ESPN's Dan Szymborski has an interesting look at baseball's MVP races and argues why Los Angeles Angels center fielder Mike Trout should still be the top candidate in the American League. Of course, we know the case against Trout: The Angels aren't going to make the playoffs, so his performance has apparently played out in a vacuum of meaninglessness. Szymborski presents a chart that lists how much a team's World Series odds would decrease minus the individual performances of key players. The top two players here are Indians right-hander Corey Kluber and Cubs slugger Kris Bryant. This is the conventional "who means the most to his team" MVP argument. Trout is nowhere to be found, since the Angels aren't contenders.

As Szymborski writes, however:

The chart points out the difficulty of actually applying this argument consistently rather than in an arbitrary fashion. Bryant still wins the NL MVP award as of today based on this measure, but Trout doesn't. But the only argument that can logically make Trout not the current best candidate for MVP also means that he's not even the 100th-best player. If MVP voting means that the meaning of the wins is paramount, then even Dariel Alvarez, who went 1-for-3 with a double in two games for the Orioles, had more of an effect on his team's World Series chances than Trout did.

Trout is universally regarded as the best all-around player in baseball. He holds a significant lead over Mookie Betts in Baseball-Reference WAR, 9.3 to 7.8 (and 8.4 to 6.9 on FanGraphs). If you asked 30 general managers which player they would want on their team, knowing they would get that player's 2016 performance, all 30 would likely take Trout. Yet he's not going to be named American League MVP -- and not just because the Angels are 62-77. Voters still love their conventional AVG, HRs, RBIs stat lines. In that department, Trout doesn't stand out from Betts or Jose Altuve:

Trout: .323, 27, 88

Betts: .316, 30, 100

Altuve: .344, 22, 92

All three have also stolen 20-plus bases, so Trout doesn't even get extra credit in that department. So why does Trout have the higher WAR? He has a big edge in on-base percentage, for starters -- .442 to Altuve's .406 and Betts' .357; he has more walks than Altuve and Betts combined. That's important; outs matter. He also has an edge over Altuve in baserunning value. FanGraphs estimates Trout has been worth 7.7 runs above average on the bases compared with Altuve's minus-0.3 (Betts is at 8.8, as he ranks second in the majors and Trout fourth). This stat includes double plays hit into, and Altuve has hit into 13 compared to Trout's four. Betts also plays in a good hitter's park, which cuts into his value (runs produced in a higher run environment are less valuable since there are more of them); Trout plays in a pitcher's park.

Defensively, Betts has been spectacular, according to Defensive Runs Saved, at plus-24. Trout is rated at plus-6, and Altuve at 0. That's what makes Trout so good: He has no weaknesses. The only chance Trout would have of winning MVP honors while playing on a bad team is to compile big home run and RBI numbers. But in a season in which home runs are way up, Trout isn't going to come close to the 41 he hit last season. He may not drive in 100 runs. Dismissed (by some).

* * * *

Only seven times has a player won an MVP award while playing on a non-winning team. To further emphasize my point about home runs and RBIs, take a look at those seven MVPs. It also helped that there wasn't always a candidate as strong as a Betts or Altuve.

2003: Alex Rodriguez, Texas Rangers (71-91)

After hitting .298/.396/.600 with 47 home runs and 118 RBIs, A-Rod was clearly the best player in the league, and the voters reluctantly crowned him MVP in a split vote -- 10 different players received first-place votes. Jorge Posada of the Yankees finished third, the highest placement for a guy on a playoff team. Rodriguez's value was pretty clear in those pre-WAR days -- he finished with 8.4 WAR, and the only others above 6.5 were pitchers Pedro Martinez (14-4, 2.22 in an abbreviated season) and Esteban Loaiza (21-9, 2.90), but both were non-factors in the voting.

1991: Cal Ripken, Baltimore Orioles (67-95)

One of the strangest MVP seasons of all time -- Ripken had hit between .250 and .264 the previous four seasons and would hit .251 and .257 the next two -- he hit .323/.374/.566 with 34 home runs, 114 RBIs, monster numbers in those pre-steroids days, especially for a shortstop. He'd also win a Gold Glove. In a controversial vote at the time, he edged out Cecil Fielder, who played for the 84-78 Tigers. Fielder himself whined vociferously about the results. The idea that people thought Fielder, who hit .261 with 44 home runs, was as valuable as Ripken was absurd then and even more so now. Ripken beat him in WAR, 11.5 to 3.7.

1989: Robin Yount, Milwaukee Brewers (81-81)

Another mixed vote, as five different players received at least three first-place votes. Yount edged out Ruben Sierra, Ripken and George Bell (whose Blue Jays made the playoffs), as voters focused on RBI guys. Ripken's WAR was 6.6, Sierra's 5.9, Yount's 5.8 and Bell's 3.3. The best candidates were Cy Young winner Bret Saberhagen of the Royals (9.7 WAR), Rickey Henderson (8.6 WAR on two teams) and Wade Boggs (8.4). Those three finished eighth, ninth and 21st in the voting.

1987: Andre Dawson, Chicago Cubs (76-85)

Widely recognized as one of the worst MVP selections, and not just because the Cubs finished in last place in the NL East. In the rabbit ball year of 1987, Dawson led the NL with 49 home runs and 137 RBIs, but he posted a low .328 OBP and ranked 18th among NL position players with 4.0 WAR. Tony Gwynn led in WAR, but his Padres were even worse than the Cubs. Ozzie Smith of the first-place Cardinals finished second in the voting and with 6.4 WAR would have been the best conventional candidate (teammate Jack Clark was third in the voting, splitting votes that helped Dawson win).

1959: Ernie Banks, Chicago Cubs (72-82)

1958: Ernie Banks, Chicago Cubs (74-80)

The power-hitting shortstop won back-to-back MVP honors, leading the NL both seasons in RBIs and in home runs in 1958. He finished second in WAR to Willie Mays in '58 and first in '59, so he was certainly a worthy candidate, impressing the voters so much that they tossed aside their usual rules. (Or did they simply pick the guy with the most RBIs?)

1952: Hank Sauer, Chicago Cubs (77-77)

Sauer wasn't bad -- 5.7 WAR -- but this was a strange choice, going for a 35-year-old slugger who hit .270 for a mediocre team. Oh, wait, he led the league in home runs (37) and RBIs (121). Robin Roberts went 28-7 for the 87-win Phillies and finished a close second in the voting. The top player for the pennant-winning Dodgers was rookie reliever Joe Black, who went 15-2. Jackie Robinson hit .308/.440/.465 with 19 home runs, 24 steals, 106 walks and 104 runs, worth 8.5 WAR. How could voters think Sauer was the more valuable player? Robinson finished a distant eighth in the voting.

Of course, we didn't have WAR in these cases to provide a starting point for argument. And it's certainly not the end-all to the discussion. In the cases here, four were players who were clearly the best in the league (Rodriguez, Ripken) or arguably the best (Banks). The other three seem odd choices in hindsight.

So it appears the AL MVP race is going to come down to Betts and Altuve (and maybe Josh Donaldson and Manny Machado). To be clear: They're not weak candidates by any means, as all four are having MVP-caliber seasons. This isn't a Juan Gonzalez campaign we're talking about. Mike Trout, however, is still the best player in the game.