MLB awards: In Rookie Of the Year voting, close counted

Ohtani and Acuna were both plausible winners of MLB awards, but statistically, they weren’t the most deserving candidates

MLB awards Rookie of the Year voters selected two runaway winners Monday, and in both cases, they got the call wrong. They were close…but wrong.

Statistically, the two most deserving candidates were not Atlanta Braves outfielder Ronald Acuna and Los Angeles Angels pitcher-hitter Shohei Ohtani, although both measured up as plausible choices. The NL result, in fact, could plausibly have been a dead heat. In that sense, the most surprising result wasn’t that Acuna and Ohtani beat out Washington Nationals outfielder Juan Soto and New York Yankee infielder Miguel Andujar, but that they did so by such decisive margins.

In the National League, Acuna received 27 of the 30 first-place votes from a writers’ panel and 144 points overall, while Soto received two first-place votes and 89 points.

In the American League, Ohtani received 25 first place votes and 137 points compared with 5 and 89 for Andujar.

The American League vote is especially interesting because of Ohtani’s unique credential: the first truly two-way player in anyone’s memory. As a batter, he made 367 plate appearances and batted .285 with 22 home runs. As a pitcher, he worked 51.3 innings and went 4-2 with a 3.31 ERA before being shut down with elbow problems.

In an awards vote, the challenge is to compare that unique multi-level performance – which has drawn comparisons to Babe Ruth’s pre-1920 work as a pitcher-outfielder – with what Andujar did playing third base for the Yankees. He hit .297 with 27 home runs in 602 plate appearances … but of course he had no pitching line whatsoever.

Generally, the best way to measure the relative contributions of players is to look at their ratio of “wins” in individual batter-pitcher confrontations. The formula for this system, which has been discussed on these pages previously, is not complicated: divide total bases by plate appearances (assigning one base for walks and hit batters). When rating pitchers – who obviously are in the business of preventing the reaching of a base, not achieving same —  reverse the polarity by subtracting the result from 1.000.

Andujar is an easy example to use. He amassed 329 total bases, translating to a .547 percentage of “wins” in his 602 plate appearances. If Andujar isn’t the AL Rookie of the Year, then the actual winner ought to beat that “win” percentage.

Particularly when we judge rookies, though, a problem emerges. This formula makes no adjustment for player usage, and there is an inherent value to playing a full season as opposed to playing only a partial season. So when we assess any player via a “rate” stat – and batter-pitcher wins is a rate stat – we have to adjust it against full-season norms.

In 2018, the average hitter who qualified for the batting title made a hair more than 600 plate appearances. The average ERA title qualifier faced about 780 batters. In every case of a potential Rookie of the Year candidate who failed to reach those appearance levels, we’ll adjust their “wins” total proportionally.

American League

Setting aside the special case that is Ohtani, here are the adjusted batter-pitcher wins for the five other players receiving votes in the AL ROY race. The table shows each player’s raw total of batter-pitcher wins, his number of plate appearances or batters faced, and his adjusted batter-pitcher wins when measured against the norms of 600 plate appearances or 780 batters faced. Pitchers, subject to the 780 batters-faced adjustment, are shown in italics

1. Name, Team, BPW, PA/BF, Adj. BPW
2. Miguel Andujar, New York Yankees, .547, 602, .547
3. Ryan Yarbrough, Tampa Bay Rays, .502, 640, .445
4. Joey Wendle, Tampa Bay Rays, .475, 541, .428
5. Gleyber Torres, New York Yankees, .520, 484, .419
6. Daniel Palka, Chicago White Sox, .520, 449, .389

So the under-appreciated guy is actually Yarbrough, who finished sixth in the actual voting. Of all the candidates, by the way, only Andujar actually met the 600 plate appearance minimum to have his rating considered in full. While he would have ranked first among this group even without any playing time adjustment, that adjustment gives him a decided edge over the field…at least absent Ohtani.

But what of Ohtani? His 367 plate appearances as a hitter and 51.1 innings as a pitcher fail to approach the “full-time” standards. Yet he did both. Here’s his breakdown, beginning with the offensive side.

In those 367 plate appearances, Ohtani earned 223 total bases, giving him a gross .610 “win” percentage. That’s better than Andujar. But since 367 is only about 61 percent of our full-time average of 600, his score reduces to a net of .373. As we consider his mound work, he stands .174 short of Andujar.

In his 51.2 innings as a pitcher, Ohtani faced 216 batters. He allowed 85 total bases, giving him a .606 “win” percentage once we reverse the polarity of the .394 average.

Ohtani’s problem is that those 216 batters faced begin to look pretty paltry measured against the MLB average of 780 for ERA qualifiers. Since he faced fewer than 28 percent of the batters needed to qualify as a full-time pitcher, his .606 raw score is reduced to .168. And that plus his .373 score as a hitter puts his total at .541. That’s a solid second, and it’s close to Andujar’s .547. But Andujar, not Ohtani, deserved the award.

National League

Because the NL race lacks a pitcher-hitter combo, the math is simpler and the result is equally close. Again, however, the voters got it wrong. Also as with the AL vote, there was a badly overlooked candidate among the final three.

Here are the adjusted batter-pitcher wins for the eight players receiving votes in the NL ROY race. The table shows each player’s raw total of batter-pitcher wins, his number of plate appearances or batters faced, and his adjusted batter-pitcher wins when measured against the norms of 600 plate appearances or 780 batters faced. Again, pitchers are shown in italics.

1. Name, Team, BPW, PA/BF, Adj. BPW
2. Juan Soto, Washington Nationals, .593, 494, .488
3. Ronald Acuna, Atlanta Braves, .595, 487, .483
4. Jack Flaherty, St. Louis Cardinals, .593, 631, .480
5. Brian Anderson, Miami Marlins, .472, 670, .472
6. Walker Buehler, Los Angeles Dodgers, .652, 550, .463
7. Harrison Bader, St. Louis Cardinals, .555        379, .337
8. Yoshi Hirano, Arizona Diamondbacks, .623, 273, .279
9. Jeff McNeil, New York Mets, .504        248, .208

Even more than in the American League, this is a breathtakingly close call, Soto’s ever-so-slight edge in workload winning out over Acuna’s equally microscopic advantage in batter-pitcher wins. In this vote, a tie would have been excusable, and a close win for either of the top two easily understandable.

Less justified was the voters’ placement of Buehler third ahead of both Flaherty – who pitched more innings and faced 80 more batters – and Anderson, who played a solid third base in Miami, the baseball equivalent of witness protection. Among NL ROY candidates, by the way, Anderson was the only one to exceed 600 plate appearances. In baseball, showing up counts; you can’t help your team from the bench.

So what do you think? Did MLB awards voters get it right? Comment below!