The biggest divergence today is Coinbase (COIN) Apr 8 Up/Down: Polymarket was pricing the favored side at 72%, while our fair value was ~1% (a 71-point gap). That scale of disagreement usually means one of three things: stale pricing, contract-leg confusion, or traders anchoring to narrative instead of settlement mechanics.
- Our largest disagreement is in two single-day equity direction markets where close-to-close data sharply constrains uncertainty.
- In NBA sides, we see one underpriced underdog (MIN) and one overpriced underdog (POR) versus sportsbook-implied baselines.
- In geopolitics, we remain below market on a May 31 uranium-custody deadline because verification and logistics are still unresolved.
- Biggest near-term risk to our calls is resolution-rule interpretation, not pure forecasting error.
Summary Comparison
| Event | Polymarket | Naly Fair Value | Gap | Our Position |
|---|---|---|---|---|
| Coinbase (COIN) Up/Down Apr 8 | 72% | 1% | -71 pts | Favored leg likely overpriced |
| Tesla (TSLA) Up/Down Apr 8 | 53% | 1% | -52 pts | Favored leg likely overpriced |
| Timberwolves vs Magic | 16% | 49% | +33 pts | MIN side underpriced |
| Trail Blazers vs Spurs | 49% | 20% | -29 pts | POR side overpriced |
| US obtains Iranian enriched uranium by May 31 | 21% | 8% | -13 pts | "Yes" side overpriced |
Live Market Prices
1) Coinbase (COIN) Up or Down on April 8?
Causal chain: If settlement is close-to-close direction, then once reliable Apr 7 and Apr 8 closes are posted, uncertainty should collapse. If market pricing remains elevated for the losing leg, that is usually a lag/arbitrage window.
Key factors:
- COIN historical prints show Apr 8 close 179.91 vs Apr 7 close 175.18.
- Daily direction contracts are mechanically sensitive to data-timestamp and leg-mapping errors.
Bayesian calculation (our frame):
- Prior: 50%
- Evidence multiplier against currently favored leg: ~99:1
- Posterior: ~1%
Alternative explanation: The quote may have represented the opposite leg at scrape time, or there was a temporary data-feed mismatch near close.
What would make us wrong: A different official settlement source, or contract rules defining a different comparison window than standard close-to-close.
Sources: https://stockanalysis.com/stocks/coin/history/ , https://polymarket.com/event/coin-up-or-down-on-april-8-2026
2) Tesla (TSLA) Up or Down on April 8?
Causal chain: Same structure as COIN: direction contracts should converge quickly once final closes lock.
Key factors:
- TSLA historical prints show Apr 8 close 354.18 vs Apr 7 close 346.65.
- Remaining variance is mostly rule interpretation, not price uncertainty.
Bayesian calculation:
- Prior: 50%
- Evidence multiplier against currently favored leg: ~60:1
- Posterior: ~1–2% (we round to 1%)
Alternative explanation: Delayed corrections or exchange-specific close definitions could affect resolution.
What would make us wrong: A non-standard settlement convention or reversed leg labeling at the time of pricing.
Sources: https://stockanalysis.com/stocks/tsla/history/ , https://polymarket.com/event/tsla-up-or-down-on-april-8-2026
3) Timberwolves vs. Magic
Causal chain: Market priced MIN like a deep longshot, but sportsbook prices and model-based matchup probabilities suggest a materially higher upset chance.
Key factors:
- ESPN matchup context listed MIN around coin-flip territory in pregame model framing.
- Book odds around MIN +205 imply much higher than 16% after no-vig adjustment.
- Injury uncertainty cuts both ways but does not justify extreme longshot pricing alone.
Bayesian calculation:
- Prior from market: 16%
- Evidence update (books + model): odds multiplier ~5.0
- Posterior: ~49%
Alternative explanation: Late lineup downgrade (especially key shot creators) could justify a big repricing down.
What would make us wrong: Confirmed late absences or final line movement materially against MIN before tip.
Sources: https://www.espn.com/nba/game/_/gameId/401811014/timberwolves-magic , https://polymarket.com/event/nba-min-orl-2026-04-08
4) Trail Blazers vs. Spurs
Causal chain: A near-50% price on POR conflicts with stronger team-quality priors and a meaningful Spurs home edge.
Key factors:
- ESPN odds context showed POR +140 / SAS -166 and Spurs -3.5, consistent with Spurs advantage.
- Season profile favored San Antonio heavily (record/home split).
- Injury variance exists, but baseline gap remains wide.
Bayesian calculation:
- Prior from market: 49%
- Evidence update (books + team strength): odds multiplier ~0.26
- Posterior: ~20%
Alternative explanation: Spurs rest/injury news close to tip can compress true edge fast.
What would make us wrong: Multiple Spurs core absences or confirmed rotation changes that flatten possession quality.
Sources: https://www.espn.com/nba/game/_/gameId/401811016/trail-blazers-spurs , https://polymarket.com/event/nba-por-sas-2026-04-08
5) US obtains Iranian enriched uranium by May 31?
Causal chain: For this market to resolve "Yes," the U.S. must obtain possession and that outcome must be officially confirmed. Diplomatic ambiguity and verification gaps make this a high-friction path before deadline.
Key factors:
- Recent AP reporting referenced ceasefire/talks but no confirmed transfer of custody.
- Reporting around IAEA access and stockpile verification highlights information gaps, not confirmed handover.
- Deadline structure (May 31) compresses execution time for a politically and logistically complex transfer.
Bayesian calculation:
- Prior from market: 21%
- Evidence multiplier from absent confirmation + logistics friction: ~0.33
- Posterior: ~8%
Alternative explanation: A sudden bilateral agreement plus explicit U.S. confirmation could reprice this sharply upward.
What would make us wrong: A formal U.S. announcement confirming possession of qualifying enriched material before May 31.
Sources: https://apnews.com/article/4b464b08aaf7220899526c0c6da4317d , https://www.washingtonpost.com/world/2026/02/27/iran-nuclear-iaea-uranium-enrichment-suspend/95f9ad04-13e5-11f1-8e8d-fe91db44677b_story.html/ , https://polymarket.com/event/us-obtains-iranian-enriched-uranium-by-may-31
Methodology
We start from market-implied prior, update with independent evidence, then convert back to posterior probability with explicit uncertainty penalties for rule ambiguity and late-breaking information. Full calibration and historical hit-rate are tracked at /track-record.
Disclaimer
This analysis is for informational purposes only and reflects probabilistic opinions, not investment advice. Prediction markets can move on liquidity, interpretation, and timing effects independent of eventual resolution.