The Playoffs and Streaks, Part 1

There is a saying in Major League Baseball that, in the playoffs, anything can happen. A full season's worth of 162 games (or 163) is cast out of the window, and eight (ten if the wild card losers count) teams equally vie for the World Series trophy. Nothing prior to the playoffs matters. The team that plays the best, fights the hardest, and is the luckiest wins the World Series.

Fans of every team in the playoffs hope that their team is the one that will catch on fire and win the trophy. Plenty of analysts have suggested that a team riding into the playoffs playing good baseball and on a winning streak is the favorite. But is that true? Is a team's success during the games before the playoffs related to success in the playoffs? Will a team that is winning at the end of the season continue to win once the playoffs start?

In order to test this interesting theory, I examined the late-season records of all playoffs teams from 2007-2012 using three arbitrary ranges. I then looked for any possible correlation between late season success and victories in the playoffs. The three sets of games were the last ten games of the season, the last fifteen games of the season, and the last month of games during the season.*

Below are the results of my research. In order to simplify the data, numeral "1" stands for winning the World Series, numeral "2" stands for winning the Championships Series (and losing in the World Series), numeral "3" stands for winning the Divisional Series (and losing in the Championship Series), numeral "4" stands for losing the Divisional Series, and numeral "5" stands for losing the Wild Card game (only applicable for the 2012 season).

Last 10 Games

A few notes of interest....

• The average (arithmetic mean) result number for teams that were over .500 is 2.97. For the other teams, it is 3.58. This certainly is solid evidence that while being over .500 is no guarantee of success, it is a better sign for a playoff bound team.
• The two teams that won nine of their last ten games both made the World Series. One of two teams to win the eight of their games made the World Series. It isn't definitive, but it seems that finishing the season very strongly during the final ten games is a solid indicator of going deep into the playoffs. Also, five of the six World Series winners won 70% of the games, while the sixth won 50%.
• Only one team with a losing record made it past the Championship Series. While winning a majority of the last ten games does not mean the team will definitely do well in the playoffs, having a losing streak at the end of the season looks to signal doom.

Last 15 Games

• Interestingly enough, when the range is expanded by five, the correlation practically disappears. Teams with a record over .500 averaged a 3.11 result while teams below .500 averaged a 3.4 result.

Last Month of Games

• Since most teams won at least half of the games in the final month of the season, it is useless to continue the previous comparison between above and below .500 teams. Instead, I divided the teams into the top 25 and the bottom 25 based on their record during the month. The top 25 had an average result of 3 while the bottom half had an average of 3.4. As was the case with the 15 game sample size, the difference is relatively insignificant. All the World Series winners won at least 65% of their games in the final month, though four teams that made the World Series failed to win at least 55% of their games.

Conclusions

While the sample size is probably too small to conclude anything definitely, it does appear that there is some correlation between how a team performs in their last ten games and how well the do in the postseason. I'll look further into the correlation by broadening the sample size (I'm planning on using data from 20 seasons) and see if the correlation continues to exist. In the second part, I will also examine whether streakiness during the regular season has any impact of a team's title hopes.

With the addition of the second wild card, the principle that winning during the final ten games shows an improved change of going deep into the playoffs is far less relevant for the Wild Card teams. Since they play in a do-or-die game, their odds of winning the World Series are cut in half. So while this finding is interesting for teams that win their division, it isn't a useful predictor of success for the Wild Cards.

*October was included with September in the few years that regular season games were played during October.