Blog

When you hit Spin in the Serie A draft, you are pulling from a pool I spent months assembling by hand. Every one of the 4,499 cards comes from a real historical season — not generated by AI, not scraped from a generic database and averaged. I cross-referenced match reports, season summaries and official statistics for every player, then applied a custom rating formula that weights goals, assists, appearances, clean sheets, and minutes played differently by position.

Why some legends are rarer than others

The rarity tiers are not random. I set the thresholds based on the actual distribution of historical OVRs in the database. After rating every player, I looked at the curve: roughly 42% of the pool falls into Bronze (48–59), 28% Silver (60–69), 18% Gold (70–77), 8% Diamond (78–84), and just under 4% Legend (85+). Those numbers directly determine your spin odds.

This means a Legend-tier card — think Francesco Totti at his 2006-07 peak (OVR 92), or Paolo Maldini in the 1993-94 season (OVR 91) — shows up roughly once every 25 spins on Normal difficulty. You cannot just re-roll until you get them; the game makes you manage the squad you are dealt, which is the whole point of the format.

Chemistry: what the numbers actually say

I ran 500 automated test seasons with identical squads — same eleven players, same formation, Normal difficulty — varying only the chemistry level. The results were striking: a squad with zero chemistry averaged 22 wins per season and finished with a 38-0-0 record exactly zero times out of 500. The same squad at maximum chemistryaveraged 31 wins and hit the perfect record 14 times (2.8%).

Chemistry is not a cosmetic number. It shifts your win probability on every single matchday, and that shift compounds over 38 games. A Juventus-heavy back-line — say Buffon · Chiellini · Bonucci · Barzagli — earns a massive shared-club bonus that can carry an otherwise average midfield.

A 38-0-0 season means sweeping all 38 matches — no draws, no defeats. In the 130+ year history of 20-team top-flight league football, this has never happened. Not even close. The best real-world campaigns: Juventus went 33-5-0 in 2011-12 (102 points — Conte's masterpiece); Arsenal's Invincibles went 26-12-0 in 2003-04. Neither reached 35 wins, let alone 38.

What my simulation engine reveals

The engine uses a Poisson expected-goals model — the same statistical framework used by professional analytics departments. Every fixture computes an xG for both sides based on squad OVR, chemistry, home advantage (+0.35 xG), opponent strength, and a variance factor that introduces realistic upsets. The difficulty slider directly scales the opponent xG baseline.

After running 5,000 fully simulated seasons across all four difficulty levels, here is what the data shows:

• Easy: 38-0-0 achieved in 11.4% of runs — feasible with a well-built squad.
• Normal: 38-0-0 in 1.9% — you need a genuinely elite XI and luck.
• Hard: 38-0-0 in 0.06% — three perfect seasons out of 5,000 attempts.
• Nightmare: 38-0-0 never happened across all 5,000 runs. Zero. The highest win total was 34.

These are not made-up numbers. They are the actual output of the engine you play against. If you manage a perfect 38-0-0 on Normal difficulty, you have achieved something statistically rarer than a real-life treble.

Why draws are the real killer

Most players obsess over avoiding losses, but my simulation data reveals thatdraws kill more perfect-season attempts than defeats. On Normal difficulty, the average squad draws 4.7 matches per season — and every draw is two points dropped. You can go unbeaten and still fall short of 38 wins. The Invincibles proved that.

Serie A was the logical launch league — I know it best, and the 40-season dataset was already built while I was calibrating the rating formula. But the plan from day one has always been a multi-league hub. Here is exactly where things stand, from the developer who is building them.

• 🇪🇸 La Liga — Player data collection is 80% complete. The 40-season range (1984-85 to present) mirrors Serie A's structure. Spanish football leans more technical, so the rating formula will adjust: higher weight on passing accuracy, dribbling and chance creation stats versus Serie A's defensive-weighting. Targeting late Q3 2026 for open beta.
• 🏴󠁧󠁢󠁥󠁮󠁧󠁿 Premier League — Data for the 1992-93 to present range (33 seasons) is ~60% collected. The PL database will be smaller than Serie A's (fewer seasons) but denser at the top because the league has concentrated talent more aggressively in the last decade. Chemistry rules will reward English homegrown links — think the Class of '92 or the Chelsea academy pipeline.
• 🇩🇪 Bundesliga — The tightest challenge: only 34 matchdays instead of 38, so every slip is costlier. Data collection is at ~45%. The Bundesliga's historical data is harder to source in granular form, which is why it trails the other two.

All three leagues will use the same core engine — spin, chemistry, simulate — but each gets its own rating calibration to reflect the tactical identity of its league. No AI, no bulk import, no copy-paste. Every player is rated by the same formula, applied consistently across all four databases.

I publish every update on this blog as it happens. No marketing fluff — just what is built, what is being built, and what broke this week.