NFL MVP market on Polymarket draws sharp money to Josh Allen

Josh Allen opened the NFL season as the favourite to win the league's Most Valuable Player award, and the prediction market on Polymarket reflects that — with Allen sitting clearly ahead of the field after Buffalo's Week 1 performance.

That much is consensus. Here is where I part from it.

The MVP award has gone to a quarterback every year since Adrian Peterson won it in 2012 as a running back for the Vikings. Fourteen straight years. At this point the market is not pricing Allen's individual season — it is pricing a structural bias so entrenched that it has become invisible. When a pattern holds that long, markets stop questioning the mechanism and start treating the outcome as a constant. That is when mispricing opens.

The specific mechanism that concerns me: the MVP vote is retrospective and narrative-driven. It rewards quarterbacks because quarterbacks carry teams, and because voters can point to a quarterback's statistics and construct a clean story. But the best versions of that story — the one where a single player visibly rescues a franchise — are actually less likely when the favourite is Allen, who plays for a genuinely strong Buffalo team. A dominant back on a mediocre team, or a pass-rusher who alters every offensive game plan in the league, carries exactly the redemption narrative that MVP voters have historically rewarded in the rare years they looked past the position.

The long-bias toward quarterbacks is real. I am not arguing it disappears. But the specific price being offered on Allen — the favourite on a good team, in a conference with legitimate competition from Lamar Jackson and others — compresses the tail risk in a way I find unattractive. The market is priced as if the non-QB scenario is decorative. I think it is underweighted.

I would be a seller of Allen at the current implied probability, and I would be looking at the field — not any single alternative, but the aggregate probability on non-quarterbacks, which the market is almost certainly treating as noise. The position that changes this view: Allen wins the first six games carrying the offence alone, Buffalo's other weapons underperform, and the narrative writes itself before November. At that point the price would be right, and I would not be in it.

The $3.12 billion in prediction market volume that Week 1 generated tells you something about the depth now available in these markets. This is no longer a thin market where a single large position moves the line. The NFL MVP contract on a liquid platform now has enough counterparty flow that a genuine mispricing can persist for weeks before sharp money closes it. That is the environment where this kind of structural argument is worth making.