of 189 pitchers in the dataset.

Exploratory analysis was performed; normality was assessed for outcome variables. A single model containing all of the predictor terms was built to compare the effects of each predictor on ERA using their p-values, and it was adjusted accordingly for multicollinearity. Year-to-year correlation for each predictor term was used to determine which were luck-based and which were skill-based. The NL variable was excluded from this analysis because, while it has a high year to year correlation, the effect of the league in which a pitcher pitches is out of his control, and so this variable is luck-based.

Predictors with a year-to-year correlation (R-squared) of 0.400 or more were put in the skill-based category, while predictors with a year-to-year correlation of 0.010 were put in the luck-based category. Predictors with an R-squared between 0.400 and 0.010 were left without a category, since they contained some element of both skill and luck. These two categories were used to make two separate models to compare the effects of skill versus the effects of luck on ERA. Each model was adjusted accordingly for multicollinearity. The skill-only model was then edited using backward elimination to create a new estimator for ERA using only significant (at the alpha = 0.05 level) predictor terms. This new model was then compared to existing ERA estimators.

Results:

After eliminating certain predictors to reduce multicollinearity, the luck-based indicators (those with a year-to-year correlation less than 0.01 or otherwise deemed as luck-indicators) turned out to be BABIP (R-squared: 0.003), HRFB (0.003), LD (0.006), and NL (N/A), and the skill-based indicators (those with a year-to-year correlation more than 0.40) were KminusBB (0.51), GBFB (0.66), Zone (0.56), Zswing (0.52), and Oswing (0.43). Variables in between the thresholds were LOB (0.13), ERA (0.11), HR9 (0.11), WHIP (0.01), IFFB (0.05), Hard (0.15), Medium (0.02), and Soft (0.11). Table 2 lists year-to-year correlations for each predictor.

The two most significant luck-based terms were BABIP (p-value in reduced full model: < 2eˆ-16) and HRFB (< 2eˆ-16). BABIP measures batting average on balls in play, or how often the pitcher allows a hit on balls that a fielder can make a play on (i.e., not home runs or strikeouts). HRFB, home run to fly ball rate, divides the total number of home runs a pitcher allows by the total number of fly balls they allow. The two most significant skill-based terms were KminusBB (p-value in reduced full model: 1.26*eˆ-15) and Oswing (0.02). KminusBB is the percentage of batters a pitcher strikes out minus the percentage of batters they walk. Oswing is the percentage of a pitcher’s pitches outside of the strike zone that batters swing at. However, in the reduced skill model, Zone (p-value: 0.05) took the place of Oswing as the second most significant skill-based indicator. Zone is the percentage of pitches a pitcher throws in the strike zone.

Other predictors that were significant in the reduced full model for ERA but not significant in any other models (if they were used in any other models) were LOB (p-value: < 2eˆ-16), GBFB (p-value: < 2eˆ-16), and NL (p-value: 0.03). LOB, left on base percentage, indicates the percent of runners that get on base that a pitcher leaves on base without allowing them to score by the end of the inning. GBFB is the number of ground balls a pitcher allows divided by the number of fly balls a pitcher allows. NL indicates whether a pitcher pitched the full year in the National or American League. Table 3 lists p-values of predictors in the reduced full model. All predictors in the reduced full model had negative coefficients except for BABIP (9.864), Zswing (0.0012), and HRFB (0.1004)–one-unit increases in these three predictors would raise ERA. The full model had the lowest mean of squared errors with 0.03. It was closely followed by the reduced full model (0.05). The skill and luck models were close, with MSE’s of 0.49 and 0.50, respectively. The reduced skill model (MSE: 0.49) stacked up fairly well against existing ERA estimators, beating out SIERA (0.51) and xFIP (0.51). However, FIP (0.33) and tERA (0.34) were much more accurate in predicting ERA. Table 1 lists the predictive values of each model.

Discussion:

My most surprising finding was that LOB, left on base percentage, correlated relatively well from year to year. It exceeded the threshold that I set for the luck category. LOB is generally considered purely luck by Sabermetricians. This result implies that some pitchers are better with runners on base than others. This could be because some thrive under the pressure of having runners close to scoring, while others choke, an explanation most Sabermetricians scoff at. Another potential explanation is that strikeout pitchers have lower LOB because when a batter puts the ball in play with a runner on base, they can advance the runner, even if they get out (like with a sacrifice bunt). Strikeout pitchers prevent batters from putting the ball in play when they strike them out. Yet another potential explanation is that worse pitchers have lower LOB’s because they give up hits and walks more frequently. All of these explanations can explain why some of the variance in LOB is not random.

The coefficients of the reduced full model were not particularly surprising. Interpreting significant coefficients, we have the variables KminusBB, BABIP, LOB, Oswing, HRFB, GBFB, and NL. Pitching in the National League should indeed lower ERA because the pitcher hits instead of another hitter like in the American League, and pitchers are not good hitters. GBFB ratio should indeed lower ERA as it increases, because ground balls are less likely to go for doubles, triples, and home runs. HRFB ratio should indeed increase ERA as it increases, because the higher it is, the more home runs a pitcher has against them. Oswing should indeed lower ERA as it increases; hitters generally do not hit well when they swing at bad pitches outside of the strike zone. LOB percentage should indeed lower ERA as it increases; the more men you leave on base without scoring, the better. BABIP increases ERA as it increases; the more hits that fall in against a pitcher, the more runs will be scored against them. KminusBB lowers ERA as it increases because as it increases, strikeouts are being maximized and walks are being limited. Zone rate was not found to be significant here, although it was significant in the skills-only model. Pitching in the zone can be good and bad; while it is generally good to throw strikes because walks will be limited, if a pitcher only throw pitches right down the middle, then the batters will start hitting them. Hard percentage was also not found to be significant in the reduced full model, probably because it does not correlated so well from year to year and as such, is primarily due to luck. Zswing was the last term, and it was found to be insignificant, which makes sense. While swinging at pitches outside of the strike zone is almost always bad, swinging at pitches in the strike zone can be good or bad depending on which pitch you are swinging at. If it is a fastball, you are more likely to get a hit, but if it is a curveball, you are less likely.

The skills-only model had slightly better predictive ability than the luck-only model. However, there are limitations on this result. The thresholds I set for the luck and skill categories were arbitrary. In addition, some significant terms were left out of the models because of the gap between the thresholds. Most importantly, LOB percentage was left out of the luck and skill models because it fell between the thresholds. It had excellent predictive ability on ERA in the reduced full model (p-value: < 2eˆ-16).

For the reduced skill model, I am not surprised that KminusBB turned out to be the most significant term (see Figure 1). Strikeouts are vital, because if the batter cannot put the ball in play, they cannot get a hit or advance potential runners. If walks are limited while strikeouts are maximized, then the hitter is left with no easy way to get on base. The reduced skill model produced a predictive value on ERA comparable to xFIP and SIERA. However, tERA and FIP fared much better. FIP only includes home runs allowed and does not account for HRFB ratio, so while this may result in better predictive value, it is less skill-based, because the amount of home runs allowed fluctuates based on HRFB ratio. Meanwhile, tERA and SIERA rely heavily on batted ball data, which I deemed insignificant in the reduced skill model (i.e., GBFB ratio) or primarily luck-based (Hard, LD, Soft, Medium, IFFB). xFIP has a comparable formula, and so a comparable predictive value. I am not saying that my model is better than these existing models. It instead serves as a simplification of the complicated relationship between luck and skill in baseball. I left out the R-squared adjusted for the ERA estimators because their full formulas with all of their predictors was not readily available, so the R-squared adjusted did not apply.

Conclusion:

To answer our initial question, we can predict ERA from batted ball and pitch data very well. Our full model had an MSE of 0.03125–it was on average off by around 0.03 points of ERA–a miniscule amount compared to the league average of around 4.00. Even our reduced full model, adjusted for multicollinearity, produced an MSE of just 0.05268. However, our second question is more complicated. While much of the variation in ERA is due to luck, we determined for our data, using arbitrary categorizations, that slightly more of the variation in ERA is due to skill. This result is probably not generalizable because of the arbitrary nature of the test, and the fact that some important predictors (i.e., LOB) were left out because they did not clearly fit in one category. Still, our reduced luck model managed a decent MSE of 0.4925. The difference between an ERA of 4.00 and an ERA of 3.51 is more significant than we would like, but it is pretty close (see Figure 2), and half a run every nine innings will usually not make or break a team’s chances. So, our reduced skill model does have some predictive ability after all, and should help to discern a pitcher’s true talent level.

 

Tables:

Table 1

Screen Shot 2018-12-19 at 12.47.54 PM.png

Table 2

Screen Shot 2018-12-19 at 12.47.38 PM.png

Table 3

Screen Shot 2018-12-19 at 12.48.10 PM.png

Figures:

Screen Shot 2018-12-19 at 12.48.21 PM.png

Screen Shot 2018-12-19 at 12.48.28 PM.png

The two scatterplots depicted in the figures are almost exactly inversely related. This is because the reduced skill model relied primarily on KminusBB.

References:

McCracken, Voros. “Pitching and Defense: How Much Control Do Hurlers Have?” Baseball Prospectus, 23 Jan. 2001, http://www.baseballprospectus.com/news/article/878/pitching-and-defense-how-much-control-do- hurlers-have/.

“GB%, LD%, FB%.” FanGraphs Baseball, http://www.fangraphs.com/library/pitching/batted-ball/.

MacAree, Graham. “Sabermetrics 101: Pitching.” Lookout Landing, 3 Mar. 2010, 12:00PM ET, http://www.lookoutlanding.com/2010/3/3/1334154/sabermetrics-101-pitching.

“Plate Discipline (O-Swing%, Z-Swing%, Etc.).” FanGraphs Baseball, http://www.fangraphs.com/library/pitching/plate- discipline-o-swing-z-swing-etc/.

Beneventano, Philip, et al. “Predicting Run Production and Run Prevention in Baseball: The Impact of Sabermetrics.” ResearchGate, June 2012, http://www.researchgate.net/profile/Bruce_Weinberg/publication/266344641_Predicting_R Run-Production-and-Run-Prevention-in-Baseball-The-Impact-of-Sabermetrics.pdf.

My Hall of Fame Ballot

How do I measure baseball greatness?

What defines a Hall of Famer? In my opinion, the two most important characteristics of baseball immortals are longevity and relative excellency. To measure both of these, I am going to use career WAR (based on the FanGraphs formula) as a reference point. Career WAR takes into account longevity, as it is a cumulative statistic, and it takes into account relative excellency, as it is adjusted based on how well the league itself performed each year.

Chipper Jones (Career WAR–84.6)

Chipper owns the highest career WAR out of any player on the ballot this year who didn’t take steroids. His 84.6 career WAR ranks sixth all-time amongst third basemen. He averaged 4.7 WAR in his eighteen years as a regular. Some might suggest this was the product of playing a lot of games and getting a lot of PAs. Although he was durable–Chipper averaged 138 games and 589 PAs in those eighteen years–he also was excellent; his career wRC+ of 141 ranks tied for ninth all-time amongst the 588 third basemen with at least 1000 PAs. Keep in mind the fact that Chipper did it for longer than most on that list, with 10614 PAs–in fact, he had the most PAs of the top 17, discounting A-Rod, a steroid user. One of the players he is tied with is A-Rod. Furthermore, he is behind Edgar Martinez, who only started 533 games at third base (as opposed to Chipper’s 1970). He has the second highest offensive runs above average, behind only A-Rod, in that group, and there is an argument to be made that using offensive runs above average is the best all-around measure because defensive statistics can be unreliable. Some other stats of note: for his career, Chipper walked more than 1000 times more than he struck out, and had a .401 career on-base percentage. He stole 150 bases at a 76.5% success rate. Chipper Jones is a surefire first-ballot Hall of Famer.

Mike Mussina (Career WAR–82.2)

Mussina has the next highest career WAR total on this year’s ballot. His total of 82.2 WAR ranks 16th all-time amongst pitchers (15th if you discount Roger Clemens, who probably used steroids). This is especially telling of Mussina’s excellency because he has the lowest career innings total amongst those 16. This is not due to durability issues, either: Mussina averaged 31 starts (and 4.7 WAR) during his seventeen full seasons. Rather, it is a matter of the era that he pitched in, which saw pitchers accruing fewer starts. The era he pitched in has adversely affected him in other ways, too. I think that the reason Mussina hasn’t received much love on the ballot is because his career ERA of 3.68 (FIP of 3.57) is considered underwhelming. I think the higher ERA is also a product of the era in which Mussina pitched: the steroid era. ERA is not a good measurement in this sense. Instead, league adjusted stats (like WAR), which take into account the higher offensive output of the era, should be used.

Jim Thome (Career WAR–69.0), Edgar Martinez (Career WAR–65.5)

I lumped these two guys together because they were incredible hitters, but poor defenders. As I said previously, though, I weigh offensive metrics for hitters much more heavily than defensive ones. The reason for this is that most, if not all, offensive events in baseball are discrete and can be measured probabilistically. Defensive events, however, are not. A player’s defensive value is determined based on the plays he made and the plays he could have made. “The plays he could have made” has always been a relatively subjective way of measurement. Some people might think it was possible for him to make that play, while others may not. However, statistics like wRC+, used to quantify offense, are purely objective. While I don’t think that defense should be entirely ignored, I think that it should be weighed far less significantly than offense.

All of that being said, we have two of the best pure hitters ever on the ballot this year.

Of the 431 players with at least 7000 PAs (longevity!), Martinez ranks 25th and Thome ranks 30th in terms of wRC+ (relative excellency!).

Thome was one of the greatest home run hitters of all time. His 612 dingers places him 8th on the list. His career OBP was .402. His career ISO of .278 ranks seventh out of those 431. Thome is a first-ballot Hall of Famer by tradition means and by advanced measures.

Martinez is different. He is adored by Sabermetricians for walking more than he struck out in his career, his astounding wRC+ of 147 (higher than even Thome’s 145), and his incredible career OBP of .418. However, he is not loved by traditional statisticians, since he was a DH for a while, has a middling home run total, and limited accolades. Nevertheless, I think he deserves to be in the Hall.

Scott Rolen (Career WAR–70.1)

While Rolen’s eight Gold Gloves signify that he was known primarily as a glove-first third baseman, he also should have been know for his offensive prowess and longevity. Don’t get me wrong, he was an excellent defender, posting positive defensive runs above average in every year except his rookie year in which he played only 37 games. But he also had a career wRC+ of 122 and an ISO of .210. He’s tied for 20th on the all-time home run list for third basemen (Chipper is fifth, by the way). Discounting his rookie audition, Rolen averaged 4.4 WAR over 16 years. Excellence and longevity.

 

That’s it for my ballot. I’ll write a post soon talking about my snubs, don’t worry! If you want to read some more about the unreliability of defensive metrics and attempts to better quantify defense, try this link:

http://www-stat.wharton.upenn.edu/~stjensen/research/safe.html

Data from FanGraphs and Baseball Reference. Picture from MLB.com. Thanks for reading!

Chad Kuhl is Throwing Heat

On the surface, Pirates’ starter Chad Kuhl appears to be having a disappointing season. However, there is more to it than just that.

Generally speaking, a slash line (ERA/FIP/xFIP/SIERA) of 5.58/4.28/4.77/4.80 isn’t very encouraging. These are the numbers that Pirates’ starter Chad Kuhl has put up to date this season through 69.1 innings. Last year, he threw 70.2 MLB innings, so we have comparable sample sizes. Yet, he seemingly hasn’t improved upon last years numbers. Yes, the strikeouts are up, from a 17.6% K-rate to a 19% this year. However, the walk rate is also up (6.6% to 9%), the ground balls are down (44.3% to 41.8%), and the home run rate has risen accordingly (0.89 HR/9 to 1.04). What, you may be wondering, do I see in this guy?

Check out his plate discipline stats.

Season O-Swing% Z-Swing% Swing% O-Contact%
2016 26.40% 68.80% 45.00% 65.70%
2017 30.50% 65.90% 46.40% 57.40%
Season Z-Contact% Contact% Zone% F-Strike% SwStr%
2016 87.50% 80.30% 43.90% 57.10% 8.90%
2017 85.30% 75.20% 45.00% 59.40% 11.40%

Improvements across the board. His chase rate has gone up while his in-zone swing rate has gone down. Hitters are making far less contact on pitches out of the zone, and even a bit less on pitches within the zone. This explains the increase in strikeouts. The walks shouldn’t be increasing, unless hitters are really going much deeper into counts, since they are making less contact. Nonetheless, this should change if Kuhl keeps things the same, because he’s throwing in the zone more often and getting more swings outside of the zone. Of the 118 pitchers who have thrown at least 60 innings this year, Kuhl’s chase rate ranks 43rd, his in-zone swing rate is tied for 48th lowest, his z-swing minus o-swing ranks 37th, and most impressive, his swinging strike rate is tied for 26th. In fact, his swinging strike rate is the same as Yu Darvish–he even has a higher chase rate than him (30.5% and 29.3%), and Darvish has a superb 26.9% strikeout rate. The underlying statistics are optimistic, so if Kuhl keeps pitching this way, the strikeouts will increase and the walks will decrease. The bigger question is, what is the driving force behind these improvements?

According to PITCHf/x data on FanGraphs, Kuhl’s average 4-seam fastball velocity has jumped from 93 last year to 95.5 this year, touching 99. Contrary to what his name might suggest, Chad Kuhl is throwing heat. In fact, all of his pitches have seen an increase in velocity (and he’s added a curveball, but he’s only thrown 38 of them and they have been largely ineffective):

Season Pitch minVel maxVel Vel
2016 SI 83.3 96.5 92.7
2016 SL 81.6 89.5 86.6
2016 FA 87.4 96.1 93
2016 CH 81.6 88.3 85.1
2017 SI 88.6 99.5 94.1
2017 FA 90.2 99.4 95.5
2017 SL 77.2 91.8 88.5
2017 CH 81.7 90.7 88
2017 CU 79.7 86.4 82.7

The velocity increase has given Kuhl more confidence in his four-seamer, and his usage of the pitch has risen to 29% this year, up from a mere 10% last year. This explains part of why the ground ball rate is dropping–the uptick in four-seamer usage has caused a drop in sinker usage (down from 57% last year to 37% this year).

In addition, while his sinker has seen an increase in arm-side run (1.6 inches more), the ground ball rate is also dropping because the sinker has seen a decrease in drop (1.1 inches less). While the drop on his sinker has decreased, the rise on his four-seamer has increased. It is now above average, ranking 52nd out of the 118 pitchers who have thrown at least 60 innings as of morning June 27th. This is in part due to a slight change in vertical release point:Brooksbaseball-Chart-16.png

This year, Kuhl is throwing more over the top with all of his pitches. This graph shows that, for his sinker, he is on average releasing the ball about two inches higher. Now, Pitch Info (which powers this graph) says that Kuhl doesn’t throw a four-seamer at all, only sinkers, as opposed to PITCHf/x. Either way, at this point, Kuhl’s “sinkers” don’t sink very much. Using Pitch Info’s data, Kuhl’s sinker has the eighth worst drop amongst the 87 starters who have thrown at least 200 sinkers this year. In that same group, the ground ball rate on Kuhl’s sinker is also eighth worst. Coincidence? I think not. His overall ground ball rate of 41.8% this year is below average, ranking 78th-lowest of the 118 pitchers who have thrown at least 60 innings this year.

All of his pitches are generating more whiffs, looking at both Pitch Info and PITCHf/x. This is probably due to the improved velocity. Using Pitch Info’s data, his slider ranks 15th in whiffs per swing out of the 87 starters who have thrown 100 sliders this year (not to mention, it ranks 10th in average velocity), and his sinker ranks 17th out of the 87 starters who have thrown 200 sinkers this year. However, his changeup still gets whiffs at a below average rate: it ranks 71st out of the 92 starters who have thrown 100 changeups this year. Although the changeup has gotten more run this year, it too has lost vertical drop and the velocity gap between it and the fastball has closed a bit. Generally, changeups are used to sit down batters of the opposite handedness, because they have arm-side run. Kuhl, a righty, has struggled against lefties this year, as they have a .445 wOBA against him, while righties have a mere .286 wOBA. At the same time though, he has gotten more strikeouts against lefties (30) than righties (29), despite having faced fewer lefties (147) than righties (163). Also, I’m not too worried that Kuhl will have struggles against lefties in the long run because his sinker has great arm-side run.

The fact that Kuhl has a diminished ability to get ground balls doesn’t bode well for his old skill set, where he relied on his control and inducing weak contact, but with an increased penchant for strikeouts, backed by improving velocity, it shouldn’t matter that much. I would still take a flyer on him; the strikeouts, walks, and platoon splits should improve, along with his ERA.

Data from FanGraphs, Brooks Baseball, and Baseball Prospectus. Picture from MLB.com. Thanks for reading!

Some Preseason Thoughts: American League

An in-depth look at the favorites to win each division and each wildcard.

So here we are, in the midst of opening week. I’m going to outline the favorites for this season, based on who was truly good last year and who made the best win-now moves this past offseason.

AL East

Favorite: Boston Red Sox

Mookie Betts accumulated 7.8 WAR and Xander Bogaerts had a career-best walk rate and ISO in their age-23 seasons. Andrew Benintendi is healthy and geared up for his first full season in the majors. Dustin Pedroia and Hanley Ramirez enjoyed bounce-back campaigns, the former garnering 5.2 WAR and playing 160 games, the latter with a 127 wRC+. Jackie Bradley Jr. proved his 2015 success was no fluke. The acquisition of Chris Sale offsets the injuries to David Price and Drew Pomeranz. They have plenty of rotation depth: Rick Porcello is a solid presence (although, some regression is to be expected) and both Eduardo Rodriquez and Steven Wright offer some sneaky upside.

Furthermore, it’s worth noting that last year, although they won 93 games, Baseball Prospectus said that they should have won 103 games (more than anyone else in the AL) based on runs scored and runs allowed, amongst other underlying statistics, and adjusted for strength of schedule. FanGraphs projects them to tie with the Astros for most wins in 2017.

AL Central

Favorite: Cleveland Indians

Carrying over from last year is a solid young core of Francisco Lindor, Tyler Naquin, and Jose Ramirez, backed by consistent veterans Carlos Santana and Jason Kipnis. The signing of Edwin Encarnacion and the return Michael Brantley further boosts this offense. As for the pitching, the big three of Carlos Carrasco, Corey Kluber, and Danny Salazar are returning, fully healthy. They are backed by a couple of solid arms in Trevor Bauer (whose excellent stuff still offers upside in his age-26 season) and Josh Tomlin. Furthermore, there are two solid prospects with major league experience: Mike Clevinger struck out more than a batter per inning in Triple-A last year, and Ryan Merritt was a postseason hero. Not to mention, a very strong bullpen composed of Andrew Miller, Cody Allen, Bryan Shaw, and the newly signed Boone Logan.

Andrew Miller’s ranks amongst the 133 qualified relievers last year:

K% BB% SIERA ERA FIP xFIP
1st 2nd 1st 2nd 3rd 1st
WAR O-Swing% Z-Swing%
2nd 2nd 1st (Lowest)

AL West

Favorite: Houston Astros

Despite a disappointing season last year, the Astros have an improving young core. Carlos Correa just put up 4.9 WAR in his first full season, which he seemed to be playing injured throughout. And he’s only 22! They’re getting a full season of Alex Bregman, who put up a 112 wRC+ in his first big league action. Jose Altuve, still only 26, had a career year, posting bests in walk rate, ISO, WAR, wRC+, OBP, Slugging, and more. George Springer, still only 27, played a full season for the first time, putting up a 124 wRC+ for the third straight season and garnering 4.5 WAR. The Astros don’t only have youngsters, though; they also improved their catching with the acquisition of veteran Brian McCann, and now have a respectable tandem of McCann and Evan Gattis. They’re set for their first full season of Yulieski Gurriel, the 32-year-old Cuban. They also signed three more solid veterans this offseason: Nori Aoki (career .353 OBP), Carlos Beltran (124 wRC+ last year), and Josh Reddick (career wRC+ of  105 and positive defensive marks).

On the pitching side of things, after struggling with shoulder issues for the last couple of years, Lance McCullers (30.1% K-rate last year) is healthy to begin the year. While Dallas Keuchel disappointed last year, his underlying metrics (3.87 FIP, 3.53 xFIP, 3.77 SIERA) suggest that he suffered from some bad luck. Joe Musgrove provided 62 solid innings in his MLB debut last year, and still has room to grow at age 24. Charlie Morton, the oft-injured veteran, showed much improved velocity last year in a small-sample and topped out at 97 this spring (he sat 91-92 in previous years) with his sinker, so he offers some sneaky upside. Collin McHugh is starting the season on the DL, but he has garnered at least 3 WAR in each of the last three seasons. The ‘Stros have some depth beyond those five: Mike Fiers can be an innings eater with the potential to showcase the swing-and-miss stuff that he showed before last year, Brad Peacock struck out a batter per inning in 117 Triple-A innings last year, and Chris Devenski, bullpen ace, started five games last year, putting up a 2.16 ERA in 108.1 innings (mostly in relief).

The bullpen is loaded. Luke Gregerson led the MLB in swinging strike rate last year, Ken Giles has a career 34% K-rate and an only 8.2% walk-rate, Will Harris and Tony Sipp are two lefties who have struck out more than a batter per inning in their careers, James Hoyt had a 2.96 SIERA last year and Michael Feliz had a 2.45. Not to mention, Chris Devenski.

Beyond the obvious depth on the MLB team, the Astros have a solid farm system. They have 9 top-100 prospects, according to KATOH, the stat-based prospect ranking system on FanGraphs. Among them are familiar names such as outfielder Kyle Tucker (119 wRC+ in A-ball last year, 188 in 69 PAs in High-A), David Paulino (1.83 ERA in 64 Double-A innings last year), Francis Martes (3.33 ERA, 2.73 FIP in 125.1 Double-A innings last year), and A.J. Reed (142 wRC+ in 296 Triple-A PAs last year).

Wildcard

Favorite: Toronto Blue Jays

While they lost Edwin Encarnacion, much firepower remains. Josh Donaldson is still Josh Donaldson. Kevin Pillar is one of the best defenders in baseball, accumulating the most defensive runs saved above average out of every outfielder the last two years. I’m expecting a bounce-back from Jose Bautista, who played through injury last season. Devon Travis, who has 4.8 WAR in only 163 career games, is fully healthy to start the season. Russell Martin provides a steady presence behind the plate. Even though his offense has declined over the last two years, Troy Tulowitzki still provides upside at shortstop and defends well. The signings of Steve Pearce (136 wRC+ last year) and Kendrys Morales (whose homerun power should play up  at Rogers Centre) should help offset the loss of Encarnacion.

Their stellar rotation from last year remains intact. Although J.A. Happ, Marco Estrada, and Aaron Sanchez are due for some regression, Marcus Stroman was unlucky last year (his ERA-FIP was the 10th highest amongst qualified pitchers last season). It’s also worth noting that Estrada, with a superb rising fastball, is known to defy his peripherals by inducing popups at a high rate. Sanchez is still young (24), so he can improve his skills before regression catches him. Either way, he led the AL in ERA in his second year in the majors and features an excellent sinker. Francisco Liriano provides some sneaky strikeout upside at the back-end of the rotation.

The bullpen is solid too, headed by Roberto Osuna, Joe Biagini (who I profiled last year), and Jason Grilli (who rebounded nicely last year).

Second Wildcard

Favorite: Seattle Mariners

 The Mariners have some upside (Mike Zunino, Mitch Haniger, and Dan Vogelbach are all former top prospects still under 27 years old), but their aging stars (Nelson Cruz is 36, Hisashi Iwakuma is 35, and Robinson Cano is 34) will have to remain effective in order for them to catch the Blue Jays. The ineffectiveness of Felix Hernandez and the injury to Drew Smyly mean the M’s will have to lean heavily on the injury-prone but high-upside James Paxton for innings.

Pennant

Favorite: Houston Astros

With unmatched depth and a stat-savvy UPenn and Northwestern educated GM in Jeff Lunhow, I pick the ‘Stros over the Red Sox (whose depth is rapidly disappearing under old-school president of baseball operations Dave Dumbrowski) and the defending pennant winners (Carlos Carrasco and Danny Salazar are too injury-prone for my liking, and their depth pales in comparison to the Astros).

About those Yankees:

The young trio of Greg Bird, Gary Sanchez, and Aaron Judge will have to really wow in order to make them a contender :(.

Data from Baseball Prospectus and FanGraphs. Picture from wordpress.com.

Thanks for reading!

 

Starlin has Really not Been Sterling

A common misconception amongst casual baseball fans is that Starlin Castro is a solid player. I’m here to debunk that myth.

Twenty-one homers, a career-high by seven. A .270 batting average. Only 118 strikeouts in 610 PAs. So, Starlin Castro is back on the map, right?

Not so fast. Castro did not just experience a sweet power-surge, as some may think. He greatly benefitted from an inflated HR/FB rate. Last year, his 15% rate was way higher than his previous career high (10.1%). Some may dismiss this stat and say that the move to Yankee Stadium helped him out, but in fact, the park factors on FanGraphs list Wrigley (106) as an easier venue for righties to hit homers than Yankee Stadium (105). As was often discussed amongst the Sabermetric community, last year the number of homers skyrocketed all around the MLB (possibly due to the balls being “juiced”), so take that with Castro’s inflated HR/FB rate and the most PAs he saw in a year since 2013, and we have a career-high in homers, even though Castro posted a fly ball rate lower than his career average and a popup rate higher than his career average. All of these homers helped lead to a career-high ISO, at .163. Because I don’t think the home run increase is sustainable, I don’t think the ISO increase is, either. He only hit 29 doubles, tying him for 51st out of the 88 players with at least 600 PAs last year. He also only hit one triple. Further undermining his ISO explosion, the league average ISO also ran up to its highest in ten years, fifth highest all-time, at .162. This was only the second time in Castro’s career that he had bested the league average ISO in a season (and he barely did so in 2016).

graph.jpg

Accompanying this power surge was an erosion of plate discipline. Castro posted the second-worst walk rate of his career, at 3.9%. This was the third-lowest walk rate amongst the 88 hitters with at least 600 PAs last year. To make matters worse, his strikeout rate (19.3%) was the worst of his career. His swinging strike rate ballooned to 11.3%, way higher than his previous career-high (9%), and firmly below average. His chase rate was its worst since 2012, ranking 11th highest in the group of 88. If there is one good thing here, he also swung at pitches in the zone at the highest rate in his career. However, this could just be a function of Castro seeing a career-high number of strikes and him choosing to have a more aggressive approach (career-high overall swing rate as well, ranking 12th in the group of 88). Yet, he still made contact on pitches in the zone at the lowest rate in his career. Either way, Castro’s .300 OBP last year undermined his improvements in the power department, and he ended up with a below-average 94 wRC+. That OBP was the seventh-lowest amongst the 88 hitters with at least 600 PAs.

Noted for his speed as a prospect, Castro never actually posted a positive BsR (base-running runs above average) value in a season. His base stealing days appear to be over, as he only attempted to steal four times last year. Although he wasn’t caught once, he still posted a -1.6 BsR.

Did he at least hit the ball hard? Well, his Hard% was the second highest of his career, but it was still just below league average. His Soft% was a couple percentage points below league average, at least. His line drive rate was the second highest of his career (right around league average). But did his exit velocity improve? Of the 61 hitters with at least 400 batted ball events last year, Castro had the 43rd highest average exit velocity, at 89 MPH. The prior year, he ranked 44/55 amongst the hitters with at least 400 batted ball events, with an 86.5 average exit velocity. While Castro experienced quite a jump there, the average of the first group was 89.9 and the average of the second group was 88.9 MPH, so the entire league experienced a jump, and Castro ended up below average both years.

Starlin Castro is definitely a below average hitter. There are certainly some good things about him, but when he has a down year on defense, like last season, his offense isn’t enough to bank on–he only accumulated 1.1 fWAR in 151 games. This was the 11th worst amongst the 88 hitters with at least 600 PAs, and the worst amongst the second basemen in that group. Despite an uptick in homers, I don’t see enough of a skills improvement for Castro to maintain a 20-per season pace. I would expect something more along the lines of 15-per season, with his increased aggressiveness giving him more shots to lift the ball out of the park. This aggressiveness, however, can be his downfall: he doesn’t walk often and he doesn’t hit the ball hard, so some bad luck on balls in play could lead to a horrific OBP.

Data from FanGraphs and Baseball Savant. Graph made courtesy of https://nces.ed.gov/. Picture: USA TODAY NETWORK/USA TODAY NETWORK/SIPA USA–via nydailynews.com

Thanks for reading!

Caught on Cotton

A few weeks ago, I wrote about a potential impact rookie hitter for 2017, Dan Vogelbach. Today I’ll evaluate a potential impact rookie pitcher, Jharel Cotton.

Just before the trade deadline this past summer, the Dodgers and Athletics struck a blockbuster deal that sent ace Rich Hill and capable right fielder Josh Reddick to the Dodgers and pitching prospects Grant Holmes, Frankie Montas, and Jharel Cotton to the A’s. The oft-overlooked Cotton was largely considered the third-best prospect the A’s received; after all, how good could a pitcher who put up a 4.90 ERA in 97.1 frames of Triple-A ball be?

Well, I’ll answer that question for you; pretty damn good. As I often write, ERA isn’t a tell-all statistic, and Cotton had some pretty good indicators that suggested he was more than just a run-of-the-mill pitcher in those 97.1 frames. Although the 1.57 homers per nine were troubling, it is encouraging that in years prior in the minors, he only allowed a rate higher than 1 homer per nine at one level. Also, that rate normalized after the trade; in 38.1 frames for the A’s Triple-A affiliate, he allowed 0.70 homers per nine, sporting a 2.82 ERA. In the majors, the rate wasn’t great, but it was still palatable at 1.23 homers per nine.

What I like most about Cotton, though, is his command and his swing-and-miss stuff. At every stop in the minors in which he threw at least 20 innings, he never walked more than 3.02 per nine. Even more exciting, in those 97.1 innings for the Dodgers Triple-A affiliate, he only allowed 2.96 walks per nine and struck out 11 per nine. That’s right. I said 11. So, he probably suffered from some bad luck there; his strand rate was just barely over 60% and his home run to fly-ball ratio was probably exorbitant. In his short stint in the majors, he only struck out 7.o6 per nine, but with only a 1.23 BB/9 rate and a 12.5% swinging strike rate, which would have been a top ten rate had he qualified.

How did he amass such a high swinging strike rate? Is it sustainable? Let’s take a look at his repertoire.

In order of usage, he throws a four-seamer (34%), a changeup (28.3%), a cutter/slider (16.4%), a two-seamer (13.5%), and a curve (7.8%). His four-seamer averages a solid 92.3, right around the average for a right-handed starter. What really makes this pitch special, though, is its rise. It would have ranked within the top 15, had he qualified. This helped Cotton to a crazy 24.4% popup rate, which was better than every single pitcher who threw 30 or more innings, except for Tyler Clippard. This would explain the miniscule BABIP that Cotton allowed (.198). While I don’t think that figure is sustainable, the rise on the fastball looks good for suppressing BABIP going forward. Let’s check the Baseball Prospectus PitchFX leaderboards for the fastball, too. Of the 228 starters who threw at least 100 four-seamers last year, Cotton’s ranked 113th in average velocity, 42nd in rise, and most impressive, second in terms of popups per balls in play.

Onto the changeup. This one’s a beauty.

 

A 40% O-Swing rate, and he only throws it in the zone 35.5% of the time! A 17.7% swinging strike rate. A 54.8% swing rate, almost as high as his fastball (55.7%). Also, it would have had top-15 drop, had he qualified. A 4.4 pVAL in only 125 pitches. If we look at the PitchFX leaderboards on Baseball Prospectus instead of FanGraphs, the prognosis is similarly positive: of the 153 starters who threw at least 100 changeups last year, Cotton’s had the 55th highest swing rate, the 48th best whiffs per swing, the 38th best drop, the 21st highest fouls per swing, and best of all, the second highest popups per balls in play rate.

The cutter/slider:

A high O-Swing rate (37.9%) and a low Z-Swing rate (62.8%) point to signs of success for this pitch and possible sustenance of its exorbitant swinging strike rate (22.2%). Here are the relevant stats from Baseball Prospectus as well: of the 89 starters who threw at least 50 cutters last year, Cotton’s had the 28th highest average velocity, the 27th highest swing rate, the third best whiffs per swing, the second best GB/FB ratio, and the best popup per balls in play ratio. Although the pitch also has the lowest fouls per swing rate in that group, the high whiff rate and excellent contact management negate any ill effects from that.

Cotton also throws a curve:

He only threw it 34 times in the majors last year, but it looks like it has good movement, and it generated a 17.7% swinging strike rate. It has above average drop and horizontal movement.

Lastly, Cotton throws a two-seamer, his most ineffective pitch, but it is good to know that he has another fastball he can turn to when he’s in need of a ground ball.

With a full starter’s repertoire, excellent contact quality management, and a robust swinging strike rate that has good foundations, look for Cotton to do big things this season in his first full year in the majors.

Data from FanGraphs and Baseball Prospectus. Video from Pitcherlist. Picture courtesy of USA Today Sports Images, via cbssports.com.

Thanks for reading!

 

 

 

Dan the Man

A promising rookie took a big step forward last year.

2016 was an excellent year for rookies. We saw 20 homers in only 229 PAs from Gary Sanchez, 20 in only 330 PAs from Ryan Schimpf, and 27 in only 415 PAs from Trevor Story. Trea Turner managed 33 steals, 13 homers, and a .388 BABIP. Tyler Naquin ran a .411 BABIP.   Not to mention, Corey Seager racked up 7.5 WAR. You get the idea. This year, Dan Vogelbach looks poised to burst onto the scene.

Vogelbach was dealt just before the trade deadline from the Cubs to the Mariners in a four-player deal that was basically just Mike Montgomery for Vogelbach, with a throw-in on each side. It was expected that the Cubs would trade Vogelbach, a notoriously bad fielder who can only play first base at best (where Anthony Rizzo is slotted for the near future), to an American League team who could deploy him at first base or DH if needed. However, there is little doubt about his offensive ability: at every stop in the minors where he logged at least 100 PAs, he put up at least a 126 wRC+ with at least a .150 ISO, 10% walk rate, and .350 OBP. Not to mention, he struck out in less than 20% of his PAs at every stop but one (where he struck out in 20.2%). Check out some of his important stats from last year, which was another solid year:

Team PA HR BB% K%
Mariners (AAA) 198 7 21.20% 17.20%
Cubs (AAA) 365 16 15.10% 18.40%
ISO BABIP AVG OBP SLG
0.182 0.263 0.240 0.404 0.422
0.230 0.362 0.318 0.425 0.548
wOBA wRC+
0.375 127
0.423 158

While his success with the Cubs’ Triple-A affiliate was partially BABIP driven, it was a larger sample, and he still maintained a solid ISO (.182) with the Mariners’ affiliate. Not to mention, he walked a ton, even more than he struck out after the trade. Either way, at each stop, he posted an OBP above .400, incredible plate discipline for a 23 year-old (now he’s entering his age-24 season). Walk percentage is an excellent predictor for future results, as it usually has an r-squared value in the .70’s from year to year. Thus, it isn’t surprising that Vogelbach received the most positive projection from Steamer this year for all incoming rookie hitters (in terms of wRC+).

Data from FanGraphs, photo by Gregg Forwerck/Getty Images, via seattlepi.com. Thanks for reading!