FAQ

A blind football draft game inspired by FIFA Ultimate Team and 82-0.com.

Pick from one of the top five leagues and choose a formation. Each round rolls a random club and season, and you pick one player from that squad to fill a slot in your locked formation. Once your XI is complete, a points-per-game model uses advanced analytics and standardised player market valuations to estimate your team's quality, which is then extrapolated over a 38-game season.

Your XI is scored by a predicted-points model trained on real team seasons across the top European leagues.

Attacking output is estimated from your players' individual expected goals (xG), expected assists (xA), and buildup involvement. Defensive quality is estimated from standardised market values, since individualized and genuinely informative defensive stats aren't reliably available.

The final score is clipped to the 0–114 range — 114 being a perfect 38-win season.

npxG (non-penalty expected goals) measures a player's ability to get into dangerous positions and take quality shots, minus penalty kicks. Chances closer to goal and at better angles are worth more than speculative long-range efforts. A high npxG player consistently creates shooting opportunities for themselves.

xA (expected assists) estimates how likely a player's passes and movements are to result in a goal for a teammate — even when those chances aren't converted. A high xA player is a reliable chance creator, regardless of whether their team finishes well.

xBuildup measures a player's involvement in attacking sequences that led to a shot, but specifically excludes their own direct contributions (shots and key passes). It captures the "invisible" work of players who move the ball into dangerous areas and help construct attacks without being the one who pulls the trigger or threads the final ball. Midfielders and ball-playing defenders tend to rank highly here.

All three stats are sourced from Understat, which tracks every shot and key pass across the top European leagues since 2014/15, and are shown on a per-game basis.

The model estimates a team's expected points per game based on player quality, but your squad is meant to reflect how players actually performed in a specific season. If a player was excellent in the games they played but missed a lot of the campaign through injury or suspension, drafting them comes at an opportunity cost.

If your players average 30 or more games played, you'll receive a bonus point for every additional average game above 30. For example: if your squad averages 31 games, you'll get one extra point on top of your score. If you field a team of true 38-game ironmen, you'll receive the maximum 8 bonus points.

You'll also receive a reward if you roll a club in a season that they won the title. To account for the unquantifiable title-winning magic that pushes a team over the line, you'll get an extra 2 points for every title winner in your squad, up to a maximum of 5 individual title winners (10 bonus points).

Market value data comes from Transfermarkt, gathered at the relevant point in each season. To make values comparable across seasons, leagues, and positions, each player's value is converted to a 'z-score' within their position group and season, so Kylian Mbappé in 2022 isn't naively compared to Virgil van Dijk in 2018.

Because Premier League players tend to carry higher valuations than counterparts in other leagues, this standardisation also means your La Liga or Bundesliga draft is fully independent and fair. World-class players in any league are rewarded, but a standout player in their own context will make a bigger difference than a raw valuation system would suggest.

Finally, young players often carry inflated valuations reflecting future potential rather than current impact. An age adjustment is applied to discount value for players who haven't yet reached their prime years. This means you can't simply draft pricey wonderkids if you roll a team who loves an ill-advised 100m signing *ahem* ChelseAndManUnited.

The standardised score you see is shown as a z-score (σ): a common statistical measure of how many standard deviations above or below the average a player sits. The Greek letter sigma (σ) represents one standard deviation. Pay special attention to this score for defenders, because this is the primary predictor of lower expected goals allowed in your team, which is crucial for a high overall score.

In blind draft mode, you pick your formation before the draft starts, and then fill in the slots as club seasons are rolled. Since you don't know what club or season you'll get in the next round, you have to choose your players carefully! You never know when you'll get Barcelona in 15/16 or Burnley in 17/18.

Each game starts with three rerolls (not included in World Cup mode). Spending one skips the current club-season and rolls a new one if you really can't stomach the players in that squad.

Instead of club seasons, each round rolls one of the 48 nations competing at the 2026 FIFA World Cup. Squads are sourced from the actual 2026 tournament rosters. Scoring is a composite index based on market value, international caps, and international goals. Since not all players in World Cup squads hail from the top five European leagues, the scoring model is simpler, rewarding market valuation, international impact and experience.

Tiers are calibrated against real points totals from the top European leagues:

🏆 80+ Points: Title Contenders — it could just be your year
🥈 62–79 Points: European Football — Qarabağ away, anyone?
🥉 54–61 Points: Top Half — comfortably mid-table
⚽ 45–53 Points: Bottom Half — looking over your shoulder
❌ 0–44 Points: Relegation Battle — watch out for the drop

Tiers are calibrated against simulated gameplay — Champions requires a top 5% draft, Semi-finalists a top 35%, and so on:

🏆 World Cup Champions — add a star to your kit
🥈 Finalists — close but no cigar
🥉 Semi-finalists — don't forget the third place game!
💜 Quarter-finalists — at least you made it to July
⚽ Round of 16 — the El Tri special
❌ Group Stage Exit — just happy to be here

Understat — xG and xA per player per game, plus other advanced metrics, on a per season basis (2014/15 onwards).
Transfermarkt via the dcaribou open dataset — market valuations, international caps, clean positions, roster data.
Wikipedia — 2026 World Cup squad lists.

This is an independent project and is not affiliated with the Premier League, FIFA, Understat, or Transfermarkt. Market values are crowd-sourced estimates, not transfer fees.

A good squad needs balance; a great attack paired with a weak defence will be punished on the predicted-points side. Defensive quality, which directly predicts your expected goals allowed, is estimated from market value, so low-quality defenders will drag the score down even if your forwards are world-class.

This is a project by Chris Bull (GitHub). Formerly a data scientist in the consulting industry, he is currently a grad student at the University of Michigan. He is a lifelong football fan and decade-plus long victim of Tottenham Hotspur Football Club. Feel free to reach out with any feedback at christopherwbull@gmail.com.