How much does the regular season matter?
You might be surprised by the answer.
Are Regular Season Head-To-Head Results Related to Playoff Results?
The Study: Heading into these Playoffs, you might be trying to find ways to predict the outcome of any number of series. One group of statistics that analysts and sports writers often quote is the “season series” between two teams—that is, the records of the two teams when they played against each other. After all, if one team consistently beat another team during the Regular Season, you’d think that they’d be more likely to do the same in the Playoffs. But, is this intuition correct?
In order to examine this question, we gathered the Points earned by every team during season series and Playoff matchups from 2013-14 to 2018-19. In both cases, we assigned 2 Points for a Win and 1 Point for any Overtime Loss; that is, although no Points are assigned to teams that lose Playoff Games, we assigned Points in order to keep the data consistent. After that, we calculated the Points differential for a Playoff matchup’s winning team (i.e. winning team’s Points minus losing team’s Points) for both the Regular Season and Playoffs. Note that this was done for all Playoff rounds, in turn resulting in 90 total matchups.
Finally, we calculated the correlation between the two differentials in order to see if there is any relationship between Regular Season and Playoff matchups’ results.
The Results: The correlation between Regular Season and Playoff matchups’ results was 0.03. Therefore, there appears to be essentially no relationship between the result of a matchup’s season series and Playoff series.
Discussion: First, if the number of matchups examined seems pretty low, keep in mind that those 90 matchups consisted of all Games played between the two teams; that is, the total number of Games used to determine the results of each matchup was greater than 540 (i.e. at least two Regular Season Games plus at least four Playoff Games per matchup). Thus, the quantity of data should be sufficient in order to give us a good idea if a matchup’s season and Playoff series’ results are related.
Otherwise, the results can be presented in another way that may make more sense: The teams that won each of the 90 Playoff matchups won the corresponding season series 42 times, tied 13 times, and lost 35 times. Therefore, based on the data, it seems more likely than not that the winner of a season series will win the corresponding Playoff matchup.
However, with the results so close and the correlation so low, the season series result is essentially an irrelevant reference point in order to predict a Playoff matchup’s result. After all, games between two teams can be bunched together, be spread apart, and happen at any point during a given Regular Season; that is, with so many other variables involved (e.g. streaks, injuries, trades, etc.), it makes sense that the result of a season series wouldn’t have a strong relationship with what happens in the Playoffs.
So, if you’re going to be making some predictions these Playoffs and you hear people talking about a matchup’s season series, you’ll know to only take this information with a grain of salt.