GUIDES · BETSLIP DOCTOR
How should I compare historical records that come from different season IDs?
Compare records from different season IDs as separate historical samples. Name each season, the wins, losses, and ties or other outcomes counted, sample size, cutoff, and captured coverage. A participant's latest completed season may not be the same season as an opponent's. You can describe their rates side by side, but do not imply the populations, schedules, or rules are identical. Historical win percentages are not probabilities for the next event.
For Bettors and cappers interpreting sports-record evidence · Updated
Label each participant's season separately
The phrase latest season can refer to different season IDs for different participants. Preserve the official season label and the period it covers for each record. Do not merge a prior completed season for one participant with a current or differently scoped season for another.
- State each season ID and the completed contests included.
- Show wins, losses, ties, and sample size under the source's outcome rules.
- Include the research cutoff and any limits in captured coverage.
Rates can be side by side without being the same sample
A rate divides an outcome count by its stated sample. Show both the numerator and denominator, not just the percentage. If season IDs, schedule length, rules, or captured periods differ, say so; a common percentage format does not make the samples identical.
Do not convert season records into a forecast
Season records describe completed contests in their own windows. They do not establish the probability of winning a future event or a betting outcome. If a season record is incomplete or unavailable, keep that limitation visible rather than filling it with an assumption.
Hypothetical example: two latest seasons, two populations
A synthetic comparison lists Subject A with 6 wins in 10 completed contests in season S-A and Opponent B with 3 wins in 6 completed contests in season S-B. The rates are 60 percent and 50 percent, but S-A and S-B remain different season IDs with different samples. The ledger reports each count, denominator, and cutoff separately rather than claiming they are one shared season. These values are invented and are not real team or player records.
Related questions
Can I compare two percentages if their season IDs differ?
You can report them side by side with each season ID, wins, sample, and coverage. Do not imply the samples are the same population or treat either rate as next-event probability.