PRO LEAGUE ORACLE

A fight model fit on scraped season results, and an arena that plays out what it predicts.

CHECKING DATA
LAYERS
50%
50%

Model prediction before the bout. No bias term, so the two sides always sum to 100%.

HEAD TO HEAD two players, one pot - the forecast is information, not a counterparty POT 0

Name both players, set the ante, then run the fight. Winner takes the pot.

LEADERBOARD

-
100
0:00
100
-

COMMENTARY

-leave-one-out accuracy
-record-only baselineback the better record
-lift over baseline
-Brier score0.25 = always saying 50/50
-McNemar p, vs baseline

CALIBRATION

When it claims 80%, does it happen 80% of the time? Bars are what actually happened; the line is what was claimed.

EVERY FIGHT, RANKED BY CONFIDENCE

Green means the model called it. Each row was predicted by a model refit from scratch without that fight.

WEAPON MATCHUP

Row beats column, as a win rate shrunk toward 50/50 by sample size. Hover a cell for the raw record.

column favouredevenrow favoured

POWER RANKING

Mean win probability against the whole field. Falls straight out of the model, so there is no second scoring system to justify.

WHAT THE MODEL LEARNED

Standardised logistic weights. Positive means the feature favours the bot that has more of it.

WHERE THE FIELD COMES FROM

loading...

Read off the from= field of each competitor's infobox, on the same 25 pages as the roster - so this cost zero extra requests. Note what it is and is not: this is where the teams are, not where the audience is.

BY REGION

EVERY COMPETITOR

PREDICTION MARKET PORT

loading...

PRO LEAGUE, IN POLYMARKET FORMAT PROPOSED - NOT TRADED

The same 24 scraped bots, emitted as Polymarket market objects priced by our model. This is a listing payload, ready to post. The prices are the model's opinion, not a market's - nobody has traded these and the label says so.

DATA LINEAGE

Every source, which Bright Data product fetched it, and what it cost us. The two refusals are listed on purpose - a lineage that only shows what worked is a sales pitch, not a lineage.

PROVENANCE

INTEGRITY CHECKS

Run against the loaded files at boot. A scraper that half-works is worse than one that fails loudly.

FAN CHATTER

Mentions and reach scraped from X, against on-field results. Reach is measured; tone is not - the X Posts dataset returns engagement counts, not sentiment, so guessing a mood score here would be inventing data. The interesting row is a bot talked about a lot that is not winning.