Loading draft…
Loading draft…
What slot #1 alone is worth: 61% playoffs · 13% title if all 12 teams drafted exactly like us (average team 50% / 8.3%).
The single-number projection (96.5% / 54.8%) is the optimistic end: it assumes the other 11 teams draft by Yahoo ADP plus noise, and it grades every roster with our own projections, so it rewards us for disagreeing with the market. With a few sharp opponents and the truth halfway between our projections and the market, slot 1 is 83–89% playoffs / 28–35% title. If all 11 draft sharply with the market's view and the market is right: 67% / 12%, still better than an average team.
| Who drafts how | Ours | Blend | Market |
|---|---|---|---|
| adp_all all 12 by noisy ADP (me too) | 54% / 10% | 54% / 10% | 54% / 10% |
| A_sym all 12 our optimizer, each on its own noisy view | 61% / 13% | 58% / 11% | 55% / 9% |
| A_bpa all 12 value-BPA (fast proxy), own views | 57% / 12% | 58% / 12% | 58% / 12% |
| C11 me optimizer; 11 opponents our optimizer (our-like views) | 75% / 25% | 72% 22% |
Slot 1; 60–100 simulated drafts × 500 seasons per scenario. Research note @170deb4 · 2026-10-18
| Optimizer | BPA | |
|---|---|---|
| Exp. category wins / 9 | 5.88 | 5.80 |
| P(win a week) | 84.1% | 83.1% |
| Matchup-win edge | +0.96 pts ±0.08 | |
We replayed the whole 12-team draft many times on the computer. In each replay the other 11 teams pick and our optimizer makes my picks; the same replay is then re-run with a simple 'take the best-valued player left' rule (BPA) for my team, so the two strategies face identical opponents.
fantasy sim preseason plays simulated leagues through a whole season to estimate P(playoffs) and P(title).In-sample: the optimizer is scored by the same projections and win model it optimizes, so its edge over BPA is an upper bound. Absolute levels are optimistic: the opponent is a league-average team drafted by ADP that never manages its roster, and weekly variance ignores schedule and category correlation. Compare the two strategies relatively; don't read 5.4 cats/week as a forecast. Variance priors and ADP noise are uncalibrated.
See How it works for the full list.
| 68% / 18% |
| C6 me optimizer; 6 such opponents, 5 ADP | 87% / 31% | 79% / 23% | 68% / 14% |
| C3 me optimizer; 3 such opponents, 8 ADP | 92% / 39% | 83% / 28% | 69% / 16% |
| C11m me optimizer; 11 opponents our optimizer on market-like views | 94% / 47% | 85% / 28% | 67% / 12% |
| C3m me optimizer; 3 market-informed optimizers, 8 ADP | 96% / 53% | 89% / 35% | 75% / 17% |
| status quo me optimizer; 11 noisy ADP (= `fantasy sim preseason`) | 97% / 55% | 91% / 37% | 78% / 20% |
P(playoffs) / P(title) for slot 1. Columns = whose view of the players is true when the season is played: our projections, the market's (ADP), or a 50/50 blend.
Distribution across simulations of my final roster's weekly win probability in each category vs a league-average team: mean and 10th/90th percentiles. Low values in a category mean the roster tends to concede it (a de-facto punt).
Who I took at each of my picks across the simulations (top 5). Early picks are nearly deterministic; later picks spread out because the board differs from draft to draft.
| Player | #1 | #24 | #36 | #37 | #60 | #61 | mine |
|---|---|---|---|---|---|---|---|
Nikola Jokić 1/1 C | ✓ | · | · | · | · | · | 100% |
Victor Wembanyama 2/2 C | ✓ | · | · | · | · | · | 0% |
Luka Dončić 3/4 PG/SG | ✓ | · | · | · | · | · | 0% |
Tyrese Maxey 10/6 PG | ✓ | · | · | · | · | · | 1% |
Jalen Johnson 12/14 SF/PF | ✓ | · | · | · | · | · | 1% |
Tyrese Haliburton 13/5 PG/SG | ✓ | 1% | · | · | · | · | 4% |
Donovan Mitchell 14/10 PG/SG | ✓ | 2% | · | · | · | · | 1% |
Scottie Barnes 15/11 SF/PF/C | ✓ | 3% | · | · | · | · | 6% |
Karl-Anthony Towns 16/12 PF/C | ✓ | 5% | · | · | · | · | 7% |
Kevin Durant 16/28 SG/SF/PF | ✓ | 8% | · | · | · | · | 0% |
Alperen Şengün 18/37 PF/C | ✓ | 12% | · | · | · | · | 0% |
Jamal Murray 21/9 PG | ✓ | 17% | · | · | · | · | 26% |
Jalen Brunson 22/21 PG | ✓ | 22% | · | · | · | · | 0% |
Amen Thompson 23/31 PG/SG | ✓ | 28% | 1% | · | · | · | 0% |
Josh Giddey 23/35 PG/SG | ✓ | 35% | 1% | 1% | · | · | 0% |
| Player | #1 | #24 | #36 | #37 | #60 | #61 | mine |
|---|---|---|---|---|---|---|---|
Austin Reaves 23/29 PG/SG/SF | ✓ | 47% | 3% | 2% | · | · | 0% |
LaMelo Ball 25/18 PG | ✓ | 59% | 7% | 5% | · | · | 5% |
Model probability that the other teams have not yet taken each key player by my first six picks, from the ADP availability model the optimizer itself uses (computed analytically from pick 1, not simulated; it ignores that I may have taken him myself earlier). sim_taken_by_me_freq is how often I actually drafted him in the simulations.
Under each name: ADP / our value rank. ✓ = certain, · = gone. "mine" = how often the optimizer drafted him in simulation.
Strategy the optimizer settled on at my first pick. A "punt" deliberately gives up one category (e.g. FT%) to be stronger in the other eight.
strategy: balanced
Three full rosters the optimizer produced: the best, median and worst simulation by expected category wins - a feel for the range of outcomes from slot 1.
Devin Booker 26/30 PG/SG |
| ✓ |
| 64% |
| 9% |
| 7% |
| · |
| · |
| 0% |
Bam Adebayo 26/17 PF/C | ✓ | 68% | 12% | 10% | · | · | 49% |
Evan Mobley 29/26 PF/C | ✓ | 72% | 16% | 13% | · | · | 3% |
Chet Holmgren 32/23 PF/C | ✓ | 79% | 24% | 21% | · | · | 15% |
Kawhi Leonard 34/20 SG/SF/PF | ✓ | 85% | 34% | 29% | · | · | 38% |
Jalen Duren 35/24 C | ✓ | 87% | 38% | 34% | · | · | 16% |
James Harden 36/32 PG/SG | ✓ | 89% | 43% | 39% | · | · | 0% |
Trey Murphy III 36/25 SG/SF | ✓ | 90% | 48% | 43% | · | · | 1% |
Walker Kessler 42/41 C | ✓ | 96% | 71% | 68% | 3% | 2% | 1% |
Anthony Davis 43/44 PF/C | ✓ | 97% | 77% | 74% | 5% | 4% | 0% |
Donovan Clingan 45/27 C | ✓ | 98% | 79% | 76% | 7% | 6% | 14% |
Derrick White 45/19 PG/SG | ✓ | 98% | 81% | 79% | 9% | 7% | 85% |
Kyrie Irving 48/33 PG | ✓ | 98% | 85% | 83% | 13% | 11% | 16% |
Kon Knueppel 53/16 SG/SF | ✓ | 99% | 92% | 90% | 26% | 24% | 34% |
Desmond Bane 53/22 SG/SF | ✓ | 99% | 93% | 91% | 29% | 27% | 58% |
Onyeka Okongwu 55/43 C | ✓ | 99% | 94% | 93% | 36% | 33% | 1% |
Nickeil Alexander-Walker 62/40 PG/SG | ✓ | ✓ | 96% | 96% | 48% | 45% | 7% |
Michael Porter Jr. 63/38 SF/PF | ✓ | ✓ | 97% | 96% | 52% | 49% | 8% |
Naz Reid 66/36 PF/C | ✓ | ✓ | 97% | 97% | 57% | 54% | 35% |
Dyson Daniels 66/45 SG/SF | ✓ | ✓ | 98% | 97% | 60% | 57% | 9% |
Kel'el Ware 77/34 C | ✓ | ✓ | 99% | 99% | 83% | 82% | 82% |
VJ Edgecombe 77/42 PG/SG | ✓ | ✓ | 99% | 99% | 85% | 83% | 35% |
Mikal Bridges 82/39 SF/PF | ✓ | ✓ | ✓ | ✓ | 92% | 91% | 74% |
strategy: balanced
strategy: balanced