Who’s Hot? Creating a Metric to Identify the ‘Hottest’ NBA Playoff Teams
Kiran Krishnamurthi
April 2022
The NBA Playoffs have arrived marking my favorite time of the year. Especially with the first round, NBA fans like myself are treated to a luxury of multiple quality and competitive games every day. After two years of some playoff abnormality influenced by the pandemic, this spring I’m looking forward to the unbeatable energy in arenas as the final 16 NBA teams battle down to the title.
The year’s NBA regular season was a roller coaster for some teams and a steady progression for others. Trades, injuries, and COVID shook NBA rosters this season, keeping fans on the edge of their seat as teams bounced up and down their respective conference standings. While ranking volatility makes for entertaining basketball as games seem more competitive and higher stakes, it matters for a team in their ability to gain confidence and progress towards becoming a true title contender. At its simplest, a team needs to win games to gain a spot in the playoffs and then win on out from there to grab the title.
A reductionist approach looks at the NBA season as games that end in wins and losses, and the team that can win the most or lose the least will be the champion at the end of it all. But all fans, both die-hards and casuals, know that a team can get “hot” or “cold” while they’re winning or losing games in a row. Win and loss streaks for that matter could be just as important to a team in building confidence towards winning a title as simply the aggregate totals of wins and losses accumulated at the end of the season. In other words, a team builds and loses ‘momentum’ throughout the regular season with win and loss streaks, and this is meaningful in understanding which teams could be most confident come the playoffs.
For this project, I attempted to create a metric I call “momentum” that seeks to capture how well a team has performed over the course of the season before they head into the post-season. With the instability and ‘streakiness’ in the intra-conference rankings, I sought to create a quantifiable metric which captures how “hot” a team is playing relative to all their prior performances.
Momentum Mathematics
Momentum Metric Equations
\(\begin{aligned} \\ & Single \ Game \ Momentum = G_{i} \times S_{i} \times D_{i} \\ & Total \ Momentum = \sum_{i=1}^{n} G_{i} \times S_{i} \times D_{i} \\ \\ \end{aligned}\)
Momentum Metric Parameters
\(\begin{aligned} \\ & 1 \ \leq \ i \ \leq 82 \\ & G_{i} = Regular \ season \ game \ number \ (out \ of \ 82) \ i \\ & S_{i} = The \ consecutive \ win \ or \ loss \ at \ a \ given \ game \ i \\ & D_{i} = The \ point \ differential \ at \ a \ given \ game \ i \\ \end{aligned}\)
Momentum Exemplified and Explained
For example, consider the game on February 10, 2022, where the Dallas Mavericks beat the Los Angeles Clippers 112 to 105. For Dallas, this was their 4th win in a row on their 56th game of the season where they won by 7 points. Thus, we can then define our variables of \(G_{i}\) as 56, \(S_{i}\) as 4, and \(D_{i}\) as 7. Therefore, the momentum to be accounted for from this specific game would be 1,568 which is simply \(56 \times 4 \times 7\). This would then get added to the sum of all the momentum scores from their previous 55 games to get the total momentum at game 56 for the Dallas Mavericks.
For \(G_{i}\), I choose to include the game number (out of 82) in the equation to account for the fact that performances towards the end of the season matter more. Often, these games are higher stakes as teams compete for playoff seeding in closely contested races. Additionally, these games are simply closer to the playoffs and thus the impact of a win or loss at this point in time is “stickier” than if it were from a game at the beginning of the season.
For \(S_{i}\), the decision to include the consecutive win or loss is to account for the compounding effect of a win or loss streak in the season. While two teams may win the same number of games at the end of the season, a team that could have achieved a handful of those wins in a lengthy streak would be considered hotter than a team that won their games in small bouts of a handful of wins. The lengthy streak feeds into the narrative that each consecutive win in a row is a confidence boost to the team and therefore supercharges their momentum as they approach the playoffs. The same can be said about consecutive losses, except instead of boosting confidence it has a demoralizing impact.
For \(D_{i}\), I consider the point margin or differential from each game in the equation to not only control for the direction of wins and losses (i.e., through positive and negative differences), but to also account for the impact of each win and loss. A blowout win is a bigger confidence boost for a team and therefore should improve their momentum more than a team who beats all their opponents by a few points. Similarly, blowout losses are even more demoralizing than a loss in a close game (from the standpoint of a team’s ability to compete) and therefore should have a greater negative impact on momentum.
Momentum for a Team Across a Season
Here is what momentum looks like for one team where their wins and losses are color coded so as to visualize the impact of the game outcome across the season on their total momentum. The below graph shows the Miami Heat’s momentum across the 82 game regular season, which shows the team to be very consistent in their wins as they gradually rose their way to the top of the Eastern conference and held their spot once they reached.
Momentum for the 8 Playoff Teams from Each Conference
Momentum Across the Final 20 Games
While some teams this season jumped out to extremely ‘hot’ starts, the end of the season has more of a lasting impression on a team’s momentum as they head into the playoffs. Thus, in the following images momentum is examined across the final 20 games of the regular season for the playoff teams in each conference to get a sense of how strong certain teams finished the season.
One important consideration in the visualizations is that I chose to omit the final 2 games of the regular season. Granted, some teams took these games quite seriously as they were fighting down to the wire for their final playoff seeding. However, in order to ensure consistency for all teams, particularly for those teams that already had their playoff seeding locked up, I omitted the games. For teams like Phoenix or Milwaukee, the decision to rest their star players and let their reserves play the final two games suggests that these two teams would’ve already ‘mentally’ moved on to the post-season at this point and therefore the results in the final two games shouldn’t really stick to their momentum as they head into the playoffs.
