KicktippAi experiment analysis
o3 (medium) 100x knowledge cutoff follow-up
match-predictions/bundesliga-2025-26/pes-squad/repeated-match/md01-fc-bayern-munchen-vs-rb-leipzig/repeat-100-knowledge-cutoff-bayern-rbl-md1
The existing report is a paired 25x comparison. This 100x run uses the same source match, prompt route, and evaluation time, but a different repeated-match dataset size, so it is published as a linked single-run follow-up instead of being folded into the paired statistical report.
Exact 6:0 Signal
Prediction Distribution
o3 (medium)
n=100
3:1
77
2:1
19
3:2
4
Run Metadata
| Field | Value |
|---|---|
| Run | repeated-match__pes-squad__o3__langfuse-o3-poc__reasoning-medium__repeat-100-knowledge-cutoff-bayern-rbl-md1__exact-time__2026-05-30t20-25-39z |
| Prompt | langfuse-o3-poc / kicktippai/predict-one-match-o3-poc, label poc |
| Reasoning effort | medium |
| Batch count | 9 |
| Evaluation time | 2025-08-22T12:00:00.0000000+02:00 |
| Selected item hash | 86fe3b148f4df8862ca2594deb2edb00925d1e2d8be8bb56a8c600ccf4c176af |
| Verified observations | 100 / 100 |
Related Analysis
The companion narrative markdown covers the 25x comparison and this 100x follow-up: knowledge-cutoff-bayern-rbl-repeated-match.md.
The paired 25x comparison remains available at 2026-05-05t23-01-38z.analysis.report.html.