What if someone would have told you that Tottenham Spurs would AVOID RELEGATION by the "skin of their teeth"??

Would You Listen?!?!

IF YOU DIDN'T..... NEXT TIME YOU WILL BECAUSE ALL THE BARNES BROTHERS DO IS WIN BABY WIN!!!!

A full breakdown of every team in the SEC, including talent analysis, coordinator concern, and schedule forecasts, with win total expectations. Enjoy!

March Madness final four recap and Championship preview...

MiracleManDan’s AI Baseball 7/10/26

Experimental :

It doesn’t seem that we should be getting so many FG plays. I questioned this one, and well, maybe so.

Orioles vs Royals today

Starters are real: Young 3.21 / 1.39 vs Avila 6.64 / 1.57 (ERA delta +3.43).

Why isn’t the model ~35–40% on KC?

The number is a 65% form / 35% LightGBM blend, and neither half treats season ERA like a human does:

Form leg (~50.5% KC) — both clubs are .400; ratings are basically even (4.20 / 4.30). This leg doesn’t use SP ERA at all.

LightGBM (~49.3% KC after calibration) — SP ERA is a feature, but for this row it barely moves the needle. Counterfactuals:

Set Avila’s ERA/WHIP = Young’s → blend unchanged (~0.0%)

Swap the two SPs → blend only −1.8% on KC

So the model is mostly saying: “even teams, Royals score more (5.3 vs 4.0 rpg), market has BAL at ~61% — we’re closer to a coin flip.” It is not saying Avila matches Young.

Should you bet it?

F5: no — already filtered out (1.8% edge, no angle).

Full game: yes per live policy — dog, edge ≥ 3%, price ≤ +175 → 1u. That angle intentionally ignores SP quality; our soft-SP sweeps found those dogs were more profitable on rate, not less.

How to think about it

You’re ...