Late last week I posted the results of a correlation analysis between UK's margin of victory and an array of team stats, both offensive and defensive. The results showed a very high correlation between margin and assists on offense, as well as margin and opponents' field goal percentage on defense. I was challenged by several of you to dive a little deeper into the true cause and effect of assists, which was a very fair challenge: while it is obvious how a lower field goal percentage by UK's opponents would translate into higher margin, it is less obvious how assists impact margin since they don't necessarily equate to more points.
I decided to take a more in-depth look at some of games at both the low end and high end of UK team assists: North Carolina, Indiana, and @ Tennessee represent low assists, while @ South Carolina, Florida, and Ole Miss represent high assists (for the statisticians in the group, I took a random sample of three games from both the first and fourth quartiles of the team assist totals); the average margins for the two groups are one and 23 points, respectively. I looked at the play-by-play and also compared a variety of team stats between the two groups. I was able to find a couple of things that I think stand out; these findings are summarized after the jump....
The first significant effect relates to baskets made and the number of those that were three-pointers: as we see from the graph below, for games in this sample with a high number of assists, UK made 3.4 more field goals on average, which included 5.7 more three-pointers - an increase in the latter by more than a factor of 3!
The average overall field goal percentage for the high and low assist groups was 50.9% and 47.6%, respectively - a modest increase - while the contrast in average three point percentage was much more dramatic: 52.8% compared to 30.7%, respectively. We should note that for both groups, almost all the three pointers were assisted, but the percentage of two pointers assisted almost doubled from the low to high group -- increasing from 25.4% to 46.9%. While the increase in assists is in part a result of more threes being set up and made, it also results from more two pointers scored off passes from teammates.
Finally, and perhaps most impressive, the increase in the number of made shots for the high assist group occurred even though the average number of possessions was almost five fewer. Overall, the increase in made shots - both two- and three-pointers - accounted for an average increase of 12.5 points for the high assist group.
I also looked at defensive statistics, and a representative result is shown in the chart below.
As we can see, for the group that had the higher number of assists, UK's opponents averaged only 0.93 points per possession, while in the games with a low number of assists the opposition averaged 1.04 points per possession; on average UK's opponents scored 11.6 fewer points for the games in the high assist group. The better defensive statistic for the high assist group would seem to be indicative of one of Cal's favorite sayings: good defense generates offense.
While this is far from an exhaustive analysis, I hope it sheds some light on the correlation between UK assists and margin of victory. I do want to add a word of thanks to everyone who commented on my last post; the reason I decided to post on this site was to get engagement from people interested in UK basketball stats to help improve the quality of my own personal work - which in this case was definitely the result. Hopefully you folks enjoy reading as well.
And as always, I'm very interested in your thoughts.