It’s only mid-May and things are starting to look dire for the Minnesota Twins. A record that is 12 games under .500 after 38 games and a 10.5 game deficit in the AL Central division standings are not what anyone had in mind for this team. The roster was projected to compete for the division title and post a winning record by a comfortable margin.
So, what’s gone wrong?
The list of topics that could be included as possible answers to that question is lengthy and ranges from team-wide factors to individual performances to just plain old bad luck. With that in mind, I wanted to dig into a handful of team level factors that I think help to explain the Twins terrible start.
Performance in close games
Optimistic Twins fans and analysts will point out that the club, so far, has dramatically underperformed the win-loss record it should have based on its number of runs scored and runs allowed. That approach, known as Pythagorean Winning Percentage, suggests the Twins should be 17-21, a full four games better than they are. The four game difference between the team’s actual record and it’s Pythagorean record is the largest negative delta of any team. While 17-21 would still be worse than anyone expected before the season, it would put the Twins close enough to division-leading Chicago that a brief hot streak could close the gap. Instead, people are starting to think about selling at the trade deadline in July.
One of the primary causes of this difference is Minnesota’s incredibly poor record in close games. Within the Twins' record are an 0-7 record in extra inning contests and a 4-8 record in games decided by 1 run, according to the detailed standings at Baseball-Reference. Those 7 extra inning games are tied with the Los Angeles Dodgers for the second most such contests of any team and only Cincinnati has played more (9 games played, 7-2 record).
In contrast, each of the Twins' AL Central division rivals (except for Chicago) have outperformed their Pythagorean expectations thanks in large part to much better results in close games. The White Sox' 24-15 record is 2 games worse than their projected 26-13 and they are 0-3 in extra innings and 4-3 in 1 run games. At 21-17, Cleveland is 1 game better than projected thanks to going 3-1 in extra inning contests. Kansas City (18-22) is also 1 game better than expected (17-23) because they are 1-0 in extras and 6-5 in close games. Even Detroit, the team that was universally predicted to finish in last place in the division is 1 game better than their Pythagorean expectations thanks to being 3-0 in extra innings and 6-5 in 1 run games.
Of course, one of the best ways to consistently win close ball games is to have a strong bullpen that can hold leads late in games and keep games close to allow for late inning comebacks by the offense. Unfortunately, the Twins have...
The worst bullpen in baseball
Coming off two consecutive seasons where the bullpen was a key strength of the team, the performance of the relievers thus far has been a dramatic disappointment. Using data from FanGraphs, we can see that Twins relievers have been charged with the most losses (13) in baseball, blown 7 saves, and allowed 23 home runs. Altogether, the team’s bullpen earned run average is 4.85, the 7th-worst number in the game. By the advanced measure Win Probability Added (WPA), which assesses how much the result of each plate appearance contributes to the team’s likelihood of winning, the Twins' relievers have hurt the team far more often than they have helped. The bullpen’s collective minus-3.60 WPA is the worst mark in MLB.
Several relievers from last season’s strong bullpen were let go in the offseason and replaced with a number of generally low-cost (and somewhat unproven) replacements. Generally speaking, bullpen performance is volatile over time, so the strategy of diversifying risk with multiple new relievers, instead of spending big on just one or two is sound. However, it has failed spectacularly. Nearly every Twin reliever (whether new or hold over from prior seasons) has failed to meet even modest expectations.
The net result is that Manager Rocco Baldelli has had very few reliable options to turn to in the late innings. This shows up in the win-loss results and the Twins have ended up losing multiple games they were highly likely to win. According to FanGraphs’ models of win expectancy -- which use historical results to give the percent chance a particular team will win based on the score, inning, outs, runners on base, and the run environment -- four of the Twins’ 25 losses have come in games where they had, at one point or another, at least a 94% chance of winning.
But, keeping games close and hanging onto leads through pitching and defense is only part of the equation. Just as important is performing offensively to add runs to extend leads or rally to erase deficits in the late innings. The Twins have not done much of either of those things, thanks in large part to...
Poor offensive performance in crucial situations
On a macro level, the Minnesota offense has been one of the better groups in baseball. By Weighted Runs Created Plus (wRC+), a comprehensive rate statistic that quantifies total offensive value, the Twins rank 5th in MLB. Twins hitters have hit the 4th-most doubles (68), 9th-most home runs (51), and the team averages the 10th-most runs per game (4.61).
They also grade out very well by the advanced metrics generated by MLB’s tracking technology, Statcast. Twins batters have the highest percentage (43.2%) of batted balls that are classified as “hard hit” -- meaning hit with an exit velocity off the bat at 95 miles per hour or higher. They also are 2nd in baseball in barrel percentage (10.1%), which is a classification applied to batted balls with optimal combinations of exit velocity and launch angle that yield batting averages over .500 and slugging percentages over 1.500.
Of course, macro level numbers such as team averages are not consistently distributed and can hide discrepancies in specific situations. For the Twins one of those discrepancies is that their offensive performance in the most critical situations has not been at the same level as their overall performance.
With runners in scoring position (i.e, on 2nd and/or 3rd base), Twins hitters are 21st with a .227 batting average and rank 20th in runs batted in. The numbers decline even further in the highest pressure situations. FanGraphs maintains a statistic called Leverage Index that quantifies how critical a situation is relative to winning the game. In situations that meet the criteria for being high leverage, the Twins are dead last in MLB in offensive production with 38 wRC+. In more traditional terms, they have batted just .144, with a .204 on base percentage, and .295 slugging percentage in high leverage situations.
Combining the two, in high leverage situations with runners in scoring position, Twins hitters are batting just . 139, with a .202 on base percentage, and .307 slugging percentage -- good for 40 wRC+. The team ranks just 25th in runs batted and has hit only 5 home runs and 2 doubles in these situations.
Since high leverage situations most often occur in the late innings, offensive numbers like I've laid out above contribute significantly to the situations the bullpen is faced with navigating. Similarly, the poor bullpen performance contributes to the situations the offense is trying to overcome. Altogether, it can become a vicious cycle.
Thanks to the bullpen struggles and the offense frequently failing to come through in the biggest spots, the Twins have been outscored 67-45 from the 7th inning on, including extra innings. That’s how you go 0-7 in extra inning games, 4-8 in one run games, and start a season well below expecations.