A statistical look at how well market odds match real-world outcomes.
96.7%
4 hr accuracy
90.4%
1 month accuracy
0.0627
Brier Score
Accuracy prior to resolution
Shows how accurate axoin.fun odds were at different points before they resolved.
Expected vs Actual
axoin.fun forecasts closely align with actual outcomes across all probability ranges.
Expected Percentage Resolved Percentage
Brier Score vs Volume
Lower is better. Brier scores measure how close predictions were to being right, not just whether they were correct.
Resolution Composition
How resolved markets on axoin.fun split between "Yes" and "No" outcomes.
Yes — 26.2%No — 73.8%
Methods
All metrics are calculated from resolved markets on axoin.fun, using price snapshots taken 1 month, 1 week, 1 day, 12 hours, and 4 hours before resolution. Accuracy reflects how often the final leading outcome matched reality, and Brier scores measure the average squared error of predictions.
Accuracy is the share of resolved markets where the leading outcome at the given timestamp matched the final result. A 96.7% reading 4 hours before resolution means that out of every 1,000 markets on axoin.fun, roughly 967 had the eventual winning side priced above 50% at that point.
A Brier score is the mean squared error between a forecast and the realised outcome (0 for a perfect call, 1 for a confident wrong call). Lower is better, and unlike pure accuracy it rewards calibration — being 95% certain on a true positive scores better than being 60% certain.
New information arrives as the resolution date approaches — polls tighten, scores update, official sources confirm. Traders react, and the price absorbs that signal. The closer you get to resolution, the less uncertainty remains for the market to price in.
Most prediction-market questions are phrased as a specific claim ("will X happen by date Y?"). The space of ways a claim can fail is larger than the space of ways it can succeed, so the base rate naturally skews toward "No".
All figures are computed from resolved markets on axoin.fun, using public price snapshots taken 1 month, 1 week, 1 day, 12 hours, and 4 hours before each market resolved.
