The same production stack priced under 4 course readings, against the bookmaker's pre-stage line. Each scenario is one opinion about what kind of day this is; the results section settles the argument.
The fork
Course profile is an input the model can't see from a map — it's set from reference races (or overridden by hand). Different readings redistribute the same probability mass across rider types.
The board
Blend win probability per scenario (cascade × LambdaRank, DNF-adjusted at the pricing layer), next to the devigged Kambi (Unibet) win line captured the morning before the stage. "Max edge" compares the most favourable scenario against the book — treat it as scenario-conditional, not a free bet.
Read "Max edge" as directional, not as a price. The blend is a set of per-rider marginals and does not sum to one book — across this field each reading totals GC engines 139%, Pure power 152%, Hills bite 127%, Hindsight (post-hoc) 135% — while the devigged book column sums to exactly 100%. Putting the two side by side flatters the model, and Max edge is inflated by about that factor: divide a rider's probability by his reading's total before calling anything an edge. Of our heads only the LambdaRank one is natively book-preserving.
| Rider | Team | GC engines | Pure power | Hills bite | Hindsight (post-hoc) | Book odds | Book devig | Max edge |
|---|---|---|---|---|---|---|---|---|
| Pogacar Tadej | UAE Team Emirates-XRG | 53.1% | 53.6% | 54.3% | 54.0% | 2.75 | 29.2% | +25.1pp |
| Evenepoel Remco | Red Bull-BORA-hansgrohe | 21.0% | 21.4% | 21.1% | 19.2% | 1.50 | 53.5% | -32.1pp |
| Tarling Joshua | Netcompany INEOS | 11.3% | 13.1% | 5.3% | 9.1% | 51.00 | 1.6% | +11.5pp |
| Lipowitz Florian | Red Bull-BORA-hansgrohe | 7.3% | 11.9% | 7.0% | 10.3% | 201.00 | 0.4% | +11.5pp |
| Seixas Paul | Decathlon CMA CGM Team | 9.7% | 10.4% | 8.1% | 8.2% | 101.00 | 0.8% | +9.6pp |
| Ganna Filippo | Netcompany INEOS | 3.9% | 4.6% | 2.2% | 3.2% | 15.00 | 5.3% | -0.7pp |
| Ayuso Juan | Lidl-Trek | 3.9% | 4.5% | 2.8% | 3.9% | 101.00 | 0.8% | +3.7pp |
| del Toro Isaac | UAE Team Emirates-XRG | 1.6% | 1.7% | 3.0% | 1.7% | 201.00 | 0.4% | +2.6pp |
| Vacek Mathias | Lidl-Trek | 2.7% | 2.9% | 1.6% | 2.2% | 201.00 | 0.4% | +2.5pp |
| Foss Tobias | Netcompany INEOS | 2.5% | 2.6% | 1.9% | 2.1% | 301.00 | 0.3% | +2.3pp |
| Arensman Thymen | Netcompany INEOS | 2.3% | 2.0% | 1.9% | 1.5% | 501.00 | 0.2% | +2.1pp |
| Pedersen Mads | Lidl-Trek | 0.9% | 2.0% | 0.6% | 1.6% | 751.00 | 0.1% | +1.9pp |
| Wærenskjold Søren | Uno-X Mobility | 1.0% | 1.6% | 0.7% | 1.3% | 1001.00 | 0.1% | +1.5pp |
| Vauquelin Kévin | Netcompany INEOS | 1.2% | 1.3% | 0.9% | 1.2% | 301.00 | 0.3% | +1.0pp |
| Jorgenson Matteo | Team Visma | Lease a Bike | 1.1% | 1.2% | 1.0% | 0.9% | 201.00 | 0.4% | +0.8pp |
| McNulty Brandon | UAE Team Emirates-XRG | 1.1% | 1.0% | 0.9% | 1.0% | 201.00 | 0.4% | +0.7pp |
| Hoole Daan | Decathlon CMA CGM Team | 0.8% | 1.1% | 0.4% | 0.8% | 601.00 | 0.1% | +0.9pp |
| # | Rider | Blend | Cascade | LambdaRank | Podium | Top 15 | Book odds |
|---|---|---|---|---|---|---|---|
| 1 | Pogacar Tadej | 53.1% | 53.8% | 38.0% | 81.1% | 90.8% | 2.75 |
| 2 | Evenepoel Remco | 21.0% | 17.2% | 16.0% | 49.3% | 82.5% | 1.50 |
| 3 | Tarling Joshua | 11.3% | 13.2% | 7.6% | 29.0% | 66.6% | 51.00 |
| 4 | Seixas Paul | 9.7% | 11.5% | 6.5% | 23.7% | 67.6% | 101.00 |
| 5 | Lipowitz Florian | 7.3% | 3.0% | 6.3% | 13.3% | 68.4% | 201.00 |
| 6 | Ayuso Juan | 3.9% | 1.9% | 3.3% | 7.9% | 53.6% | 101.00 |
| 7 | Ganna Filippo | 3.9% | 5.2% | 2.5% | 18.9% | 54.6% | 15.00 |
| 8 | Vacek Mathias | 2.7% | 1.9% | 2.2% | 7.9% | 51.1% | 201.00 |
| 9 | Foss Tobias | 2.5% | 1.2% | 2.1% | 5.3% | 30.1% | 301.00 |
| 10 | Arensman Thymen | 2.3% | 1.4% | 1.8% | 5.5% | 33.6% | 501.00 |
| 11 | del Toro Isaac | 1.6% | 1.2% | 1.2% | 4.4% | 33.2% | 201.00 |
| 12 | Vauquelin Kévin | 1.2% | 1.2% | 0.9% | 3.7% | 32.1% | 301.00 |
| 13 | McNulty Brandon | 1.1% | 1.4% | 0.7% | 4.7% | 32.9% | 201.00 |
| 14 | Jorgenson Matteo | 1.1% | 1.2% | 0.7% | 4.6% | 23.8% | 201.00 |
| 15 | Wærenskjold Søren | 1.0% | 1.4% | 0.6% | 4.6% | 36.8% | 1001.00 |
| 16 | Skjelmose Mattias | 1.0% | 1.1% | 0.7% | 5.6% | 49.8% | 501.00 |
| 17 | Gee Derek | 0.9% | 0.7% | 0.7% | 4.8% | 26.8% | 601.00 |
| 18 | Pedersen Mads | 0.9% | 1.3% | 0.5% | 4.0% | 28.5% | 751.00 |
| 19 | Hoole Daan | 0.8% | 1.1% | 0.5% | 6.5% | 41.2% | 601.00 |
| 20 | Asgreen Kasper | 0.7% | 1.6% | 0.3% | 3.3% | 31.2% | 751.00 |
| 21 | Vermeersch Florian | 0.7% | 1.3% | 0.4% | 4.2% | 36.6% | 751.00 |
| 22 | Van Wilder Ilan | 0.7% | 0.7% | 0.5% | 3.6% | 35.1% | 401.00 |
| 23 | Martinez Lenny | 0.7% | 1.4% | 0.3% | 4.4% | 35.9% | 601.00 |
| 24 | van der Poel Mathieu | 0.6% | 1.0% | 0.3% | 10.4% | 26.0% | 201.00 |
| 25 | Armirail Bruno | 0.6% | 0.5% | 0.4% | 1.6% | 26.5% | 301.00 |
| 26 | Wright Fred | 0.5% | 0.6% | 0.4% | 3.8% | 24.8% | 1501.00 |
| 27 | Grégoire Romain | 0.5% | 0.4% | 0.4% | 2.5% | 14.7% | 1001.00 |
| 28 | Castrillo Pablo | 0.4% | 0.5% | 0.2% | 1.5% | 15.8% | 1501.00 |
| 29 | Affini Edoardo | 0.3% | 0.4% | 0.2% | 3.2% | 34.5% | 601.00 |
| 30 | Healy Ben | 0.3% | 0.7% | 0.1% | 1.5% | 8.5% | 501.00 |
| # | Rider | Blend | Cascade | LambdaRank | Podium | Top 15 | Book odds |
|---|---|---|---|---|---|---|---|
| 1 | Pogacar Tadej | 53.6% | 53.7% | 35.3% | 81.3% | 90.8% | 2.75 |
| 2 | Evenepoel Remco | 21.4% | 15.9% | 15.4% | 48.3% | 80.7% | 1.50 |
| 3 | Tarling Joshua | 13.1% | 13.4% | 8.5% | 30.0% | 67.1% | 51.00 |
| 4 | Lipowitz Florian | 11.9% | 6.6% | 9.2% | 23.1% | 74.7% | 201.00 |
| 5 | Seixas Paul | 10.4% | 11.7% | 6.5% | 22.8% | 65.1% | 101.00 |
| 6 | Ganna Filippo | 4.6% | 5.6% | 2.8% | 21.2% | 56.7% | 15.00 |
| 7 | Ayuso Juan | 4.5% | 1.9% | 3.6% | 8.2% | 51.2% | 101.00 |
| 8 | Vacek Mathias | 2.9% | 2.1% | 2.1% | 8.9% | 52.0% | 201.00 |
| 9 | Foss Tobias | 2.6% | 1.1% | 2.1% | 5.0% | 29.5% | 301.00 |
| 10 | Pedersen Mads | 2.0% | 5.1% | 0.5% | 10.7% | 49.8% | 751.00 |
| 11 | Arensman Thymen | 2.0% | 1.0% | 1.5% | 4.7% | 31.1% | 501.00 |
| 12 | del Toro Isaac | 1.7% | 1.2% | 1.3% | 4.4% | 31.5% | 201.00 |
| 13 | Wærenskjold Søren | 1.6% | 3.6% | 0.5% | 12.8% | 35.4% | 1001.00 |
| 14 | Vauquelin Kévin | 1.3% | 1.0% | 0.9% | 4.1% | 30.1% | 301.00 |
| 15 | Jorgenson Matteo | 1.2% | 1.5% | 0.7% | 5.2% | 25.3% | 201.00 |
| 16 | Hoole Daan | 1.1% | 1.4% | 0.6% | 7.6% | 45.9% | 601.00 |
| 17 | McNulty Brandon | 1.0% | 1.1% | 0.6% | 4.4% | 34.1% | 201.00 |
| 18 | Skjelmose Mattias | 1.0% | 1.0% | 0.6% | 5.8% | 49.5% | 501.00 |
| 19 | Vermeersch Florian | 0.9% | 1.8% | 0.4% | 5.3% | 36.8% | 751.00 |
| 20 | Gee Derek | 0.9% | 1.0% | 0.6% | 4.0% | 23.9% | 601.00 |
| 21 | van der Poel Mathieu | 0.8% | 1.5% | 0.3% | 9.8% | 25.2% | 201.00 |
| 22 | Armirail Bruno | 0.7% | 0.6% | 0.5% | 3.8% | 35.4% | 301.00 |
| 23 | Van Wilder Ilan | 0.7% | 0.7% | 0.5% | 3.6% | 35.0% | 401.00 |
| 24 | Martinez Lenny | 0.6% | 1.4% | 0.2% | 4.0% | 36.7% | 601.00 |
| 25 | Wright Fred | 0.5% | 0.8% | 0.3% | 3.4% | 24.4% | 1501.00 |
| 26 | Asgreen Kasper | 0.5% | 0.7% | 0.3% | 4.3% | 32.4% | 751.00 |
| 27 | Grégoire Romain | 0.5% | 0.4% | 0.3% | 2.5% | 14.0% | 1001.00 |
| 28 | Affini Edoardo | 0.4% | 0.5% | 0.3% | 4.2% | 40.9% | 601.00 |
| 29 | Matthews Michael | 0.4% | 0.9% | 0.1% | 2.3% | 23.8% | 1001.00 |
| 30 | Castrillo Pablo | 0.4% | 0.5% | 0.2% | 1.9% | 16.4% | 1501.00 |
| # | Rider | Blend | Cascade | LambdaRank | Podium | Top 15 | Book odds |
|---|---|---|---|---|---|---|---|
| 1 | Pogacar Tadej | 54.3% | 56.4% | 42.3% | 79.8% | 90.4% | 2.75 |
| 2 | Evenepoel Remco | 21.1% | 15.8% | 18.0% | 48.3% | 82.7% | 1.50 |
| 3 | Seixas Paul | 8.1% | 8.2% | 6.3% | 17.5% | 64.0% | 101.00 |
| 4 | Lipowitz Florian | 7.0% | 3.4% | 6.4% | 14.0% | 69.2% | 201.00 |
| 5 | Tarling Joshua | 5.3% | 4.9% | 4.3% | 16.2% | 54.5% | 51.00 |
| 6 | del Toro Isaac | 3.0% | 2.2% | 2.6% | 9.6% | 49.8% | 201.00 |
| 7 | Ayuso Juan | 2.8% | 2.0% | 2.4% | 7.4% | 51.2% | 101.00 |
| 8 | Ganna Filippo | 2.2% | 2.7% | 1.6% | 12.1% | 45.1% | 15.00 |
| 9 | Arensman Thymen | 1.9% | 1.5% | 1.6% | 5.6% | 30.5% | 501.00 |
| 10 | Foss Tobias | 1.9% | 1.1% | 1.7% | 4.1% | 27.4% | 301.00 |
| 11 | Vacek Mathias | 1.6% | 1.7% | 1.2% | 7.2% | 44.0% | 201.00 |
| 12 | Gee Derek | 1.0% | 0.9% | 0.8% | 5.1% | 25.3% | 601.00 |
| 13 | Jorgenson Matteo | 1.0% | 0.7% | 0.8% | 2.7% | 24.8% | 201.00 |
| 14 | Vauquelin Kévin | 0.9% | 1.0% | 0.7% | 3.6% | 28.4% | 301.00 |
| 15 | Skjelmose Mattias | 0.9% | 0.7% | 0.7% | 3.9% | 39.0% | 501.00 |
| 16 | McNulty Brandon | 0.9% | 1.2% | 0.6% | 4.4% | 29.7% | 201.00 |
| 17 | Martinez Lenny | 0.8% | 0.9% | 0.6% | 3.2% | 35.0% | 601.00 |
| 18 | Wærenskjold Søren | 0.7% | 1.6% | 0.3% | 8.4% | 22.6% | 1001.00 |
| 19 | Van Wilder Ilan | 0.7% | 0.8% | 0.5% | 2.9% | 34.5% | 401.00 |
| 20 | van der Poel Mathieu | 0.7% | 1.3% | 0.4% | 9.3% | 21.0% | 201.00 |
| 21 | Pedersen Mads | 0.6% | 0.9% | 0.4% | 3.4% | 22.9% | 751.00 |
| 22 | Grégoire Romain | 0.5% | 0.5% | 0.4% | 3.2% | 14.7% | 1001.00 |
| 23 | Hoole Daan | 0.4% | 1.3% | 0.1% | 2.0% | 21.7% | 601.00 |
| 24 | Pidcock Tom | 0.4% | 0.8% | 0.2% | 4.4% | 30.6% | 1501.00 |
| 25 | Johannessen Tobias Halland | 0.3% | 0.3% | 0.3% | 2.4% | 18.5% | 2501.00 |
| 26 | Castrillo Pablo | 0.3% | 0.4% | 0.2% | 1.5% | 14.8% | 1501.00 |
| 27 | Vermeersch Florian | 0.3% | 0.7% | 0.2% | 2.9% | 18.7% | 751.00 |
| 28 | Healy Ben | 0.3% | 0.7% | 0.1% | 1.5% | 7.5% | 501.00 |
| 29 | Bernal Egan | 0.3% | 0.8% | 0.1% | 1.5% | 11.9% | 2501.00 |
| 30 | Armirail Bruno | 0.3% | 0.4% | 0.2% | 1.0% | 14.1% | 301.00 |
| # | Rider | Blend | Cascade | LambdaRank | Podium | Top 15 | Book odds |
|---|---|---|---|---|---|---|---|
| 1 | Pogacar Tadej | 54.0% | 51.1% | 40.8% | 83.2% | 92.3% | 2.75 |
| 2 | Evenepoel Remco | 19.2% | 14.9% | 15.4% | 51.6% | 83.2% | 1.50 |
| 3 | Lipowitz Florian | 10.3% | 5.3% | 8.9% | 23.0% | 75.9% | 201.00 |
| 4 | Tarling Joshua | 9.1% | 8.7% | 6.8% | 22.4% | 57.7% | 51.00 |
| 5 | Seixas Paul | 8.2% | 9.2% | 5.8% | 18.4% | 65.8% | 101.00 |
| 6 | Ayuso Juan | 3.9% | 1.9% | 3.4% | 8.4% | 51.6% | 101.00 |
| 7 | Ganna Filippo | 3.2% | 3.4% | 2.3% | 13.5% | 45.7% | 15.00 |
| 8 | Vacek Mathias | 2.2% | 1.7% | 1.7% | 6.5% | 41.6% | 201.00 |
| 9 | Foss Tobias | 2.1% | 1.0% | 1.8% | 4.1% | 28.3% | 301.00 |
| 10 | del Toro Isaac | 1.7% | 1.3% | 1.4% | 4.9% | 34.9% | 201.00 |
| 11 | Pedersen Mads | 1.6% | 4.9% | 0.4% | 10.2% | 45.4% | 751.00 |
| 12 | Arensman Thymen | 1.5% | 1.1% | 1.2% | 4.8% | 29.6% | 501.00 |
| 13 | Wærenskjold Søren | 1.3% | 3.5% | 0.4% | 12.2% | 31.9% | 1001.00 |
| 14 | Vauquelin Kévin | 1.2% | 1.1% | 0.9% | 4.9% | 32.3% | 301.00 |
| 15 | McNulty Brandon | 1.0% | 1.3% | 0.7% | 5.0% | 34.7% | 201.00 |
| 16 | Skjelmose Mattias | 1.0% | 1.2% | 0.7% | 6.1% | 51.6% | 501.00 |
| 17 | Jorgenson Matteo | 0.9% | 0.5% | 0.8% | 2.3% | 23.7% | 201.00 |
| 18 | Gee Derek | 0.8% | 1.0% | 0.5% | 4.9% | 25.3% | 601.00 |
| 19 | Hoole Daan | 0.8% | 1.2% | 0.4% | 6.8% | 39.3% | 601.00 |
| 20 | van der Poel Mathieu | 0.7% | 2.0% | 0.2% | 11.5% | 30.7% | 201.00 |
| 21 | Vermeersch Florian | 0.7% | 1.5% | 0.3% | 5.0% | 36.6% | 751.00 |
| 22 | Asgreen Kasper | 0.6% | 1.6% | 0.3% | 2.9% | 28.7% | 751.00 |
| 23 | Van Wilder Ilan | 0.6% | 0.7% | 0.4% | 4.2% | 36.8% | 401.00 |
| 24 | Armirail Bruno | 0.6% | 0.7% | 0.5% | 4.1% | 33.3% | 301.00 |
| 25 | Martinez Lenny | 0.6% | 1.6% | 0.2% | 5.0% | 39.5% | 601.00 |
| 26 | Grégoire Romain | 0.4% | 0.4% | 0.3% | 2.4% | 14.9% | 1001.00 |
| 27 | Wright Fred | 0.4% | 0.6% | 0.3% | 3.0% | 21.5% | 1501.00 |
| 28 | Castrillo Pablo | 0.3% | 0.6% | 0.2% | 1.6% | 15.5% | 1501.00 |
| 29 | Affini Edoardo | 0.3% | 0.3% | 0.2% | 2.9% | 29.6% | 601.00 |
| 30 | Matthews Michael | 0.3% | 0.8% | 0.1% | 2.2% | 19.5% | 1001.00 |
Angle
There is no break to price on a time trial, so the conditional this table answers is different: P(win | the winner is NOT in the GC top 9 on actual standings after stage 15). The book gives that pool 14% of the unconditional market, almost all of it Ganna; our GC-engines read holds 41% raw blend mass outside the GC nine. Ineos' time-trial unit dominates the table — Tarling, Ganna, Foss, Arensman and Vauquelin are five of the first six rows — with the Lidl-Trek and UAE rouleurs (Vacek, McNulty, Vermeersch) behind. Read it as the race for third: on every one of our boards the two GC leaders hold over 70% of the win mass between them.
| Rider | Team | GC engines win p | P(win | not GC top 9) | Book odds | Book devig |
|---|---|---|---|---|---|
| Tarling Joshua | Netcompany INEOS | 11.3% | 27.6% | 51 | 1.6% |
| Ganna Filippo | Netcompany INEOS | 3.9% | 9.6% | 15 | 5.3% |
| Vacek Mathias | Lidl-Trek | 2.7% | 6.7% | 201 | 0.4% |
| Foss Tobias | Netcompany INEOS | 2.5% | 6.2% | 301 | 0.3% |
| Arensman Thymen | Netcompany INEOS | 2.3% | 5.6% | 501 | 0.2% |
| Vauquelin Kévin | Netcompany INEOS | 1.2% | 3.1% | 301 | 0.3% |
| McNulty Brandon | UAE Team Emirates-XRG | 1.1% | 2.7% | 201 | 0.4% |
| Jorgenson Matteo | Team Visma | Lease a Bike | 1.1% | 2.7% | 201 | 0.4% |
| Wærenskjold Søren | Uno-X Mobility | 1.0% | 2.4% | 1001 | 0.1% |
| Gee Derek | Lidl-Trek | 0.9% | 2.2% | 601 | 0.1% |
| Pedersen Mads | Lidl-Trek | 0.9% | 2.1% | 751 | 0.1% |
| Hoole Daan | Decathlon CMA CGM Team | 0.8% | 1.9% | 601 | 0.1% |
| Asgreen Kasper | EF Education-EasyPost | 0.7% | 1.8% | 751 | 0.1% |
| Vermeersch Florian | UAE Team Emirates-XRG | 0.7% | 1.8% | 751 | 0.1% |
Angle
Shortest book prices for the stage, with our blend win probability (and rank) under each scenario. 'free' = outside the excluded GC pool (top 9 by the latest GC-model blend (tdf_gc.json)), i.e. the type the peloton might let up the road.
| Rider | Book odds | Devig | GC engines | Pure power | Hills bite | GC status |
|---|---|---|---|---|---|---|
| Evenepoel Remco | 1.50 | 53.5% | 21.0% (#2) | 21.4% (#2) | 21.1% (#2) | GC #3 |
| Pogacar Tadej | 2.75 | 29.2% | 53.1% (#1) | 53.6% (#1) | 54.3% (#1) | GC #1 |
| Ganna Filippo | 15.00 | 5.3% | 3.9% (#7) | 4.6% (#6) | 2.2% (#8) | free |
| Tarling Joshua | 51.00 | 1.6% | 11.3% (#3) | 13.1% (#3) | 5.3% (#5) | free |
| Seixas Paul | 101.00 | 0.8% | 9.7% (#4) | 10.4% (#5) | 8.1% (#3) | GC #4 |
| Ayuso Juan | 101.00 | 0.8% | 3.9% (#6) | 4.5% (#7) | 2.8% (#7) | GC #6 |
| Lipowitz Florian | 201.00 | 0.4% | 7.3% (#5) | 11.9% (#4) | 7.0% (#4) | GC #2 |
| Vacek Mathias | 201.00 | 0.4% | 2.7% (#8) | 2.9% (#8) | 1.6% (#11) | free |
| del Toro Isaac | 201.00 | 0.4% | 1.6% (#11) | 1.7% (#12) | 3.0% (#6) | GC #5 |
| Jorgenson Matteo | 201.00 | 0.4% | 1.1% (#14) | 1.2% (#15) | 1.0% (#13) | free |
Angle
Model column is the GC-engines read RENORMALIZED over the field (blend / 139.1%), directly comparable to the devigged book. One disagreement dwarfs the rest and it is the day's headline: the book makes Evenepoel the 1.5 favourite (53.5%) with Pogacar at 2.8 (29.2%); every one of our reads inverts that — Pogacar ~53% raw to Evenepoel's ~21%. The book is pricing the man who just beat Pogacar up an 11 km wall, the reigning world TT champion, on a course with a real climb in it. The model is pricing two seasons of head-to-head dominance everywhere else, and it does not know Solaison happened — its snapshot features end at the Dauphine and its within-race state carries standings, not yesterday's watts. After stage 15 settled a 16x disagreement our way, this is the sharpest test yet of the same prior in the opposite direction: this time WE are the ones fading fresh evidence for structural form. Below the pair, the pattern is the stage-15 one again — the model spreads real mass to the strong GC men the book has at three digits — Seixas (101), Lipowitz (201), Ayuso (101) — and to Tarling at 51, while among the book's own favourites it is colder only on Ganna.
| Rider | Team | Model (renormalized) | Book devig | Book odds | Difference |
|---|---|---|---|---|---|
| Pogacar Tadej | UAE Team Emirates-XRG | 38.2% | 29.2% | 2.75 | +9.0pp |
| Tarling Joshua | Netcompany INEOS | 8.1% | 1.6% | 51 | +6.5pp |
| Seixas Paul | Decathlon CMA CGM Team | 7.0% | 0.8% | 101 | +6.2pp |
| Lipowitz Florian | Red Bull-BORA-hansgrohe | 5.3% | 0.4% | 201 | +4.9pp |
| Ayuso Juan | Lidl-Trek | 2.8% | 0.8% | 101 | +2.0pp |
| Vacek Mathias | Lidl-Trek | 2.0% | 0.4% | 201 | +1.6pp |
| Ganna Filippo | Netcompany INEOS | 2.8% | 5.3% | 15 | -2.5pp |
| Evenepoel Remco | Red Bull-BORA-hansgrohe | 15.1% | 53.5% | 1.5 | -38.4pp |
After the stage
| Pos | Rider | Team | GC engines win p | Pure power win p | Hills bite win p | Hindsight (post-hoc) win p |
|---|---|---|---|---|---|---|
| 1 | Evenepoel Remco | Red Bull-BORA-hansgrohe | 21.0% | 21.4% | 21.1% | 19.2% |
| 2 | Pogacar Tadej | UAE Team Emirates-XRG | 53.1% | 53.6% | 54.3% | 54.0% |
| 3 | Skjelmose Mattias | Lidl-Trek | 1.0% | 1.0% | 0.9% | 1.0% |
| 4 | Seixas Paul | Decathlon CMA CGM Team | 9.7% | 10.4% | 8.1% | 8.2% |
| 5 | del Toro Isaac | UAE Team Emirates-XRG | 1.6% | 1.7% | 3.0% | 1.7% |
| 6 | Cattaneo Mattia | Red Bull-BORA-hansgrohe | 0.1% | 0.1% | 0.0% | 0.1% |
| 7 | Arensman Thymen | Netcompany INEOS Cycling Team | 2.3% | 2.0% | 1.9% | 1.5% |
| 8 | Armirail Bruno | Team Visma | Lease a Bike | 0.6% | 0.7% | 0.3% | 0.6% |
| 9 | Castrillo Pablo | Movistar Team | 0.4% | 0.4% | 0.3% | 0.3% |
| 10 | Tarling Joshua | Netcompany INEOS Cycling Team | 11.3% | 13.1% | 5.3% | 9.1% |
| Scenario | Blend win p | Book devig |
|---|---|---|
| GC engines | 20.95% | 53.49% |
| Pure power | 21.35% | 53.49% |
| Hills bite | 21.14% | 53.49% |
| Hindsight (post-hoc) | 19.21% | 53.49% |
Reflection
The road picked the branch none of our boards led with: Evenepoel won the time trial by 28 seconds over Pogacar, and the book's 1.50 was right. All three reads gave the winner ~21% (GC engines 20.95%, Pure power 21.35%, Hills bite 21.14%) against a devigged 53.5%, and all three made Pogacar the ~53% pick. After stage 15 settled a 16x disagreement our way — same two riders, same question, our side — the sharpest test of the Tour went the other way: the book was pricing Solaison's watts, we were pricing two seasons of head-to-heads, and the watts won.
The hindsight run closes the usual escape hatch. Rebuilt from the five DB time trials whose finish-gap curves most resemble today's (Perugia'24, Peyragudes'25, Giro'26 s10, Alicante'22, Combloux'23 — GC-decided TTs, exactly today's shape), the board is unchanged: Pogacar 54.0%, Evenepoel 19.2%. Every ITT carries the identical profile fingerprint, so the only lever hindsight has is finish structure, and finish structure wasn't the constraint. The miss lives in the rider prior: the model's snapshot features end at the Dauphine and its within-race state carries standings, not yesterday's performances. It literally cannot know Solaison happened.
Put stages 15 and 16 side by side and the deeper finding appears: they are the same property of the model wearing opposite signs. Our Evenepoel number barely moved across the two days (~10% on Solaison, ~15% renormalized here) while the book swung from 0.8% to 53.5%. A static form prior fades archetype stories AND fades fresh physical evidence — profitable on Saturday when the book's 'TT-shaped men can't win 9% walls' story lost, costly on Tuesday when the book's update was simply correct. Carry-forward rule: when a 'book-knows-news' gap traces to a direct on-road demonstration on comparable terrain days earlier — not narrative, watts — discount our side of the disagreement.
Below the big two, the quieter reads mostly aged well. The road agreed with the Hills bite direction — climbers swept the top five (Skjelmose 3rd at 1:04, Seixas 4th, del Toro 5th) and the pure engines were buried (Tarling 10th at 2:02, Ganna 11th) — and Hills bite was the only board that had demoted Tarling (11.3→5.3%) and promoted del Toro, so the TT axis does re-sort when pushed; the probe answered yes. Tarling at 11.3% was the biggest single miss below the top two, while the model's coldness on Ganna (3.9% vs book 5.3%) was vindicated. Seixas 4th at 101 book odds is another mark for the standing edge. And the day's structural news travels to tomorrow: Lipowitz crashed out of GC #2, the field is down to 166, and the GC cliff now sits after Pidcock in 8th (+12:58), with 10 minutes of air below him.