Future Analytics

[RetroVol]

We had some discussion recently about CKC's adoption of analytics this year and the effect it has had on some players. We also, of course, look at the oldest of analytics -- stats -- and make predictions, comment, etc. I was just listening to this podcast and starting at around 29:20 one of the speaker talked about how the Charlotte hornets are using AI analysis of game video to study how potential prospects "move in space." Do they generate space for themselves and other players. Is the way a player moves a fit for the way Charlotte wants to play. Wow. Just wild. They also mentioned how AI Fantasy League agents are dominating because "if a player is running slower down the court than normal, average fans won't notice, but the AI will and will incorporate that into its actions immediately." I claim no particular knowledge about any of this, but I do spend some time playing around with AI and thought it was fascinating.

What's next?

 

[BruisedOrange]

We had some discussion recently about CKC's adoption of analytics this year and the effect it has had on some players. We also, of course, look at the oldest of analytics -- stats -- and make predictions, comment, etc. I was just listening to this podcast and starting at around 29:20 one of the speaker talked about how the Charlotte hornets are using AI analysis of game video to study how potential prospects "move in space." Do they generate space for themselves and other players. Is the way a player moves a fit for the way Charlotte wants to play. Wow. Just wild. They also mentioned how AI Fantasy League agents are dominating because "if a player is running slower down the court than normal, average fans won't notice, but the AI will and will incorporate that into its actions immediately." I claim no particular knowledge about any of this, but I do spend some time playing around with AI and thought it was fascinating.

What's next?

Thanks for posting! Very interesting.

My first thought is:
If we could get this NIL spending under control (with caps, tiers, and incentives), then schools and conferences would have more money to spend on dedicated sports AI that could help every player and every team improve.

 

[lvocd]

My first thought:
Can A.I. make a darn layup?

 

[Remy]

Back when I was in the work force before retirement the next big thing we were told was being a data scientist and analytics. Having read something on the subject way back then ( I suspect AI will make data scientists go the way of the dinosaur) one of the things about good analytics rather than just scrambling the same data into the same packages is one is supposed to look at data relationships not previously done. Now this is where you have to think like quantum stuff, outside the box, or magic. Examples they gave then were if you knew your best sales day of the week did you ever see the relationship of the color of cars that show up on your best day? Your worst day? extend that to brands of cars as well. Just take oddball data and do comparative analysis was the thing. What was found was many times there were no correlations at all. But occasionally you could find gold nuggets you never knew were there, My next door neighbor is an Army officer at the Pentagon now but for two years he was getting his PHD from Georgia Tech in data science. And that exercise was all about using various tools to do better choices on analyzing data and shortening intervals. Again if you know how to properly set up a scrape for AI to use to mine data the better info AI will provide to you. Yes, I set up rudimentary scrapes using grok these days when I want AI to work for me rather than the reverse.

 

[chuckiepoo]

Back when I was in the work force before retirement the next big thing we were told was being a data scientist and analytics. Having read something on the subject way back then ( I suspect AI will make data scientists go the way of the dinosaur) one of the things about good analytics rather than just scrambling the same data into the same packages is one is supposed to look at data relationships not previously done. Now this is where you have to think like quantum stuff, outside the box, or magic. Examples they gave then were if you knew your best sales day of the week did you ever see the relationship of the color of cars that show up on your best day? Your worst day? extend that to brands of cars as well. Just take oddball data and do comparative analysis was the thing. What was found was many times there were no correlations at all. But occasionally you could find gold nuggets you never knew were there, My next door neighbor is an Army officer at the Pentagon now but for two years he was getting his PHD from Georgia Tech in data science. And that exercise was all about using various tools to do better choices on analyzing data and shortening intervals. Again if you know how to properly set up a scrape for AI to use to mine data the better info AI will provide to you. Yes, I set up rudimentary scrapes using grok these days when I want AI to work for me rather than the reverse.

New vocabulary: Scrape