The Traps That the Twins Need to Avoid

It’s June 1st, and the best record in the American League belongs to the Minnesota Twins. As winners of seven of their last eight games — all against teams who expected to contend coming into the season — the Twins now stand at 30-19, a half game ahead of the Royals in the AL Central; they’re also three and a half games ahead of Detroit and six and a half ahead of Cleveland. That’s a pretty great start for a team that we gave basically no chance of contending this year, as our pre-season forecasts had them as a 74 win team, and gave them just a 3% chance of reaching the postseason.

After two months of playing .600 baseball, our projections now expect the Twins to finish with 81 wins, and give them a 27% chance of reaching the postseason. On the one hand, that’s a huge jump, and a roughly one-in-four chance of making the playoffs in what was supposed to be a rebuilding year is quite the accomplishment. On the other hand, those 27% odds still put them behind eight other teams, and those odds are buoyed by their chances of reaching the less-valuable Wild Card game; Minnesota’s estimated 12% chance of winning their division ranks just 10th in the AL.

The Twins strong start is both a blessing and a curse. Watching winning baseball is a lot more fun than losing baseball, and surprising playoff runs can invigorate a fan base, driving significant revenue gains from increased attendance and television ratings. Even just putting a watchable product on the field can help a franchise avoid a financial death spiral, and even if the Twins cool off in the second half, their strong start should help keep people interested in baseball through most of the summer.

But there’s a potential downside here too, because while the Twins front office should absolutely be enjoying their strong start to the season, they need to be realistic about what it does and does not mean. And it does not mean that the 2015 Twins are actually a good baseball team.

The Twins have a .612 winning percentage, second best in baseball, but their BaseRuns expected winning percentage is .424, sixth-worst in baseball.

BaseRuns looks at the expected run values of every play each team is involved in, and creates estimated runs scored and runs allowed totals from those events, under the assumption that teams are going to distribute those events across types of context in a normal way. The way to dramatically outperform (or underperform, in the A’s case) your BaseRuns expectation is to cluster all of your positive events together, making sure you don’t strand your own runners and do strand opposing runners, while also performing far better in high-leverage situations than when the game is already decided.

The Twins have done exactly that in the first two months of the season. With the bases empty, the Twins have a .284 wOBA, but that jumps to .333 with men on base, and .351 with men in scoring position. That kind of sequencing is why BaseRuns looks at the Twins offense and sees a team that should be scoring about 3.9 runs per game, but the Twins are actually scoring 4.6 runs per game. Likewise, the pitching staff is expected to be giving up 4.6 R/G, but is only actually at 4.2 R/G, so the Twins have outscored their opponents by 21 runs while BaseRuns expects that they should have been outscored by 34.

This is a great example of why I’m not a big fan of pythagorean expected record; often times, statistically inclined writers have used pythag to strip out the sequencing in the conversion of runs to wins, but that ignores the fact that there is an awful lot of sequencing that goes into those runs themselves, and unfortunately for teams like the Twins, there’s just no evidence that teams have the ability to consistently and sustainably beat their BaseRuns expected records; the model is very good, and teams generally fall within a pretty narrow band around their expected record.

We have BaseRuns data going back to 2002, and in the 13 full seasons that we’ve tracked BaseRuns data for, no team has ever exceeded their BaseRuns expected winning percentage by more than 70 points; the 2008 Angels had a .617 Win% versus a .544 BaseRuns Win%, which translated out to 12 wins above their expected record; the 2005 Diamondbacks and 2004 Yankees matched that total, so three times out of 390 team-seasons a team has managed to steal a dozen extra wins by stacking their events in the most optimal way possible.

The Twins are currently beating their BaseRuns winning percentage by 188 points, which translates to 30 extra wins over a 162 game season. To say that they can’t keep this up might be the understatement of the year. The reality is that the Twins are not actually playing like a very good team, and sooner than later, their record is going to reflect that. You just can’t keep winning games the way the Twins have been winning them in April and May, and they don’t appear to have enough talent to win games the old fashioned way of just being better than their opponents. Pretty soon, the Twins are likely going to go back to looking like the building-for-the-future club that everyone expected.

With the trade deadline just two months away, though, the team’s strong start will likely still have them in the midst of the playoff race when the buy/sell decisions have to be made. Even if they play just .450 ball the rest of the way, as our projections suggest, that prorates out to a 23-27 record over their next 50 games, which would leave them at 53-46 headed into the final stretch of the season. Yes, the Twins could just implode and make it clear that the team isn’t actually a contender, but that’s not what regression to the mean usually looks like, and it’s more likely that the Twins are hanging on in the middle of the Wild Card race when the trade deadline rolls around.

So Terry Ryan and his staff are going to have to do some difficult balancing. On the one hand, you don’t want to throw away an opportunity to make a postseason run when it presents itself, even if you don’t think you really have a playoff-caliber roster. After all, it was less than a year ago that I wrote a piece entitled “It’s Time for the Royals to Trade James Shields“; they ignored me and went to the World Series instead. While playing the probabilities is still an organization’s best bet, it’s not like these forecasts are set in stone, and a reasonable chance for the Twins to give their fans a special season isn’t worth throwing away.

But at the same time, this team isn’t actually all that good, and probably are going to fade out of the race by the time the season ends. The Twins don’t have enough good assets for the future to start trading them in to make upgrades to a team that probably won’t be good enough to keep up with the real contenders in the AL over the next four months, and even with the best record in the AL on June 1st, the Twins should still be prioritizing the future over the present. You don’t want to trade a guy who could have a real impact on your team in the next few years for a guy who might just help you finish this year with 82 wins instead of 81.

The Twins strong start is going to take them out of the seller’s market, but it also shouldn’t push them strongly into the buyer’s market either. If they can take some low-cost flyers on guys like Aaron Harang or Marlon Byrd, hey, okay, knock yourself out, and see what happens. It’s worth throwing a few minor chips into the pot to see if the team starts actually playing better and steals a Wild Card spot. But it’s not worth giving up significant pieces of the future to try and keep up with the big boys who are winning by actually outplaying their opponents rather than out-sequencing them.

There are ditches on both sides of this road, and the Twins front office is going to navigate away from both going too strongly and too weakly after their new-found chance at a 2015 playoff spot. Stay the course, make some minor upgrades if they don’t cost much, and maybe this turns into a great story. Just don’t get tricked into thinking that their record is reflective of the fact that this is a team that’s ready to win. Smoke and mirrors don’t last, and it’s up to the front office to realize that they’re better off building a team that doesn’t need them.