Jim Cook

Builder. Solo dev behind MA4B ( — transfer portal analytics for college basketball. 59-feature projection model trained on 3,791 transfers.

[Jim Cook @jimcook57]

@SteaknShake Such a deal. Got a hat too!

[Jim Cook @jimcook57]

@KySportsRadio Sankey needs to go.

[Jim Cook @jimcook57]

@KySportsRadio @MattNorlander If true, we go to the mattresses. And by we, I mean Mark Mitchell, Mark Pope, J Batts, UK, Kentucky AG, and Kentucky Governor, and BBN. Sankey needs to go.

[Jim Cook @jimcook57]

@CoachMarkPope What have you learned about portal recruiting this cycle that will change how you approach next spring?

[Jim Cook @jimcook57]

@MrsTylerKSR

[Jim Cook @jimcook57]

@KySportsRadio Donovan!

[Jim Cook @jimcook57]

From last year to this year, what Kentucky lost in shooting, defense, and ball security, they gained in shot creation and playmaking. Whether it's an upgrade depends entirely on what Pope builds around them. Surround them with shooters (Momcilovic, Celiscar, Bashir) and the playmaking becomes elite. Surround them with more non-shooters and the lane gets clogged. Overtall, it's a lateral move with different strengths.

[Jim Cook @jimcook57]

I would say to all of those who are panicing: The portal ENTRY window closes tonight. The COMMITMENT window runs to Memorial Day. The NBA draft withdrawal deadline is mid-June. Momcilovic (49% from 3) and Graves (41%) are both testing the draft with portal eligibility. International players have no deadline at all. Pope needs 3 adds — a wing, a stretch 4, a backup C. That's it. We have two legit playmakers already. The building hasn't even started yet. Stop grading a half-built house.

[Jim Cook @jimcook57]

This is the right energy. And the numbers say there's reason for it. Momcilovic (49% 3P) and Graves (41% 3P) are both testing the draft with portal eligibility — draft withdrawal deadline is mid-June. Commitment window runs to Memorial Day. International options have no deadline. The offseason is long for a reason.

[Jim Cook @jimcook57]

The fit question is real. Our model ranked Diallo and Wilkins as top system fits for Kentucky's offense even though they weren't top-10 portal names. If
the staff is prioritizing fit over ranking, that's defensible at guard. The problem is they haven't found that fit at the 4 yet. That's the gap.

[Jim Cook @jimcook57]

@KySportsRadio The next couple of weeks, plus mid-June when draft-testing guys like Momcilovic (49% 3P) and Graves (41% 3P) decide whether to come back to college. That's
the real deadline.

[Jim Cook @jimcook57]

@BRamseyKSR The numbers back you up. Momcilovic (49% 3P) and Graves (41% 3P) are testing the draft with portal eligibility. Mid-June withdrawal deadline. Plus
international options have no deadline at all. Freeman hurts today, but this roster isn't close to finished.

[Jim Cook @jimcook57]

The portal entry window closes tonight. The commitment window doesn't close until Memorial Day. Momcilovic (49% from 3, Impact 8.7) and Graves (41%, Impact 7.7) are both testing the NBA draft with portal eligibility. If either comes back in June, they're the best available player in the country. It's April 21, not August 21.

[Jim Cook @jimcook57]

@steipete @MilosDragicevic Our problem isn’t with anthropic, it is with OpenAI, xai and the others. If they built better models we would say Anthropic who?

[Jim Cook @jimcook57]

@DrewFranklinKSR @KySportsRadio Will miss Chandler and sorry to see him go. More backcourt money for Stokes and Blackwell. There is no scenario where Chandler starts and we field a final four roster.

[Jim Cook @jimcook57]

Wright's BPR of 7.1 is top 25 nationally — comparable to Blackwell (7.2). He's not just the best PG in the portal, he might be a top 5 overall player.
For Kentucky specifically: 41% from three fits Pope's Perimeter Sniper system perfectly. Pope built that system at BYU — Wright ran it there last season.
The Arkansas angle is interesting. Calipari runs a different system. Wright's shooting might be underutilized there.

[Jim Cook @jimcook57]

@AlexFinn Running 600B+ parameter models locally on a Mac Studio M3 Ultra. GLM-5.1 at 18.9 tok/s, MiniMax-M2.5 at 26 tok/s — both via mlx-lm, no cloud, no API keys. GLM writes like a senior analyst. MiniMax writes like a senior analyst in a hurry. 512GB of unified memory is a beast.

[Jim Cook @jimcook57]

Ran all 8 through our projection model. Quick data hits:
Thomas Jr. grades highest — 6.1 BPR, projects 13ppg. Best pure PG on the board.

Miles Byrd is the sleeper — only 10.4ppg but 4.3 Defensive Impact. Best defender on this list by far.

Haralson concern: 20% from three. Tough fit for Pope's system.

Solid list. Pope is targeting the right profiles.

[Jim Cook @jimcook57]

Interesting call. Freeman's 16.5ppg and 7.2rpg at Syracuse are legit. One concern our model flags: 30% from three. Pope has spent 5 years running a Perimeter Sniper system — he needs his 4 to space the floor. If Pope wants a stretch-4 who shoots, J.P. Estrella (Tennessee, 40% from 3, 6-11) is the model's #1 PF fit for Kentucky. Freeman is the better scorer. Estrella is the better system fit. Depends on what Pope prioritizes.

[Jim Cook @jimcook57]

Love the energy but a couple issues: Quaintance is projected #14 in the NBA draft — he's not coming back to college. Dioubate entered the portal and left Kentucky. He's gone. Blackwell + Stokes + Freeman is a great core though. Our model has Blackwell as the #1 portal fit for UK out of 533 players. Haggerty's tricky — 23ppg but his teams are a combined 0-for-winning. Our model flags high-volume scorers who change schools repeatedly.