Just about everything that could go right for the 2016 Texas Rangers did. Flying in the face of preseason projections that called for them to finish around .500, the Rangers rode stellar seasons from Adrian Beltre, Cole Hamels and Elvis Andrus (among others) all the way to 95 wins, the AL West crown and the league’s best record. At the same time, however, they relied on a combination of metrics that confounded sabermetricians all season long: a 36-11 record in one-run games — the best mark in MLB history — and on top of that, a better run differential than we’d expect from their underlying stats, too.
In other words, number-crunchers suspected the Rangers had been massively, historically lucky last season — and were probably due for a downturn in 2017. But even the statheads didn’t see Texas’s fortunes reversing quite as much as they have. After that record-setting mark in close games a year ago, the Rangers somehow have the league’s second-worst winning percentage in one-run contests this year, and they’ve also scored fewer runs (and allowed more) than their statistics would predict. All the things that went right last season are now going wrong; as a result, Texas is in fourth place with little chance of making the playoffs.
Here’s the funny thing, though: Deep down, the Rangers’ 2017 squad is probably every bit as good as its 2016 iteration, and possibly a little better, despite the huge decline in winning percentage. According to wins above replacement (WAR),21 Texas had the 16th-best team in baseball last season with 32.4 total WAR. (For a sense of scale, the average team has about 33 WAR,22 meaning the 95-win Rangers were basically a middle-of-the-road ballclub after stripping away their good fortune in high-leverage situations.) This season, the Rangers are on pace for 32.8 total WAR over 162 games — ever so slightly more than they posted a year ago, and good for 13th-most in MLB.
Some Texas players have been even better than last year (Andrus) and some disappointingly worse (Rougned Odor), but on the whole they’ve played basically the same at an individual level. The only real difference between the two versions of the Rangers has been in sequencing — that is, scoring more (and/or allowing fewer) runs than the underlying stats would predict, because events were timed right within an inning — and performance in close games. As Texas is learning, those categories can be extremely fluky, and teams who overachieve in them one year tend to see their records dramatically regress to the mean in the following season.
We can measure how much a team benefited from each kind of good fortune with a couple of metrics: Base runs, which estimate how many runs a team “should have” scored and allowed based on its underlying stats, and the Pythagorean expectation, which estimates how many games a team “should have” won based on its run differential. Deviations from each estimate come, respectively, from fortunate sequencing and good luck in close games. And by those standards, last year’s Rangers — who beat their estimated record by a staggering 13 wins — were one of the luckiest teams in baseball history. According to TheBaseballGauge.com, they tied for the third-biggest difference between a team’s actual and expected records during the expansion era.23
After the luck has gone
MLB teams with the most extra wins of luck from sequencing and close games, 1961-2017
| ACTUAL RECORD | WINS OF LUCK | ||||||
|---|---|---|---|---|---|---|---|
| YEAR | TEAM | WINS | LOSSES | SEQUENCING | CLOSE GAMES | TOTAL | CHG. IN WINS NEXT SEASON |
| 2008 | Angels | 100 | 62 | +5 | +12 | +17 | -3 |
| 1969 | Mets | 100 | 62 | +6 | +8 | +14 | -17 |
| 1963 | Dodgers | 99 | 63 | +6 | +7 | +13 | -19 |
| 1987 | Cardinals | 95 | 67 | +10 | +3 | +13 | -19 |
| 2006 | Athletics | 93 | 69 | +6 | +8 | +13 | -17 |
| 2007 | D-backs | 90 | 72 | +2 | +11 | +13 | -8 |
| 2016 | Rangers | 95 | 67 | 0 | +13 | +13 | -15 |
| 1973 | Reds | 99 | 63 | +7 | +5 | +12 | -1 |
| 2012 | Orioles | 93 | 69 | 0 | +11 | +12 | -8 |
| 2013 | Yankees | 85 | 77 | +6 | +6 | +12 | -1 |
| 1961 | Reds | 93 | 61 | +1 | +10 | +11 | 0 |
| 1962 | Reds | 98 | 64 | +6 | +5 | +11 | -12 |
| 1972 | Mets | 83 | 73 | 0 | +11 | +11 | -4 |
| 1977 | Orioles | 97 | 64 | +2 | +9 | +11 | -7 |
| 1982 | Red Sox | 89 | 73 | +7 | +4 | +11 | -11 |
| 1984 | Mets | 90 | 72 | -1 | +12 | +11 | +8 |
| 1985 | Angels | 90 | 72 | +5 | +6 | +11 | +2 |
| 1985 | White Sox | 85 | 77 | +8 | +2 | +11 | -13 |
| 1987 | Twins | 85 | 77 | +5 | +6 | +11 | +6 |
| 1989 | Astros | 86 | 76 | +3 | +7 | +11 | -11 |
Even among those lucky teams, the Rangers’ fall this season has been especially steep. They’re on pace for a 15-win decline from 2016, in part because their luck has actually turned in the opposite direction. So far this season, they’ve managed to win one fewer game than they would have with neutral sequencing and normal luck in close games. And this fickle twist of fate is basically the difference between Texas comfortably returning to the playoffs and being on the outside looking in: The Rangers are currently two games out of the American League’s second wild-card spot, with five teams ahead of them to leapfrog as well. If they had last year’s luck again, they’d be five games clear of the New York Yankees for the first wild-card.
Instead, the Rangers serve as yet another reminder of just how unpredictable baseball can be. A team can boast practically identical talent in consecutive seasons,24 and still see their record fluctuate wildly from one year to the next. In the face of such a cruel and random universe, the best a team and its fans can do is to relish the breaks when they go their way, and enjoy the small blessings of a season when they don’t.
Check out our latest MLB predictions.