The Relationship Between Season Series Records and Winning Playoff Series

When discussing Stanley Cup Playoff predictions, it is common to hear the mention of the “season series” as an argument for predicting the winner of a playoff series, where the season series is another word for head-to-head records within a season. For example, if team A beats team B three out of four times in a season, team A had the better head-to-head matchup. There are times when teams split the season series with equal wins, but one team won all games in regulation, whereas the opposing team needed extra time for at least one of their wins. Because of this, points percentage will be used in this analysis, as it will increase the sample of playoff matchups where one team had a different points percentage than the other team. The season range will be in the salary cap era from 2005-2006 to 2023-2024, with the exclusion of the 2019-2020 season due to an incomplete season and a significant break before the 2020 Stanley Cup Playoffs, causing confounding variables. Additionally, season series and playoff results will be analyzed for rounds one, two, and three. The Stanley Cup Finals will not be included, as teams of opposite conferences play each other a maximum of two times in one season since 2005-2006, making it too small of a sample size.

Overview

Since 2005-2006, and excluding 2019-2020, there have been a total of 252 playoff series; 144 first round series, 72 second round series, and 36 third round series. Of the 252 total series, 36 series are disregarded from this analysis due to having an equal season series record with their opponent, resulting in 216 series analyzed. Of these 216 series, 129 of the teams with the better head-to-head points percentage won their playoff series, for a 59.7% win rate. In terms of round by round, 66.1% of first round teams won when they had a better season series record, 49.2% of teams with a better head-to-head record won in the second round, and 54.8% of teams with the better head-to-head record won in the third round. Without any further context, these numbers are not overly impactful, but the significance of season series can be better discovered when comparing with other similar variables, the best one in this context being total points.

Of the 252 playoff series, there have been 11 instances of two teams facing each other with the same number of regular season points. Of the remaining 241 series, the team with more points, typically also indicating the team with home-ice advantage, won 134 times, for 55.6% of the time; 61.2% of first round series, 52.1% of second round series, and surprisingly, only 38.7% of third round series. The head-to-head points percentage outperformed points as an indicator in rounds one and three by a significant margin, whereas in round two the two variables have a similar win rate. However, this is still not enough to make a conclusion, as combining these two variables into conditional variables will better help to understand the importance.

Combining the situation of teams with more points than their opponent and a better head-to-head record results in a sample of 128 series out of the 252. The sheer drop in the sample is because this condition removes series where either the two teams have the same number of points, the two teams have the same head-to-head points percentage, or the team with fewer points has a better head-to-head record. Of these 128 series, the team with more points and the better head-to-head points percentage won 79 times for a 61.7% win rate. By round, the winning rate under this condition was 71.1% in the first round, 50.0% in the second round, and 43.8% in the third round. It is important to note that due to these stricter conditions, the third round sample size falls to only 16 total series, making it rather inconclusive when analyzing round by round in this scenario. It is still interesting to note that only the first round seems to show any benefit to having an advantage in both points and season series.

Comparison of series win rates among teams with more points, a better head-to-head record, and more points and a better head-to-head record combined

What is likely more impactful is to view the relationship between teams with fewer points in the standings, but a better head-to-head matchup, as this would gauge if a favourable head-to-head matchup can compensate for fewer points. Of the 252 series, 80 met the criteria where one team had fewer points, but had a better head-to-head record compared to their opponent. To reiterate, based off of points, teams with fewer points won 39.8% of the time in the first round, 47.9% in the second round, 61.3% of the time in the third round, and 44.4% overall. Going by head-to-head points percentage, the team with the worse record won 33.9% of the time in the first round, 51.8% of the time in the second round, 45.2% of the time ion the third round, and 40.3% of the time overall. However, teams with a lower point total but better head-to-head record improved drastically. In the first round, teams under this criteria won 25 out of 43 series for a 58.1% win rate, an enormous improvement over the benchmark 33.9% win rate, and an even lower 28.9% win rate among team with fewer points and a worse head-to-head record. In round two, the win rate is even at 50%, with these teams having won 12 out of 24 series. The sample for third round series is only 13, but these teams won nine of them for a 69.2% win rate. Combined, teams with fewer points but a better head-to-head record won 46 out of 80 times for a 57.5% win rate, a vast improvement over the 44.4% win rate of teams with fewer points.

Comparison of series win rates among teams with fewer points, a worse head-to-head record, and fewer points but a better head-to-head record combined

Season Series Based on Dominance

In addition to looking at a binary variable of simply who won or who lost the season series, it is also important to analyze the degree to which a team won or lost. When looking only at head-to-head matchups where the season series winner had a points percentage greater than or equal to 70.0%, there are a total of 147 out of 216 cases Of 90 in round one, the team with the higher points percentage won 58 times for a 64.4% win rate. Surprisingly, round two saw 19 out of 43 teams with a head-to-head points percentage greater than or equal to 70.0% win, for a win rate of just 44.2%. Finally, the third round sample is only 14 games, of which eight were won by the season series winner. Combined, 85 of the 147 playoff series were won by the team that “dominated” the season series, for a win rate of 57.8%. As a result, there is no significant difference based on isolating for dominant season series wins, and in fact, the performance was slightly worse when compared to looking at the total picture with no restrictions.

Conclusion

The results show that head-to-head records based on points percentage is a slightly better predictor variable than the benchmark of the team with more points. The importance of season series is amplified for teams that have fewer points than their opponent, but a better head-to-head points percentage. Teams that have fewer points won 44.4% of series, but was drastically improved to a 57.5% win rate if that team had a better head-to-head record. The first round saw the most significant correlation with season series, as in all cases, a team with a better season series record had a better win rate in the first round than the benchmark of the team with more points, which was not always the case with rounds two and three. This is most likely due to the fact that as teams go further into the playoffs, statistics from the regular season continue to have less significance, and statistics from previous rounds of the post-season become more significant and relevant.

The most significant finding was the fact that teams with fewer points and a worse head-to-head record won only 29.9% of the time, but improved to a 58.1% first round win rate if the team with fewer points won the season series.

It was determined that the degree to which a team won the season series had no significant impact on the series win rates when compared to the findings without filtering by teams with a head-to-head points percentage greater than or equal to 70.0%.

It is conclusive to say that season series does matter, and that there is something to be said about head-to-head matchups, proven by the vast difference in performance among teams with fewer points with a worse head-to-head record compared to teams with fewer points but a better head-to-head record. It can also be concluded that the season series is only important for the first round, and that it should not be taken into consideration for any further rounds. As with most variables, season series on its own is not a strong predictor variable, but may be more impactful when used in conjunction with other predictor variables.