Saturday, January 04, 2025

Correlation, Causation, and Inference in Big Time College Sports

The past day or two I have been puzzling about whether the results of the Rose Bowl, a college football game where Ohio State overwhelmed Oregon in the first half, and the Illinois basketball game against Oregon on Thursday evening, played on Oregon's home court, where the Illini set the record for the largest point spread victory by a visiting team in NCAA basketball, are somehow related.  In each case Oregon was the higher rated team going into the game.  Ohio State had played Oregon previously during the regular season.  The game was close and played on Oregon's home field.  Oregon won by a small margin.  Illinois had played other higher rated teams tough, notably Tennessee, even though that ended as a loss for the Illini.  So, ahead of time, one might reasonably have predicted that these games would be close.  That they each ended up as blowouts was quite a surprise.

As a fan, I care more about Illini basketball than I do about Illini football, though I did watch the Citrus Bowl.  And I don't generally watch college football on TV, but I did watch most of the first half of the Rose Bowl.  With basketball, not only did I watch the game against Oregon, but I also watched on YouTube the postgame interviews with the head coaches.  Dana Altman, the head coach of Oregon, bemoaned the lack of defensive effort his team showed against Illinois.  He said that Illinois was a good offensive team, but not that good.  They looked unstoppable because Oregon didn't play defense.  Defense is mainly an effort thing, and Altman was taking to task the effort of his own players.  But he never said why the effort level was so poor.  Nate Bittle, the Oregon Center, and supposedly quite a good player, appeared for 21 minutes, while he averages over 25 minutes a game.  He was out quite a bit during the second half.  Again, one wonders why. 

Might it be that the Oregon football team, particularly the defensive unit, had a low level of effort in the Rose Bowl?  If so, one again would want to know why that was.  And, maybe, the underlying explanation for the basketball team's poor defensive performance is similar to or even identical to the underlying explanation for the football team's poor defensive performance.  

Now a bit on probability that I'm taking from Daniel Kahneman.   In his book, Thinking, Fast and Slow, he warns the reader about making inferences from small samples.   Fans are prone to do that and, in particular, if the team has struggled some in the past then to take account only of the recent successes, as if the team "figured it out" and is now on the path to greater glory.  But it is possible for there to be an outlier great performance that doesn't become a repeat event.  I am particularly reminded of Brandon Paul's great performance against Ohio State where he shot the lights out.  Further, if the opponent is not up to snuff, for whatever reason, it would seem that the outlier becomes somewhat more likely, especially if that is unknown by by the other team, which might otherwise lessen its own effort to keep the situation from becoming too embarrassing. 

There are now several sports pundits who produce their own ratings of college sports teams.  I'm behind the times on this.  I used to follow the college basketball ratings that Jeff Sagarin produced, but he no longer seems to be doing this.  I only mention this here because I'm guessing that no matter which rating you follow, they don't use information from other college sports.  In other words, Oregon losing to Ohio State in the Rose Bowl won't impact at all how to account for Oregon losing to Illinois in men's basketball on January 2.  But might there be relevant information content that should be accounted for?

Likewise, it seems there is now a huge industry of sports betting.  In that, the odds set by the bookmakers depend not just on past team performance, but also on the latest poop about the team, such as if a player has been suspended, or injured, or has some emotional issue to cope with.  But do the teams disclose all that might matter here?  During Covid, rules were imposed that forced disclosure of players who had tested positive and when that was.  I believe it to be the norm, however, that many "minor" injuries are not disclosed.  So, what I'm really wondering here is whether at Oregon there was some illness or emotional trauma that impacted players on both the football and the basketball teams but it was kept hidden.  A big time gambler, who might be a donor to the athletic program at Oregon, could have made a lot of money if in possession of this information when the rest of the world wasn't.  I mention this only because it gives some possible explanation for why the information wasn't disclosed.

But what I'm really interested in is the perception of how good Ohio State football is as well as how good Illinois basketball is.  It would seem that each of them would be overrated if there were such private information about Oregon that wasn't disclosed.   Performance of these teams in the upcoming game(s) will speak to how good they are.  I'm wondering whether anyone will find out if there is important information about the Oregon teams that wasn't disclosed.  I'm unaware of mechanism that might reveal that now.

Of course, everything I've said here is speculation. It's possible that Oregon football under performed simply because they hadn't played in a while (they had a bye in the first round of the playoffs) and Oregon basketball just had a bad hair day.  I have no way of knowing which it is.  It's just that the juxtaposition of two highly unlikely events happening in tandem creates a situation that raises one's eyebrows.   And I'm surprised that nobody else has brought up the possibility, though maybe they have and I'm just ignorant of it.