For more about what's going on in this post check out the beginning from last week.
Did it ever feel like Michael Thompson was the only player on the floor for Northwestern in losses this season? Did you miss seeing Drew Crawford step up in big spots? There might be a reason for these feelings.
Out of all the teams I've looked at, I think the correlation numbers for Northwestern, because the team was so heavily focused on offensive efficiency all season, are the most revealing of all.
Let's start with the numbers from the entire season. Here are the correlations for a few key players:
- Drew Crawford: 0.578
- John Shurna: 0.499
- Luka Mirkovic: 0.251
- Alex Marcotullio: 0.214
- Michael Thompson: 0.084
- Davide Curletti: -0.088
These numbers seem to fit with the overall flow of Northwestern's season. When Crawford was performing at a high level the Wildcats were hard to beat. Some good examples? Games against Georgia Tech (181.4 offensive rating), at Iowa (158.3), Minnesota (121.7) and at Boston College (144.4). There were just two real counter examples: at Minnesota (166.7, lost by 11) and at Michigan (134.5, lost by nine). When Crawford played poorly so did the Wildcats. His poorest games of the season are ones most Northwestern fans would rather forget (including a 22.2 offensive rating in the loss at Illinois).
Other the other side of the equation were Thompson and Curletti. Thompson's consistency is hurting him a bit here with the correlations. He didn't have a single terrible game all season, with his worst being a 69.5 offensive rating in the bad loss at Penn State. Thus it's hard to see his impact in these numbers. That said, the Wildcats will surely miss his steadying force at the point guard spot next season.
Curletti's numbers are wonky because of his limited possessions. He used four of fewer possessions in 26 of Northwestern's 34 games this season. I merely included him as a foil to Mirkovic's results.
But what if we look at just Big Ten play? In all of the other cases conference play has been very revealing, so maybe someone was more important than Crawford once the Big Ten season rolled around.
- Alex Marcotullio: 0.613
- Drew Crawford: 0.516
- John Shurna: 0.381
- Michael Thompson: 0.324
- Luka Mirkovic: 0.035
- Davide Curletti: -0.201
Well hello there emerging Marcotullio. At the end of the season the sophomore guard's usage rate shot up. This allowed him to have more of an impact on the offensive end. He played only five minutes in the loss to Penn State, but it's still factored in here. But basically when his threes were going down Northwestern was a better team. This seems quite obvious, but it's pretty incredible how strongly it's reflected in the data. As third and fourth options Marcotullio and Crawford were the keys for Northwestern down the stretch.
I hope you've enjoyed the series and found that you've gotten something out of these statistics. I think like many advanced basketball statistics, they certainly need some context around them in order to be understood properly, but they come with meaningful conclusions. At least I sure hope they do, because otherwise I've wasted 1,156 lines in an Excel spreadsheet.
Next up later this week? The Chicago Five Tournament begins.