For entertainment only — MMA analysis only, not financial advice.

Published pick

Matthieu Duclos vs Luis Dias De Assis

Winner
Matthieu Duclos
Win chance
52.9%
Strength
Slight lean
Evidence
4/7 lenses · partial
Agreement
1 pick · 2 opponent · 1 neutral · 3 abstain

Winner revised in an earlier published version

Generated automatically by The Oracle. No individual human reviews each prediction before publication.

Model version
kimi-k3
Source cutoff
Publication lock

Win probabilities are estimates, not guarantees. Hit rate is not profitability. Correction policy

Full blueprint outline

See exactly what Sharp opens

The structure is public. Matchup-specific conclusions remain behind the verified access check.

  1. 01 · Finish projection

    Likely method & round

    A conditional finish lane plus the assumptions that could overturn it.

  2. 02 · Matchup analysis

    Evidence and matchup analysis

    Supporting and conflicting evidence, tactical interactions, and the uncertainty behind the pick.

  3. 03–04 · Decision support

    Paths, evidence & limits

    Both fighters' win paths, supporting and contradictory receipts, uncertainty, and declared limitations.

Checking saved status…

This is the original pre-fight pick retained after the event. Its settled result appears in the public Track Record.

Publication context

How to read this forecast

This immutable automated forecast covers Matthieu Duclos vs Luis Dias De Assis at UFC Fight Night: Hooker vs. Parnasse on September 5, 2026. Its public winner, probability, source cutoff, and publication lock remain attached to the matchup for later grading.

A probability is an estimate, not a promise. The 52.9% estimate still allows the opponent a meaningful upset path, and hit rate must not be treated as evidence of betting profitability.

Publication details

See when and how this pick was checked

+
Prediction engine
The Oracle by Blueprint MMA
Engine version
kimi-k3
Information checked through
Aug 30, 2026, 9:53 PM UTC
Published
Aug 30, 2026, 10:12 PM UTC
Page data generated
Sep 11, 2026, 7:03 PM UTC
Data revision
10f9254ce9b80bfe619567aafc293a4d6f308065166c3ef700ddb61c538cbe5d

Checks made before publication

  • • The event and matchup must agree.
  • • Each fighter and fight must be uniquely identified.
  • • The winner and win chance must be saved before the fight.
  • • The website and public data must show the same pick.

Each fight uses one saved, source-linked research file. The engine weighs supporting and contradictory evidence, records uncertainty, and publishes its estimated win chance with the original evidence cutoff.

Authorship, limitations, and corrections

Generated automatically by The Oracle. No individual human reviews each prediction before publication.

Blueprint MMA publishes the report and the Oracle Engine generates it. Sources can be incomplete, outdated, or wrong, and late injuries, cancellations, weigh-in changes, or market movement can change the fight after the displayed cutoff.

Blueprint MMA never equates hit rate with profitability and makes no profit guarantee. Review the methodology, correction policy, settled proof, and current UFC prediction board.