Just How Lucky Can One Team Be?

Last September 16, the Texas Rangers trailed the Oakland Athletics by one in the bottom of the ninth, with Oakland closer Ryan Madson on the mound. After Rangers leadoff batter Ian Desmond grounded out, the rally began. Trade-deadline pickup Carlos Beltrán doubled, and Delino DeShields pinch ran. A wild pitch moved DeShields to third, and an Adrián Beltré walk put runners on the corners. Joey Gallo, who pinch ran for him, stole second, setting the scene for Jonathan Lucroy, another trade-deadline addition, to drive in DeShields and Gallo — the tying and winning runs — with a single down the left-field line.

The Rangers, who’d been behind by four runs as late as the sixth, walked off, becoming the third team since 1974 to win eight times in a season after trailing through eight innings. Just another night in the life of one of the luckiest teams of all time.

Rangers fans are probably tired of reading that, and understandably so. Analysts labeled the Rangers lucky for much of last year, but the team’s luck never turned: The Rangers ran right off the ledge, and they never looked down. They won an AL-leading 95 games and an AL West title before falling to Toronto in an ALDS sweep.

Those 95 wins aren’t going away. Rather than try to tarnish them, we should celebrate their improbability: The Rangers’ refusal to lose, despite all the signs that suggested they would, made them one of baseball’s best stories. But we can’t ignore that the 2016 Rangers had truly fantastic timing. And the way they won last year tells us something about the 2017 team, because that type of timing is impossible to sustain.

The case against the 2016 Rangers’ true talent boils down to run differential, the difference between a team’s runs scored and allowed. Like most so-called advanced stats, run-differential-based metrics rely on an intuitive concept: To win one game, a team has to outscore its opponent. To win many games, a team has to outscore its opponents regularly. If a team outscores its opponents regularly, it will finish the season with many more runs scored than allowed.

Unless that team is the 2016 Rangers. The Rangers outscored their opponents by only eight runs, but they won 28 times more than they lost.

Prior to last year’s Rangers, only one team in history had ever won more than 90 games with a run differential that low: the similarly confounding 2012 Orioles, who outscored their opponents by seven runs but won 93 times. To find another team with at least 95 wins — the 1977 Orioles — we have to raise the minimum run differential to 66 runs, 58 more than the Rangers’.

The Rangers beat the odds by being incredibly clutch. FanGraphs offers a stat called Clutch, which measures how much better or worse players and teams perform in high-leverage situations than they do overall. The Rangers were the second-clutchest team in the database, which goes back to 1974. Their Clutch score suggests that they won 13 extra games because their production came at opportune times.

Partly as a result, the Rangers recorded the best-ever winning percentage in one-run games, the outcomes of which are largely independent of team quality.

Clutchness is valuable, but it isn’t repeatable. The 2012 Orioles appear on both of those tables, and we know what happened to them in 2013: Their Clutch rating fell from 10.6 to 3.2, their one-run winning percentage plummeted from .763 to .392, and they lost eight wins from their total — not enough to make them a bad team, but enough to make them miss the playoffs. The fate of one team tells us only so much, but if we look at a larger group of teams that resembled last year’s Rangers, we can estimate how much regression might be in store for the 2017 team.

Baseball Prospectus has a stat called third-order winning percentage, which estimates a team’s winning percentage based on its underlying performance and the quality of its opponents. The 2016 Rangers’ third-order record pegged them as a 79–83 team — or, to be as precise as Baseball Prospectus, a 79.1–82.9 team. That 15.9-win difference between the Rangers’ actual and estimated records is the greatest in BP’s records, which go back to 1950.

If we take the 50 teams with the biggest differences between their actual and third-order records (excluding 2016), we find that they bested their third-order records by 11.1 wins, on average. As one would expect, given that large an advantage, those teams tended to do well, averaging a .557 winning percentage — a 90-win pace over a 162-game schedule. But in the season immediately following those 50 teams’ big third-order overperformances, the difference between their actual and estimated records was completely wiped away; they won exactly as often as their stats said they should. And with that good luck or good timing eliminated, they lost, on average, eight wins over a 162-game season, just as the Orioles did from 2012 to 2013.

It’s hard to watch a team defy the stats for six months and not believe there was more than luck involved: smart managing, maybe, or a good late-inning bullpen, or team-wide intestinal fortitude. And maybe there was some skill involved in the Rangers’ extreme success in stretching their runs. But if defying one’s third-order record were mostly skill, we would expect the teams that did it best to keep doing it to a degree. And historically, they haven’t.

All of which helps explain why the preseason projections have the 2017 Rangers finishing around where the 2013 Orioles did: third place. Compared with the active Astros and Mariners, the Rangers had a quiet winter. They lost their second-most-valuable player, Desmond, to free agency, along with Colby Lewis, Mitch Moreland, and Beltrán. Off the field, the Rangers took two more blows: Thad Levine, who’d served as the team’s assistant GM for more than a decade, became the GM of the Twins, and the new CBA limits spending in the international market, which the Rangers have mined for more talent than any other team. Their highest-profile additions, Mike Napoli, Andrew Cashner, and Tyson Ross are, respectively, a 35-year-old DH coming off a one-win season; an injury-prone pitcher who was near replacement-level last season; and a starter who got into one big league game last year before having surgery to correct thoracic outlet syndrome, which will keep him out until at least May or June.

Although most of the players who contributed to those 95 wins will still be in Arlington, the projections don’t expect their almost-miraculous 2016 timing to carry over into this year. So in light of the Rangers’ likely regression and the increased competition from their AL West rivals, what can fans of Jon Daniels’s and Jeff Banister’s team hang their hopes on this season? Three things:

Double the Darvish

The Rangers’ erstwhile ace spent 101 days on the DL last season recovering first from 2015 Tommy John surgery and then a shoulder strain suffered soon after his return. Altogether, Texas got 17 starts and 100 1/3 innings from one of the best pitchers in baseball. When active, Darvish pitched more or less like his old self, ssetting a new high in strikeout-minus-walk rate and average fastball speed. An extra half-season of health from Darvish might mean more to the Rangers than any free-agent starter they could have added on this winter’s weak market.

More From Midseason Acquisitions

July trade targets Lucroy and Jeremy Jeffress and August signee Carlos Gómez combined for 3.0 wins above replacement in a combined 349 plate appearances plus batters faced for the Rangers. There’s no telling how Gómez will follow a strange bust-and-boom season that saw him released by one Texas team only to star with another, but the three players should more than make up for any drop-off in their rate stats with the Rangers by totaling more playing time.

Progress From Young Players

In addition to banking on the ageless Adrián Beltré continuing to play like he’s in his prime, the Rangers will hope that their young players and post-hype prospects make major strides. Maybe not-yet-22-year-old Nomar Mazara will up his patience or power and walk or elevate more often in his sophomore season. Maybe DeShields’s winter weight loss will yield a return to his 2015 form. Maybe Gallo’s improvement in his second season at Round Rock was the prelude to a survivable strikeout rate at the major league level. Maybe Jurickson Profar can top 90 games and a .338 slugging percentage.

The projections, of course, already know that the Rangers will get more from their midseason additions, and that their pre-prime players might have improved production ahead. Even so, the stats think they’re no better than several other AL wild-card contenders — which tells us that one way or another, luck and timing might make the difference for this Rangers team, too.

Thanks to Rob McQuown of Baseball Prospectus for research assistance.

An earlier version of this story misspelled the name of a former Rangers outfielder. He is Carlos Beltrán, not Carlos Beltáan.