Prabhat M

Math & Stats @ Purdue (May 2027)
Quant research, sports analytics, prediction markets

[Prabhat M @PrabhatM27]

The #WorldCup2026 is the most open field the markets have priced recently: no team above ~16% to win it (the favourite was ~20% in 2010 & 2022, ~18% in 2018). The top 4 are barely half the field combined.

I'm tracking what Kalshi, Polymarket & the bookmakers reveal about it 🧵

[Prabhat M @PrabhatM27]

The most valuable goal at this World Cup is the one nobody's watching.
For a team fighting for a best-third spot, dragging your GD from -1 to level is worth +28 points of reaching the Round of 32. And you can score it in a game that is already won or lost. 🧵 #WorldCup2026

[Prabhat M @PrabhatM27]

One model, one tournament, and the table shifts every matchday. But it holds for the final group games this week: don't take your foot off the gas at 3-0.

Methodology and the sim: https://t.co/lWH91HKt56

[Prabhat M @PrabhatM27]

The most valuable goal at this World Cup is the one nobody's watching.
For a team fighting for a best-third spot, dragging your GD from -1 to level is worth +28 points of reaching the Round of 32. And you can score it in a game that is already won or lost. 🧵 #WorldCup2026

[Prabhat M @PrabhatM27]

The catch is it's the model's conditional curve, and the effect lives entirely at the -1 to 0 hinge. Above level GD, a goal is nearly worthless; below -2 you're probably out anyway. It's a cliff, not a slope.

[Prabhat M @PrabhatM27]

The World Cup group stage looked like carnage. Spain held 0-0 by Cape Verde, Portugal by DR Congo, and 38% of the model's strong favourites dropped points.

So did the betting market panic and dump the giants? The opposite. 🧵

[Prabhat M @PrabhatM27]

When a goal goes in, does the betting market move enough? I built a fair-value win-probability model for every second of a World Cup match to check. Across 33 goals, the market consistently moves less than it should. 🧵 #WorldCup2026

[Prabhat M @PrabhatM27]

Mid-tournament, the model conditions on games played, and the market's France lean could be conviction or over-loyalty.

Live model vs market + the microstructure:
https://t.co/sPJypKrEyf

[Prabhat M @PrabhatM27]

That's the 48-team format doing two things at once. The best-thirds safety net lets a minnow park the bus and bank a point against a giant (the chaos), but it also catches the giants when they slip: almost every shocked favourite still qualifies.

[Prabhat M @PrabhatM27]

The market's boldest call: it now prices France a full 6pp above our model to win it all (19.7% vs 13.6%), all-in on Mbappé's form team. And this is the market our cross-venue research shows is the fastest, sharpest price-discovery engine of the two exchanges.

[Prabhat M @PrabhatM27]

Through all of it, the title-winning odds moved just ±3pp. And the market didn't flee the favourites; it re-sorted them: it punished the genuine strugglers (Spain −3.3pp, Portugal −3.1, Brazil −2.1) and rewarded the standouts (France +3.4, Argentina +2.6).

[Prabhat M @PrabhatM27]

The World Cup group stage looked like carnage. Spain held 0-0 by Cape Verde, Portugal by DR Congo, and 38% of the model's strong favourites dropped points.

So did the betting market panic and dump the giants? The opposite. 🧵

[Prabhat M @PrabhatM27]

Across 10 matches and 33 goals, the market's move averages about 5pp smaller than fair value. Small and soccer-specific, the model is calibrated to the pre-game price.

Full method + the live dashboard:
https://t.co/sPJypKrEyf

[Prabhat M @PrabhatM27]

When a goal goes in, does the betting market move enough? I built a fair-value win-probability model for every second of a World Cup match to check. Across 33 goals, the market consistently moves less than it should. 🧵 #WorldCup2026

[Prabhat M @PrabhatM27]

So this is not a free trade. The market moving less isn't the market being slow; it's the market being smart. Telling "smart" from genuinely "slow" needs a model that sees in-game dominance (live expected goals), and that is the next build.

[Prabhat M @PrabhatM27]

Does the World Cup draw actually matter? I re-ran the real draw thousands of times against a fair re-draw, scored on odds to reach the Round of 32.

The "group of death" is a top-team myth, the draw only makes or breaks the teams on the edge. 👇

[Prabhat M @PrabhatM27]

For the favourites it's noise — Spain, France, England, Argentina all swing under half a point. They advance from any group.

For the bubble it's the whole tournament: the draw is worth +24pp to Bosnia's knockout odds, +21 to Egypt, +18 to Iran.

[Prabhat M @PrabhatM27]

Does the World Cup draw actually matter? I re-ran the real draw thousands of times against a fair re-draw, scored on odds to reach the Round of 32.

The "group of death" is a top-team myth, the draw only makes or breaks the teams on the edge. 👇