Hubert Davis Catch-all

  • Thread starter Thread starter LeoBloom
  • Start date Start date
  • Replies: 2K
  • Views: 33K
  • UNC Sports 
Calheel said:
I know several professional sports bettors and, to the man, they all tell me that they don't have the time and/or it is a waste of time to actually watch the games.
Click to expand...
Voulgaris said the same thing. I seriously doubt bettors watch much basketball, or if they do, it's purely for fun. You can't make money in sports betting being a scout.
 
duluoz said:
That’s hyperbolic. You don’t need eyeballs on every hour of every game in order to have the eye test factor into oddsmaking. You should know that.
Click to expand...
Part of the eye test is looking at scores and stats.

I'd bet that there are people on this thread who could set Vegas like opening lines fairly close to what the metrics spit out. Of course much easier, efficient, and cost prohibitive to let AI do it.

Again, setting lines is not that much like seeding a tournament.
 
Calheel said:
Have not seen the line posted, but I would have to imagine the odds of him returning are very high at this point
Click to expand...

Estimated odds (based on media and insider talk)

Not official betting odds, but a realistic estimate:
  • ~80–90% chance Davis is the UNC coach next season
  • ~10–20% chance UNC makes a change
Most analysts say “barring a catastrophic finish,” he’s safe for now.

✅ Bottom line: Unless UNC totally crashes in March, Davis is very likely back in Chapel Hill next season.
 
Heelsandeers said:
Is earning a higher seed in the NCAAT than predictive metrics would indicate a sign of good or bad coaching?
Click to expand...
Predictive metrics are based on efficiency, which is the result of a number of inputs, including coaching.

So getting a higher seed than the predictive metrics suggest wouldn't indicate anything specific about the coaching acumen of the head coach. If you could show a consistent pattern regarding one or two specific stats/inputs across multiple teams who got higher seeds than the predictive metrics would indicate, you would likely have evidence of a particular input the seeding committee values above the predictive metrics. But that would likely be the extent of it.
 
SnoopRob said:
Predictive metrics are based on efficiency, which is the result of a number of inputs, including coaching.

So getting a higher seed than the predictive metrics suggest wouldn't indicate anything specific about the coaching acumen of the head coach. If you could show a consistent pattern regarding one or two specific stats/inputs across multiple teams who got higher seeds than the predictive metrics would indicate, you would likely have evidence of a particular input the seeding committee values above the predictive metrics. But that would likely be the extent of it.
Click to expand...

What determines the value of the coaching input?

Is it just high efficiency = good coaching

Medium efficiency = ok coaching? Etc
 
Heelsandeers said:
What determines the value of the coaching input?

Is it just high efficiency = good coaching

Medium efficiency = ok coaching? Etc
Click to expand...
I think you'd need a completely different metric to evaluate coaching.

I think it'd be more than a bit subjective no matter how you formulated it, but if I were tasked with building such a model it would begin with evaluating every player (on every team) and formulating an "talent-based expected wins" total based on player talent & projected playing time and then coaching would be the delta between those talent-based expectations and the real life outcomes of games across a season.

It'd still be a fairly noisy stat as a lot of things could impact it outside of coaching, but that'd be where I'd start. I'd also not want to stake my credibility on how well the model seemed to perform.

Coaching seems much more art than science and that extends to being a difficult input to measure in any systematic way.
 
You must log in or register to reply here.
Back
Top