Receipts: Mako Bets had a winning week, +57.53 units, but the losses are real and most samples are small
Across every model, the 7 days since 2026-10-27 netted +57.53 units. Several models lost, and many have far too few graded bets to judge.
The honest headline: a winning week, with caveats
All models combined finished the week at +57.53 units. That is a winning week. It does not prove the models work. One week is short, and many rows below are forward tests that have not reached their read points. Some of these rows also look like overlapping variants of the same scanner, so the combined total is not 25 independent results.
Best models this week
The top row by week units is the +EV scanner (EV1-X) at +29.40 units, going 32-103 on the week. That is a low win rate made up by longer prices. It has 157 graded bets for the season and is a forward test to be read at 500 bets, so it is not proven. College basketball first half was next at +22.36 units on 74-46, a +18.6% season return over 120 graded bets. That is under the 150-bet small-sample line, and its picks are admin-only until it passes. College football picks not shown on the site made +16.06 units on 32-14 this week. Over the season it sits at +2.35 units on 302 graded bets, a +0.8% return. That is a thin edge at best.
| Model | This week | Week units | Season W-L | Season units | Season return | Graded | Status |
|---|---|---|---|---|---|---|---|
| +EV scanner (EV1-X) | 32-103 | +29.40 | 34-123 | +13.05 | +8.3% | 157 | Forward test from Oct 24; read at 500 bets |
| College basketball first half | 74-46 | +22.36 | 74-46 | +22.36 | +18.6% | 120 | Forward test; picks admin-only until it passes |
| College football picks not shown on the site | 32-14 | +16.06 | 146-155 | +2.35 | +0.8% | 302 | Logged, not shown on the site |
| +EV scanner (EV1-X) | 34-52 | +15.61 | 34-52 | +15.61 | +18.1% | 86 | Forward test from Oct 24; read at 500 bets |
| +EV scanner (EV1-X) | 18-7 | +15.26 | 18-7 | +15.26 | +61.0% | 25 | Forward test from Oct 24; read at 500 bets |
| +EV scanner, power de-vig (EV1-XP) | 15-2 | +13.71 | 15-2 | +13.71 | +80.6% | 17 | Shadow of EV1-X from Oct 24; read with it |
| L1 rushing and receiving unders | 29-19 | +12.98 | 65-63 | +10.77 | +8.4% | 128 | Forward test in progress |
| +EV scanner, power de-vig (EV1-P) | 47-41 | +10.45 | 87-63 | +40.23 | +26.8% | 150 | Shadow of EV1 from Oct 24; read with it |
| L2 receptions usage model | 51-31 | +6.69 | 100-58 | +18.73 | +11.9% | 158 | Passed holdout read 2 (design version); live forward test |
| Fair Odds: books above the exchanges (FP1) | 11-11 | +6.34 | 11-11 | +6.34 | +28.8% | 22 | Forward test from Oct 23; read at 300 bets |
| Fair Odds: books above Kalshi (FP1-X) | 5-0 | +5.22 | 5-0 | +5.22 | +104.4% | 5 | Forward test from Oct 24; read at 300 bets |
| +EV scanner (EV1) | 41-50 | +4.62 | 83-76 | +54.10 | +34.0% | 159 | Forward test from Oct 23; read at 500 bets |
| +EV scanner, power de-vig (EV1-XP) | 43-42 | +4.52 | 54-56 | -1.13 | -1.0% | 110 | Shadow of EV1-X from Oct 24; read with it |
| Fair Odds: books above Kalshi (FP1-X) | 9-10 | +3.83 | 9-10 | +3.83 | +20.2% | 19 | Forward test from Oct 24; read at 300 bets |
| College football picks shown on the site | 24-20 | +2.28 | 66-83 | -21.15 | -14.2% | 149 | Live forward test |
| +EV scanner, power de-vig (EV1-XP) | 19-24 | +0.67 | 19-24 | +0.66 | +1.6% | 43 | Shadow of EV1-X from Oct 24; read with it |
| Following sharp Polymarket wallets (PM2) | 6-6 | -2.25 | 6-6 | -2.25 | -18.7% | 12 | Forward test from Oct 26; read at 300 sharp bets from 5+ wallets, or Feb 15 |
| L4 passing yards usage model | 7-11 | -4.74 | 17-21 | -5.84 | -15.4% | 38 | Failed design test; shown by choice |
| QB1 passing yards | 2-7 | -5.25 | 3-9 | -5.97 | -49.7% | 12 | Experimental: untested model |
| +EV scanner, power de-vig (EV1-XP) | 17-23 | -6.50 | 18-26 | -8.59 | -19.5% | 44 | Shadow of EV1-X from Oct 24; read with it |
| USE-NBA usage model | 71-79 | -9.78 | 460-435 | -0.57 | -0.1% | 895 | Failed its tests in every market (holdout or design); shown by choice |
| L3 rushing and receiving yards usage models | 82-87 | -13.17 | 169-144 | +7.82 | +2.5% | 313 | Failed design test; shown by choice |
| Fair Odds: books above Kalshi (FP1-X) | 11-28 | -13.77 | 13-33 | -16.92 | -36.8% | 46 | Forward test from Oct 24; read at 300 bets |
| Fair Odds: books above Kalshi (FP1-X) | 37-59 | -17.47 | 48-66 | -12.60 | -11.1% | 114 | Forward test from Oct 24; read at 300 bets |
| +EV scanner (EV1-X) | 44-80 | -19.00 | 54-121 | -47.82 | -27.3% | 175 | Forward test from Oct 24; read at 500 bets |
| Parlay Builder suggestions (PB1) | 10-39 | -20.51 | 10-39 | -20.51 | -41.9% | 49 | Forward test from Oct 28; read at 300 parlays, or Feb 15 |
Worst models this week
The worst row was the Parlay Builder suggestions (PB1) at -20.51 units, going 10-39. It started Oct 28 and has 49 graded parlays, far short of its 300-parlay read. Parlays lose often, so this is early, but it is a real loss. The next worst was a +EV scanner (EV1-X) row at -19.00 units on 44-80, with a season mark of -47.82 units over 175 graded bets. A Fair Odds: books above Kalshi (FP1-X) row lost 17.47 units on 37-59 for the week. L3 rushing and receiving yards usage models lost 13.17 units on 82-87 and USE-NBA usage model lost 9.78 units on 71-79. Both are labeled as having failed their tests and are shown by choice. L3 is still +7.82 units for the season, but its status says not to lean on that.
Where the sample is too small
We flag anything under 150 graded bets as too small to judge. That covers most of the table, including the best-looking lines. The +EV scanner, power de-vig (EV1-XP) row at 15-2 and +80.6% has 17 graded bets. Fair Odds: books above Kalshi (FP1-X) at 5-0 has 5. Neither tells you anything yet. Larger samples are more informative. USE-NBA usage model has 895 graded bets and sits at -0.57 units, a -0.1% return, which is roughly break-even. College football picks shown on the site is -21.15 units at -14.2% over 149 graded bets, just under the small-sample line.
Questions
Did Mako Bets win or lose this week?
Combined, the models won +57.53 units over the 7 days since 2026-10-27. Individual models ranged from big wins to big losses.
Which Mako Bets model was best this week?
The +EV scanner (EV1-X) row at +29.40 units was the top by week units. It is a forward test with 157 graded bets, so it is not yet proven.
How many bets does a model need before the record means something?
We treat anything under 150 graded bets as a small sample. Many models are also waiting on their own read points, such as 300 or 500 bets.