Expected Goals
Posted: Thu Oct 05, 2023 6:16 am
Limestone's 2023 defensive unit proved to be a formidable force, ranking as the 14th best in the nation. The team boasted impressive percentile rankings across several key statistics, particularly when adjusting for the strength of the opposing offenses. For instance, the team ranked in the 98th percentile for opponent-adjusted shot-on-goal rate, opponent-adjusted turnover rate, and opponent-adjusted shots allowed per possession. This clearly indicates Limestone's ability to counter even the strongest offenses, a characteristic that places them in the upper echelons of the national landscape.
Examining the team's performance throughout the season, there were notable fluctuations that pinpoint the keys to their defensive strength. The team's best stretch from May 1 to May 21, saw them exhibit a stellar adjusted defensive efficiency of 16%. For context, this essentially saying that for 4 games in May, they had the best defense in the country. During this period, they managed to keep the adjusted shooting percentage of opponents to 24%, a feat that underscores their defensive prowess. Conversely, the worst stretch from Feb 25 to Mar 15, saw these figures increase significantly, with shooting percentages reaching 33%, and adjusted defensive efficiency falling to 32%, placing them in the 23rd percentile. And these stats are taken only from stretches of games when the outcome was in doubt; so there are no garbage-time goals skewing the numbers.
Another key aspect of Limestone's defense is their performance relative to the length of the possessions they faced. The team's defense was at its best during possessions lasting 40 to 59 seconds, allowing goals on only 24.1% of these possessions. This is a significant 5.3 percentage points better than the average. However, when possessions lasted longer than 60 seconds, their efficiency dipped quite a bit, allowing goals on 23.6% of possessions, just 0.4 percentage points better than average. (On average, possessions that last longer than 60 seconds are less likely to result in goals than shorter possessions. Survivorship bias.) This suggests that the longer the opposing offense probes the Limestone defense, the more likely the defense is to be breached. This was a great unit, but if you are a particularly patient offense who can work the defense for a long period of time, they became essentially average.
Finally, a critical statistical threshold for Limestone is the shots-per-possession ratio of their opponents. To reduce the impact of mismatches, we are looking at games against conference foes and teams with similar LaxElo rankings. When opposing offenses average more than 1.02 shots per possession, Limestone's win rate drops considerably. This is evidenced by their 0 - 4 record in such games, where they allowed opponents to score on a whopping 37% of their possessions. On the flip side, when opponents took fewer than 1.02 shots per possession, Limestone's record was an unbeaten 8 - 0, with an opponent efficiency of just 21%. This stark contrast emphasizes the importance of limiting the number of shots the opposition can take per possession whether the extras come from backed-up misses or rebounds. And it dovetails nicely with the longer-possessions nugget highlighted above. Teams were much more effective against Limestone when the possession had a chance to develop, even if that meant an early miss or two.
This was originally published in the Fri Sep 29, 2023 edition of Expected Goals, my daily newsletter. If you want 5-10 minutes of team and player analysis in your inbox every morning, sign up here.
Examining the team's performance throughout the season, there were notable fluctuations that pinpoint the keys to their defensive strength. The team's best stretch from May 1 to May 21, saw them exhibit a stellar adjusted defensive efficiency of 16%. For context, this essentially saying that for 4 games in May, they had the best defense in the country. During this period, they managed to keep the adjusted shooting percentage of opponents to 24%, a feat that underscores their defensive prowess. Conversely, the worst stretch from Feb 25 to Mar 15, saw these figures increase significantly, with shooting percentages reaching 33%, and adjusted defensive efficiency falling to 32%, placing them in the 23rd percentile. And these stats are taken only from stretches of games when the outcome was in doubt; so there are no garbage-time goals skewing the numbers.
Another key aspect of Limestone's defense is their performance relative to the length of the possessions they faced. The team's defense was at its best during possessions lasting 40 to 59 seconds, allowing goals on only 24.1% of these possessions. This is a significant 5.3 percentage points better than the average. However, when possessions lasted longer than 60 seconds, their efficiency dipped quite a bit, allowing goals on 23.6% of possessions, just 0.4 percentage points better than average. (On average, possessions that last longer than 60 seconds are less likely to result in goals than shorter possessions. Survivorship bias.) This suggests that the longer the opposing offense probes the Limestone defense, the more likely the defense is to be breached. This was a great unit, but if you are a particularly patient offense who can work the defense for a long period of time, they became essentially average.
Finally, a critical statistical threshold for Limestone is the shots-per-possession ratio of their opponents. To reduce the impact of mismatches, we are looking at games against conference foes and teams with similar LaxElo rankings. When opposing offenses average more than 1.02 shots per possession, Limestone's win rate drops considerably. This is evidenced by their 0 - 4 record in such games, where they allowed opponents to score on a whopping 37% of their possessions. On the flip side, when opponents took fewer than 1.02 shots per possession, Limestone's record was an unbeaten 8 - 0, with an opponent efficiency of just 21%. This stark contrast emphasizes the importance of limiting the number of shots the opposition can take per possession whether the extras come from backed-up misses or rebounds. And it dovetails nicely with the longer-possessions nugget highlighted above. Teams were much more effective against Limestone when the possession had a chance to develop, even if that meant an early miss or two.
This was originally published in the Fri Sep 29, 2023 edition of Expected Goals, my daily newsletter. If you want 5-10 minutes of team and player analysis in your inbox every morning, sign up here.