The Unicorn and the Statistic

June 29, 2021
 
In late May, Neil Paine at FiveThirtyEight attempted to contextualize Shohei Ohtani’s season by showing his combined WAR as a pitcher and a hitter. Here’s the table he used, under the heading, "Most Valuable by WAR":
 
RANK
PLAYER
TEAM
WAR AS BATTER
WAR AS PITCHER
TOTAL
1
Gerrit Cole
NYY
0.00
2.81
2.81
2
Max Muncy
LAD
2.79
0.00
2.79
3
Xander Bogaerts
BOS
2.72
0.00
2.72
4
Nick Castellanos
CIN
2.72
0.00
2.72
5
Zack Wheeler
PHI
0.18
2.52
2.70
6
Byron Buxton
MIN
2.64
0.00
2.64
7
V. Guerrero Jr.
TOR
2.57
0.00
2.57
8
John Means
BAL
0.00
2.42
2.42
9
Jacob deGrom
NYM
0.40
1.99
2.38
10
Shohei Ohtani
LAA
1.46
0.84
2.30
 
That’s an old table: spin-king Cole got torched by the Red Sox on Sunday, Byron Buxton is doing his annual ping-pong on the IL, and Shohei single-handedly carried the Angels to a comeback win against the Rays. An updated list, as of Monday morning:
 
RANK
PLAYER
WAR AS BATTER
WAR AS PITCHER
TOTAL
1
Shohei Ohtani
3.2
1.3
5.5
2
Jacob deGrom
0.6
4.4
5.0
3
V. Guerrero Jr.
4.6
0.0
4.6
4
Zack Wheeler
0.1
3.8
3.9
5
F. Tatis Jr
3.8
0.0
3.8
 
Shohei Ohtani now ranks as the most valuable player in baseball, according to this method. As a fan of Ohtani, I’m pleased that he’s currently topping the charts.
 
But is method accurate? Does this give us a reasonable picture of Ohtani’s season?
 
 
*            *            *
 
Let’s start with the pitching, and compare Shohei Ohtani with the player who currently sits at #2: Mets pitcher Jacob deGrom:
FanGraphs
Player
Pitching WAR
deGrom
4.4
Ohtani
1.3
 
According to FanGraphs, deGrom has been significantly better on the mound this season. Color me surprised: I didn’t know Jacob deGrom was having much of a season. Must be that pesky New York media underreporting his performance.
 
Jacob deGrom has been absolutely terrific, and Ohtani, just as a pitcher, doesn’t compare to the Mets ace.
 
But they are not a million miles apart, either. For one thing, Ohtani and deGrom have actually faced the same number of batters this season:
 
Player
Batters Faced
ER
H
HR
BB
K
deGrom
287
6
30
3
11
122
Ohtani
274
17
39
6
31
82
 
deGrom has certainly been better than Ohtani in quality, but the quantity of their pitching is close. And just from a quality perspective, Ohtani is hardly a slouch. Jacob deGrom’s K/9 is 14.1…Ohtani is at 12.4. deGrom’s ERA is an absurd 0.69, but Shohei’s ERA is a very respectable 2.58.
 
If Jacob deGrom were pitching the same number of innings as Cole or Beiber, one would expect a significant margin to exist between the pair. But deGrom has missed starts, and he’s been pulled from games a few times. He’s been brilliant, but he hasn’t been a workhorse.
 
FanGraphs WAR says that deGrom merits an edge, 4.4 Pitching WAR to 1.3. But another WAR version – the Baseball-Reference versions – has them much closer:
 
Player
FanGraphs
BB-Ref
Name
Pitching WAR
Pitching WAR
deGrom
4.4
4.0
Ohtani
1.3
2.4
 
Well…why?
 
Because FanGraphs uses Fielding-Independent Pitching to calculate pitching WAR, while Baseball-Reference uses Adjusted ERA (ERA+). FIP gives more weight to strikeouts, homeruns, and walks allowed, while ERA+ considers the number of runs scored against a pitcher.
 
Ohtani has walked a lot more hitters than deGrom, and he has struck out fewer hitters, and he has allowed more homeruns, so FanGraphs argues that deGrom is significantly better. Baseball-Reference says, "well, it’s six runs allowed versus seventeen runs …we’re just going to work from that." That narrows things a little.
 
This is an important distinction, and it is up to you to decide how you prefer to measure pitching effectiveness. FanGraphs works its math from specific events - how many strikeouts did a pitcher collect, how many homeruns and walks did they allow – and translates those events into wins above replacement. Baseball-Reference begins from the outcome of those events: Baseball-Reference says that from all of the events, how many runs scored? What does that mean in the context of wins?
 
It’s up to you.
 
 
*            *            *
 
Which gets us to the hitting side of the equation.
 
Mercifully, FanGraphs and Baseball-Reference agree on their assessments of Ohtani and deGrom:
 
Player
FanGraphs
BB-Ref
Name
Offensive WAR
Offensive WAR
deGrom
0.6
0.6
Ohtani
3.2
3.1
 
deGrom has had an exceptionally lucky season at the plate. A .189 hitter coming into the season, deGrom is currently hitting (checks notes) .414 on the season. He’s turned into Rod Carew circa 1977.
 
We can break that down further: deGrom has had 29 plate appearances, collecting 11 single and one double. That is a very fine season…it is a season that would make Zack Greinke proud, if Zack Grienke cared about nonsense like a pitcher’s hitting.
 
But Ohtani has been just a touch more productivewith the bat:
 
Player
PA
1B
2B
3B
HR
BB
SB
deGrom
29
11
1
0
0
0
0
Ohtani
292
25
17
4
25
32
11
 
This is an unfair comparison, of course. Jacob deGrom has had some good luck on ground balls finding their way into the outfield. Shohei Ohtani has a slugging percentage that would make Babe Ruth jealous: he is one of the very best offensive players in the major leagues.
 
So let’s look at real hitters. FanGraphs says that Ohtani’s offense is comparable to Baltimore CF Cedric Mullins, and Reds LF Nick Castellanos:
 
Player
PA
HR
SB
wRC+
WAR
Castellanos
309
14
2
165
3.3
Mullins II
330
13
12
151
3.2
Ohtani
292
25
11
174
3.2
 
This seems odd. Castellanos and Mullins have better batting averages than Shohei, but Ohtani walks more often, and he has a significantly better slugging percentage. So how are they this close on offensive production?
 
Easy. Positional adjustment:
 
Player
Off
Def
WAR
Castellanos
25.9
-3.5
3.3
Mullins II
20.9
-0.7
3.2
Ohtani
27.7
-6.3
3.2
 
 
Ohtani is viewed – rightly – as the most productive offensive player of the trio, but he is dinged by WAR’s defensive adjustment. While Mullins’ 20.9 runs are measure against the offensive output of his peer centerfielders, and Castellanos’ offensive output is translated into wins for a corner outfielder, Ohtani’s defensive adjustment considers him a designated hitter, because for most of his at-bats he is a designated hitter.
 
Which is perfectly reasonable. And also absolutely unreasonable.
 
Shohei Ohtani is a DH five days a week because he is a starting pitcher every sixth day.
 
Shohei every bit as qualified to be considered a pitcher as Jacob deGrom: they’ve faced the same number of hitters, pitched nearly the same number of innings.
 
But Jacob deGrom - who isn’t anyone’s idea of major league hitter - is currently getting a lot of credit from WAR, for some lucky singles and a double. Shohei Ottani – one of the ten best hitters in baseball – gets demerits.
 
What is the justification for that?
 
But – devil’s advocate - what are you going to do?
 
 
*            *            *
 
What should we do? How should we understand Ohtani’s offensive contributions?
 
Is it reasonable to contextualize his hitting among the game’s designated hitters?
 
Of course not. It’s preposterous. Shohei Ohtani isn’t Nelson Cruz or J.D. Martinez, and it is ludicrous to reduce how we interpret his offensive production because he takes the hill every sixth day.
 
Phrased differently: Ohtani is only a DH because he is a capable starting pitcher. He is fast, and his baseball instincts are off the charts, and he has a canon arm. If he wasn’t a starting pitcher, he’d likely be a hybrid of Aaron Judge and Mookie Betts as a defensive player. He would rack up significant contributions as a defensive player. Instead, he is asked to rack up contributions as a starting pitcher.
 
But how to measure Ohtani’s contributions is a harder question. A decent compromise might be to adjust his offense within the context of a better position player than DH. Perhaps his contributions as a hitter should be interpreted as a centerfielder or a middle infielder.
 
But that might undersell him, too.
 
At the same time, it’s nonsense to credit him for defensive work that he doesn’t do. And Ohtani is being credited for the outs he makes as a starting pitcher. So how do you answer the question of positional adjustment?
 
Maybe the problem is the statistic itself.
 
WAR aims to understand players through the players that are like them, the players that might replace them. How many more runs does Vladimir Guerrero produce over a replacement-level first baseman? How many wins does Jacob deGrom get the Mets over a replacement-level starting pitcher?
 
WAR finds convincing answers to those questions. I am not trying to drag us into another ‘WAR is terrible’ argument. WAR gets us close to a good answer.
 
But there is no replacement level for Shohei Ohtani, because there is no parallel to what he is doing right now. If he were injured, the Angels their best starting pitcher and their best hitter. That isn’t one replacement the Angels would have to scramble for, but two. He is doing the work of two players.
 
WAR can’t accurately measure Ohtani’s contributions because the bedrock of the metric imagines a narrower band of capacities than Ohtani contributes. Like fantasy baseball servers, WAR wasn't designed to anticipate a player capable of being an elite starting pitcher and an elite hitter.
 
There is no 'replacement level' for Ohtani because his replacement doesn’t exist.
 
 
*            *            *
 
Just to chime in on another angle: there is no parallel for what Shohei Ohtani is doing right now. Even Babe Ruth isn’t a comparable.
 
Babe Ruth had seasons where he was a very, very good pitcher, and Babe Ruth had a few passable years with the bat. Given a century of distance, the casual fan tends to blur those two facts, imagining that they happened in parallel.
 
They didn’t. Babe Ruth’s best years as a pitcher came before he was understood as anything except an oddball hitter. His seasons as the game’s greatest hitter came after he had largely ceased to be an effective pitcher.
 
Ruth’s best pitcher-and-hitter seasons were 1918 and 1919:
 
Player
PA
HR
OPS+
IP
ERA
ERA+
Ruth '18
317
11
192
166.1
2.22
122
Ruth '19
432
29
217
133.1
2.97
102
 
Ruth was not a full-time hitter each year (though he paced the league in homeruns both seasons), and he wasn’t a full-time pitcher, at least not within the contexts of that era.
 
And he wasn’t an elite pitcher. He was good in 1918 and then a league-average starter in 1919, when he set the single-season HR mark in 1919. He was a spot-starter then, but his arm wasn’t what it had been in his prime.
 
Shohei Ohtani is excelling at both sides of the game:
 
Player
PA
HR
OPS+
IP
ERA
ERA+
Ruth '18
317
11
192
166.1
2.22
122
Ruth '19
432
29
217
133.1
2.97
102
Ohtani '21
297
26
172
59.1
2.58
178
 
And he has speed. That can get glossed over. Ohtani has eleven stolen bases, and he leads the league with four triples. Ruth wasn’t fast.
 
Ohtani won’t eclipse Ruth’s offensive production from 1919, but he is comfortably ahead of Ruth’s production as a starting pitcher. If he stays healthy and effective, Shohei Ohtani’s 2021 season will be the greatest dual-role season baseball has ever seen.
 
And if we can get the metrics aligned right, we might understand it as the greatest season ever.
 
 
David Fleming is a writer living in southwest Virginia. He welcomes comments, questions, and suggestions here and at dfleming1986@yahoo.com. 
 
 
 

COMMENTS (22 Comments, most recent shown first)

Jaytaft
As incredible as his hitting stats are, they'd likely be better if a healthy Trout was in the lineup.
9:46 AM Jul 5th
 
murrayj
If you leave the DH adjustment of -6, but add in the pitcher hitting adjustment of +5, then his positional adjustment is -1. This would result in about 4.1 WAR hitting to go with his 1.6 WAR pitching: 5.7 WAR

That would give him the highest WAR of any player and put hi on pace for a 10-11 WAR season which is an all time great season. That seems much more accurate
3:46 PM Jul 4th
 
CharlesSaeger
As a note simply because I looked and wanted to throw it out there: of the other six games Ruth missed, 5 of them were against fellow southpaw starters. That might have influenced even the last game, the second time he sat against New York's George Mogridge of these six, or getting hit on the hand in the opener (I had missed that this was a double-header). I did a quick check of 1919, and a similar story.

None of this is really relevant to WAR, of course.
5:29 PM Jul 1st
 
DaveFleming
Thanks for the correction, Charles. Noted. You're absolutely right. I just glanced at his games and PA's and assumed he wasn't really playing full-time.
9:18 PM Jun 30th
 
CharlesSaeger
"Ruth was not a full-time hitter each year"

That's simply not true, and prompts an overlong response from me, mostly because I took a look at Ruth's May 1918 some time ago.

In 1918, Ruth was a pitcher-only for the first 18 games of the year. In the 19th game (May 6th in New York), Ruth started at first base. From then on, Ruth was a full-time pitcher/outfielder (occasionally a first baseman). Ruth played only 95 games that year because he:

* Missed 10 games when was a pitcher/pinch hitter in the first 18 games, the only time frame when this was indeed true;
* Missed 10 games in the latter half of May due to some kind of injury (checks Boston Globe, it was tonsillitis);
* The season was cut short due to the economic problems stemming from the Great War; the Red Sox played only 126 games, with the last game before the World Series coming on September 2nd. You can see a baseball fatigue if you look through the sport sections of the year; write-ups get shorter, and small bits like game times get dropped.

Aside from his tonsillitis bout, Ruth didn't have a gap out of the lineup for more than 3 games after May 6th. There were only two gaps of more than one game. He was out two games on July 3rd and 4th (morning game; he played the afternoon), quitting the team after some kind of dispute with Ed Barrow (it sounds like Ruth was asked to do something at bat and didn't, like he refused to bunt or something, and Ruth went back to his dad's house in Baltimore). He was out three games from August 25th to the 30th, after his father died from a fight with George Ruth Sr.'s brother-in-law. So, that's 10 games out for being a pitcher/pinch hitter, 10 games out for tonsillitis, 2 games out for a temper tantrum, and 3 games out his dad's funeral. Otherwise, he took off no more than a game at a time six times, one of which was the first game of a doubleheader, and another was the last game of the season (Ruth pitched the Series opener, so sitting him was understandable, given that the Red Sox had clinched).

1919 is even more clear: Ruth played 130 games, and the Red Sox played 138 games. Again, there were economic pressures shortening the season.
5:51 PM Jun 30th
 
ksclacktc
Thanks.

Personally, I usually find the Win Shares lists matchup better with my eye test. Thus not as surprising as WAR can sometimes be. Perhaps the lower level comparison of WS vs WAR (replacement level is lower). It appears that you are compared to the 24,25,26 players on a roster-not the lowest level starters. And, thus more accurate for Ohtani.
9:28 AM Jun 30th
 
DaveFleming
In terms of Win Shares leaders this year, the best I could pick out:

20 - Shohei
18 - Vlad Jr.
17 - Tatis Jr, Xander, Devers
15- Muncy
14 - Correa, Mullins, Castellanos, Trea Turner, Winker
13 - deGrom, Acuna, Schwarber
9:15 AM Jun 30th
 
DaveFleming
Just following up on Win Shares:

WS has Ruth as:

Year / Batting / Defense / Pitching / Total
1918 - 24.7 - 2.5 - 12.7 - 39.9
1919 - 33.5 - 2.0 - 7.9 - 43.4

Ohtani is currently at
2021 - 14.5 - 0.0 - 5.5 - 20.0

If Shohei has a good start tonight in the Bronx, that might push him up a little bit, and it'd be the halfway point in the season. He's certainly challenging Ruth, but at least according to Win Shares, it's still very close.
9:07 AM Jun 30th
 
ksclacktc
Can someone post a link to Win Shares leaders, as I saw posted earlier. I can't find it anywhere?
8:31 AM Jun 30th
 
ksclacktc
I think this should be the most talked about story in sports, and once again baseball is failing to capitalize on marketing a star player.

And, no WAR doesn't capture everything that he does as many have pointed out. He is as WS shows the MVP in baseball so far this year.
8:23 AM Jun 30th
 
W.T.Mons10
Is he saving the Angels a roster spot? Since he only starts every 6th game, they have to carry a sixth starter, Jose Quintana and his 7.12 ERA.
9:15 PM Jun 29th
 
W.T.Mons10
"But is method accurate?" Not unless 3.2+1.3 somehow actually equals 5.5.
8:43 PM Jun 29th
 
tjmaccarone
I agree that WAR is undervaluing Ohtani, but for a different reason. Ohtani basically saves a team a roster spot by being the starting DH and a starting pitcher. That allows a team to stick an extra above-replacement level player on the roster and get some value from him, while if Ohtani were broken into two players, they wouldn't have that extra roster spot.
8:04 PM Jun 29th
 
DaveFleming
"[Y]ou can't credit a player for his abilities, only what he does - how he has actually helped.?"

Right.

WAR's route to answering that question - how much does a player help his team win games - is to imagine that player against a replacement-level player. How many more wins does this player get us over a typical Triple-A player.

Shohei Ohtani is doing the work of two players, and the metric is trying to answer that question by saying, "Pitcher Ohtani is work x over a replacement-pitcher, and DH-Ohtani is worth y over a replacement-DH...let's add those wins together and call it a day."

But that's not really what Ohtani IS. He is not two separate players, but one player taking on two roles that have historically been distinct. The metric is trying to juggle the challenge of interpreting that contribution within it's framework of replacement value, but I'm not convinced it is doing it accurately.

For anyone interested, there is a Reddit thread that is discussing this very question right now. Some of the BB-Ref staffers are chiming in:

https://www.reddit.com/r/baseball/comments/oa8flw/does_shohei_ohtanis_batting_war_get_positionally/
2:54 PM Jun 29th
 
elwarren
To add to my below comment regarding current Win Shares totals, Ohtani leads the majors. Two individuals are close behind, both batter/fielders: Boegarts and V. Guerrero Jr, who each currently have 18 WS. Remarkably, only Vlady has more WS attributable to Hitting alone (16.8 vs. Ohtani's 13.9)!
2:43 PM Jun 29th
 
hotstatrat
Not sure Ohtani should get credit for taking up less roster space than either a pure position player or pure pitcher, Rallymonky, because if he bats for himself as a pitcher, the Angels need a couple extra players for pinch hitting.

I really don't see the problem here, David. As you started to imply, you can't credit a player for his abilities, only what he does - how he has actually helped.​
2:40 PM Jun 29th
 
elwarren
Seems like this may be an instance in which comparing players' Win Shares would be useful, as WS does not use the positional adjustment schema for the purpose of comparing like positional players (it is indifferent to this aspect, which I see as a positive). According to this site, Ohtani currently has 19 WS while deGrom has 13.

Breakdown is:

Ohtani:
Pitching: 5.5
Hitting: 13.9
Fielding: null

deGrom:
Pitching: 11.8
Hitting: 1.4
Fielding: 0
2:32 PM Jun 29th
 
DaveFleming
And thanks, ksclacktc. That's a very kind thing to say, and I appreciate it.
2:18 PM Jun 29th
 
DaveFleming
Ooooh...that's a good point, R Monkey! Ohtani does have a cost when he starts, with the Angels giving up the DH. I hadn't considered that. Thanks for mentioning that.
2:02 PM Jun 29th
 
Rallymonkey5
Now I see Ohtani is -6, same as Cruz and J.D. Martinez. I don’t know if he’s being treated as a DH while pitching in AL games, or if that’s a function of rounding and different playing time.

The vast bulk of his time is as a DH, he’s played a few more games than Cruz, and Martinez has mixed in some OF with his DH time.
1:55 PM Jun 29th
 
Rallymonkey5
The position adjustment values Ohtani as a DH when he is the DH, and as a pitcher when he is on the mound. The position adjustment column is the sum of those, so at the end of the year it won’t be as much a negative as, for example, Nelson Cruz.

That might overstate his value a bit though. When deGrom hits in a national league game, the alternative is another crappy pitcher hitting. The Angels choose to let Ohtani hit, but they have an alternative better than that, they can use a real DH. And once Ohtani leaves the game they can’t get the DH back. This came back to bite them last week against the Giants, when they ran out of players and had 2 pitchers in a lineup while the Giants had none.

I’m not 100% sure how bbref handles that, but it would be reasonable to use the position adjustment of the DH for him in games where the DH is allowed.

The one place where Ohtani picks up some value not measured by WAR is allowing the team an extra roster spot, since he’s doing two jobs but only taking one place.
1:50 PM Jun 29th
 
ksclacktc
Absolutely super article Dave. You remain my favorite writer on this site!
1:28 PM Jun 29th
 
 
©2021 Be Jolly, Inc. All Rights Reserved.|Powered by Sports Info Solutions|Terms & Conditions|Privacy Policy