Stat Chat: Better shots for lesser teams

Juliette Pope/The Miscellany News

More than a few interesting questions loom over the 2020 baseball season. Last year, the dynamic Houston Astros won 100 games for the third time in a row. They posted the best run differential (runs scored minus runs allowed) in either league for the second straight year, and clinched their second World Series berth in three years. In last year’s World Series, the Astros faced the Washington Nationals, who, after narrowly eliminating the Milwaukee Brewers 4-3 in the wild card game, possessed the worst record among the remaining playoff teams. It was a bit of a surprise that the Nats made it to the World Series at all, since it was their first such appearance in franchise history (even including their time as the Montreal Expos). 

If it was a bit of a surprise that the Nats made it to the Series, it was a huge one that they beat the Astros. If you had put money on the Nats winning the Series at the beginning of the postseason, you would have seen a profit 15 times what you bet. On the other hand, money on the Astros at the beginning of the postseason would have returned a profit just over two times what you bet. Against all odds, the underdog Nats squeaked past the ‘Stros, winning four games to their three, for their first title in franchise history. 

While the Nationals lost their best position player, Anthony Rendon, a free agent after the Series, the Astros were the ones faced with a truly devastating offseason. They lost their best pitcher, Gerrit Cole, to the Yankees, a team they only narrowly defeated on their way to the World Series last year. They also lost Wade Miley and Collin McHugh, a couple of solid back-end arms who tossed a combined 242 innings of 2.5 fWAR ball for them last year. Needless to say, the Astros began the offseason with some glaring holes to address. Generally, the person to address them would be their General Manager. However, Jeff Luhnow, the mastermind behind the Astros late 2010’s resurgence, was dismissed this January (along with A.J. Hinch, the manager on the field) after his involvement in a sign stealing scandal.  Basically, catchers communicate with their pitchers about which pitch to throw through finger signals. Astros’ employees would spend time in the middle of each game reviewing opponents’ pitcher-catcher signals from footage earlier in the game, and they would communicate their findings to the players. When the Astros had a runner on second base with a clear view of the signals in real time, he would indicate to the batter which pitch was about to be thrown. The ‘Stros were also fined a record $5 million dollars.

But they weren’t the only team to face fallout from the scandal. Alex Cora, who led the Boston Red Sox to a championship in his 2018 managerial debut, was also fired due to his role in the scandal as an Astros’ coach in 2017, when the sign stealing allegedly began (although the question remains as to whether he continued the illicit activities while with the Red Sox). 

I’m also looking forward to watching the Sox this season, despite being a lifelong Yankees fan. The Red Sox had won three straight division titles from 2016-18, winning the World Series in 2018, only to see a third-place finish in their division in 2019. They seem intent on entering a rebuilding phase, a rarity for the playoff stalwarts, having traded away their perennial all-star Mookie Betts and solid southpaw David Price to the Dodgers in the offseason.

I expect the Dodgers will also have me glued to my TV screen. They’ve won seven division titles in a row, but have failed to win a World Series since 1988. This failure may be due to the sign stealing scandal; the Dodgers made the World Series in both 2017 and 2018, but fell first to the Astros and then the Red Sox, the two teams implicated. With a retooled roster that now includes Betts and Price, and two weakened foes in the Red Sox and the Astros, there is nothing stopping the Boys in Blue…although, even Rendon-less, the Nats might have something to say about that; I wouldn’t discount them after they ousted the Dodgers in the Division Series en-route to their World Series win last season. But I also have another reason for not selling the Nationals short.

My theory is that a shorter MLB season, which we are bound to have this year, will lead to more surprising outcomes. The fewer opportunities a team has to prove itself, the less reflective their win-loss record will be of their true talent level, and so, the harder it will be to predict their record. Since win-loss records will not be as reflective of teams’ true talent levels, it stands to reason that a lesser team has a better chance of beating a superior one in a shortened season.

Just how surprising might these outcomes be? How much harder will it be to predict win-loss records for a shortened season? We can look to one modern precedent: from 1994-95, all players engaged in a strike against the owners in an effort to obtain higher pay, and the 1994 season was cut short by 48 games. How much worse was the 1993 season at predicting the 1994 season than, say, the 2018 season was at predicting the 2019 season (both full-length)?

I decided to use previous seasons’ win totals as stand-ins for teams’ true talent levels to predict their performance for the next season. After all, barring some major offseason changes, a huge 162-game sample of head-to-head outcomes serves as a pretty good indicator of future talent. Even for a team that undergoes major offseason changes, such as the Astros’ and their starting rotation, the changes would have to be drastic to cause more than a rumble in the projections. The ‘Stros are still predicted to be the best team in the American League next year, despite losing their ace and managers. 

That being said, here’s the math behind the theory: For each additional win in 2018, teams won on average 0.7964 (R2 = 0.543, p < 0.005) more games in 2019. However, for each additional win in 1993, teams only won on average 0.2851 (R2 = 0.208, p < 0.05) more games in 1994. While 1994 is our only modern precedent for a 114-game season, we can double check other 162 game seasons that happened back-to-back to make sure 2018-19 wasn’t an anomaly. I used the previous season’s win totals (2013-18) to predict win totals for each season from 2014-19, since 2013 was the first year that the MLB shifted to their current divisional alignment. On average, each additional win in the previous season led to 0.5525 (R2 = 0.271, p < 0.005) more wins in the next, still nearly twice as many as the 1993-94 rate.

There is one more thing that we need to take into account: One win in a 114-game season is worth a heckuva lot more than one in a 162-game season. I divided each “win-rate” by the number of games in their respective seasons to try to account for this discrepancy. 0.5525/162 came out to 0.0034, meaning each win in a previous season added 0.0034 wins per game the next season, while 0.2851/114 came out to 0.0025 wins per game. So, the first rate still comes out to be significantly higher. If previous season’s wins really are indicative of true talent level, my analysis suggests that true talent level has less of an impact on wins in a shorter season.

What does this mean for the opportunistic Dodgers? Their astronomical true talent is less likely to manifest itself in a shortened season. The net result from the proposed changes to the MLB rules in the face of the pandemic might be positive for them, since they have more depth to accommodate the sudden introduction of a designated hitter in the National League and a 30-man roster. This might nullify their losses in moving to a shortened season. But then again, the 1994 season also had teams deal with some changes, so the comparison should still work. 1994 was the first time that there were three divisions in each league and saw the introduction of the extra wild card playoff spot, meaning more teams would be in contention through the end of the season (despite the strike canceling the playoffs that year, teams still competed like the playoffs were going to happen). The Boys in Blue might have to wait until 2021—as long as they can re-sign Mookie Betts, and the Astros and Red Sox are still reeling from the fallout of the sign stealing scandal—for their best shot at capturing their first World Series’ crown in over 30 years.

Leave a Reply

Your email address will not be published.