79%Probability of reaching the shortlistGPHG Index Score 79 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
73%Probability of reaching the shortlistGPHG Index Score 73 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
66%Probability of reaching the shortlistGPHG Index Score 66 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
65%Probability of reaching the shortlistGPHG Index Score 65 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
63%Probability of reaching the shortlistGPHG Index Score 63 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
59%Probability of reaching the shortlistGPHG Index Score 59 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
Royal Oak Offshore Selfwinding Flying Tourbillon Chronograph
58%Probability of reaching the shortlistGPHG Index Score 58 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
55%Probability of reaching the shortlistGPHG Index Score 55 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
52%Probability of reaching the shortlistGPHG Index Score 52 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
49%Probability of reaching the shortlistGPHG Index Score 49 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
49%Probability of reaching the shortlistGPHG Index Score 49 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
49%Probability of reaching the shortlistGPHG Index Score 49 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
49%Probability of reaching the shortlistGPHG Index Score 49 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
47%Probability of reaching the shortlistGPHG Index Score 47 / 100
Model's leading pick for 2026 — the official shortlist is not out yet.
Why this watch?
This edition is scored by the preselection model, which estimates which entrants reach the shortlist. It does not export per-watch factor contributions the way the winner model does, so there is no honest breakdown to show here yet. What the model does use →
The 2026 shortlist has not been announced. These are estimates of which entrants reach it, from a weaker model than the one that ranks a published shortlist. This is a statistical prediction based on historical GPHG data. It is not a GPHG result and not betting odds.
Does it actually work?
Scored on 84 shortlists the model had never seen, 2020–2025.
37%The model's favourite wins
vs
17%A blind guess would
In plain language: pick the model's top watch in a category and it wins that category about 37% of the time — against one time in six for a name drawn out of the six finalists at random. That is roughly 2.2× better than chance, on 31 hits from 84 shortlists. It is an edge, not a crystal ball.
Counted more generously — the favourite winning any GPHG prize that year, including the cross-category awards it might also be up for — the model hits 41% against a 21% baseline. Both measures are shown because they answer different questions; the headline above is the stricter one.
Every figure here is out of sample: for each edition the model was retrained with that year removed, then asked to rank it. The blind-guess baseline is measured from the same shortlists under the same rule, never assumed to be one-in-six. Shortlists whose winner we could not match to a finalist are excluded from both sides. Ranking quality across all shortlisted watches is 0.62 ROC-AUC. Full definitions →
Called it, and missed
Every category the model has been graded on, 2020–2025.
✓31favourites that won
×53favourites that didn't
Individual editions swing hard around that average — 2025, the most recent one we can mark, went 3 of 15. A six-watch shortlist and a jury of people is not a stable thing to predict, and the record is published either way.
Reaching the final six is the hard part — and Chaumet has done it 18 times across 47 entries between 2007 and 2024 without once being called up. The jury is a different filter from the Academy.
Volume is not the same as pedigree. MB&F has entered just 14 watches, but 73% of its shortlist appearances have turned into a prize — ahead of Voutilainen on 63%. Among brands with ten or more shortlists, nobody converts better.
The model's favourite has won 4 of 5 Tourbillon shortlists, and 0 of 6 in Petite Aiguille. Technical categories reward the same establishment names year after year; the decorative ones are far closer to a coin toss. Small samples — a handful of editions each.
