NeuPortal

Live public experiment

Our models vs the markets

We forecast real-world events — football, crypto price moves, politics, weather, esports — and score ourselves against the markets people actually trade on: Polymarket, SX Bet, Binance. The model's probabilities are locked before each event, the market's price is frozen at the same instant, and both get scored against reality afterwards. Every call stays on the record — including the ones we lose. Two tracks are live: sealed crypto price distributions, and World Cup match odds.

1 Every forecast is written in full, SHA-256 hashed and anchored to the Bitcoin blockchain before it is published.
2 We publish a distribution, not a price call: a core 50% zone and a wider 80% band, each with its stated probability.
3 The outcome is the Binance close of the minute containing the stated resolution time. No other venue, no re-pick.
4 Coverage is the score. The outcome should land inside the core zone about half the time — much more is cowardice, much less is overconfidence, and both are failures.
Forecasts scored · distinct days 448over 46 days
Core 50% coverage (target 50%) 53%
Wide 80% coverage (target 80%) 89%
Median error · sealed & waiting 2.34%104 pending

These assets move together. Eight forecasts sealed on one morning are not eight independent claims — when Bitcoin leaves its band, most of the rest leave theirs on the same candle. That is why the count of distinct days sits next to the count of forecasts, and why coverage is broken out below rather than pooled into one flattering number.

By horizon. This is the breakout that matters, and until 7 September 2026 it was missing — only the per-asset split below was published. The assets differ by a few points. The horizons do not.

24h core 67% of 50% wide 93% of 80% mean error +0.5% 180 scored
7d core 72% of 50% wide 86% of 80% mean error +3.2% 160 scored
30d core 3% of 50% wide 86% of 80% mean error +28.2% 108 scored

The 30-day band failed over this period. Its core zone should hold about half the outcomes and held a twentieth, and every one of the misses fell the same way — the market rose further than the band allowed for, ninety times out of ninety. That is a method failure on a trending regime, not a run of bad luck, and the 30-day horizon is being reworked. Read those cards as a record of what went wrong, not as a forecast to use.

BNB 50% core · 109 scored BTC 48% core · 115 scored ETH 57% core · 115 scored SOL 57% core · 109 scored

📈 Crypto — sealed price distributions vs Binance spot

Newest first · showing 171–180 of 552 sealed forecasts

BNB · 30d resolves 30 Sept, 10:24 UTC
Core 50%Wide 80%MedianOutcome
Sealed at $686 $666 – $768 $606 – $877 $715 …

🔒 sealed 31 Aug, 10:24 UTC · sha256 c0cd608ed392 · bias long — awaiting result

SOL · 24h resolves 1 Sept, 10:24 UTC
Core 50%Wide 80%MedianOutcome
Sealed at $103 $101 – $106 $98.68 – $110 $104 $102

🔒 sealed 31 Aug, 10:24 UTC · sha256 06dc46f71201 · bias long — inside the core 50%, median off by -1.10%

SOL · 7d resolves 7 Sept, 10:24 UTC
Core 50%Wide 80%MedianOutcome
Sealed at $103 $98.29 – $115 $90.19 – $126 $105 $105

🔒 sealed 31 Aug, 10:24 UTC · sha256 06dc46f71201 · bias long — inside the core 50%, median off by -0.15%

SOL · 30d resolves 30 Sept, 10:24 UTC
Core 50%Wide 80%MedianOutcome
Sealed at $103 $93.73 – $131 $79.08 – $167 $112 …

🔒 sealed 31 Aug, 10:24 UTC · sha256 06dc46f71201 · bias long — awaiting result

BTC · 24h resolves 1 Sept, 10:22 UTC
Core 50%Wide 80%MedianOutcome
Sealed at $78,426 $77,675 – $79,980 $76,191 – $81,767 $78,740 $77,975

🔒 sealed 31 Aug, 10:22 UTC · sha256 446da580efc7 · bias long — inside the core 50%, median off by -0.97%

BTC · 7d resolves 7 Sept, 10:22 UTC
Core 50%Wide 80%MedianOutcome
Sealed at $78,426 $76,588 – $83,815 $72,596 – $88,987 $79,987 $79,421

🔒 sealed 31 Aug, 10:22 UTC · sha256 446da580efc7 · bias long — inside the core 50%, median off by -0.71%

BTC · 30d resolves 30 Sept, 10:22 UTC
Core 50%Wide 80%MedianOutcome
Sealed at $78,426 $74,667 – $91,426 $66,041 – $104,091 $82,120 …

🔒 sealed 31 Aug, 10:22 UTC · sha256 446da580efc7 · bias long — awaiting result

ETH · 24h resolves 1 Sept, 10:22 UTC
Core 50%Wide 80%MedianOutcome
Sealed at $2,447 $2,410 – $2,522 $2,345 – $2,594 $2,466 $2,456

🔒 sealed 31 Aug, 10:22 UTC · sha256 a4978f665d40 · bias long — inside the core 50%, median off by -0.41%

ETH · 7d resolves 7 Sept, 10:22 UTC
Core 50%Wide 80%MedianOutcome
Sealed at $2,447 $2,354 – $2,660 $2,173 – $2,877 $2,500 $2,490

🔒 sealed 31 Aug, 10:22 UTC · sha256 a4978f665d40 · bias long — inside the core 50%, median off by -0.41%

ETH · 30d resolves 30 Sept, 10:22 UTC
Core 50%Wide 80%MedianOutcome
Sealed at $2,447 $2,204 – $3,072 $1,884 – $3,803 $2,600 …

🔒 sealed 31 Aug, 10:22 UTC · sha256 a4978f665d40 · bias long — awaiting result

All tracks

One protocol — locked calls, frozen benchmarks, public scoring — across every domain we forecast:

Methodology, exactly

Don't trust our timestamps — verify them

This is an educational, transparency-first project by NeuPortal — entertainment and research, not betting or financial advice. Probabilities are for understanding events, not acting on them.