I'm bored | Page 2 | The Boneyard
.-.

I'm bored

I like the answers about analyzing strategy and providing feedback about what works. But it sounds tricky. I suspect the current models are not trained on this information.... yet.

Analyzing 2-3 hours of video takes a decent amount of processing. That's problem one. but conceptually easy to solve. Just big fast computers with the best GPUs available. Latest NVIDIA stuff.

Then you need to train AI to understand basketball. Most of the AI models/platforms common people are using are pre-trained models trained on lots of stuff. Basketball is likely not one of them. I mean, AI probably knows all the language, down-screen, horns, A-set, etc.... And it could probably explain in words each concept. Or find video examples of basically all basketball concepts. Both defensive and offensive. That information is all over the internet and thus AI knows it. And knows how to find it.

But can you give AI a video of an offensive set (or defensive) and actually make sense out of it. Can AI identify what sets/plays/defense/etc... are being run. I'm sure all this could be done. But I suspect current models have not been trained with this information. One version of a down-screen or a horns set can look a bit different from another. Can it identify if defense is "switching everything" or not. These things are not easily identifiable. Again, all possible, but I suspect current models are not trained to on this.
 
Asked AI if UConn played well against Kansas. I think the Chief’s briefs may soon be replaced. Or is he really an AI imposter? Was surprised that it said Ross played some minutes at pg? Don’t recall seeing that. He was playing the 2 at times from what I recall.

AI Overview


+3



Yes, UConn played well enough to win against Kansas, securing a 61-56 road victory despite not always playing at their best. The Huskies excelled in key areas, particularly in the second half, where they dominated the rebounding battle, limited Kansas's shooting, and had strong performances from players like freshman Braylon Mullins.
You can watch this video to see highlights of the game and Mullins' performance:

UConn Huskies
https://www.youtube.com/watch?v=6eMP_t2dCbw&t=248
YouTube • Dec 3, 2025
  • Second-half dominance: UConn out-rebounded Kansas 25-10 in the second half and held the Jayhawks to just 25% shooting in the final 20 minutes.
  • Strong individual performances: Braylon Mullins was a breakout star, scoring 17 points, including clutch free throws to seal the win. Other players like Solo Ball also stepped up, with Jayden Ross effectively handling point guard duties when needed.
  • Key plays: The Huskies' defense was sound, especially in the final moments, and Mullins' put-back of a missed shot was a crucial play that helped secure the win.
  • Overall performance: While the offense wasn't at its peak, the team demonstrated resilience and the ability to win in a tough environment, a hallmark of championship-level teams.
 
DW1361 said:
Asked AI if UConn played well against Kansas. I think the Chief’s briefs may soon be replaced. Or is he really an AI imposter? Was surprised that it said Ross played some minutes at pg? Don’t recall seeing that. He was playing the 2 at times from what I recall.

AI Overview


+3



Yes, UConn played well enough to win against Kansas, securing a 61-56 road victory despite not always playing at their best. The Huskies excelled in key areas, particularly in the second half, where they dominated the rebounding battle, limited Kansas's shooting, and had strong performances from players like freshman Braylon Mullins.
You can watch this video to see highlights of the game and Mullins' performance:

UConn Huskies

YouTube • Dec 3, 2025
  • Second-half dominance: UConn out-rebounded Kansas 25-10 in the second half and held the Jayhawks to just 25% shooting in the final 20 minutes.
  • Strong individual performances: Braylon Mullins was a breakout star, scoring 17 points, including clutch free throws to seal the win. Other players like Solo Ball also stepped up, with Jayden Ross effectively handling point guard duties when needed.
  • Key plays: The Huskies' defense was sound, especially in the final moments, and Mullins' put-back of a missed shot was a crucial play that helped secure the win.
  • Overall performance: While the offense wasn't at its peak, the team demonstrated resilience and the ability to win in a tough environment, a hallmark of championship-level teams.
Expand ...

Maybe we need to start a new post game thread authored by AI with a fancy name like Chief and Richz have. Any suggestions?
 
.-.
MexicoHusky said:
I like the answers about analyzing strategy and providing feedback about what works. But it sounds tricky. I suspect the current models are not trained on this information.... yet.

Analyzing 2-3 hours of video takes a decent amount of processing. That's problem one. but conceptually easy to solve. Just big fast computers with the best GPUs available. Latest NVIDIA stuff.

Then you need to train AI to understand basketball. Most of the AI models/platforms common people are using are pre-trained models trained on lots of stuff. Basketball is likely not one of them. I mean, AI probably knows all the language, down-screen, horns, A-set, etc.... And it could probably explain in words each concept. Or find video examples of basically all basketball concepts. Both defensive and offensive. That information is all over the internet and thus AI knows it. And knows how to find it.

But can you give AI a video of an offensive set (or defensive) and actually make sense out of it. Can AI identify what sets/plays/defense/etc... are being run. I'm sure all this could be done. But I suspect current models have not been trained with this information. One version of a down-screen or a horns set can look a bit different from another. Can it identify if defense is "switching everything" or not. These things are not easily identifiable. Again, all possible, but I suspect current models are not trained to on this.
Expand ...

I remember Ross taking it up the court off the inbounds at least twice. So that leads me to infer AI must recognize that’s the hallmark of a point guard… maybe it has been trained on some bball. Or maybe someone else typed that on the internet already and AI just scooped it up directly.
 
storrsroars said:
AI will be used in college sports for performance optimization, recruiting, and strategic analysis through tools that track athlete metrics, analyze game footage, and create personalized training plans. Beyond the field, AI will help with academic and career-related tasks, data-driven fan engagement, and automated aspects of officiating.

Athlete performance and health
  • Performance tracking: AI systems with sensors and cameras can track an athlete's movement, physiology, and performance in real-time to provide immediate feedback.
  • Personalized training: AI can analyze vast datasets to create customized training programs that target specific strengths and weaknesses, as well as help monitor nutrition and identify injury risks.
  • Injury prevention: By analyzing training data, game data, and other health metrics, AI can help predict and prevent injuries before they occur.
Recruiting and strategy
  • Player evaluation: AI can evaluate a recruit's statistics, physical qualities, mental toughness, and consistency from video footage and data, helping coaches build a more comprehensive profile.
  • Recruiting models: AI can model ideal player profiles that fit a specific team's system, helping coaches narrow in on targets more effectively.
  • Game strategy: Predictive analytics allow coaches to analyze historical data, opponent behavior, and game conditions to optimize strategies and make in-game adjustments.
Operations and other support
  • Game analysis: AI-powered platforms can automatically tag and analyze game footage, breaking down plays, routes, and formations in seconds, freeing up coaches to focus on coaching.
  • Academic and career support: AI can assist with academic monitoring, degree progress, and even career readiness, providing a more holistic view of a student-athlete's success.
  • Automated officiating: AI could be used to assist with officiating to reduce human error and improve fairness in calls.
  • Fan engagement: AI can help sports media departments create new and exciting fan experiences, from personalized content to virtual reality scenarios.
😉
Expand ...
Yeah, but will high school athletes geniuses still all wanna go to Duke, UNC or Calipari/Pitno U (wherever they surface).
 
thekollieman said:
I remember Ross taking it up the court off the inbounds at least twice. So that leads me to infer AI must recognize that’s the hallmark of a point guard… maybe it has been trained on some bball. Or maybe someone else typed that on the internet already and AI just scooped it up directly.
Expand ...
yeah, that's interesting. Not sure how it knew. But online comments, online analysis and quotes from Hurley and or others seems totally possible.

Knowing what a screen is, what a switch is, identifying an offensive set all from an images/videos... and maybe audio commentary seems like something it would not know how to do. Maybe... I just doubt it.
 
you could also take all the data the assistant coaches are gathering during the games. I'm not sure what the format of that data is. Might just be hand written notes on paper. Regardless, it could be scanned if necessary. Given to AI. See if AI comes up with some other thoughts on what worked. What didn't work. But I suspect that analysis would mirror what the coaching staff already knows.
 
FfldCntyFan said:
Better yet, how do we know if any boneyarders are real?
Expand ...
There is no way that we are all real. I will never believe that. Or maybe there are some that are two or three different people so if someone has three different user names does that mean that two of those users are not real? The questions this thread can create are endless. You would need AI to figure out the math on those possibilities.
 
.-.
Hunt for 7 said:
There is no way that we are all real. I will never believe that. Or maybe there are some that are two or three different people so if someone has three different user names does that mean that two of those users are not real? The questions this thread can create are endless. You would need AI to figure out the math on those possibilities.
Expand ...
It kills me that, after all these years, people still think Fishy is a real person. LOL
 
nomar said:
Great comments. Here are my original thoughts.

AI is already starting to reshape college sports, and over the next few years its impact will grow a lot. Here are the biggest ways it’s likely to be used:​


1. Performance Analysis & Player Development​

  • Automated film breakdown: AI can tag plays, recognize formations, identify tendencies, and provide instant scouting insights that used to take hours of staff time.
  • Personalized training plans: Models can analyze biometric data (GPS trackers, wearables, force plates) to tailor workouts and skill development to each athlete.
  • Technique correction: Computer vision tools can break down shooting mechanics, pitching form, sprint gait, etc., offering real-time feedback.

2. Injury Prevention & Sports Medicine​

  • Workload monitoring: AI can spot patterns that predict fatigue or injury risk before humans notice them.
  • Return-to-play modeling: Data from rehab sessions can help trainers know when an athlete is ready to compete safely.

3. Recruiting & Roster Management​

  • Talent identification: AI can compare prospective recruits’ stats and film to historical data to find undervalued players.
  • Transfer portal navigation: With thousands of players moving each year, AI tools can help coaches identify the best fit.

4. Game Strategy & Scouting​

  • Opponent analysis: AI can recognize strategic tendencies (e.g., “they run this play 80% of the time on 3rd and short”).
  • Predictive modeling: Some teams may use simulation tools to test game plans or anticipate how opponents will react.

5. Fan Engagement & Business Operations​

  • Personalized content: AI-generated highlights, tailored social media clips, or real-time game stats for fans.
  • Ticketing and attendance optimization: Predicting demand, dynamic pricing, and improving game-day logistics.

6. Academic Support​

  • While still regulated heavily for fairness, AI tutors and study tools will probably become part of athlete academic support centers.

7. Compliance & Administration​

  • Monitoring NIL activities, flagging potential violations, and managing complex NCAA rulesets.
If you want, I can also:

  • give predictions for 5–10 years from now,
  • compare AI use in D1 vs D2 vs D3,
  • or describe how specific sports (football, basketball, baseball, etc.) may use AI differently.
Expand ...
Compliance and administration do we work for the same company. Every metric we have to meet that is related to compliance and administration is completely managed by AI.

Edit: Another thought came to mind: I find myself in many depositions and asked to testify in many civil trials, there are some go too moves that AI could never use so that will be interesting. I mean AI could never say in a deposition when asked a difficult question I really need to use the restroom.

I think AI could be the lawyer but not the person being asked to testify either in open court or in a deposition. But the day is coming very soon that many jobs will fall victim to AI. It will be a tragedy when it really starts to happen. We could see 30% or more of white collar jobs go away.

If I knew this day was coming I would have gotten into the plumbing HVAC business because I think robots are still a generation away from being able to go to a job site and complete manual labor like fixing a compressor or chiller.
 
Last edited:
80shusky said:
AI is garbage, shouldn't be used for anything, and is going to crash the economy.
Expand ...
It will crash civilization, itself.

It will happen at the intersection of auto-correct and AI.
 
AI could replace announcers with the option to choose their appearance, personality and voice to name a few. Raftery could announce every UConn game, if desired full of “onions” quotes.
 
Some of us once got together to do some pro bono landscaping work at Quarry Ridge. Although it wasn’t technically “approved” landscaping work, per se.

Although that could have been a simulation.
 
.-.
John said:
we won't have to worry about college sports once Skynet takes over.
Expand ...
Biggest issue with all this AI is that no one wants it. It is all defensive. If every company spends $100M for AI, where is the advantage? You just wasted money to stay with status quo. It’s annoying.

I also, firmly think, that guys like Elon saw T2 and were like, wouldn’t sky net be cool?
 
JohnFSilver said:
Biggest issue with all this AI is that no one wants it. It is all defensive. If every company spends $100M for AI, where is the advantage? You just wasted money to stay with status quo. It’s annoying.

I also, firmly think, that guys like Elon saw T2 and were like, wouldn’t sky net be cool?
Expand ...

I don't know. Maybe some people like the little things like medical research, diagnosis and treatment; safer, cleaner, cheaper and more efficient energy; being able to ask questions to AI so you don't have to ask your spouse ...
 
It will be used to strengthen or come up with more reasons to keep UConn out of the P4 😞

1. “Basketball school energy”​

UConn is too good at basketball. Some fans argue that the school treats football like a highly inconvenient hobby between March Madness runs.

2. Football inconsistency​

The football program has had some rough stretches, and critics say a school should be consistently competitive (or at least not chronically rebuilding) before joining the P4.

3. Lack of stable conference identity​

UConn has hopped conferences a few times, so rival fans like to claim they’re “nomads” who can’t sit still long enough to build long-term P4 rivalries.

4. Geography doesn’t help​

For leagues with heavy southern or midwestern footprints, adding a Northeast outlier is a travel headache waiting to happen.

5. Stadium atmosphere jokes​

Other fanbases love to poke fun at UConn’s football attendance, making it a meme that P4 commissioners want “atmospheres that don’t sound like a library at finals week.”

6. “They’ll steal every basketball trophy”​

P4 schools—especially in the Big 12 or ACC—might jokingly claim they don’t want UConn in their league because they’d instantly vacuum up conference titles in both men’s and women’s hoops.

7. Media markets vs. passion​

Yes, Connecticut is part of the NYC media footprint… but critics argue that the actual fans in the region don’t move TV needles the way P4 execs want.

8. UConn memes are too powerful​

“Blue-blood basketball powers shouldn’t also be allowed to ask for football sympathy points.”
(Message boards say this with passion.)

 
nomar said:
Great comments. Here are my original thoughts.

AI is already starting to reshape college sports, and over the next few years its impact will grow a lot. Here are the biggest ways it’s likely to be used:​


1. Performance Analysis & Player Development​

  • Automated film breakdown: AI can tag plays, recognize formations, identify tendencies, and provide instant scouting insights that used to take hours of staff time.
  • Personalized training plans: Models can analyze biometric data (GPS trackers, wearables, force plates) to tailor workouts and skill development to each athlete.
  • Technique correction: Computer vision tools can break down shooting mechanics, pitching form, sprint gait, etc., offering real-time feedback.

2. Injury Prevention & Sports Medicine​

  • Workload monitoring: AI can spot patterns that predict fatigue or injury risk before humans notice them.
  • Return-to-play modeling: Data from rehab sessions can help trainers know when an athlete is ready to compete safely.

3. Recruiting & Roster Management​

  • Talent identification: AI can compare prospective recruits’ stats and film to historical data to find undervalued players.
  • Transfer portal navigation: With thousands of players moving each year, AI tools can help coaches identify the best fit.

4. Game Strategy & Scouting​

  • Opponent analysis: AI can recognize strategic tendencies (e.g., “they run this play 80% of the time on 3rd and short”).
  • Predictive modeling: Some teams may use simulation tools to test game plans or anticipate how opponents will react.

5. Fan Engagement & Business Operations​

  • Personalized content: AI-generated highlights, tailored social media clips, or real-time game stats for fans.
  • Ticketing and attendance optimization: Predicting demand, dynamic pricing, and improving game-day logistics.

6. Academic Support​

  • While still regulated heavily for fairness, AI tutors and study tools will probably become part of athlete academic support centers.

7. Compliance & Administration​

  • Monitoring NIL activities, flagging potential violations, and managing complex NCAA rulesets.
If you want, I can also:

  • give predictions for 5–10 years from now,
  • compare AI use in D1 vs D2 vs D3,
  • or describe how specific sports (football, basketball, baseball, etc.) may use AI differently.
Expand ...
Damn!!

To think all the team needed was uniforms and an opponent back in my day. This post would have exposed my faults and thus make practice harder for me.

And to think we used to gauge our performances on crowd feedback and cheerleader responses.

AI may make coaches, admin and the team overall smarter. The core athlete is still the similar to my childhood days; most rely on talent not smarts.

Just saying AI could make a player more knowledgeable but we all know that when the game starts we rarely use our brain because the opponent won't let you think anyway.
 
.-.
storrsroars said:
AI will be used in college sports for performance optimization, recruiting, and strategic analysis through tools that track athlete metrics, analyze game footage, and create personalized training plans. Beyond the field, AI will help with academic and career-related tasks, data-driven fan engagement, and automated aspects of officiating.

Athlete performance and health
  • Performance tracking: AI systems with sensors and cameras can track an athlete's movement, physiology, and performance in real-time to provide immediate feedback.
  • Personalized training: AI can analyze vast datasets to create customized training programs that target specific strengths and weaknesses, as well as help monitor nutrition and identify injury risks.
  • Injury prevention: By analyzing training data, game data, and other health metrics, AI can help predict and prevent injuries before they occur.
Recruiting and strategy
  • Player evaluation: AI can evaluate a recruit's statistics, physical qualities, mental toughness, and consistency from video footage and data, helping coaches build a more comprehensive profile.
  • Recruiting models: AI can model ideal player profiles that fit a specific team's system, helping coaches narrow in on targets more effectively.
  • Game strategy: Predictive analytics allow coaches to analyze historical data, opponent behavior, and game conditions to optimize strategies and make in-game adjustments.
Operations and other support
  • Game analysis: AI-powered platforms can automatically tag and analyze game footage, breaking down plays, routes, and formations in seconds, freeing up coaches to focus on coaching.
  • Academic and career support: AI can assist with academic monitoring, degree progress, and even career readiness, providing a more holistic view of a student-athlete's success.
  • Automated officiating: AI could be used to assist with officiating to reduce human error and improve fairness in calls.
  • Fan engagement: AI can help sports media departments create new and exciting fan experiences, from personalized content to virtual reality scenarios.
😉
Expand ...

^^ THIS ^^

Well done post!

I use AI extensively and suggest that if you are not using it in your professional and personal lives, you start ‘playing’ with it yesterday. I’m a big fan of the recent Gemini 3 release. I addded more units to my Google holdings based on the experience. Before you ask, I did not ask Gemini if I should buy more Google, lol.
 
You must log in or register to reply here.

Online statistics

Members online
490
Guests online
6,873
Total visitors
7,363
Totals may include hidden visitors.

Forum statistics

Threads
165,622
Messages
4,447,464
Members
10,319
Latest member
Fan1952


Top Bottom