clock menu more-arrow no yes

Filed under:

Is it better to be lucky than good?

New, 15 comments

Being a modern, SABR-minded baseball fan can make “unsustainable” success complicated.

Los Angeles Angels of Anaheim v Minnesota Twins
Ervin Santana stares into the void
Photo by Hannah Foslien/Getty Images

The Twins represent one of the biggest surprises of the 2017 MLB season, behind perhaps the unexpected success of the Brewers and Diamondbacks and the surprise ineptitude of the Giants. (The Mets’ implosion feels predestined, somehow.)

Although the Twins no longer sit atop the American League Central, they can claim a winning record (44-41, 1.5 games behind Cleveland) through July 6th. That’s a huge improvement over last season, when they were 29-55 (22 games behind first-place Cleveland) through the same date.

The 2017 Twins have nearly the exact record the 2015 vintage had at this point in the season; the ‘15 team boasted a 44-39 mark through July 6.

There is one key difference between the 2015 team — also a surprise second-wild-card contender for most of the season — and this year’s team. At this date in 2015, the 43-39 Twins had outscored their opponents 340-336, a +4 run differential. This year, the 43-41 Twins have been outscored 388-441, a staggering -53 run differential.

By winning percentage (.512), the Twins rank 11th in baseball, surrounded by the 44-40 Royals and the 44-42 Rays. By run differential, however, they’re 25th, sandwiched by the Athletics, who lack any clear direction, and the Braves, who have been actively trying to lose for a couple years.

The Twins boast the largest difference between actual record and Pythagorean won-loss record (Pythag win/loss, a Bill James creation, estimates a team’s “true” winning percentage using runs scored and allowed) in MLB.

Because hits and outs are essentially distributed randomly, run differential provides a better snapshot at a team’s true talent than record, the thinking goes.

10 Luckiest Teams by Pythag Diff.

Rank Team W-L Pythag W-L Difference
Rank Team W-L Pythag W-L Difference
1 Twins 44-41 38-47 +6
2 Padres 36-49 30-55 +6
3 Orioles 40-45 35-50 +5
4 Rockies 50-38 46-42 +4
5 Royals 44-40 40-44 +4
6 Braves 41-43 37-47 +4
7 Athletics 38-48 35-51 +3
8 Astros 58-28 56-30 +2
9 Blue Jays 40-45 38-47 +2
10 Angels 44-45 43-46 +1

It’s perhaps not a surprise that the Twins’ Pythagorean plus/minus is six, the same number of pitching appearances as one Chris Gimenez.

(Since James introduced Pythag won/loss, numerous offshoots — Fangraphs’ BaseRuns and Baseball Prospectus’ First-/Second-/Third-Order Winning Percentages, most notably — have expanded on his research, baking in underlying statistics, park adjustments and strength of schedule. Without getting too lost in the weeds, let’s just say that all of these metrics rate the Twins about the same, if not slightly lower than Pythag.)

The Twins have managed a pretty good record despite a poor run differential for three main reasons: (1) they’ve managed to shoehorn a lot of those runs-against into some severe blowouts (see the red bars below).

Image via Baseball-Reference

(2) They have a 10-5 record, with a +5 run differential, in 1-run games despite having one of the worst bullpens in Major League Baseball.

Twins Bullpen Ranks

Team/Rank ERA FIP xFIP WAR K% K-BB%
Team/Rank ERA FIP xFIP WAR K% K-BB%
Twins 4.95 4.94 4.75 -0.2 18 9.6
MLB Rank 27th 29th 29th 28th 30th 30th

And (3), they’ve benefited ever-so-slightly on the pitching side from Cluster Luck, a delightfully named statistic developed by mathematician Ed Feng.

(The Twins have also had an above-average defense, but it has regressed slightly over the last month from its lofty heights at the beginning of the season. If you want to argue that the defense is a large part of the team’s “luck,” I wouldn’t disagree too strenuously.)

Cluster Luck, again, starts with the assertion that clutch hitting is a myth — which of course people may disagree with, despite the fact that no evidence of it exists. Cluster Luck drills one level deeper than run differential to say, “Sure, the Twins are getting outscored; but based on their/their opponents’ hit distribution, should they be?”

The 2015 Twins benefited greatly from Cluster Luck; it was the primary reason the lineup kept scoring runs and the FSN broadcasters kept talking about how clutch they were and how vital Torii Hunter was to clubhouse chemistry. Whenever a big hit was needed, the ‘15 Twins delivered.

This year’s team is far less lucky than two years ago, and on the hitting side has actually been slightly unlucky (8.5 runs below expected). The pitching staff sits at +3.4 runs above expected entering Friday’s game, and Thursday’s 6-4 win over Baltimore provides a prime example of how Cluster Luck can turn a game — and potentially a season.

The Orioles outhit the Twins 11-8 but lost 6-4. Why? The Twins clustered four of their hits in one inning, while the Orioles never had more than two hits in a single inning.

Orioles vs. Twins Hits Per Inning 7/6/17

Team 1 2 3 4 5 6 7 8 9
Team 1 2 3 4 5 6 7 8 9
Orioles 0 1 2 2 1 1 2 1 1
Twins 0 1 4 0 0 2 0 1 x

Fangraphs’ Win Probability chart from Thursday’s game demonstrates just how damaging a four-hit inning — coupled with a walk, a HBP, a steal and an extremely fortuitous bounce — can be.

Image via Fangraphs

I know that all this data can be overwhelming. It overwhelms me sometimes.

Many fans argue that advanced statistics and analytics devalue or cheapen the game — that they turn baseball from a first love into a formula.

Throw in the modern fans’ obsession with “fungible assets,” prospects, team-building, byzantine roster machinations — this time of year, all the chatter is about buyers and sellers, hoarding assets, tweaking the 40-man, seemingly everything but actual baseball — and modern fandom can feel overwhelmingly egg-headed.

But I think having this glut of information actually helps me cherish this Twins team. We’re not supposed to be this good, and yet we are.

Byron Buxton should not be valuable with a .592 OPS, and yet he is. Robbie Grossman should not be able to reach base without ever swinging, and yet he does. Ervin Santana should not have carved out an improbable late-career renaissance based on little more than guile, and yet he has. Miguel Sano should not be able to strike out more than 13 of the time and be elite, and yet he is.

I love that Brandon Kintzlerone of nine pitchers in baseball history with more than 22 saves and fewer than 25 strikeouts through the first half of the season — continues to close out games despite possessing none of the traits or abilities one would want in a closer. Call it moxy or grit, call it unsustainable or over-performing his peripherals — whatever. I don’t care. It’s fun to watch him sweat and wriggle through the ninth knowing he can’t count on swings-and-misses.

I love hard-hit balls finding our outfielders’ gloves, and infielders with little track record of fielding prowess making “!!!” plays.

I love watching a team over-perform. I love watching a team getting hits in bunches or holding on to a lead they have no business hanging on to.

Sure, it’d be nice to cheer for a juggernaut, and I hope some day I will. But in the meantime, I’ll take this: knowing that it can’t last, it won’t last, not for too much longer, but hoping all the same that it will — and, as we inch toward Game 162, cherishing each and every day the Twins remain in the mix.

Between 2011 and 2016, the Twins were in first for a total of 17 days.

This season, they’ve claimed the top spot for 50. It may be unsustainable and it may be fluky, but each one of those glorious 50 days counts just the same.

You be good. I’ll take the luck.