Boston Red Sox Send Big Prospect Package to San Diego Padres for… a Closer?

Boston Red Sox Send Big Prospect Package to San Diego Padres for… a Closer?

I wanted to begin by saying that we’re living in some kind of Bizarro world because Padres GM A.J. Preller fleeced Red Sox President of Baseball Operations Dave Dombrowski, who is a Hall of Fame-caliber executive. However, that would imply that Dombrowski didn’t realize what he was doing. I’m sure he understood exactly how much he was paying for Craig Kimbrel. With the tremendous depth in the Red Sox farm system, he probably felt that the exorbitant price was worth it.

The Red Sox sent CF Manuel Margot, SS Javier Guerra, and a couple of lottery tickets in exchange for one of the best relievers in baseball. Margo and Guerra are both top-100 or even top-50 prospects in all of baseball. This is an absurdly high price to pay for a reliever who will pitch 60-70 innings for the next 2-3 seasons, depending on whether or not the Red Sox pick up his option for 2018. The Red Sox easily paid more for 2-3 years of Kimbrel than the Angels paid for 5 years of Andrelton Simmons, who has a historically good glove at shortstop. Put another way, the Red Sox paid more for 6-9 WAR than the Angels did for 15-20 WAR, with the upside for significantly greater than that. That is awful.

Trades should only ever be evaluated based on what is known at the time. Process over results. Until somebody gets that crystal ball working, that’s the only fair, logical way to do it. That being said, let’s pretend for a moment that results are how we should evaluate this trade. If even one of these players going to San Diego becomes so much as a league-average regular, this trade will be a disaster. It would be much better to have six cost-controlled seasons of an average regular than 2-3 seasons of an elite closer. It’s highly likely that the result will be much better than that.

It appears that Dombrowski is overreacting to the Royals success as a result of their elite bullpen, and his own failures in building a competent bullpen in Detroit, as if that were the reason why the Tigers never won a World Series during his tenure there. I’m not going to go in-depth here, but the Tigers failure in bringing a championship to Detroit was only in small part because of the bullpen. Along the same lines, the Royals tremendous success the past couple of seasons was mostly the result of other factors besides the monster bullpen. The Astros had one of the best bullpens in baseball in this past season and didn’t even win their division, nor did they get past the ALDS. The Yankees had two of the best relievers in baseball in Dellin Betances and Andrew Miller, and they couldn’t even win their division either. Bullpens are important, but you need a lot more to be a good baseball team.

Here’s the thing: The Red Sox are trading from a tremendous amount of depth in order to fill a significant need. Last season, the Sox bullpen was tied for eighth-worst in the majors by ERA-, dead last by FIP-, and fourth worst by DRA, a new advanced metric from Baseball Prospectus. Their now former closer, Koji Uehara, is 41 years old and has a history of being injury prone. Only he and Junichi Tazawa were above replacement level in the bullpen. Acquiring Craig Kimbrel should improve this bullpen by 2-3 wins a year. Furthermore, if Joe Kelly is moved to the bullpen like he should be, this could end up being a very potent bullpen. I’ve always believed that Kelly could be an elite reliever.

I always cite the volatility of relievers whenever a team overpays for one, but Kimbrel has proven to be an anomaly so far. He’s been worth approximately 2-3 WAR each full season in Atlanta. His year in San Diego might look like he declined significantly, but I would disagree. While he did pitch in the most pitcher-friendly ballpark in the game, it was in front of one of its worst defenses. His 2.88 RA9 looks pretty bad when you see that he had a sub-2.00 RA9 the past three seasons. Besides the defense, of course, his 13.6% HR/FB is more than twice his career rate going into 2015. Pitching in as big of a ballpark as Petco Park, that’s incredibly fluky. His groundball rate was actually up while his line drive rate was down. His cFIP was roughly the same as 2014, which was only slightly worse than it was in 2013.

If we look at Kimbrel’s PITCHf/x data on Brooks Baseball, we see that his fastball and curveball have not lost any velocity, nor have they lost significant movement. You can argue that Kimbrel declined slightly in 2015, but it’s tough to say given his track record that he isn’t more or less the same pitcher he’s always been.

Margot and Guerra are still a long ways off from reaching the majors, and given the team’s depth at those positions they were unlikely to ever reach the Show in a Red Sox uniform. The team has Xander Bogaerts under control for four more years with Deven Marrero still in the system1. Mookie Betts just put up a 6 WAR season and is under control for five more years. Behind him are Jackie Bradley Jr. and 2015 1st-rounder Andrew Benintendi. If you’re going to overpay for a player, this is the best possible situation to do so.

With the Red Sox desperate need for starting pitching, I’m shocked that they weren’t able to get anything for this package. I would’ve parted with a #3 starter for this package, maybe even a #2! It’s hard to imagine that Dombrowski didn’t at least try to do so. Claiming that he was done trading for the offseason, it’s possible that the plan is to fix the bullpen regardless of the cost, and then pay for Zack Greinke, David Price, or Johnny Cueto. It’s really not a bad plan. It would’ve cost the Sox even more in prospects to get an ace. Instead, they just have to spend money. That’s a much more plentiful resource than prospects in today’s game, especially for the Red Sox.

From the Padres perspective, this is far and away the best move that A.J. Preller has made in his short time as GM. The team is clearly punting 2016 in order to rebuild, which is the right decision. They traded a closer that they won’t need, especially in Petco Park, for a tremendous return in prospects which will fill the team’s gaping holes in center field and shortstop. This is the best return a team has gotten compared to what they gave away since, coincidentally, the Red Sox traded away two months of Andrew Miller for Eduardo Rodríguez. I just cannot give Preller enough credit for this trade.

Normally I would excoriate a team for doing what the Red Sox did, and they do deserve to be criticized for giving up so much for a closer. However, the team has tremendous depth in their farm system, are set at center field and shortstop for years to come, had a big need in the bullpen, and the players traded away are far from reaching the majors. The Sox are going to have to obtain an ace through free agency for this trade to look better. There’s no way I ever do this trade if I’m Dombrowski, but If you squint just enough, the trade becomes defensible.

  1. Yoan Moncada was drafted as a shortstop but has already been moved to 2nd base. He’s too big to be a shortstop. 
Atlanta Braves Send Andrelton Simmons to the Los Angeles Angels in Bizarre Trade

Atlanta Braves Send Andrelton Simmons to the Los Angeles Angels in Bizarre Trade

Well, at least I don’t have to worry about my Mets getting Andrelted anymore.

I wasn’t too surprised when ESPN’s Jonah Keri reported that the Braves were listening to offers on Andrelton Simmons. The Braves are rebuilding and Simmons is an All-Star caliber player thanks to being the best defender the position of shortstop has seen since Ozzie Smith1.

We all know about the shortcomings of the advanced defensive metrics, but nobody argues over how they rate Simmons. His defensive mastery is apparent to even the most casual of baseball fans. He has 80 range with an 80 arm. That’s insane. In 2013, he became the first player in the live-ball era to achieve a dWAR over 52.

Simmons’ value is almost entirely tied to his glove. To be fair, that’s still a lot of value. His bat is still a problem. You’ll live with it because of the defense, but Angels fans will have trouble watching when he comes up to the plate. He’s a well below average hitter. In 2015, he hit .265/.321/.338, which is good for an 82 wRC+. At least that was a bounce back from an even worse 2014, when it appeared that the Braves hitting coach was giving him bad advice. He’s one of the best contact hitters in baseball, but he doesn’t walk much and doesn’t hit for any power.

After his best offensive season in 2013, I was optimistic that Simmons’ bat would improve even further, especially since he was only 24 years old. Strangely, he hit 17 HR that year. With a couple more seasons of data available, it appears to be because he hit an inordinate amount of flyballs. His 39.1 FB% that year is 8 percentage points higher than his career average. You could argue that he should go back to that, but low power, high contact hitters like Simmons are generally best off keeping the ball on the ground.

There’s always a chance that Simmons will figure something out at the plate. Even if he doesn’t, the Angels have the best defensive shortstop in baseball locked up on a team-friendly deal through 2020 that I loved at the time. That doesn’t mean I’m a big fan of the trade, though.

On the surface, the deal makes sense. Their current shortstop, Erick Aybar, had only one year left until free agency and is already declining. He was even worse than Simmons offensively in 2015, and his defense is declining to the point where he may only have a year or two left at shortstop. If his decline continues, Aybar is going to have a very hard time finding suitors in free agency.

While having a cost-controlled, 3-4 WAR shortstop through 2020 is great, the Angels greatly weakened an already poor farm system to get it done. Sean Newcomb is one of the best left-handed pitchers in the minors and projects to be a number two starter. Chris Ellis is a RHP who is expected to be a number four starter. You don’t have to be a prospect guy at this point to say that the Angels now have the worst farm system in baseball by a mile.

The Angels have good rotation depth that they leveraged in this trade. The problem now is that they still don’t have a 3rd baseman, catcher, or left fielder. Their starting 2nd baseman last season, Johnny Giavotella, is a poor fielder at the position. They have no way to address those problems in the trade market. Owner Arte Moreno is going to have to open his wallet to fill those holes. C.J. Wilson will be coming off the books after next season, but right now he still owes him $20 million, Albert Pujols $165 million, and Josh Hamilton almost $60 million to play for the Rangers because he hurt his feelings.

Stranger still is while the Angels would need a shortstop after 2016, Simmons is an odd fit for this Angels club. Over the span of the past three seasons, Angels pitchers have had the lowest groundball rate in all of baseball. The Dodgers would’ve been a much better destination. They generate lots of groundballs and they have the farm system depth to make the trade. They could’ve offered a better deal, too. They could’ve moved Corey Seager to 3rd base where he’d likely be better off, and play Simmons at shortstop.

I’m suspicious that new Angels GM Billy Eppler did this on his own. My guess, and this is only a guess, is that this was Mike Scioscia’s doing. It’s a decent trade but it’s not very well thought out. Again, it’s an odd fit for the club and the team now has nothing to trade to fill in the rest of its holes.

The Braves have been doing a great job in their rebuilding efforts. They’re clearly punting 2016, and it’s the right decision. However, the return for Simmons was a little light. It’s fine, but they could’ve done better. Furthermore, it’s odd to trade away a defensive wizard like Simmons when you’re clearly building around pitching.

I’ve always been very high on the value of Simmons relative to what he was getting paid. Defense doesn’t slump, and it’s one of the few things in baseball that has any kind of consistency to it. Even if he is still a 3-4 WAR player in the final year of his contract, that’s still a steal for the $15 million he’ll be making that year. There’s still upside to him too. He’s still capable of putting up a 7 WAR season like he did in 2013, or at least something close to that.

In terms of trade assets, the Braves only had Simmons and Freddie Freeman. Believe it or not, one could make the argument that Freeman isn’t better than Simmons. Yes, he’s a way better hitter, but the gap in defensive value between the two is like the distance from the Earth to the Sun. You could fairly disagree and say Freeman is better and have a good argument. Still, Freeman is making way more money than Simmons. The Braves likely would’ve had to eat a good chunk of Freeman’s salary in order to get the same return that they got for Simmons.

Thanks to the rise in sabermetrics, teams have never been better at valuing defense. That’s why Jason Heyward is going to get >$150 million. That being said, I don’t believe the Braves valued Simmons’ defense enough. They should’ve gotten more from him, especially on such a cheap deal.

It’s a decent but strange trade for both sides. I still don’t believe that either team should’ve gone through with it. The Angels now have no other option but to spend lots of money to patch up their holes. The Braves could’ve gotten more for Simmons. Heck, his trade value wouldn’t be much lower even a year from now.

  1. Please don’t tell me, “But what about Omar Vizquel?” I don’t understand how anybody who has seen both shortstops play can claim that Vizquel even comes close to Simmons. I’d listen to an argument comparing Simmons to Mark Belanger, though. 
  2. Kevin Kiermaier just became the second. 
2015 Year-End Awards: MVP

2015 Year-End Awards: MVP

Feel free to check out my picks for relievers and rookies, as well as my Cy Young picks.

Last year I made fake MVP picks here, and explained why I did so here.

My MVP picks are simply the top ten players in each league this season. I assess players solely by what can be objectively proven and evaluated, completely independent of team performance. That’s the only logical, fair way to do it. If you want to throw in narratives, soft factors, and basically anything that’s unprovable, then you can make an argument for literally any player.

I’m fairly confident that Bryce Harper will win NL MVP. There has been some push back among fans, though that has mainly been from Cubs fans pushing for Anthony Rizzo and Pirates fans pushing for Andrew McCutchen. Of course, the arguments against Harper are the “value” ones, such as how valuable can he be if his team didn’t make the playoffs. It’s a popular argument for fans and writers to defend voting for inferior players. It’s not Harper’s fault the Nationals didn’t make the playoffs. It was injuries and Matt Williams that derailed the Nationals season. In fact, Harper was one of the only good things the Nationals had going.

I saw on Twitter a while back that some die-hard Cubs fan was defending Anthony Rizzo as the deserving MVP. He stated that Rizzo must’ve been worth 20 wins to the Cubs. He then admitted to not being familiar with the Wins Above Replacement model. Surprise, surprise. Naturally, he still didn’t back off his statement when a few people tried to educate him about it, because of course he didn’t. The WAR record for the live-ball era unsurprisingly belongs to Babe Ruth, who had 14.1 WAR in 1923. This fan was basically claiming that Rizzo was roughly 50% better than Babe Ruth was in his best season!

It’s funny, Harper is getting all this support for MVP, and rightfully so, even though he wasn’t on a playoff team, but Mike Trout never got that same support in 2012 and 2013. I’m 100% positive that history will judge the writers who didn’t vote for Trout poorly, but I figure that we’re at least a decade, probably much more, away from that.

I hope that I’m wrong, but I don’t think the change in approach is the result of fans and writers setting their emotionally driven, pre-drawn conclusions aside in favor of logical, fact-based driven analysis. I believe that it’s simply the fact that people are drawn to high offense players. It’s fun, it’s exciting, and hitting is the easiest part of a position player’s game to accurately evaluate and understand.

Bryce Harper was inarguably the best hitter in baseball this season. In 2012 and 2013, Trout was not. It was clearly Miguel Cabrera. If Miguel Cabrera were on terrible teams in 2012 and 2013, but Trout were on great teams, would the majority of fans and writers still have picked Cabrera? Obviously this is just pure speculation, but I believe so. All the fans championing Cabrera over Trout in 2013 because the Tigers were good and the Angels were not would likely be saying the complete opposite under my proposed scenario.

Let’s take it a step further. Bryce Harper was a ~10 WAR player this past season almost entirely from his bat. Let’s say that Harper’s 10 WAR was like 2012 Mike Trout’s season. Let’s pretend that Harper wasn’t the best hitter in baseball and that a third of his value came from baserunning and excellent defense at a premium position. Would fans and the media still be as strongly in favor of Harper for MVP even though his team didn’t make the playoffs? I doubt it.

Fans and the media are still in the habit of treating the MVP like the best hitter award, as if defense, position, and baserunning don’t count. Last year I was horrified when Víctor Martínez was SECOND in the AL MVP voting. He didn’t even make my fake ballot that year! He had no positional value, no defensive value, and was one of the worst baserunners in the majors. An argument could certainly be made that he deserved some down-ballot votes, but I just could not convince myself to put him above other players who played the field and played it well, despite V-Mart’s offensive advantage.

Even those players who win MVPs without being the best hitter in their league still have to be one of the best offensively and be on a playoff team. Last season, Trout wasn’t the best hitter in baseball1, but he did hit .287/.377/.561 with 36 HR and made the playoffs. He also led the league in RBI and Runs scored, and that helps because there are voters who still believe that those are useful metrics for player performance when in reality they’re nothing more than accounting stats. Josh Donaldson will likely win the AL MVP this season even though he wasn’t the best hitter. But again, he was one of the best hitters and his team won the division. If his team didn’t make the playoffs, I don’t believe Donaldson would’ve stood a chance to win MVP.

Ben Lindbergh of the now defunct Grantland wrote an interesting piece where he investigated who really contributed the most to a winning team this season. In other words, he took the most objective approach that I’ve seen to what is normally subjectively assessed as “value.” He concluded that Anthony Rizzo was the most “valuable” by his methodology. He still stated that Harper was his MVP pick.

It was certainly an interesting read. My problem with it is that I’m strongly against assessing players on factors beyond their control. Players can’t control when their teams are good or bad. They can’t control when they come up with RISP, nor can they control when they’ll get hits in those situations. Hitting with RISP is not a skill. They can’t control when they’re “clutch” or when they’re not. That’s not a skill either.

This is all why I treat the MVP as the Best Player Award. If being the best player doesn’t make you the most valuable, then the MVP isn’t an award worth winning. If you want to say one player is more “valuable” than another even though he isn’t better, that’s not an argument I’m interested in having.

I assess players based on who is better than whom. Period. Anybody who disagrees are free to argue over their silly little “value” award with whomever they’d like. Just not me.

Like in my previous awards posts, you can click on the ballot in order to take you to a Fangraphs page with a table of all my selections so you can compare them yourself. Since there are both pitchers and hitters on the ballot, you’ll have to click on the “Batting” and “Pitching” buttons to switch back and forth. WAR values for pitchers will be from the Deserved Run Average (DRA) version at Baseball Prospectus. Position players will be from Baseball Reference. The reason for that being that I prefer Defensive Runs Saved to Ultimate Zone Rating when evaluating defense. I also find it helpful that Baseball Reference offers the offensive and defensive splits for WAR. The defensive component can be a little shaky, so it’s useful to be able to see how it’s gauging the player’s fielding.

You’ll also notice that Yoenis Céspedes isn’t on either ballot. Getting traded to another league killed his chance to make a ballot. He just doesn’t have enough playing time in either league.

Continue reading

2015 Year-end Awards: Cy Young

2015 Year-end Awards: Cy Young

For my Rookie of the Year and reliever awards, click here!

Below are my fake ballots for the Cy Young awards. I evaluate pitchers for factors within their control as much as possible. To help with this, I will be using two of the most advanced pitching stats that came out this year from Baseball Prospectus: Deserved Run Average (DRA) and Contextual Fielding Independent Pitching (cFIP). Going purely by runs allowed ignores factors beyond a pitcher’s control, such as batted ball luck, quality of his defense, and park factors. That list is not exhaustive either, and you can learn more by clicking on the DRA or cFIP links above.

It’s important to try to discern what parts of a pitcher’s performance are within his control and what aren’t. If you disagree, then whichever starter has the lowest RA9 is the deserving winner, end of story1. Not ERA, because that credits pitchers for defensive errors, which is a factor beyond their control.

I don’t factor in hitting performance for the Cy Young award. Some may disagree with me, but I believe it to be purely a pitching award. All WAR values will be taken from Baseball Prospectus’ DRA-based model. You can click on either ballot and it will take you to a Fangraphs page where you can compare the players yourself. For DRA and cFIP stats, you’ll have to find it for yourself here.

Continue reading

2015 Year-end Awards: Managers, Rookies, and Relievers

2015 Year-end Awards: Managers, Rookies, and Relievers

I’ll be covering all the major awards with one exception, including the newer Trevor Hoffman and Mariano Rivera awards. These picks will also count towards the voting for the Baseball Bloggers Alliance’s own awards, if they’re still doing them. I don’t know.

Manager of the Year: Abstain

I’m going to get the one exception out of the way first. I do not believe that the game of baseball lends itself to a Coach of the Year type of award. The only thing we can objectively assess about a manager is his lineup construction and in-game tactics. The problem exasperates itself when you consider the fact that even the best managers in this regard are barely passable, and most of them are terrible. Too much small ball, too many bad lineups, too much bad bullpen management, and too much obstinacy to learn from mistakes.

We also seldom have any idea how good a manager is at player development and personnel management, which are major parts of the job. A beat writer may be able to give us some insight into these things, but it’s like I’ve mentioned before which is that they only get to intimately know one clubhouse, so they have no basis for comparison.

There’s also the problem of how team performance factors into the award. You never see a manager on a bad team get considered. Why? Like the MVP, it’s an individual award. Two years ago, Cleveland Indians manager Terry Francona won the award, and I do believe he’s an excellent manager, but would he have gotten any consideration if he was managing the Astros two years ago? Would the Astros even have done any better than last place in the AL that year? While the manager probably has some input into whom the organization signs, trades, and drafts, ultimately it’s the GM’s decision to make. It’s entirely possible that if we could objectively prove who the best manager of the year was, he could turn out to be on a terrible team.

What about a good manager on a good team that just suffered a bad season as a result of being bitten by the bad luck monster? You could say this fits Bruce Bochy this year. This season, Mike Matheny will likely finish in the top three for the award, and despite the points I’ve already made, I can confidently say he’s one of the worst managers in all of baseball. Yet, he’s still going to get more votes than far superior managers such as Bruce Bochy and Chip Hale, who were just unlucky enough to not have made the playoffs.

The biggest indictment of this award happened just last year. Matt Williams won NL Manager of the Year, and for what? “Leading” a loaded Nationals team to 96 wins? This past season, he showed himself to be one of the worst managers I’ve ever seen, which is saying a lot. Not only was he terrible at what can be objectively assessed, he was also terrible at personnel management and behind-the-scenes stuff. You can read all about it in this outstanding piece by the Washington Post’s Barry Svrluga. It’s an excellent piece of reporting. The bottom line is that if he was this bad this year, then he was no better last year when he won NL Manager of the Year.

Before I move on to the reliever awards, I want to make it clear that I did not consider the number of saves when making my choices. Like I said on my stats page, it’s the worst stat in baseball. Also, you can click on the ballot if you want to see all the stats on the players at Fangraphs. I will be making use of the latest, most advanced pitching stats from Baseball Prospectus: Deserved Run Average (DRA) and Contextual Fielding Independent Pitching (cFIP). All WAR values will be from Baseball Prospectus’ DRA based method, as opposed to Fangraphs’ FIP method and Baseball Reference’s RA9 method.

Pitchers are evaluated based on factors within their control as much as possible. Using cFIP and DRA helps with that. Those stats control for almost everything you can think of.

Continue reading

Some Thoughts on the Mets and Dodgers NLDS

Some Thoughts on the Mets and Dodgers NLDS

To get the obvious out of the way, I’m not going to go in-depth on the Chase Utley slide. As a lifelong Mets fan, I have a hard time being objective about it. As a fan, I was angry about what happened and the ensuing run scoring that it enabled. However, the run scoring in that inning was enabled by manager Terry Collins as much as anyone. More on that later.

Not surprisingly, you see a lot of overlap between the people who supported Jonathan Papelbon choking Bryce Harper, and those that thought that Utley’s slide was a good, hard baseball play. With respect to the slide, I’ll defer to this excellent piece written by SB Nation’s Grant Brisbee1.

In terms of Ultey still being called safe, the umpires appeared to have gotten that one right, believe it or not. Yes, I’m just as shocked as you are. I’ll try to explain this as best I can.

It’s not a neighborhood play because it wasn’t a routine, bang-bang double play. In other words, if the odds of turning a double play are deemed low, the neighborhood play does not apply. Let me insert a quick reminder here that I’m not the one who makes up the rules.

Utley was not held accountable for not touching the bag because he was called out before he reached it. The rules say that you can’t hold it against a player for not touching a base because he got called out anyway. That’s why Utley was allowed back on the base. It was the correct ruling.

Some have argued that there is a rule, and I apologize that I can’t find it in order to properly cite it, that allows the umpires to make a judgement call saying that Utley made an improper slide, and not only call him out, but call out the runner on 1st base as well. As much as I want to prop up this rule to justify my angry feelings as a fan, it’s hard to hold it against the umpires for exercising their right to do this. Ironically, umpires don’t like to make judgment calls like this. It opens them up to too much criticism because of the subjectivity involved in interpreting and enforcing this rule. More importantly, there is no precedence for enforcing this rule. Is it really fair to expect the umps to choose Game 2 of the 2015 NLDS as the first time? Dodgers fans would’ve been just as upset as Mets fans are now.

I’m disappointed with MLB for taking the Roger Goodell approach to handing down discipline. Utley was only suspended because of the public outrage over the incident. There’s no precedence for suspending a player over a dirty slide. On a personal level, I’m happy he got suspended, but more objectively, I don’t see how this suspension can stand on appeal.

The thing that nobody is talking about is how a two-game suspension doesn’t negate what happened. I’m not talking about Ruben Tejada’s leg, either. The outcome of Game 2 is set in stone. In Utley’s eyes, what he did led to the Dodgers winning a playoff game in a best of five series. I’m sure that he, and any other player for that matter, believe that’s a worthwhile trade-off. If you were to anonymously poll major league players and ask them if they’d accept a two-game suspension in exchange for one playoff game win, the vast majority of them would accept that.

(Yup, ~550 words is me not having much to say!)

Continue reading

Lucas Duda, Yoenis Céspedes, and the Truth About Hitter Platoon Splits

Lucas Duda, Yoenis Céspedes, and the Truth About Hitter Platoon Splits

There has been quite a bit of commentary this season regarding hitters’ platoon splits. As a Mets fan, I’ve mostly heard comments about the splits of Lucas Duda and Yoenis Céspedes. The widespread belief is that Duda can’t hit lefties, and Céspedes has reverse splits. That may be what the results indicate so far, but in terms of true talent, both of those comments are false.

Duda has been criticized for not being able to hit lefties for his entire major league career. Part of the reason for this is that he seldom gets the opportunity to hit against left-handed pitching. He has faced a LHP in only ~25% of his career plate appearances. Regardless of what a lefty’s true talent is versus LHP, he’s not going to get any better by not facing them. Manager Terry Collins really should’ve taken advantage of the Mets non-contending years, which consist of Duda’s entire career until now, to give Duda more plate appearances against lefties.

When it comes to any player at all, their in-season platoon splits are always cited in decision-making regarding lineup construction and pinch-hitting, and in the analysis of those decisions. Here’s the problem:

In-season platoon splits are worthless. They suffer from small sample sizes.

This is why you’ll see a player’s in-season splits vary dramatically from year to year. For example, in 2013, Duda had a .362 wOBA vs. RHP and a .284 wOBA vs. LHP, with a split of 78 points. In 2014, Duda had a .394 wOBA vs. RHP and a .241 wOBA vs. LHP, with a split of 153 points. In 2015, Duda had reverse splits! He had a .354 wOBA vs. RHP and a .375 wOBA vs. LHP, with a split of -21 points! These are enormous changes in platoon splits. No player’s true talent varies that wildly.

Because there are so many more right-handed pitchers than there are left-handed ones, it takes a long time for hitters to accumulate enough plate appearances versus lefties to accurately gauge their true talent platoon splits. For left-handed hitters, they need to accumulate at least 1,000 PA against left-handed pitching1 in order to have an accurate measure of their true talent platoon splits. It takes 6-7 seasons of playing full-time to get that many plate appearances! For right-handed hitters, they need 2,200 PA against lefties! That’s ~15 seasons of playing full-time! Short of that, you have to regress the actual platoon split heavily towards the league’s mean platoon split.

For calculating the regression towards the mean, platoon splits are frequently converted to percentages. For example, in his career, Duda has a .364 wOBA vs. RHP and a .297 wOBA vs. LHP, so you could say that he’s ~19% better versus righties than lefties. That’s a huge split. This may not be very intuitive, but how a lefty hits against right-handed pitching informs how he should hit versus left-handed pitchers. That means there’s no way that Duda is a true talent .297 wOBA hitter versus left-handed pitchers. We need to regress his stats to the league mean in order to determine what his true platoon splits are, and therefore discover what his true talent versus lefties is.

A great tool for doing this is with this calculator that Ian Malinowski of D Rays Bay developed. I think it needs to be updated, and it doesn’t seem to work for reverse splits, but it should be good enough to estimate Lucas Duda’s true talent versus lefties and righties. For more about how these calculations are done, check out this article by Fangraphs’ Matt Klaassen. Well-known sabermetrician Tom Tango also wrote about this here. A more in-depth discussion of this topic can be found in The Book, of which Tom Tango is one of the authors.

I can’t find Duda’s current projections anywhere, so for the purpose of this exercise, I used his wOBA for the last three seasons, which comes out to .355. If we use the calculator mentioned above, His 19% split drops to ~16%. That’s still a big split relative to the league average, but obviously not as large as his results indicate. Applying this projected split to his .355 wOBA gives us a .369 wOBA vs. RHP and a .313 wOBA vs. LHP. We now see that Duda’s true talent against lefties is significantly better than we’ve observed. He’s only a slightly below average hitter against lefties, and significantly better against lefties compared to other left-handed hitting position players, who averaged a .298 wOBA against left-handed pitchers in 14,481 PA this season.

Of course this doesn’t mean that manager Terry Collins shouldn’t platoon him against tough lefties, like the one the Mets will face in Game 1 of the NLDS, Clayton Kershaw. If Michael Cuddyer is healthy, it’d be best to put him at 1st base against Kershaw. If we perform the same calculations for Cuddyer that I did for Duda, and going by his Steamer projected .320 wOBA going into this season2, we get a .350 wOBA versus LHP. Even if you want to argue that Cuddyer’s present true talent is lower than the .320 wOBA that Steamer projected, which is an argument I would agree with, it would have to be dramatically lower to not make Cuddyer the better choice against Kershaw. His .306 wOBA this season would indicate that his true talent indeed is not dramatically lower than a .320 wOBA.

Moving on to Céspedes, he’s had a strange career with respect to his platoon splits. Over 4 seasons and 2,435 PA, Céspedes has actually hit righties better than lefties. He has hit for a .348 wOBA vs. RHP and a .337 wOBA vs. LHP. That’s approximately a -4% split. In 2015, that split was absurdly pronounced. He hit for a .383 wOBA vs. RHP and a .313 wOBA vs. LHP. That’s approximately a -21% split! There’s no way on earth that his career splits are real, let along his 2015 ones.

As a result, there are analysts who have criticized managers for not substituting a RHP for a LHP late in games against Céspedes this season. Heck, I’m sure managers have been guilty of making the same mistake.

There was one time when now former Nationals manager Matt Williams was criticized for letting a right-handed bullpen pitcher face Céspedes. Now there are many, many, many, many, many, things that can be fairly criticized during Williams’ time with the Nationals, but that wasn’t one of them. Williams actually made the right decision3.

No right-handed hitter has true talent reverse splits.

Performing the same calculations for Céspedes that we did for Duda gives us a 3.8% split. That’s still small for a right-hander, but unlike his actual -4% split, it actually sounds reasonable. His average wOBA from the last three seasons is .338. Applying this split to that number gives us a .347 wOBA vs. LHP and a .335 wOBA vs. RHP. Obviously, that .347 wOBA vs. LHP is way better than the .313 wOBA he put up against them this season. It’s also significantly better than the .324 wOBA that right-handed hitting position players put up against left-handed pitchers this season in 33,799 PA.

Of course, nobody is going to advocate benching Céspedes against Kershaw with the season he’s had. However, that would be the wrong argument to make. Céspedes’ .367 wOBA in 2015 is in no way representative of his true talent. He enjoyed a fair amount of BABIP and HR/FB luck this season. The right reason to play Céspedes against Kershaw is because of the reasons I stated in the previous paragraph.

While I’m on the subject, I’ll end with a note on Yankees manager Joe Girardi’s decision to bench Jacoby Ellsbury against Dallas Keuchel in the AL Wild Card game. It was the right decision, but again, it was for the wrong reasons. Girardi cited Ellsbury’s ineffectiveness against left-handed pitching this season. He’s hit for a .295 wOBA vs. LHP this season. We apply Ellsbury’s career 4.5% split in a significant sample size to his .329 wOBA Steamer projection and that number improves to a .319 wOBA vs. LHP4. Obviously that’s significantly better than that .295 wOBA he put up. He’s actually negligibly worse than Gardner, whose true splits using his .331 wOBA from the last seasons gives us a .323 wOBA vs. LHP. Carlos Beltrán is a switch-hitter with virtually no platoon splits, so his Steamer projected .331 wOBA is what it is. As a right-handed hitter, Chris Young was a no-brainer. Applying his career 17.7% split in a significant sample size of games to his Steamer projected .311 wOBA gives us a .350 wOBA vs. LHP.

So Young and Beltrán were clearly better choices than Gardner or Ellsbury going up against the tough lefty, Dallas Keuchel. All the fancy numbers and calculations in the previous paragraph, and the right reason to play Gardner over Ellsbury was because he’s been ineffective for a couple of months due to injury. Simply put, Gardner is healthy, Ellsbury is not. Claiming, however, that Gardner was the right choice because his wOBA vs. LHP was 42 points better in 2015 only is specious reasoning. It’s the right decision for the wrong reasons.

If all of this went over your head, that’s fine. It doesn’t mean you’re not smart, or anything like that! This kind of analysis isn’t for everyone. The take home message here is simple: Never draw conclusions on a hitter’s platoon splits based on single season’s worth of plate appearances. A hitter’s true talent splits are likely far closer to the league-average splits than his single-season results would indicate.

  1. Noted sabermetrician Mitchel Lichtman once tweeted that the number should be 1,400 PA. 
  2. Like with Duda I don’t have his current projections. I chose not to go by his average wOBA of the last three seasons because it’s inflated from playing in Colorado. 
  3. To be fair, I bet that Williams wasn’t aware of Céspedes’ in-season reverse splits, and that if he had, he would’ve made the wrong decision and put in a LHP to face him. Given Williams’ track record, he doesn’t deserve the benefit of the doubt. 
  4. Here I’m declining to use Ellsbury’s past three-season average because his 2015 numbers are deflated as a result of injury.