Monday, October 4, 2021

Does the Panthers' Preseason Success Matter?

The vibes have rarely been better for the Florida Panthers. After ripping off a 79 points in 56 a game campaign to set a franchise record for points percentage, the Panthers have opened their preseason with 4 wins in 4 games, outscoring their opponents 18-11 in the process. Two schools of thought can emerge from any pre-season results. One camp might point to success or failure in these September and early October contests and suggest the results are a sign of things to come. Others might simply write off these games as meaningless warm-ups for the real thing.

I was curious (and bored) enough to see if there was any noticeable correlation between what teams showed in the pre-season, and how they fared in the regular season. So, I turned to NaturalStatTrick and extracted data from the pre-season and regular season of every season that actually had a preseason (excluding 2012 and 2021 in the process). The stats I'm using are all 5v5 only and have been score and venue adjusted as per NST's formula.

Preseason Goals Won't Predict Regular Season Goals

Score one for those who write off the pre-season: there's no noticable correlation between goal differential in the pre-season and in the regular season. In some teams' cases, such as the Panthers themselves, pre-season success on the scoreboard has actually tended to come in years where the team floundered in the regular season. Time to panic, Cats fans?

Florida's regular season success is negatively correlated with pre-season success.

Preseason Shot Attempts Mean A Little

Notice the 2013-14 Sabres?

Those in the know are well away that shot attempts tend to be a better predictor of future goal differentials than past goal differentials are. Perhaps that makes it less surprising to see that a bit of a correlation shows itself when we change our focus from just goals to shot attempts. The correlation is not very strong, with an r-squared value of just 0.15. That puts it barely above the commonly used 0.10 threshold of statistical significance. Shot attempts do a bit better job of predicting regular season goals than preseason goals, but still correlate together weakly, at an r-squared value of 0.05.

Expected Goals are a Little Worse

Expected goals values in the preseason trend similarly to shot attempts in predicting both expected goals and actual goals in the regular season. The correlation is just a bit weaker, at r-squared of 0.12 for preseason xG to regular season xG, and 0.04 for preseason xG to regular season goals

Pre-season doesn't mean a lot. So what?

Even if an undefeated pre-season doesn't guarantee that Stanley C. Panther will lift the Stanley Cup come June, a good pre-season still gives fans reasons to get excited about seeing their team hit the ice again, and victory in the pre-season still tastes sweet. The goal of pre-season was never to show teams a litmus of their abilities; it serves as preparation for the intensity of the regular season. The tests players face in the heat of battle are difficult to prepare for without having a sparring partner. You can say "it's just pre-season", but these games still pit two groups of immensely talented, highly competitive professional athletes against one another, and it's very easy to get excited by that. With that in mind, the games don't have to tell us anything, as long as they remind us that regular season hockey is just around the corner.  

No comments:

Post a Comment