#statistical-modeling

The 2024 AL MVP stats model keeps giving me a number that looks like a spreadsheet error from 1998. I'm watching a team with a 4‑10 pitching staff and the model spits out 78 MPH average spin on a guy that normally throws a 92‑high screwball in a two‑ball count. I keep pointing out that his sinker velocity dips to 76 MPH on a change‑up but the model just averages the splitters. Did anyone else feel that the model's default of 0.5 for off‑speed is off the charts? I'm fed up. Every analyst I see uses the same 1.25 RPM overestimation for a 180‑RPM wrist flick. It's the same thing: a ball's life is defined by the pitcher's hands, not a spreadsheet's greed for numbers. I hate that people think I don't understand pitching mechanics. I'm literally doing real time PITCHFLEX analysis. If you think there's a new velocity‑data‑in‑field problem, I'll show you the raw data. I don't need your sympathy. I need a solution.

To sum it up:

• 2024 AL MVP model over‑values splitters
• Ignores real off‑speed data
• Default off‑speed factor is a bad assumption

Help: Do you see the same inconsistency? Have you fixed it or do you just let the spreadsheets decide the game? My frustration is at a new level and I'm not talking about the pitcher or the staff.

I'm not saying I invented the algorithm. I'm saying that the algorithm isn't working. I want to know why. Or at least someone who can prove the numbers are fine.

Because of that I feel like the baseball community is ignoring the data that actually matters…