Tuesday, November 28, 2017

Leadoff Hitters, 2017

I will try to make this as clear as possible: the statistics are based on the players that hit in the #1 slot in the batting order, whether they were actually leading off an inning or not. It includes the performance of all players who batted in that spot, including substitutes like pinch-hitters.

Listed in parentheses after a team are all players that started in twenty or more games in the leadoff slot--while you may see a listing like "COL (Blackmon)" this does not mean that the statistic is only based solely on Blackmon's performance; it is the total of all Colorado batters in the #1 spot, of which Blackmon was the only one to start in that spot in twenty or more games. I will list the top and bottom three teams in each category (plus the top/bottom team from each league if they don't make the ML top/bottom three); complete data is available in a spreadsheet linked at the end of the article. There are also no park factors applied anywhere in this article.

That's as clear as I can make it, and I hope it will suffice. I always feel obligated to point out that as a sabermetrician, I think that the importance of the batting order is often overstated, and that the best leadoff hitters would generally be the best cleanup hitters, the best #9 hitters, etc. However, since the leadoff spot gets a lot of attention, and teams pay particular attention to the spot, it is instructive to look at how each team fared there.

The conventional wisdom is that the primary job of the leadoff hitter is to get on base, and most simply, score runs. It should go without saying on this blog that runs scored are heavily dependent on the performance of one’s teammates, but when writing on the internet it’s usually best to assume nothing. So let's start by looking at runs scored per 25.5 outs (AB - H + CS):

1. COL (Blackmon), 7.9
2. HOU (Springer), 7.0
3. STL (Carpenter/Fowler), 6.7
Leadoff average, 5.5
ML average, 4.6
28. CHA (Garcia/Anderson/Sanchez), 4.4
29. KC (Merrifield/Escobar), 4.3
30. SD (Margot/Pirela), 3.9

That’s Leury Garcia for the White Sox, in case you were wondering. One of my favorite little tidbits from the 2017 season was their all-Garcia outfield: Willy, Leury, and Avasail. Sadly, one of my favorite tidbits from the last few season was no more as Kansas City finally decided that Esky Magic had run its course. Alcides Escobar only got 25 starts leading off, with Whit Merrifield leading the way with 115. The Royals still made plenty of appearances in the trailer portions of these lists.

The most basic team independent category that we could look at is OBA (figured as (H + W + HB)/(AB + W + HB)):

1. COL (Blackmon), .399
2. STL (Carpenter/Fowler), .374
3. HOU (Springer), .374
Leadoff average, .333
ML average, .327
28. CIN (Hamilton), .295
29. TOR (Pillar/Bautista), .287
30. KC (Merrifield/Escobar), .282

Even if we were to park-adjust Colorado’s .399, they’d be at .370, so it was a fine performance by Blackmon and company (mostly Blackmon, with 156 starts), but not the best in the league. There’s no good reason I don’t park-adjust, although my excuse is that park adjustments don’t apply (or at least can’t be based off the runs park factor) for some of the metrics presented here. Of the categories mentioned in this post, R/G, OBA, 2OPS, RG, and LE could be if one was so inclined.

The next statistic is what I call Runners On Base Average. The genesis for ROBA is the A factor of Base Runs. It measures the number of times a batter reaches base per PA--excluding homers, since a batter that hits a home run never actually runs the bases. It also subtracts caught stealing here because the BsR version I often use does as well, but BsR versions based on initial baserunners rather than final baserunners do not. Here ROBA = (H + W + HB - HR - CS)/(AB + W + HB).

This metric has caused some confusion, so I’ll expound. ROBA, like several other methods that follow, is not really a quality metric, it is a descriptive metric. A high ROBA is a good thing, but it's not necessarily better than a slightly lower ROBA plus a higher home run rate (which would produce a higher OBA and more runs). Listing ROBA is not in any way, shape or form a statement that hitting home runs is bad for a leadoff hitter. It is simply a recognition of the fact that a batter that hits a home run is not a baserunner. Base Runs is an excellent model of offense and ROBA is one of its components, and thus it holds some interest in describing how a team scored its runs. As such it is more a measure of shape than of quality:

1. STL (Carpenter/Fowler), .341
2. COL (Blackmon), .336
3. PHI (Hernandez), .327
5. SEA (Segura/Gamel), .319
Leadoff average, .295
ML average, .288
28. CIN (Hamilton), .265
29. KC (Merrifield/Escobar), .248
30. TOR (Pillar/Bautista), .241

With the exception of Houston, the top and bottom three are the same as the OBA list, just in different order (HOU was eighth at .313, their 38 homers out of the leadoff spot tied with Colorado; Minnesota, the Mets, Cleveland, and Tampa Bay also got 30 homers out of the top spot).

I also include what I've called Literal OBA--this is just ROBA with HR subtracted from the denominator so that a homer does not lower LOBA, it simply has no effect. It “literally” (not really, thanks to errors, out stretching, caught stealing after subsequent plate appearances, etc.) is the proportion of plate appearances in which the batter becomes a baserunner able to be advanced by his teammates. You don't really need ROBA and LOBA (or either, for that matter), but this might save some poor message board out there twenty posts, by not implying that I think home runs are bad. LOBA = (H + W + HB - HR - CS)/(AB + W + HB - HR):

1. COL (Blackmon), .354
2. STL (Carpenter/Fowler), .352
3. PHI (Hernandez), .332
4. HOU (Springer), .330
Leadoff average, .303
ML average, .298
28. CIN (Hamilton), .268
29. KC (Merrifield/Escobar), .253
30. TOR (Pillar/Bautista), .251

The next two categories are most definitely categories of shape, not value. The first is the ratio of runs scored to RBI. Leadoff hitters as a group score many more runs than they drive in, partly due to their skills and partly due to lineup dynamics. Those with low ratios don’t fit the traditional leadoff profile as closely as those with high ratios (at least in the way their seasons played out, and of course using R and RBI incorporates the quality and style of the hitters in the adjacent lineup spots rather then attributes of the leadoff hitters’ performance in isolation):

1. MIA (Gordon), 2.5
2. TEX (DeShields/Choo/Gomez), 2.2
3. CIN (Hamilton), 2.2
Leadoff average, 1.5
ML average, 1.0
27. SD (Margot/Pirela), 1.3
28. KC (Merrifield/Escobar), 1.2
29. MIN (Dozier), 1.1
30. CLE (Kipnis/Lindor/Santana), 1.1

Cleveland only settled on a permanent leadoff fixture (Lindor) late in the season, but all three of their 20+ game leadoff men were of the same general type. They didn’t really have a player who saw regular time who fit anything like the leadoff profile. It worked OK; their .328 OBA was lower than the leadoff average, but as we’ll see later their 5.2 RG was above average.

A similar gauge, but one that doesn't rely on the teammate-dependent R and RBI totals, is Bill James' Run Element Ratio. RER was described by James as the ratio between those things that were especially helpful at the beginning of an inning (walks and stolen bases) to those that were especially helpful at the end of an inning (extra bases). It is a ratio of "setup" events to "cleanup" events. Singles aren't included because they often function in both roles.

Of course, there are RBI walks and doubles are a great way to start an inning, but RER classifies events based on when they have the highest relative value, at least from a simple analysis:

1. CIN (Hamilton), 1.7
2. MIA (Gordon), 1.6
3. TEX (DeShields/Choo/Gomez), 1.3
Leadoff average, .8
ML average, .6
28. TB (Dickerson/Kiermaier/Smith/Souza), .5
29. BAL (Smith/Beckham/Jones/Rickard), .5
30. COL (Blackmon), .5

Both Tampa Bay and Baltimore followed the Cleveland pattern of using multiple leadoff hitters, although one of the four for each (Mallex Smith and Joey Rickard) fit more of a traditional profile. The Rays got 5.1 RG out of this hodgepodge, which is above-average; the Orioles’ 4.3 was not. For the record, I’m basing my assessment of Joey Rickard’s traditional leadoff style bona fides on his career minor league line (.280/.388/.392), and not his major league line (.255/.298/.361) for which the only stylistic interpretation is “bad”.

Since stealing bases is part of the traditional skill set for a leadoff hitter, I've included the ranking for what some analysts call net steals, SB - 2*CS. I'm not going to worry about the precise breakeven rate, which is probably closer to 75% than 67%, but is also variable based on situation. The ML and leadoff averages in this case are per team lineup slot:

1. WAS (Turner/Goodwin), 37
2. CIN (Hamilton), 36
3. MIA (Gordon), 25
4. NYA (Gardner), 21
Leadoff average, 8
ML average, 2
28. CHN (Jay/Zobrist/Schwarber), -2
29. COL (Blackmon), -5
30. HOU (Springer), -7

A lot of the leaders and trailers are flipped on this list from the overall quality measures.

Shifting back to said quality measures, first up is one that David Smyth proposed when I first wrote this annual leadoff review. Since the optimal weight for OBA in a x*OBA + SLG metric is generally something like 1.7, David suggested figuring 2*OBA + SLG for leadoff hitters, as a way to give a little extra boost to OBA while not distorting things too much, or even suffering an accuracy decline from standard OPS. Since this is a unitless measure anyway, I multiply it by .7 to approximate the standard OPS scale and call it 2OPS:

1. COL (Blackmon), 979
2. HOU (Springer), 887
3. MIN (Dozier), 865
Leadoff average, 763
ML average, 755
28. TOR (Pillar/Bautista), 678
29. KC (Merrifield/Escobar), 658
30. CIN (Hamilton), 651

Along the same lines, one can also evaluate leadoff hitters in the same way I'd go about evaluating any hitter, and just use Runs Created per Game with standard weights (this will include SB and CS, which are ignored by 2OPS):

1. COL (Blackmon), 7.8
2. HOU (Springer), 6.4
3. MIN (Dozier), 6.1
Leadoff average, 4.8
ML average, 4.6
28. TOR (Pillar/Bautista), 3.7
29. CIN (Hamilton), 3.6
30. KC (Merrifield/Escobar), 3.5

Allow me to close with a crude theoretical measure of linear weights supposing that the player always led off an inning (that is, batted in the bases empty, no outs state). There are weights out there (see The Book) for the leadoff slot in its average situation, but this variation is much easier to calculate (although also based on a silly and impossible premise).

The weights I used were based on the 2010 run expectancy table from Baseball Prospectus. Ideally I would have used multiple seasons but this is a seat-of-the-pants metric. The 2010 post goes into the detail of how this measure is figured; this year, I’ll just tell you that the out coefficient was -.230, the CS coefficient was -.597, and for other details refer you to that post. I then restate it per the number of PA for an average leadoff spot (750 in 2017):


1. COL (Blackmon), 45
2. HOU (Springer), 26
3. MIN (Dozier), 25
Leadoff average, 2
ML average, 0
28. CIN (Hamilton), -17
29. TOR (Pillar/Bautista), -18
30. KC (Merrifield/Escobar), -24

Esky Magic has residual effects, apparently. I don’t recall seeing the same teams in the leaders and trailers list for 2OPS, RG, and LE before, but they are all very similar in terms of their construction, with 2OPS arbitrarily but logically tilted towards OBA and LE attempting to isolate run value that would be contributed if all plate appearances came in a leadoff situation. RG represents the approximate run value of a player’s performance in an “average” situation on an average team.

The spreadsheet with full data is available here.

Wednesday, November 15, 2017

Hypothetical Ballot: MVP

My heart’s just not in writing this, since it’s the first time in the brilliant career of Mike Trout that he will not top my AL MVP ballot. This is made a little better by noting that, prorated to 150 games, he contributed 90 RAR to my choice’s 76, but you add no value to your team when you’re sitting at home. I should note that there have been many less deserving MVP winners than Mike Trout would be in 2017.

Jose Altuve leads Aaron Judge by four RAR, not considering baserunning or fielding. Judge’s fielding metrics are better than one might expect from a man of his size--5 FRAA, 6 UZR, 9 DRS--but Altuve ranks as average, and per Fangraphs was worth another run on the basepaths, so I don’t think it’s enough to bump him. It’s well within any margin of error and Judge would certainly be a fine choice as MVP.

The other candidate for the top spot is Corey Kluber--with 81 RAR, he’d would my choice by default. But while Kluber’s RAR using his peripherals (78) and DIPS (71) are good, they are basically a match for Altuve’s 77 (after baserunning). Our analytical approach for evaluating hitters is much more like using pitcher’s peripherals than their actual runs allowed, and there should be some consideration that some of the value attributed to a pitcher is actually due to his fielders (if you’re not making an explicit adjustment for that). For me, if a pitcher doesn’t clearly rank ahead of a hitter, he doesn’t get the benefit of the doubt on a MVP ballot.

The rest of the ballot is pretty straightforward by RAR, mixing the top pitchers in, as the top-performing hitters were all solid in the field and didn’t change places:

1. 2B Jose Altuve, HOU
2. RF Aaron Judge, NYA
3. SP Corey Kluber, CLE
4. SP Chris Sale, BOS
5. CF Mike Trout, LAA
6. SP Carlos Carrasco, CLE
7. 3B Jose Ramirez, CLE
8. SP Luis Severino, NYA
9. SP Justin Verlander, DET/HOU
10. SS Carlos Correa, HOU

In the NL, Joey Votto has a two-run RAR lead over Giancarlo Stanton, but Fangraphs has him at a whopping -10 baserunning runs to Stanton’s -2. BP has the same margin, but -8 to 0. Their fielding numbers (FRAA, UZR, DRS) are almost identical--(10, 11, 7) for Votto and (9, 10, 7) for Stanton. I’m not sure I’ve ever determined the top spot on my ballot on the basis of baserunning value before, but even if you were to be extremely conservative and regress it by 50%, it makes the difference. Paul Goldschmidt will get a lot of consideration, but even as a good baserunner and fielder I don’t think his 52 RAR offensively gets him in the picture for the top of the ballot.

Max Scherzer’s case is similar to Kluber’s, except with an even more pronounced gap between his actual runs allowed-based RAR (77), his peripherals (71), and DIPS (61).

The rest of my ballot follows RAR, as there were no players who made a huge difference in the field. I might be more inclined to accept an argument that Buster Posey was more valuable than the statistics suggest in a season in which San Francisco didn’t have the second-worst record in the league. But one player is missing from my ballot who will be high on many (although not in the top three of the BBWAA vote) is Nolan Arenado, and I feel that deserves a little explanation.

Rather than comparing Arenado to every player on my ballot, let’s look at my last choice, Marcell Ozuna. Ozuna starts with a ten run lead in RAR (56 to 46). Arenado is widely regarded as an excellent fielder, but the metrics aren’t in agreement--1 FRAA, 7 UZR, 20 DRS. Ozuna’s figures are (5, 11, 3). If you believe that Arenado is +20 fielder, than he would rank about dead even with Kris Bryant at 66 RAR (bumping Bryant from 61 on the strength of his own (especially) baserunning and fielding). It’s certainly not out of the realm of possibility. But if you only give Arenado credit for 10 fielding runs, that only pulls him even with Ozuna, before giving Ozuna any credit for his fielding.

I was going to write a bit more about how it might be easy for writers to consider Coors Field but understate how good of a hitter’s park it is (116 PF). If you used 110 instead, then Arenado starts at 51. But given that he didn’t finish in the top three, I don’t think there’s any evidence of not taking park into account. As discussed, there are perfectly reasonable views on Arenado’s fielding value that justify fourth-place. I’m not sure Arenado would be 11th or 12th or 13th if I went further, as Tommy Pham, Corey Seager, Clayton Kershaw, Gio Gonzalez, and Zack Greinke are all worthy of consideration for the bottom of the ballot themselves:

1. RF Giancarlo Stanton, MIA
2. 1B Joey Votto, CIN
3. SP Max Scherzer, WAS
4. 3B Kris Bryant, CHN
5. CF Charlie Blackmon, COL
6. 3B Anthony Rendon, WAS
7. 1B Paul Goldschmidt, ARI
8. 3B Justin Turner, LA
9. SP Stephen Strasburg, WAS
10. LF Marcell Ozuna, MIA

Tuesday, November 14, 2017

Hypothetical Ballot: Cy Young

The starting pitchers (and they’re the only ones that can possibly accrue enough value to be serious Cy Young candidates if you subscribe to the school of thought that all innings are created equal and the only leveraging effect appropriate to credit to relief aces is that they are used in close games) did a very nice job of separating themselves by RAR into groups of five or six, with a five run gap to the next pitcher. This makes a very convenient cut point to define the ballot candidates:

AL: Kluber 81, Sale 71, Carrasco 62, Verlander 61, Severino 59, Santana 59, Stroman 52

NL: Scherzer 77, Gonzalez 66, Kershaw 64, Strasburg 62, Greinke 57, Ray 51

Corey Kuber has a clear edge over Chris Sale in RAR, but it’s closer (78 to 73) in RAR based on eRA, and in RAR based on DIPS theory (assuming an average rate of hits on balls in play), Sale flips the standard list almost perfectly (80 to 71). My philosophy has always been that the actual runs allowed takes precedence, and while DIPS can serve to narrow the difference, Kluber is still outstanding when viewed in that light (e.g. this is not a Joe Mays or Esteban Loaiza situation). I don’t think it comes close to making up the difference. FWIW, Baseball Prospects WARP, which attempts to account for all matter of situational effects not captured in the conventional statistical record, sees Kluber’s performance as slightly more valuable (8.0 to 7.6).

The rest of my AL ballot goes in order except to flip Severino and Verlander. Severino had significantly better marks in both eRA (3.15 to 3.73) and dRA (3.40 to 4.12). Santana had an even more marked disparity between his actual runs allowed and the component measures (3.82 eRA, 4.75 dRA) which also triggers confirmation bias as he and Jason Vargas’ first-half performances were quite vexing to this Cleveland fan.

1. Corey Kluber, CLE
2. Chris Sale, BOS
3. Carlos Carrasco, CLE
4. Luis Severino, NYA
5. Justin Verlander, DET/HOU

In the NL, I was a little surprised to see that in some circles, Clayton Kershaw is the choice for the award and may well win it. Tom Tango pointed out that Kershaw’s edges over Scherzer in both W-L (18-4 to 16-6) and ERA (2.31 to 2.51) give him a clear edge in the normal thought process of voters. I have been more detached than normal this season from the award debates as you might hear on MLB Tonight, and so seeing a 13 run gap in RAR I didn’t even consider that there might be a groundswell of support for Kershaw. With respect to ERA, Scherzer has a lower RRA (based on runs allowed, adjusting for park, and crudely accounting for bullpen support) and Kershaw’s raw .20 ERA lead drops to just .08 (2.41 to 2.49) when park-adjusting.

What’s more is that Scherzer has a larger edge over Kershaw in eRA (2.71 to 3.26) and dRA (3.20 to 3.73) than he does in RRA (2.56 to 2.72)--leading in all three with a 25 inning advantage. Scherzer led the NL in RRA and eRA, was a narrow second to his teammate Strasburg (3.07 to 3.20) in dRA, and was just seven innings off the league lead (albeit in seventh place). For Cy Young races in non-historic pitcher seasons, I don’t think it gets much more clear than this.

As a final note on Kershaw v. Scherzer, perhaps some of the pro-Kershaw sentiment goes beyond W-L and ERA and into the notion that Kershaw is the best pitcher in baseball. I don’t think this is relevant to a single season award, and I think it would have a much more obvious application to the AL MVP race, where not only is Mike Trout the best player in baseball, but the best by a tremendous margin, and was easily the most valuable player on a rate basis in the league (NOTE: I am not advocating that Trout should be the MVP, only that he has a better case using this argument than Kershaw). But it may be time to re-evaluate Kershaw as the best pitcher in baseball as a fait accompli. Over the last three seasons, Scherzer has pitched 658 innings with a 2.84 RRA and 211 RAR; Kershaw has pitched 557 innings with a 2.37 RRA and 206 RAR. At some point, the fact that Scherzer has consistently been more durable than Kershaw should factor into the discussion of “best”.

Strasburg placed second to Scherzer in eRA, and as discussed bettered him in dRA, recording one more out than Kershaw did. That’s enough for me to move him into second over teammate Gonzalez as well, who had an even larger peripheral gap than Kershaw (basing RAR on eRA, Strasburg beats Gonzalez 62 to 56; on dRA, 56 to 37), so I see it as:

1. Max Scherzer, WAS
2. Stephen Strasburg, WAS
3. Clayton Kershaw, LA
4. Gio Gonzalez, WAS
5. Zack Greinke, ARI

Monday, November 06, 2017

Hypothetical Ballot: Rookie of the Year

Let's assume for a moment that you care about post-season awards (I've lost at least 30% of you).

Then let's suppose you care about my opinion about how should win them (another 50% gone...I'd guess higher but you are reading this blog after all).

So for the 20% of you left (those percentages were based on the original population for those checking math), do you care about who I think should finish second - fifth on a ballot for awards which will almost certainly and deservedly be unanimously decided? Especially if the awards in question are Rookie of the Year?

I didn't think so. Rookie of the Year is the least interesting award for a number of reasons, including but not limited to:

1. For as much as people like to argue about what "valuable" means, RoY has even more contentious questions: how much should perceived future potential outweigh current year performance, and how should players who are new to the Major Leagues but veterans of high-level professional baseball in other countries (or in segregated leagues when the award was young) be treated?

2. While winning a RoY award might become a part of the standard broadcaster rundown and a line in the Baseball Register (RIP), it rarely takes on any significance beyond that. In contrast, Cy Youngs and MVPs enter a feedback loop of subjective awards when they are cited in Hall of Fame discussions.

Because of this, particularly #2, downballot selections on the Cy Young or MVP ballot take on a little more importance, even if the winner of the award gets so many first place votes that the downballot choices don't factor into the outcome. Award shares, top 5 MVP finishes, etc. may not be that important in the grand scheme of things--but compared to who finished fourth in the RoY voting, they might as well be the list of pennant winners.

This is a long-winded way of saying that:

1) Aaron Judge and Cody Bellinger are going to, and should, easily win the 2017 RoYs
2) The 20% of you who might theoretically have some interest in this post really aren't going to care who I say should be fourth. It was more fun to write this explanation than to write a detailed breakdown of why I think
Yuli Gurriel deserved to rank ahead of Trey Mancini while still having a respectable word count for a blog post:

AL:
1. RF Aaron Judge, NYA
2. SP Jordan Montgomery, NYA
3. 1B Yuli Gurriel, HOU
4. RF Mitch Haniger, SEA
5. 1B Trey Mancini, BAL

NL:
1. 1B Cody Bellinger, LA
2. SS Paul DeJong, STL
3. SP German Marquez, COL
4. SP Kyle Freeland, COL
5. SP Trevor Williams, PIT