The Market — Sharpest Bettors & Mispriced Players

Read this before the numbers make sense

• It's an AMM, not a sportsbook. The market runs on an LMSR (logarithmic market-scoring rule) — a bot that always quotes a price and takes the other side. There's no order book, no bid/ask, no waiting for a counterparty. You buy "Yes" or "No" shares and the price moves as you trade. SOLID
• Brier = forecast accuracy. Lower is better. It's the squared error between what you priced something at and whether it happened. Call a coin flip 50/50 forever and you sit at 0.25; nail everything and you approach 0. It rewards being calibrated and confident, not lucky.
• ROI & net worth = profit. Different question entirely — did your trades make units? You can forecast beautifully and still go broke by sizing badly, and vice-versa. Keep the two apart in your head.
• CLV doesn't exist here. In real betting, "closing-line value" is the tell for skill because the true probability is never known. Here every market resolves to the truth — the season actually plays out — so we grade against reality directly. CLV is deliberately omitted. COLOR

"Volume" in the tables below is Σ|cost| across trades, not a share count. Bot/strategy accounts (Bayes, Galton, Kelly, Pascal, Treasury, Tversky, results-bot) are excluded — this is a human leaderboard.

Sharpest bettor — and the trap in the top line

Sorted by Brier (accuracy). The crown belongs to the robust forecaster with a real book behind the number — so we gate the leaderboard at ≥20 resolved markets. Thinner books are shown separately, because a great Brier on a dozen near-locks isn't the same skill.

⭐ The robust leaderboard — 20+ resolved markets

The thin-book leaderboard — under 20 markets TINY n

Great Briers live up here too, but on tiny samples — a couple of near-certain bets will do it. Read the market count out loud.

Money, not accuracy

• Net-worth leader — pine at 5,035 units (also a solid 0.150 Brier on 35 markets). SOLID
• Best return — kit kat at +59% ROI (+156 units on 265 invested). Small stake, ruthless efficiency. SOLID

De-dupe note: dannyridel4 is almost certainly the same human as dannyridel; Varnite/varnite and blue water/bluewater16 are likely alts too. They're flagged, not merged — if the crown-holder is double-listed, treat the alt as the same person.

Most over- and underrated players

Take the market's implied ranking (from live LMSR prices) and lay it next to where each player actually sits after five weeks of real buzz data. The gaps are the story. SOLID (framing is ordinal — prices vs. actual rank)

⭐ Overrated — the market's favorites who aren't

⭐ Underrated — the market slept on them

Where the market actually nailed it — the 32+ props

Eleven "will player X score 32+ in a game?" markets. The pool gave exactly one player a real shot — Kian D at 93% — and he dropped 44. It priced the other ten longshots at 7–32%, and none of them cleared 32 all season. Perfect read. SOLID

The market can't call a single game

This is the load-bearing result of the whole segment. Across 60 resolved game markets (moneylines + totals), the pregame line was right exactly 50% of the time — a literal coin flip. Even on the 16 games where it had a real favorite (≥55%), it went just 38%. SOLID

Method note: by lock, live trading dragged the favorite to a median ~72% — but that's the result leaking into the price in real time, not a forecast. These numbers use the pregame line, before tip.

⭐ So the "upsets" aren't upsets — reframe them

The market's four "biggest misses" were all 59–61% pregame favorites that lost. A 60% favorite is supposed to lose about 40% of the time. These are coinflips landing on the underdog — soft lines, not shocking upsets. Call them that on air.

The early bets beat the line 69%

Here's the twist that saves the bettors' honor. Take the sharp cohort — the top forecasters by Brier — and look only at positions placed more than 6 hours before tip, so there's zero live leakage. Those early bets sided with the eventual winner in 69% of 49 game markets. And when the sharps split from the pregame favorite, they were right 69% of the time. COLOR (small cohort, early-bet subset)

The cleanest example — dan.k.memes vs anurag

The pregame market favored anurag at 60%. The sharps loaded the other side — dan.k.memes — early. Dan.k.memes won, and by lock the market had swung all the way to 99.6% on him. The smart money was on it before the room caught up. COLOR

The biggest movers of the season

Where the price ran the furthest from open to last trade. All-star markets dominate — the ballot narrative moved hard as the season revealed who was real. SOLID (direction), note exact open %

Method note: the "open %" is the first recorded trade, not always a genuine opening price — the raw opening field is sparse. Directions and magnitudes are trustworthy; treat the exact endpoints as soft.