Ok, seems simple enough. Let’s look at how the Sixers do at the end of quarters in comparison to the rest of the game. Because quarters are easy, and 12 minutes is nicely divisible by four, we’ll go with minutes 9-12 of each quarter as the “end of quarter” group.
What’s Good vs. Bad?
I imagine that there are many valid ways to do this. However, what I first thought of was using win probability data. What we want is to look at the average rate of change of win probability within each three-minute segment of each quarter. That way, we’ll have 12 non-end-of-game segments and four end-of-game segments to compare. I’m a fan of using win probability because it takes into account the time left in the game. Slowly bleeding points up 15 with two minutes to go in the game shouldn’t be considered the same as slowly bleeding points up 15 with seven minutes to go in the second quarter. The data used for this includes all regular season and playoff games this season.
Here are the slopes for the section lines, aggregated across all quarters. These are adjusted for the fact that the axis is counting down and not up. The way to read these is “for every minute of game time elapsed within a section, the Sixers’ percent chance of winning changed by ____”.
Now, before the pitchforks come out, keep in mind that this is not garbage time adjusted. The last few minutes of games that are not close are not particularly useful to determine a team’s quality, either good or bad.
Fire the Coaches or Lifetime Deals?
Given that these are obviously the only two options, I lean towards the latter. Missed amongst the slopes is that you should pay attention to the y-axis, or the percent win probability. Those are all good! Winning games is the point, regardless if the team gives up a little ground at the end of the second quarter or not. Plus, I don’t think that the end of quarters actually means anything. I’ll get around this offseason to updating that linked piece by the way.