Optimal Prior Per Category · Per Date
Each cell = mean of actual event outcomes for that category on that date (= what a perfect fixed prior would be). Your prior value is shown alongside so you can compare. Blue = low outcome, yellow = moderate, red = high. ✓ = date where a fixed prior can reach top-5, ✗ = unreachable (live-price agents dominate).
🌍 Geo Pool — optimal category mean per date
| Category | A prior | B prior | C prior | D prior | 06-09 ✗ | 06-11 ✗ | 06-12 ✗ | 06-14 ✗ | 06-21 ✗ | 06-22 ✓ | 06-23 ✗ | 06-24 ✗ |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| general | 0.14 | 0.12 | 0.22 | 0.14 | 0.14 n=56 | 0.17 n=64 | 0.16 n=54 | 0.13 n=60 | 0.14 n=56 | 0.09 n=58 | 0.11 n=47 | 0.11 n=58 |
| geo_other | 0.11 | 0.09 | 0.12 | 0.10 | 0.09 n=39 | 0.08 n=46 | 0.10 n=50 | 0.13 n=48 | 0.15 n=51 | 0.08 n=36 | 0.14 n=45 | 0.12 n=46 |
| iran_diplomatic | 0.10 | 0.08 | 0.10 | 0.08 | 0.09 n=18 | 0.08 n=19 | 0.10 n=24 | 0.14 n=22 | 0.14 n=16 | 0.03 n=15 | 0.01 n=16 | 0.01 n=6 |
| commodities | 0.22 | 0.18 | 0.17 | 0.10 | 0.10 n=3 | 0.34 n=3 | 0.16 n=3 | 0.31 n=3 | 0.20 n=5 | 0.13 n=3 | 0.07 n=5 | 0.03 n=4 |
| crypto | 0.09 | 0.07 | 0.09 | 0.08 | — | 0.14 n=4 | 0.00 n=1 | 0.00 n=1 | 0.10 n=1 | 0.07 n=2 | 0.15 n=4 | 0.15 n=2 |
🌐 Global Pool — optimal category mean per date
| Category | A prior | B prior | C prior | D prior | 06-09 ✓ | 06-11 ✗ | 06-12 ✗ | 06-14 ✓ | 06-21 ✗ | 06-22 ✗ | 06-23 ✗ | 06-24 ✗ |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| general | 0.14 | 0.12 | 0.22 | 0.14 | 0.23 n=836 | 0.23 n=880 | 0.22 n=889 | 0.25 n=879 | 0.18 n=680 | 0.13 n=104 | 0.16 n=123 | 0.10 n=99 |
| earnings_move | 0.33 | 0.33 | 0.33 | 0.05 | 0.29 n=35 | 0.37 n=27 | 0.39 n=33 | 0.33 n=27 | 0.31 n=16 | — | 0.00 n=2 | 0.00 n=5 |
| earnings_eps | 0.54 | 0.54 | 0.54 | 0.52 | 0.48 n=33 | 0.58 n=19 | 0.60 n=20 | 0.58 n=19 | 0.67 n=9 | — | — | 0.50 n=6 |
| sports | 0.15 | 0.12 | 0.15 | 0.10 | 0.29 n=7 | 0.29 n=7 | 0.33 n=6 | 0.00 n=3 | 0.00 n=3 | — | — | — |
| commodities | 0.22 | 0.18 | 0.17 | 0.10 | 0.07 n=59 | 0.15 n=52 | 0.18 n=45 | 0.21 n=24 | 0.33 n=24 | 0.00 n=2 | 0.00 n=5 | 0.00 n=2 |
Select priors to compare
Reg cost0.8 TAO
Dates covered3/8
Hit rate38%
Geo Pool Rank
Global Pool Rank
Prior APrior BPrior CPrior D● = top-5 hit · green band = achievable date · dashed = rank 5
5-day combined window · 238 geo agents · 238 global agents
Prior A Geo specialist
GEO#11/2380.0308 · 0.802 TAO/day
GLOBAL#23/2380.1547 · 0.759 TAO/day
Prior B Ultra-low geo
GEO#11/2380.0308 · 0.802 TAO/day
GLOBAL#30/2380.1566 · 0.735 TAO/day
Prior C Global high
GEO#27/2380.0344 · 0.745 TAO/day
GLOBAL#19/2380.1537 · 0.774 TAO/day
Prior D Global low
GEO#11/2380.0308 · 0.802 TAO/day
GLOBAL#24/2380.1552 · 0.756 TAO/day
Per-Date Detail — brier / rank / gap to top-5
| Date | Top-5 threshold | Geo Pool | Global Pool | ||||||
|---|---|---|---|---|---|---|---|---|---|
| Prior A | Prior B | Prior C | Prior D | Prior A | Prior B | Prior C | Prior D | ||
06-09 ✓ achievable | G≤0.019L≤0.179 geo:116 glb:971 | 0.0210#15/221+0.002 | 0.0209#15/221+0.002 | 0.0238#44/221+0.005 | 0.0205#15/221+0.002 | 0.1799#8/223+0.001 | 0.1826#17/223+0.003 | 0.1728#2/223★T5 | 0.1805#13/223+0.001 |
06-11 ✗ unreachable floor geo#28 glb#30 | G≤0.024L≤0.170 geo:136 glb:986 | 0.0284#36/217+0.005 | 0.0291#41/217+0.005 | 0.0301#53/217+0.006 | 0.0292#41/217+0.005 | 0.1863#42/217+0.017 | 0.1898#47/217+0.020 | 0.1787#32/217+0.009 | 0.1891#46/217+0.020 |
06-12 ✗ unreachable floor geo#6 glb#28 | G≤0.035L≤0.167 geo:132 glb:995 | 0.0374#6/221+0.002 | 0.0378#7/221+0.003 | 0.0388#12/221+0.004 | 0.0374#6/221+0.002 | 0.1798#34/224+0.013 | 0.1830#40/224+0.016 | 0.1741#28/224+0.008 | 0.1840#43/224+0.017 |
06-14 ✓ achievable | G≤0.030L≤0.195 geo:134 glb:953 | 0.0378#24/228+0.008 | 0.0387#29/228+0.009 | 0.0418#45/228+0.012 | 0.0391#32/228+0.009 | 0.1982#9/228+0.003 | 0.2025#27/228+0.007 | 0.1882#1/228★T5 | 0.2008#19/228+0.006 |
06-21 ✗ unreachable floor geo#18 glb#27 | G≤0.015L≤0.135 geo:129 glb:732 | 0.0527#18/208+0.037 | 0.0539#21/208+0.039 | 0.0550#22/208+0.040 | 0.0537#20/208+0.038 | 0.1527#41/216+0.017 | 0.1547#58/216+0.019 | 0.1534#47/216+0.018 | 0.1555#61/216+0.020 |
06-22 ✓ achievable | G≤0.012L≤0.106 geo:114 glb:106 | 0.0110#4/188★T5 | 0.0094#1/188★T5 | 0.0185#45/188+0.006 | 0.0103#3/188★T5 | 0.1085#11/188+0.003 | 0.1080#9/188+0.002 | 0.1167#72/188+0.011 | 0.1077#8/188+0.002 |
06-23 ✗ unreachable floor geo#9 glb#6 | G≤0.025L≤0.126 geo:117 glb:131 | 0.0290#41/178+0.004 | 0.0285#38/178+0.003 | 0.0329#60/178+0.007 | 0.0280#35/178+0.003 | 0.1319#13/178+0.006 | 0.1325#13/178+0.006 | 0.1338#17/178+0.007 | 0.1288#7/178+0.002 |
06-24 ✗ unreachable floor geo#13 glb#14 | G≤0.025L≤0.090 geo:116 glb:112 | 0.0292#33/170+0.004 | 0.0285#25/170+0.003 | 0.0342#63/170+0.009 | 0.0281#21/170+0.003 | 0.1008#35/170+0.011 | 0.0995#32/170+0.010 | 0.1116#90/170+0.022 | 0.0953#20/170+0.005 |
3/8 dates achievable with fixed priors — Jun-09 global (C #2), Jun-14 global (C #1), Jun-22 geo (B#1 D#3 A#4). On all other dates the top-5 agents use live Polymarket prices or per-event LLM, which no fixed prior can match.
How to read the table: each cell is the mean actual outcome for that category on that date. If it differs significantly from your prior value, your prior would have a higher brier score for that category. Low values (blue) = outcomes were mostly NO. High values (red) = outcomes were mostly YES.
Daily cron: data refreshes at 09:05 UTC each day via
collect_daily_events.pythenrun_walkforward.py at 09:15.