The same production stack priced under 2 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 Aitana shatters 68%, Cohesive GC group 72% — 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 | Aitana shatters | Cohesive GC group | Book odds | Book devig | Max edge |
|---|---|---|---|---|---|---|
| Roglic Primoz | Red Bull-BORA-hansgrohe | 8.4% | 10.0% | 17.00 | 3.8% | +6.2pp |
| Gall Felix | Decathlon CMA CGM Team | 7.9% | 4.6% | 2.60 | 24.8% | -16.9pp |
| Onley Oscar | Netcompany INEOS | 6.6% | 6.7% | 11.00 | 5.9% | +0.8pp |
| Mas Enric | Movistar Team | 4.4% | 2.7% | 6.50 | 9.9% | -5.6pp |
| Carapaz Richard | EF Education-EasyPost | 3.3% | 3.9% | 34.00 | 1.9% | +2.0pp |
| Vine Jay | UAE Team Emirates-XRG | 2.1% | 3.1% | 4.75 | 13.6% | -10.4pp |
| Skjelmose Mattias | Lidl-Trek | 2.1% | 3.1% | 101.00 | 0.6% | +2.5pp |
| Johannessen Tobias Halland | Uno-X Mobility | 2.7% | 2.6% | 21.00 | 3.1% | -0.3pp |
| van Aert Wout | Team Visma | Lease a Bike | 0.4% | 2.5% | n/a | – | – |
| Kuss Sepp | Team Visma | Lease a Bike | 2.3% | 1.0% | 26.00 | 2.5% | -0.2pp |
| Almeida Joao | UAE Team Emirates-XRG | 1.3% | 2.0% | 51.00 | 1.3% | +0.7pp |
| Nordhagen Jørgen | Team Visma | Lease a Bike | 1.6% | 1.0% | 151.00 | 0.4% | +1.2pp |
| Romeo Ivan | Movistar Team | 1.2% | 1.6% | 401.00 | 0.2% | +1.5pp |
| Rodriguez Carlos | Netcompany INEOS | 1.4% | 0.9% | 151.00 | 0.4% | +1.0pp |
| Riccitello Matthew | Decathlon CMA CGM Team | 1.3% | 0.7% | 81.00 | 0.8% | +0.5pp |
| Tejada Harold | XDS Astana Team | 1.3% | 1.3% | 101.00 | 0.6% | +0.7pp |
| Bisiaux Léo | Decathlon CMA CGM Team | 0.9% | 1.3% | 151.00 | 0.4% | +0.8pp |
| Widar Jarno | Lotto-Intermarché | 1.3% | 1.1% | 51.00 | 1.3% | +0.0pp |
| Berthet Clément | Groupama-FDJ United | 0.9% | 1.2% | 101.00 | 0.6% | +0.6pp |
| Tulett Ben | Team Visma | Lease a Bike | 0.8% | 1.2% | 151.00 | 0.4% | +0.8pp |
| Bilbao Pello | Bahrain Victorious | 1.0% | 1.2% | 201.00 | 0.3% | +0.8pp |
| Fortunato Lorenzo | XDS Astana Team | 0.7% | 0.5% | 23.00 | 2.8% | -2.1pp |
| Torres Pablo | UAE Team Emirates-XRG | 0.3% | 0.3% | 19.00 | 3.4% | -3.1pp |
| # | Rider | Blend | Cascade | LambdaRank | Podium | Top 15 | Book odds |
|---|---|---|---|---|---|---|---|
| 1 | Roglic Primoz | 8.4% | 7.0% | 12.8% | 25.6% | 76.4% | 17.00 |
| 2 | Gall Felix | 7.9% | 7.2% | 11.9% | 33.8% | 83.5% | 2.60 |
| 3 | Onley Oscar | 6.6% | 6.9% | 9.6% | 29.7% | 80.2% | 11.00 |
| 4 | Mas Enric | 4.4% | 2.2% | 7.0% | 22.7% | 77.5% | 6.50 |
| 5 | Carapaz Richard | 3.3% | 2.6% | 5.1% | 14.7% | 73.5% | 34.00 |
| 6 | Johannessen Tobias Halland | 2.7% | 1.8% | 4.3% | 7.6% | 35.4% | 21.00 |
| 7 | Kuss Sepp | 2.3% | 2.4% | 3.3% | 17.2% | 63.7% | 26.00 |
| 8 | Skjelmose Mattias | 2.1% | 2.2% | 3.1% | 9.9% | 67.4% | 101.00 |
| 9 | Vine Jay | 2.1% | 2.3% | 3.0% | 5.5% | 15.0% | 4.75 |
| 10 | Nordhagen Jørgen | 1.6% | 2.1% | 2.3% | 5.7% | 20.5% | 151.00 |
| 11 | Rodriguez Carlos | 1.4% | 0.8% | 2.3% | 4.9% | 20.1% | 151.00 |
| 12 | Riccitello Matthew | 1.3% | 1.4% | 1.9% | 5.3% | 36.8% | 81.00 |
| 13 | Almeida Joao | 1.3% | 3.3% | 1.4% | 7.7% | 19.8% | 51.00 |
| 14 | Widar Jarno | 1.3% | 1.2% | 1.9% | 5.8% | 41.1% | 51.00 |
| 15 | Tejada Harold | 1.3% | 2.0% | 1.7% | 8.6% | 54.7% | 101.00 |
| 16 | Romeo Ivan | 1.2% | 0.8% | 1.9% | 3.9% | 15.9% | 401.00 |
| 17 | Bilbao Pello | 1.0% | 1.0% | 1.4% | 5.0% | 21.0% | 201.00 |
| 18 | Berthet Clément | 0.9% | 1.1% | 1.4% | 3.4% | 35.2% | 101.00 |
| 19 | Buitrago Santiago | 0.9% | 0.6% | 1.4% | 3.9% | 16.0% | 41.00 |
| 20 | Bisiaux Léo | 0.9% | 1.4% | 1.1% | 4.9% | 32.2% | 151.00 |
| 21 | Tulett Ben | 0.8% | 0.8% | 1.2% | 3.3% | 14.2% | 151.00 |
| 22 | Omrzel Jakob | 0.8% | 0.9% | 1.1% | 4.6% | 47.2% | 81.00 |
| 23 | Castrillo Pablo | 0.7% | 0.8% | 1.1% | 1.4% | 5.9% | 201.00 |
| 24 | Fortunato Lorenzo | 0.7% | 0.4% | 1.1% | 2.2% | 11.5% | 23.00 |
| 25 | Svestad-Bårdseng Embret | 0.6% | 0.8% | 0.9% | 1.9% | 21.6% | 201.00 |
| 26 | Fisher-Black Finn | 0.5% | 0.7% | 0.8% | 2.2% | 9.8% | 751.00 |
| 27 | Vermaerke Kevin | 0.4% | 0.2% | 0.7% | 1.5% | 9.7% | 201.00 |
| 28 | van Aert Wout | 0.4% | 1.6% | 0.4% | 3.7% | 10.0% | n/a |
| 29 | Zana Filippo | 0.4% | 0.2% | 0.7% | 0.9% | 13.5% | 201.00 |
| 30 | Landa Mikel | 0.4% | 0.1% | 0.6% | 0.9% | 11.5% | 101.00 |
| # | Rider | Blend | Cascade | LambdaRank | Podium | Top 15 | Book odds |
|---|---|---|---|---|---|---|---|
| 1 | Roglic Primoz | 10.0% | 6.0% | 14.9% | 18.8% | 71.7% | 17.00 |
| 2 | Onley Oscar | 6.7% | 5.6% | 9.5% | 17.7% | 72.0% | 11.00 |
| 3 | Gall Felix | 4.6% | 2.9% | 6.8% | 13.2% | 62.0% | 2.60 |
| 4 | Carapaz Richard | 3.9% | 2.1% | 5.9% | 12.0% | 63.2% | 34.00 |
| 5 | Vine Jay | 3.1% | 5.2% | 3.8% | 9.6% | 23.1% | 4.75 |
| 6 | Skjelmose Mattias | 3.1% | 2.3% | 4.5% | 8.7% | 61.0% | 101.00 |
| 7 | Mas Enric | 2.7% | 1.1% | 4.2% | 8.0% | 53.4% | 6.50 |
| 8 | Johannessen Tobias Halland | 2.6% | 1.6% | 3.8% | 6.9% | 36.3% | 21.00 |
| 9 | van Aert Wout | 2.5% | 2.2% | 3.6% | 5.8% | 15.5% | n/a |
| 10 | Almeida Joao | 2.0% | 5.2% | 2.0% | 7.9% | 18.5% | 51.00 |
| 11 | Romeo Ivan | 1.6% | 2.1% | 2.1% | 5.2% | 17.1% | 401.00 |
| 12 | Tejada Harold | 1.3% | 1.7% | 1.7% | 7.3% | 47.5% | 101.00 |
| 13 | Bisiaux Léo | 1.3% | 1.8% | 1.6% | 5.4% | 35.0% | 151.00 |
| 14 | Berthet Clément | 1.2% | 0.9% | 1.8% | 3.0% | 37.8% | 101.00 |
| 15 | Tulett Ben | 1.2% | 1.2% | 1.7% | 4.6% | 17.6% | 151.00 |
| 16 | Bilbao Pello | 1.2% | 1.7% | 1.5% | 6.9% | 25.2% | 201.00 |
| 17 | Widar Jarno | 1.1% | 1.1% | 1.6% | 4.3% | 39.6% | 51.00 |
| 18 | Nordhagen Jørgen | 1.0% | 1.2% | 1.3% | 4.1% | 17.8% | 151.00 |
| 19 | Kuss Sepp | 1.0% | 1.4% | 1.2% | 7.6% | 45.2% | 26.00 |
| 20 | Fisher-Black Finn | 0.9% | 1.6% | 1.1% | 5.5% | 14.4% | 751.00 |
| 21 | Rodriguez Carlos | 0.9% | 0.8% | 1.2% | 2.2% | 15.4% | 151.00 |
| 22 | Omrzel Jakob | 0.9% | 0.9% | 1.2% | 3.6% | 46.7% | 81.00 |
| 23 | Laurance Axel | 0.8% | 1.2% | 1.0% | 5.1% | 18.3% | 1001.00 |
| 24 | Riccitello Matthew | 0.7% | 0.6% | 1.1% | 2.6% | 28.5% | 81.00 |
| 25 | Buitrago Santiago | 0.7% | 0.6% | 1.0% | 3.5% | 22.5% | 41.00 |
| 26 | Svestad-Bårdseng Embret | 0.7% | 0.7% | 1.0% | 1.9% | 21.7% | 201.00 |
| 27 | Laporte Christophe | 0.7% | 2.2% | 0.6% | 8.4% | 11.8% | n/a |
| 28 | Vansevenant Mauri | 0.7% | 0.7% | 0.9% | 1.9% | 17.7% | 1001.00 |
| 29 | Castrillo Pablo | 0.6% | 1.0% | 0.8% | 2.0% | 7.0% | 201.00 |
| 30 | Pedersen Mads | 0.6% | 1.1% | 0.7% | 2.0% | 3.3% | 751.00 |
After the stage
| Pos | Rider | Team | Aitana shatters win p | Cohesive GC group win p |
|---|---|---|---|---|
| 1 | Mas Enric | Movistar Team | 4.4% | 2.7% |
| 2 | Onley Oscar | Netcompany INEOS Cycling Team | 6.6% | 6.7% |
| 3 | Roglic Primoz | Red Bull-BORA-hansgrohe | 8.4% | 10.0% |
| 4 | Gall Felix | Decathlon CMA CGM Team | 7.9% | 4.6% |
| 5 | Rodriguez Cristian | XDS Astana Team | 0.4% | 0.2% |
| 6 | Skjelmose Mattias | Lidl-Trek | 2.1% | 3.1% |
| 7 | Widar Jarno | Lotto-Intermarché | 1.3% | 1.1% |
| 8 | Tejada Harold | XDS Astana Team | 1.3% | 1.3% |
| 9 | Carapaz Richard | EF Education-EasyPost | 3.3% | 3.9% |
| 10 | Buitrago Santiago | Bahrain Victorious | 0.9% | 0.7% |
| Scenario | Blend win p | Book devig |
|---|---|---|
| Aitana shatters | 4.35% | 9.91% |
| Cohesive GC group | 2.74% | 9.91% |
Reflection
Enric Mas won at Alto de Aitana and took the race lead, with Oscar Onley on the same time, Roglic at 12 seconds and Gall at 18. The GC group came apart exactly as a summit finish on an open classification should make it: a winning group of two, 18,640 seconds of racing, and the oracle returned gc_day on the tiny_group rule.
The key was exact at full weight — the first time this campaign the macro item was keyed at 1.0 and the road agreed. This is the stage-7 correction working as intended: start from the macro item and demand a positive reason to demote it. On stage 7 the chase-incentive test said nobody had a reason to spend and I keyed gc_day anyway; here four men were separated by 2:20 with a summit finish in front of them, so the incentive was real and the macro was right. Scorecard moves to 4/8 exact, 7/8 covered, mean weight 0.54.
The board's retrieval was the best of the campaign. `aitana_shatters` put the actual top four — Roglic 8.4%, Gall 7.9%, Onley 6.6%, Mas 4.4% — in its own top four, in a different order (they finished 3, 4, 2, 1). Getting the quartet without the ordering is precisely what these heads are for; the ordering inside a four-man summit-finish group is close to a coin toss. The cohesive-group branch was slightly worse, dropping Mas to #7 at 2.7%.
The book had a bad day and the model had a good one. Gall was the favourite at 2.60 (24.8% devig) and finished 4th; the winner Mas was only its third choice at 6.50 (9.9%). More striking, its second favourite Jay Vine at 4.75 — 13.6% devig — finished 119th. That is the second consecutive stage where the book's price on Vine has been badly wrong (85th at 8.5% devig on stage 7), and it moves the break-card ledger back toward the model after stage 7 handed the market a clean win.
The retrieval channel earned its keep again: the keyed kernod board held 10 of the actual top 15 against production's 9 on the shatters scenario and 10 against 8 on the cohesive one. Ledger across scored stages: st2 -1, st4 level, st5 +1, st6 -1, st7 +3, st9 +1/+2.
The GC heads now disagree in a way that is worth recording. Against a book reading Mas 33.3% / Roglic 30.6% / Gall 15.3% / Onley 10.9%, the state-aware outlook head returns 33.9% / 29.7% / 4.2% / 2.1% — it matches the market almost exactly on the two leaders and then collapses on the chasers, while the blend does the opposite, putting Mas at 7.5% behind Roglic, Onley and Gall because the main heads never see the standings. Neither head is right alone: the outlook knows who is winning and over-concentrates, the blend knows who is good and ignores the race. This is the clearest in-race evidence yet for the dedicated-GC direction — a corrector over the shared board rather than either head on its own.
Carry forward. (1) Pogacar's stage-8 abandon has turned a procession into a four-man race and every downstream output now has to be read on the new classification, not the old one. (2) Rest day 2026-08-31; stage 10 is 2026-09-01, a reduced hilly sprint where the book leads Brennan 2.75, van Aert 4.00, Pedersen 7.00 and then falls off a cliff to Meeus at 19. (3) The chase-incentive test is now 1-for-1 since it replaced the macro-frequency prior — keep applying it rather than counting macro streaks.