Updated daily through the season. Spot something off? Tell us →Updated daily. Tell us →
SharpXI© Open the screener
Free · graded in public

Today’s player report: stat-based picks, sent the moment the markets open.

Referees, form, and the market price — free. Read every round from 422,147 player records.

One email a round. Confirm by clicking the link we send. Unsubscribe in one click.

Manuel LocatelliTo be carded 18/54.75 Juventus v AC Milan 5 of 10 games 1 of 10 for a typical starter

Or open the tool · open the screener · see the track record

Widest gaps on the Sun 6 Sep slate

Premier League, Bundesliga, Championship, La Liga, Ligue 1 and Serie A · each player's last 10 games against what a typical starter in that league manages

2026-09-06
  1. AROUND EVENSa harder line, a bigger return
  2. Armand Laurienté 1+ shots on target
    Winger · Bologna v Sassuolo Tough match-up give up the 5th-fewest shots on target of 20 1.0 a game
    EVS2.07 BetMGM 6 of 10 games 2 of 10 for a typical starter Shot map → Pro
  3. Edu Expósito 2+ shots
    Central midfielder · Espanyol v Sevilla Even match-up give up the 8th-most shots of 20 2.0 a game
    11/82.38 BetMGM 6 of 10 games 3 of 10 for a typical starter Shot map → Pro
  4. Gleison Bremer 1+ shots
    Centre-back · Juventus v AC Milan Even match-up give up the 8th-fewest shots of 20 1.5 a game
    EVS2.05 BetMGM 8 of 10 games 5 of 10 for a typical starter Shot map → Pro
  5. LONGER PRICES3/1 and up, record still checked
  6. James Tarkowski 1+ shots on target
    Centre-back · Everton v Man Utd Even match-up give up the 8th-fewest shots on target of 20 0.5 a game
    4.904.90 BetMGM 5 of 10 games 3 of 10 for a typical starter Shot map → Pro
  7. ALSO ON THIS SLATEranked on record
  8. Fabian Rieder 2+ fouls committed
    Attacking midfielder · Eintracht Frankfurt v FC Augsburg Tough match-up give up the 3rd-fewest fouls of 18 1.6 a game
    7/42.75 bet365 5 of 10 games 2 of 10 for a typical starter Foul map → Pro
  9. Bright Osayi-Samuel To be carded
    Full-back · Birmingham City v Wolves Soft match-up give up the 1st-most cards of 24 0.4 a game
    13/53.60 bet365 4 of 10 games 1 of 10 for a typical starter Player stats → Pro
  10. Jon Ander Olasagasti 1+ shots on target
    Central midfielder · Málaga CF v Levante UD Soft match-up give up the 1st-most shots on target of 20 0.6 a game
    11/102.10 bet365 5 of 10 games 2 of 10 for a typical starter Shot map → Pro
  11. Luke Shaw 1+ fouls won
    Full-back · Everton v Man Utd Even match-up give up the 8th-most fouls of 20 1.1 a game
    11/102.10 bet365 8 of 10 games 6 of 10 for a typical starter Foul map → Pro

Baseline = the share of player-games where a starter playing 60+ minutes met the same line, over finished fixtures in that league's most recent full season. Hit rates and match-ups are free — the model's verdict is the paid tool. See every player on the slate →

How this weekend's referees compare

Appointed officials

A cards bet is partly a bet on the official. Each official's career card rate against the average for that official's league, pulled toward that average where we hold fewer matches.

More cards than the league

Fewer cards than the league

Career figures across every season we hold, against that league's average over the same matches. A referee's tendency is one input the model applies to a player's own booking rate — it does not decide a card on its own. All officials →

617,862
predictions graded on matches the model never trained on · all 7 leagues
90/119
markets that beat guessing · across the leagues
5
seasons of data per league

From the live track record — how accurate the numbers are, not how much they won. Wins and losses alike: the model grades itself in public.

Why only 7 with a graded record? Because a market only ships once it beats guessing in that league. 5 seasons of match-by-match data sit behind each one. Of 119 markets tested, 90 earn a verdict — and the 29 that don't are marked on the track record, not quietly dropped.

7 competitions, each with a public graded record

The long version, with the actual statistics, lives on the methodology page — and the guides walk through the ideas one at a time, in plain English. Who's behind this, and where to get help if the fun stops.