NFL · 2026 Week 3
2026 Week 3 Half PPR QB Rankings
Final pre-kickoff board · published September 24, 2026, 22:49 UTC · first kickoff September 25, 2026, 00:15 UTC
Half PPR, 1QB · 32 players projected · updated September 24, 2026
See how these rankings performed
Sportsbook prices reached 45 of 145 fantasy-relevant players (31%), so the market half of this board is thinner than a full week's.
How these rankings work
How these rankings are built
We start with preseason player projections and draft position — where players were being selected in fantasy drafts entering the season.
We then update those expectations for Week 3 using recent usage, availability and market signals. Market signals are current sportsbook and prediction-market prices that can reflect changing expectations for a player's workload, performance, or availability.
We then simulate the week 4,000 times and rank players by the projected fantasy points those simulations produce.
This board is priced in Half PPR. The order and the finish percentages below are specific to it: another format is a separate simulation of the same projections, not this table with new points.
What the percentages mean
For every simulation, we also record where each player finishes at his position.
- Top 3
- how often he finishes among the top three at his position
- Top 12
- how often he finishes in the top 12
- Top 24
- how often he finishes in the top 24
A WR with a 58% Top 12 rate finished as a top-12 WR in 58% of the simulated weeks.
How often a strong start lasts
Since 1999, 56% of QBs inside the top 12 through Week 2 finished the season inside the top 12. Measured over 324 QB-seasons from 1999-2025, in the same scoring rule as this board.
What stood out in Week 3
32 QBs ranked for Week 3. Josh Allen leads at 21.4 projected points.
Tier breaks
2.9 projected points separate QB1 Josh Allen from QB2 Tyler Shough.
2.5 projected points separate QB30 Marcus Mariota from QB31 Cooper Rush.
Biggest differences from the draft market
Behind
- Drake MayeQB18Our projection is lower than the outside forecastTop 3: 7% · Top 12: 37% · Top 24: 77% -15 vs. QB3
- Justin HerbertQB22Our projection is lower than the outside forecastTop 3: 6% · Top 12: 33% · Top 24: 76% -8 vs. QB14
- Lamar JacksonQB6Our projection is lower than the outside forecastTop 3: 14% · Top 12: 48% · Top 24: 83% -4 vs. QB2
| # | Pos | Player | Pos rank | Proj | Outcome profile |
|---|---|---|---|---|---|
| 1 | QB | Josh AllenBUF Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 26% of the time, 4th to 12th 38%, 13th to 24th 26%, and outside the top 24 10%. Across the simulated weeks he finishes QB8 at the median, QB3 - QB16 in the middle half, and QB1 - QB25 in the middle 80%.
Top 3: 26%Top 12: 64%Top 24: 90% |
QB1 | 21.4 | Across the simulated weeks at the position, he finishes in the top 3 26% of the time, 4th to 12th 38%, 13th to 24th 26%, and outside the top 24 10%. Across the simulated weeks he finishes QB8 at the median, QB3 - QB16 in the middle half, and QB1 - QB25 in the middle 80%. |
| 2 | QB | Tyler ShoughNO Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 17% of the time, 4th to 12th 35%, 13th to 24th 32%, and outside the top 24 16%. Across the simulated weeks he finishes QB12 at the median, QB5 - QB21 in the middle half, and QB2 - QB27 in the middle 80%.
Top 3: 17%Top 12: 53%Top 24: 84% |
QB2 | 18.5 | Across the simulated weeks at the position, he finishes in the top 3 17% of the time, 4th to 12th 35%, 13th to 24th 32%, and outside the top 24 16%. Across the simulated weeks he finishes QB12 at the median, QB5 - QB21 in the middle half, and QB2 - QB27 in the middle 80%. |
| 3 | QB | Bryce YoungCAR Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 16% of the time, 4th to 12th 35%, 13th to 24th 33%, and outside the top 24 16%. Across the simulated weeks he finishes QB12 at the median, QB5 - QB21 in the middle half, and QB2 - QB28 in the middle 80%.
Top 3: 16%Top 12: 51%Top 24: 84% |
QB3 | 18.1 | Across the simulated weeks at the position, he finishes in the top 3 16% of the time, 4th to 12th 35%, 13th to 24th 33%, and outside the top 24 16%. Across the simulated weeks he finishes QB12 at the median, QB5 - QB21 in the middle half, and QB2 - QB28 in the middle 80%. |
| 4 | QB | Patrick MahomesKC Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 15% of the time, 4th to 12th 37%, 13th to 24th 34%, and outside the top 24 15%. Across the simulated weeks he finishes QB12 at the median, QB6 - QB20 in the middle half, and QB3 - QB27 in the middle 80%.
Top 3: 15%Top 12: 51%Top 24: 85% |
QB4 | 17.9 | Across the simulated weeks at the position, he finishes in the top 3 15% of the time, 4th to 12th 37%, 13th to 24th 34%, and outside the top 24 15%. Across the simulated weeks he finishes QB12 at the median, QB6 - QB20 in the middle half, and QB3 - QB27 in the middle 80%. |
| 5 | QB | Jalen HurtsPHI Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 15% of the time, 4th to 12th 33%, 13th to 24th 33%, and outside the top 24 19%. Across the simulated weeks he finishes QB13 at the median, QB6 - QB22 in the middle half, and QB2 - QB28 in the middle 80%.
Top 3: 15%Top 12: 48%Top 24: 81% |
QB5 | 17.5 | Across the simulated weeks at the position, he finishes in the top 3 15% of the time, 4th to 12th 33%, 13th to 24th 33%, and outside the top 24 19%. Across the simulated weeks he finishes QB13 at the median, QB6 - QB22 in the middle half, and QB2 - QB28 in the middle 80%. |
| 6 | QB | Lamar JacksonBAL Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 14% of the time, 4th to 12th 34%, 13th to 24th 35%, and outside the top 24 17%. Across the simulated weeks he finishes QB13 at the median, QB6 - QB22 in the middle half, and QB3 - QB28 in the middle 80%.
Top 3: 14%Top 12: 48%Top 24: 83% |
QB6 | 17.4 | Across the simulated weeks at the position, he finishes in the top 3 14% of the time, 4th to 12th 34%, 13th to 24th 35%, and outside the top 24 17%. Across the simulated weeks he finishes QB13 at the median, QB6 - QB22 in the middle half, and QB3 - QB28 in the middle 80%. |
| 7 | QB | Jared GoffDET Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 13% of the time, 4th to 12th 37%, 13th to 24th 33%, and outside the top 24 16%. Across the simulated weeks he finishes QB12 at the median, QB6 - QB20 in the middle half, and QB3 - QB28 in the middle 80%.
Top 3: 13%Top 12: 51%Top 24: 84% |
QB7 | 17.4 | Across the simulated weeks at the position, he finishes in the top 3 13% of the time, 4th to 12th 37%, 13th to 24th 33%, and outside the top 24 16%. Across the simulated weeks he finishes QB12 at the median, QB6 - QB20 in the middle half, and QB3 - QB28 in the middle 80%. |
| 8 | QB | C.J. StroudHOU Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 12% of the time, 4th to 12th 33%, 13th to 24th 36%, and outside the top 24 19%. Across the simulated weeks he finishes QB14 at the median, QB7 - QB22 in the middle half, and QB3 - QB28 in the middle 80%.
Top 3: 12%Top 12: 46%Top 24: 81% |
QB8 | 16.9 | Across the simulated weeks at the position, he finishes in the top 3 12% of the time, 4th to 12th 33%, 13th to 24th 36%, and outside the top 24 19%. Across the simulated weeks he finishes QB14 at the median, QB7 - QB22 in the middle half, and QB3 - QB28 in the middle 80%. |
| 9 | QB | Brock PurdySF Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 13% of the time, 4th to 12th 32%, 13th to 24th 36%, and outside the top 24 19%. Across the simulated weeks he finishes QB14 at the median, QB7 - QB22 in the middle half, and QB3 - QB29 in the middle 80%.
Top 3: 13%Top 12: 45%Top 24: 81% |
QB9 | 16.8 | Across the simulated weeks at the position, he finishes in the top 3 13% of the time, 4th to 12th 32%, 13th to 24th 36%, and outside the top 24 19%. Across the simulated weeks he finishes QB14 at the median, QB7 - QB22 in the middle half, and QB3 - QB29 in the middle 80%. |
| 10 | QB | Dak PrescottDAL Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 13% of the time, 4th to 12th 32%, 13th to 24th 35%, and outside the top 24 20%. Across the simulated weeks he finishes QB14 at the median, QB7 - QB23 in the middle half, and QB3 - QB29 in the middle 80%.
Top 3: 13%Top 12: 45%Top 24: 80% |
QB10 | 16.5 | Across the simulated weeks at the position, he finishes in the top 3 13% of the time, 4th to 12th 32%, 13th to 24th 35%, and outside the top 24 20%. Across the simulated weeks he finishes QB14 at the median, QB7 - QB23 in the middle half, and QB3 - QB29 in the middle 80%. |
| 11 | QB | Jordan LoveGB Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 10% of the time, 4th to 12th 34%, 13th to 24th 40%, and outside the top 24 17%. Across the simulated weeks he finishes QB14 at the median, QB7 - QB22 in the middle half, and QB4 - QB28 in the middle 80%.
Top 3: 10%Top 12: 43%Top 24: 83% |
QB11 | 16.3 | Across the simulated weeks at the position, he finishes in the top 3 10% of the time, 4th to 12th 34%, 13th to 24th 40%, and outside the top 24 17%. Across the simulated weeks he finishes QB14 at the median, QB7 - QB22 in the middle half, and QB4 - QB28 in the middle 80%. |
| 12 | QB | Joe BurrowCINQQuestionable Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 12% of the time, 4th to 12th 29%, 13th to 24th 35%, and outside the top 24 25%. Across the simulated weeks he finishes QB16 at the median, QB8 - QB25 in the middle half, and QB3 - QB30 in the middle 80%.
Top 3: 12%Top 12: 40%Top 24: 75% |
QB12 | 15.8 | Across the simulated weeks at the position, he finishes in the top 3 12% of the time, 4th to 12th 29%, 13th to 24th 35%, and outside the top 24 25%. Across the simulated weeks he finishes QB16 at the median, QB8 - QB25 in the middle half, and QB3 - QB30 in the middle 80%. |
| 13 | QB | Trevor LawrenceJAX Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 9% of the time, 4th to 12th 31%, 13th to 24th 40%, and outside the top 24 21%. Across the simulated weeks he finishes QB16 at the median, QB8 - QB23 in the middle half, and QB4 - QB29 in the middle 80%.
Top 3: 9%Top 12: 40%Top 24: 79% |
QB13 | 15.5 | Across the simulated weeks at the position, he finishes in the top 3 9% of the time, 4th to 12th 31%, 13th to 24th 40%, and outside the top 24 21%. Across the simulated weeks he finishes QB16 at the median, QB8 - QB23 in the middle half, and QB4 - QB29 in the middle 80%. |
| 14 | QB | Geno SmithNYJ Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 7% of the time, 4th to 12th 32%, 13th to 24th 40%, and outside the top 24 20%. Across the simulated weeks he finishes QB15 at the median, QB9 - QB23 in the middle half, and QB4 - QB28 in the middle 80%.
Top 3: 7%Top 12: 40%Top 24: 80% |
QB14 | 15.3 | Across the simulated weeks at the position, he finishes in the top 3 7% of the time, 4th to 12th 32%, 13th to 24th 40%, and outside the top 24 20%. Across the simulated weeks he finishes QB15 at the median, QB9 - QB23 in the middle half, and QB4 - QB28 in the middle 80%. |
| 15 | QB | Baker MayfieldTB Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 9% of the time, 4th to 12th 27%, 13th to 24th 41%, and outside the top 24 23%. Across the simulated weeks he finishes QB16 at the median, QB9 - QB24 in the middle half, and QB4 - QB29 in the middle 80%.
Top 3: 9%Top 12: 36%Top 24: 77% |
QB15 | 15.1 | Across the simulated weeks at the position, he finishes in the top 3 9% of the time, 4th to 12th 27%, 13th to 24th 41%, and outside the top 24 23%. Across the simulated weeks he finishes QB16 at the median, QB9 - QB24 in the middle half, and QB4 - QB29 in the middle 80%. |
| 16 | QB | Matthew StaffordLAR Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 9% of the time, 4th to 12th 28%, 13th to 24th 39%, and outside the top 24 24%. Across the simulated weeks he finishes QB16 at the median, QB8 - QB24 in the middle half, and QB4 - QB30 in the middle 80%.
Top 3: 9%Top 12: 37%Top 24: 76% |
QB16 | 15.1 | Across the simulated weeks at the position, he finishes in the top 3 9% of the time, 4th to 12th 28%, 13th to 24th 39%, and outside the top 24 24%. Across the simulated weeks he finishes QB16 at the median, QB8 - QB24 in the middle half, and QB4 - QB30 in the middle 80%. |
| 17 | QB | Deshaun WatsonCLE Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 8% of the time, 4th to 12th 28%, 13th to 24th 39%, and outside the top 24 25%. Across the simulated weeks he finishes QB17 at the median, QB9 - QB24 in the middle half, and QB4 - QB29 in the middle 80%.
Top 3: 8%Top 12: 36%Top 24: 75% |
QB17 | 14.9 | Across the simulated weeks at the position, he finishes in the top 3 8% of the time, 4th to 12th 28%, 13th to 24th 39%, and outside the top 24 25%. Across the simulated weeks he finishes QB17 at the median, QB9 - QB24 in the middle half, and QB4 - QB29 in the middle 80%. |
| 18 | QB | Drake MayeNE Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 7% of the time, 4th to 12th 30%, 13th to 24th 40%, and outside the top 24 23%. Across the simulated weeks he finishes QB16 at the median, QB9 - QB24 in the middle half, and QB4 - QB29 in the middle 80%.
Top 3: 7%Top 12: 37%Top 24: 77% |
QB18 | 14.8 | Across the simulated weeks at the position, he finishes in the top 3 7% of the time, 4th to 12th 30%, 13th to 24th 40%, and outside the top 24 23%. Across the simulated weeks he finishes QB16 at the median, QB9 - QB24 in the middle half, and QB4 - QB29 in the middle 80%. |
| 19 | QB | Kirk CousinsLV Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 8% of the time, 4th to 12th 28%, 13th to 24th 39%, and outside the top 24 25%. Across the simulated weeks he finishes QB17 at the median, QB9 - QB25 in the middle half, and QB4 - QB30 in the middle 80%.
Top 3: 8%Top 12: 36%Top 24: 75% |
QB19 | 14.8 | Across the simulated weeks at the position, he finishes in the top 3 8% of the time, 4th to 12th 28%, 13th to 24th 39%, and outside the top 24 25%. Across the simulated weeks he finishes QB17 at the median, QB9 - QB25 in the middle half, and QB4 - QB30 in the middle 80%. |
| 20 | QB | Jacoby BrissettARI Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 7% of the time, 4th to 12th 28%, 13th to 24th 40%, and outside the top 24 25%. Across the simulated weeks he finishes QB17 at the median, QB9 - QB25 in the middle half, and QB5 - QB29 in the middle 80%.
Top 3: 7%Top 12: 35%Top 24: 75% |
QB20 | 14.5 | Across the simulated weeks at the position, he finishes in the top 3 7% of the time, 4th to 12th 28%, 13th to 24th 40%, and outside the top 24 25%. Across the simulated weeks he finishes QB17 at the median, QB9 - QB25 in the middle half, and QB5 - QB29 in the middle 80%. |
| 21 | QB | Bo NixDEN Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 8% of the time, 4th to 12th 27%, 13th to 24th 38%, and outside the top 24 27%. Across the simulated weeks he finishes QB17 at the median, QB9 - QB25 in the middle half, and QB4 - QB30 in the middle 80%.
Top 3: 8%Top 12: 35%Top 24: 73% |
QB21 | 14.4 | Across the simulated weeks at the position, he finishes in the top 3 8% of the time, 4th to 12th 27%, 13th to 24th 38%, and outside the top 24 27%. Across the simulated weeks he finishes QB17 at the median, QB9 - QB25 in the middle half, and QB4 - QB30 in the middle 80%. |
| 22 | QB | Justin HerbertLAC Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 6% of the time, 4th to 12th 27%, 13th to 24th 43%, and outside the top 24 24%. Across the simulated weeks he finishes QB17 at the median, QB10 - QB24 in the middle half, and QB5 - QB29 in the middle 80%.
Top 3: 6%Top 12: 33%Top 24: 76% |
QB22 | 14.3 | Across the simulated weeks at the position, he finishes in the top 3 6% of the time, 4th to 12th 27%, 13th to 24th 43%, and outside the top 24 24%. Across the simulated weeks he finishes QB17 at the median, QB10 - QB24 in the middle half, and QB5 - QB29 in the middle 80%. |
| 23 | QB | Drew LockSEA Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 7% of the time, 4th to 12th 25%, 13th to 24th 39%, and outside the top 24 29%. Across the simulated weeks he finishes QB18 at the median, QB10 - QB26 in the middle half, and QB5 - QB30 in the middle 80%.
Top 3: 7%Top 12: 32%Top 24: 71% |
QB23 | 14.0 | Across the simulated weeks at the position, he finishes in the top 3 7% of the time, 4th to 12th 25%, 13th to 24th 39%, and outside the top 24 29%. Across the simulated weeks he finishes QB18 at the median, QB10 - QB26 in the middle half, and QB5 - QB30 in the middle 80%. |
| 24 | QB | Malik WillisMIA Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 5% of the time, 4th to 12th 26%, 13th to 24th 43%, and outside the top 24 26%. Across the simulated weeks he finishes QB18 at the median, QB11 - QB25 in the middle half, and QB6 - QB29 in the middle 80%.
Top 3: 5%Top 12: 31%Top 24: 74% |
QB24 | 13.9 | Across the simulated weeks at the position, he finishes in the top 3 5% of the time, 4th to 12th 26%, 13th to 24th 43%, and outside the top 24 26%. Across the simulated weeks he finishes QB18 at the median, QB11 - QB25 in the middle half, and QB6 - QB29 in the middle 80%. |
| 25 | QB | Daniel JonesIND Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 5% of the time, 4th to 12th 24%, 13th to 24th 42%, and outside the top 24 29%. Across the simulated weeks he finishes QB19 at the median, QB11 - QB26 in the middle half, and QB5 - QB30 in the middle 80%.
Top 3: 5%Top 12: 30%Top 24: 71% |
QB25 | 13.5 | Across the simulated weeks at the position, he finishes in the top 3 5% of the time, 4th to 12th 24%, 13th to 24th 42%, and outside the top 24 29%. Across the simulated weeks he finishes QB19 at the median, QB11 - QB26 in the middle half, and QB5 - QB30 in the middle 80%. |
| 26 | QB | Cam WardTEN Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 5% of the time, 4th to 12th 23%, 13th to 24th 41%, and outside the top 24 31%. Across the simulated weeks he finishes QB20 at the median, QB12 - QB26 in the middle half, and QB6 - QB30 in the middle 80%.
Top 3: 5%Top 12: 28%Top 24: 69% |
QB26 | 13.1 | Across the simulated weeks at the position, he finishes in the top 3 5% of the time, 4th to 12th 23%, 13th to 24th 41%, and outside the top 24 31%. Across the simulated weeks he finishes QB20 at the median, QB12 - QB26 in the middle half, and QB6 - QB30 in the middle 80%. |
| 27 | QB | Aaron RodgersPIT Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 4% of the time, 4th to 12th 23%, 13th to 24th 43%, and outside the top 24 29%. Across the simulated weeks he finishes QB19 at the median, QB12 - QB26 in the middle half, and QB6 - QB30 in the middle 80%.
Top 3: 4%Top 12: 28%Top 24: 71% |
QB27 | 13.0 | Across the simulated weeks at the position, he finishes in the top 3 4% of the time, 4th to 12th 23%, 13th to 24th 43%, and outside the top 24 29%. Across the simulated weeks he finishes QB19 at the median, QB12 - QB26 in the middle half, and QB6 - QB30 in the middle 80%. |
| 28 | QB | Carson WentzMIN Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 5% of the time, 4th to 12th 22%, 13th to 24th 40%, and outside the top 24 34%. Across the simulated weeks he finishes QB20 at the median, QB12 - QB27 in the middle half, and QB6 - QB31 in the middle 80%.
Top 3: 5%Top 12: 26%Top 24: 66% |
QB28 | 12.7 | Across the simulated weeks at the position, he finishes in the top 3 5% of the time, 4th to 12th 22%, 13th to 24th 40%, and outside the top 24 34%. Across the simulated weeks he finishes QB20 at the median, QB12 - QB27 in the middle half, and QB6 - QB31 in the middle 80%. |
| 29 | QB | Jameis WinstonNYG Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 5% of the time, 4th to 12th 19%, 13th to 24th 37%, and outside the top 24 38%. Across the simulated weeks he finishes QB21 at the median, QB13 - QB28 in the middle half, and QB6 - QB31 in the middle 80%.
Top 3: 5%Top 12: 25%Top 24: 62% |
QB29 | 12.3 | Across the simulated weeks at the position, he finishes in the top 3 5% of the time, 4th to 12th 19%, 13th to 24th 37%, and outside the top 24 38%. Across the simulated weeks he finishes QB21 at the median, QB13 - QB28 in the middle half, and QB6 - QB31 in the middle 80%. |
| 30 | QB | Marcus MariotaWAS Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 3% of the time, 4th to 12th 14%, 13th to 24th 35%, and outside the top 24 49%. Across the simulated weeks he finishes QB24 at the median, QB17 - QB29 in the middle half, and QB9 - QB32 in the middle 80%.
Top 3: 3%Top 12: 16%Top 24: 51% |
QB30 | 10.3 | Across the simulated weeks at the position, he finishes in the top 3 3% of the time, 4th to 12th 14%, 13th to 24th 35%, and outside the top 24 49%. Across the simulated weeks he finishes QB24 at the median, QB17 - QB29 in the middle half, and QB9 - QB32 in the middle 80%. |
| 31 | QB | Cooper RushATL Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 1% of the time, 4th to 12th 7%, 13th to 24th 28%, and outside the top 24 64%. Across the simulated weeks he finishes QB27 at the median, QB22 - QB30 in the middle half, and QB14 - QB32 in the middle 80%.
Top 3: 1%Top 12: 7%Top 24: 36% |
QB31 | 7.8 | Across the simulated weeks at the position, he finishes in the top 3 1% of the time, 4th to 12th 7%, 13th to 24th 28%, and outside the top 24 64%. Across the simulated weeks he finishes QB27 at the median, QB22 - QB30 in the middle half, and QB14 - QB32 in the middle 80%. |
| 32 | QB | Tyson BagentCHIQQuestionable Drivers show how expected points change when each signal is removed. Effects can overlap, so they do not add up to the final projection.
Why isn’t matchup a separate factor?We don’t currently show matchup as a separate factor. We tested a standalone matchup adjustment built from historical defensive performance, and it did not improve predictions consistently out of sample.Matchup information is still reflected indirectly through our outside weekly projection and betting-market inputs — sportsbook odds and prediction markets — when available.We only show a separate factor when it adds reliable predictive value on its own. How to read
Across the simulated weeks at the position, he finishes in the top 3 2% of the time, 4th to 12th 7%, 13th to 24th 23%, and outside the top 24 68%. Across the simulated weeks he finishes QB29 at the median, QB22 - QB32 in the middle half, and QB13 - QB54 in the middle 80%.
Top 3: 2%Top 12: 9%Top 24: 32% |
QB32 | 7.1 | Across the simulated weeks at the position, he finishes in the top 3 2% of the time, 4th to 12th 7%, 13th to 24th 23%, and outside the top 24 68%. Across the simulated weeks he finishes QB29 at the median, QB22 - QB32 in the middle half, and QB13 - QB54 in the middle 80%. |
Simulated finishes
This board is ranked by projected fantasy points. The map shows where those same players finished at their own position across the 4,000 simulated weeks the projections came from.
Box: the middle half of a player's simulated finishes. Line: his median. Thin bar: the middle 80%. Two players whose boxes overlap are two players this board does not separate.
Week 3 QB outcome map: the positional finish each player reaches across the 4,000 simulated weeks, as a median, a middle half and a middle 80%. The board itself is ranked by projected fantasy points.
Projected rank
Median finish
136122436
-
QB1 Josh AllenQB8Josh Allen is ranked QB1 on projected fantasy points. Across the simulated weeks he finishes QB8 at the median, QB3 - QB16 in the middle half, and QB1 - QB25 in the middle 80%.
-
QB2 Tyler ShoughQB12Tyler Shough is ranked QB2 on projected fantasy points. Across the simulated weeks he finishes QB12 at the median, QB5 - QB21 in the middle half, and QB2 - QB27 in the middle 80%.
-
QB3 Bryce YoungQB12Bryce Young is ranked QB3 on projected fantasy points. Across the simulated weeks he finishes QB12 at the median, QB5 - QB21 in the middle half, and QB2 - QB28 in the middle 80%.
-
QB4 Patrick MahomesQB12Patrick Mahomes is ranked QB4 on projected fantasy points. Across the simulated weeks he finishes QB12 at the median, QB6 - QB20 in the middle half, and QB3 - QB27 in the middle 80%.
-
QB5 Jalen HurtsQB13Jalen Hurts is ranked QB5 on projected fantasy points. Across the simulated weeks he finishes QB13 at the median, QB6 - QB22 in the middle half, and QB2 - QB28 in the middle 80%.
-
QB6 Lamar JacksonQB13Lamar Jackson is ranked QB6 on projected fantasy points. Across the simulated weeks he finishes QB13 at the median, QB6 - QB22 in the middle half, and QB3 - QB28 in the middle 80%.
-
QB7 Jared GoffQB12Jared Goff is ranked QB7 on projected fantasy points. Across the simulated weeks he finishes QB12 at the median, QB6 - QB20 in the middle half, and QB3 - QB28 in the middle 80%.
-
QB8 C.J. StroudQB14C.J. Stroud is ranked QB8 on projected fantasy points. Across the simulated weeks he finishes QB14 at the median, QB7 - QB22 in the middle half, and QB3 - QB28 in the middle 80%.
-
QB9 Brock PurdyQB14Brock Purdy is ranked QB9 on projected fantasy points. Across the simulated weeks he finishes QB14 at the median, QB7 - QB22 in the middle half, and QB3 - QB29 in the middle 80%.
-
QB10 Dak PrescottQB14Dak Prescott is ranked QB10 on projected fantasy points. Across the simulated weeks he finishes QB14 at the median, QB7 - QB23 in the middle half, and QB3 - QB29 in the middle 80%.
-
QB11 Jordan LoveQB14Jordan Love is ranked QB11 on projected fantasy points. Across the simulated weeks he finishes QB14 at the median, QB7 - QB22 in the middle half, and QB4 - QB28 in the middle 80%.
-
QB12 Joe BurrowQB16Joe Burrow is ranked QB12 on projected fantasy points. Across the simulated weeks he finishes QB16 at the median, QB8 - QB25 in the middle half, and QB3 - QB30 in the middle 80%.
-
QB13 Trevor LawrenceQB16Trevor Lawrence is ranked QB13 on projected fantasy points. Across the simulated weeks he finishes QB16 at the median, QB8 - QB23 in the middle half, and QB4 - QB29 in the middle 80%.
-
QB14 Geno SmithQB15Geno Smith is ranked QB14 on projected fantasy points. Across the simulated weeks he finishes QB15 at the median, QB9 - QB23 in the middle half, and QB4 - QB28 in the middle 80%.
-
QB15 Baker MayfieldQB16Baker Mayfield is ranked QB15 on projected fantasy points. Across the simulated weeks he finishes QB16 at the median, QB9 - QB24 in the middle half, and QB4 - QB29 in the middle 80%.
-
QB16 Matthew StaffordQB16Matthew Stafford is ranked QB16 on projected fantasy points. Across the simulated weeks he finishes QB16 at the median, QB8 - QB24 in the middle half, and QB4 - QB30 in the middle 80%.
-
QB17 Deshaun WatsonQB17Deshaun Watson is ranked QB17 on projected fantasy points. Across the simulated weeks he finishes QB17 at the median, QB9 - QB24 in the middle half, and QB4 - QB29 in the middle 80%.
-
QB18 Drake MayeQB16Drake Maye is ranked QB18 on projected fantasy points. Across the simulated weeks he finishes QB16 at the median, QB9 - QB24 in the middle half, and QB4 - QB29 in the middle 80%.
-
QB19 Kirk CousinsQB17Kirk Cousins is ranked QB19 on projected fantasy points. Across the simulated weeks he finishes QB17 at the median, QB9 - QB25 in the middle half, and QB4 - QB30 in the middle 80%.
-
QB20 Jacoby BrissettQB17Jacoby Brissett is ranked QB20 on projected fantasy points. Across the simulated weeks he finishes QB17 at the median, QB9 - QB25 in the middle half, and QB5 - QB29 in the middle 80%.
-
QB21 Bo NixQB17Bo Nix is ranked QB21 on projected fantasy points. Across the simulated weeks he finishes QB17 at the median, QB9 - QB25 in the middle half, and QB4 - QB30 in the middle 80%.
-
QB22 Justin HerbertQB17Justin Herbert is ranked QB22 on projected fantasy points. Across the simulated weeks he finishes QB17 at the median, QB10 - QB24 in the middle half, and QB5 - QB29 in the middle 80%.
-
QB23 Drew LockQB18Drew Lock is ranked QB23 on projected fantasy points. Across the simulated weeks he finishes QB18 at the median, QB10 - QB26 in the middle half, and QB5 - QB30 in the middle 80%.
-
QB24 Malik WillisQB18Malik Willis is ranked QB24 on projected fantasy points. Across the simulated weeks he finishes QB18 at the median, QB11 - QB25 in the middle half, and QB6 - QB29 in the middle 80%.
Show QB25 - QB32
-
QB25 Daniel JonesQB19Daniel Jones is ranked QB25 on projected fantasy points. Across the simulated weeks he finishes QB19 at the median, QB11 - QB26 in the middle half, and QB5 - QB30 in the middle 80%.
-
QB26 Cam WardQB20Cam Ward is ranked QB26 on projected fantasy points. Across the simulated weeks he finishes QB20 at the median, QB12 - QB26 in the middle half, and QB6 - QB30 in the middle 80%.
-
QB27 Aaron RodgersQB19Aaron Rodgers is ranked QB27 on projected fantasy points. Across the simulated weeks he finishes QB19 at the median, QB12 - QB26 in the middle half, and QB6 - QB30 in the middle 80%.
-
QB28 Carson WentzQB20Carson Wentz is ranked QB28 on projected fantasy points. Across the simulated weeks he finishes QB20 at the median, QB12 - QB27 in the middle half, and QB6 - QB31 in the middle 80%.
-
QB29 Jameis WinstonQB21Jameis Winston is ranked QB29 on projected fantasy points. Across the simulated weeks he finishes QB21 at the median, QB13 - QB28 in the middle half, and QB6 - QB31 in the middle 80%.
-
QB30 Marcus MariotaQB24Marcus Mariota is ranked QB30 on projected fantasy points. Across the simulated weeks he finishes QB24 at the median, QB17 - QB29 in the middle half, and QB9 - QB32 in the middle 80%.
-
QB31 Cooper RushQB27Cooper Rush is ranked QB31 on projected fantasy points. Across the simulated weeks he finishes QB27 at the median, QB22 - QB30 in the middle half, and QB14 - QB32 in the middle 80%.
-
QB32 Tyson BagentQB29Tyson Bagent is ranked QB32 on projected fantasy points. Across the simulated weeks he finishes QB29 at the median, QB22 - QB32 in the middle half, and QB13 - QB54 in the middle 80%.
Simulated positional finish. The scale is compressed to the right, so equal distances are not equal numbers of places.
The axis compresses lower-ranked finishes so the top of the board stays readable. Use the printed ranges for exact finish positions.
How these are made
This board was built from 2026 usage for 67% of the board, draft position for 100%, an outside projection for 59%, sportsbook props for 11%, an official status for 81%, over 4,000 simulated weeks. The projection and the odds are both unconditional: for a player carrying an official Questionable or Doubtful, the weeks he sits are in the average and in the odds. Ranks follow projected points for Half PPR, 1QB.
Availability is a model input: players ruled out, inactive, on a reserve list or without a game are left off the weekly board, and an official Questionable or Doubtful is simulated as playing 71% and 7% of the time. Status evidence was captured September 24, 2026 22:48 UTC. Model rankings-v1.4.0, board captured 2026-09-24.
LMD Rankings are produced by Let's Mock Draft's own statistical and market model. They are not analyst rankings: consensus boards are not an input, and nothing is edited by hand.