Early games kick at 10:00 PT. Pickens, DeVonta Smith, Wicks and Skattebo are OUT. St. Brown is active with a rib injury. The London game is almost over and the Bills-Vikings legs play Monday, so those are flagged.
Four findings. They are observations from a small sample, not hard rules yet. Each one is a place where our money is either working or leaking.
The near-lock tier is real. It only missed on Michigan (-625) and Harden. This is where hammers live.
Regular lines priced near even money have been hitting 62%. That's the edge you were feeling, Ace.
Lines above a player's average paid off far more than I graded. Small sample (18), so small stakes only.
60s-tier picks bought down to alt lines hit 58%, which is less than Kalshi charges for them. This is where lottos die.
Green bars are what actually hit, white ticks are my average grade. Above 70% my grades sort winners well. Below 70% the bars flatten out: a leg I grade 45 has hit about as often as one I grade 65.
You asked what separates the 50-60% picks that hit from the ones that don't. These are 122 legs in that zone, split by market. The white line marks 50%.
Mid-tier props on the second night of a back-to-back almost all missed. Only Bam 15+ survived.
Points + rebounds + assists lines in the coin-flip zone. Three ways to come up short.
Assists depend on teammates making shots. Pure coin flip in this zone.
Mid-tier CFB legs, plus Michigan at -625 and Ohio State both lost. Spreads went 0 of 2.
The strongest coin-flip category we have. Big men and wings on regular lines.
Every NBA moneyline we graded cashed, including Philly and Portland as near coin flips.
Threes made went 5 of 6. NFL player props went 6 of 7 in this zone.
11 misses by exactly one, 6 hits that landed right on the number. Buy the half-point only when it's cheap.
You're right that two 55% legs can beat seven 80% legs. What decides it is the edge on each leg: how often it hits compared with what the price assumes. Here are three ways to build the same $100, using our actual hit rates.
NBA is still the engine: 155 legs, and 85% on the top tier. NFL props have been excellent on a small sample, which matters today. College football's top tier hit only 4 of 6.
Solid borders are rules we've set. Dashed borders are new observations we're testing before they become rules.
Sample: 211 graded legs with verified results, Nov 4 to 7, 2026. Legs with no box score were left out. Small samples swing; treat patterns as leads, not laws.