BUF/MIA wind dropped 4mph — what 10k sims say about Allen now
The market hasn't fully repriced. We dug into the new distribution and there's a measurable edge.
Two days ago Caesars opened BUF/MIA at 49.5, with the over juiced to -115. The wind forecast called for sustained 18mph gusts in Orchard Park, which is the kind of weather that pushes Allen's passing distribution about a standard deviation to the left.
This morning the forecast updated. Winds are now 14mph, and the total ticked to 50.5. The market moved. But it didn't move enough.
What changed in 10k iterations
We resampled the slate at 8am this morning using the updated weather and totals. The most interesting number is Allen's ceiling: it moved from 34.2 to 38.6, which is bigger than the total movement implies because reduced wind affects deep-ball variance disproportionately.
Quick math: Allen's implied passing TD rate jumped from 5.2% to 5.8% per attempt. Across 33 attempts that's a tenth of a TD in expected value, but the tail moved further — the chance of 3+ passing TDs rose from 22% to 34%.
And critically, the field hasn't adjusted. Allen ownership on DK projects at 18%, almost unchanged from Wednesday. His leverage score is now 0.94, the highest on the slate among QBs priced above 8k.
Why this is a stack play, not just an Allen play
The Allen–Diggs correlation in our sim is 0.62, the highest on the slate. When wind drops, that correlation tightens further because deep targets get cleaner. A naked Allen captures some upside; an Allen+Diggs stack captures the whole left tail of the joint distribution.
Bring-back is interesting too. Tyreek Hill's ceiling moved +0.8, also disproportionately because his deep-ball variance contracted with wind. A standard QB+WR+bring shape here looks like the highest-EV construction on the slate at current ownership.
The three lineups we built around this edge
We ran 10,000 iterations through Build with QB+WR+bring locked, fade thresholds at 20% ownership, and a min-uniqueness of 3. The top three lineups by ceiling all leaned into the Allen+Diggs stack with either Cook or Achane completing the bring-back.
Late swap rule: if Allen ownership ticks above 24% by 11:30am, pivot to Hurts. The ceiling EV gap closes fast above that threshold.
That's the take. If you want to run this with your own priors — different weather assumptions, different ownership projections — that's what the rest of the app is for.
Builds models, breaks them, writes about it. Cashed a Milly Maker once. Will not let you forget it.