Does the team which enters the playoffs with the hottest hand perform better than teams who limp into the postseason? It’s a time honored trope, and every year we discuss it: who would you rather be? The team on fire, or the team limping into the playoffs? Personally, I don’t know how much it matters. I recall the 2007 Colorado Rockies barreling into the playoffs: they won 14 of their last 15 games, then swept the Phillies and the Diamondbacks. They were promptly swept by the Red Sox in four games. Looking at the Indians: the hottest team in recent memory was the 2017 Indians, who lost a heartbreaking ALDS to the New York Yankees.
So my priors on this are pretty set: end of season records really don’t matter much. But to investigate this I wanted to look at the same data set since the playoffs expanded. I debated on what defined the "hot hand" record? I felt a full month was a bit too long, as even in a month-long period teams can experience fairly large swings. I settled on 15 games: five series, a little more than two weeks. Close enough to matter, but not so many games that you lose the closeness to the playoffs.
You can find the data set here, one thing which might be helpful is the order I looked at and added the teams to the data set. I started with the World Series winner, then the World Series loser. I then looked at who the World Series loser beat in the Championship Series, and then who the loser of the Championship Series beat in the Division Series. I then looked at the remaining team. I performed the same process for the World Series winner. This way you can look at whether the hot hand team won their playoff series easily.
Overall, the data indicated a slight bias against the "hot hand" succeeding in the playoffs. In fact the World Series winner entered the postseason the weakest in the sample more often than with the ‘hot hand’ which surprised me. Overall the average number of wins entering the postseason was 8.64 wins, which was slightly lower than the World Series loser (by a game); the average best record entering the postseason was about 12 wins.
The exceptions are a mismatch of the usual suspects and some unexpected teams. The team with the worst record entering the postseason to eventually win the World Series is the 2000 New York Yankees, who limped into the postseason after losing 13 of their final 15 games. Other teams with poor records to finish the regular season and yet win include the 1997 Florida Marlins (six wins), 2006 St. Louis Cardinals (five wins), and the 2014 San Francisco Giants (six wins).
Only four teams with the best record headed into the postseason eventually won the World Series: the 1998 New York Yankees, the 2008 Philadelphia Phillies, the 2017 Houston Astros* and the 2019 Washington Nationals. An additional seven teams won the most games headed into the postseason, and made the World Series: the 1999 Atlanta Braves, 2002 San Francisco Giants, 2007 Colorado Rockies, the 2011 Texas Rangers, and the 2018 Los Angeles Dodgers.
*I am contractually obligated to mention: the 2017 Houston Astros cheated.
A few interesting tidbits to note:
The average number of games won by a playoff team was just under nine games; I split the sample set for the old four team format and the new five team one, there was little difference between the two, nor was there a large difference between the average playoff team and the average World Series team.
Only one World Series featured two teams with losing records entering the postseason, the 1997 World Series
The best record to enter the playoffs in this sample was the 2007 Colorado Rockies, who not only won 14 of their last 15 games, but also won 10 games in a row before being swept by the Boston Red Sox
The 2013 Cleveland Indians carried the longest winning streak into the postseason in the sample at 10 games.
Only one team who won 13 of their last 15 games into the playoffs won a pennant in the sample: the 2011 Texas Rangers...they’re also the only team to lose two World Series in a row in the sample.
Honestly the data matched my priors neatly in this particular data set. 15 games is somewhat arbitrary, I chose it because I felt like the right series number (five three game series, a tad more than the playoffs themselves), and is about 10% of a teams’ entire season. I’d like to run this exercise again at some point using a different number of games instead. Overall, little correlation existed between how a team fared entering the postseason compared to how they finished.