Sexy Football, Measured — This Time I Was (Mostly) Right

I have a football-geeky friend. We argue about football all the time — this player vs that player, this team vs that team. We split the difference on almost everything, and we both use AI to back up our own version. Last time the data proved me wrong.

This round the question was: has Slot’s Liverpool played sexy football?

My claim: only at the beginning, then it stopped. His claim: it has, full stop.

What is sexy football?

You can’t settle an argument about a fuzzy word. So first we made “sexy” measurable. Four ingredients, scored for every league match:

1. Chances created (non-penalty xG) — 35%
2. Passes into the box area (deep completions) — 25%
3. Winning the ball high (pressing intensity, PPDA) — 25%
4. Goals — 15%

Each match gets one number: 50 is average, 15 points is one standard deviation. A “sexy spell” is when the five-match rolling average stays at 55 or higher.

The data

I wanted to use StatsBomb open data. Plot twist: it contains no Liverpool matches from Slot’s time at all. The only Liverpool season in there is 2015/16. So the per-match numbers come from Understat instead — all 76 of Slot’s league matches.

The StatsBomb season wasn’t wasted, though. I used 2015/16 to check that the index works: it jumps right at Klopp’s first match in October 2015. A metric that detects Klopp arriving is a metric I trust.

The verdict

Liverpool under Slot was sexy in spells — four of them, about 106 days in total, and every single one in the 2024/25 title season. The peak was the 6–3 at Spurs in December 2024. In 2025/26 the rolling index never reached the sexy zone once. The season average dropped from 54.8 to 45.2.

So: I was right that it faded, and right that year two wasn’t sexy. But I was wrong about “only at the beginning” — the sexiest stretch came mid-season, December to February. My friend was right about year one, wrong about year two.

We split the difference again. Of course we did.

Bonus round: Slot vs Klopp

Since we had the index, we compared eras on one shared scale: Klopp’s first two years, Klopp’s last two, and Slot.

Klopp, first 2 years (Oct 2015 – May 2017)

Klopp, last 2 years (2022/23 – 2023/24)

Slot (2024/25 – 2025/26)

Era	Mean index	Days in sexy spells
Klopp, first 2 years	48.0	50
Klopp, last 2 years	53.2	202
Slot	48.6	87

Late Klopp wins, and it isn’t close. Early Klopp pressed the hardest of the three but created the least. Slot creates almost as much as late Klopp but presses far less — control football, which the index prices as less sexy.

What I learned

• Define the fuzzy word first. Half of our argument dissolved the moment “sexy” became four numbers with weights.
• Check the dataset before you promise anything. The source I planned to use had no data for the question I asked.
• A metric needs a sanity check. Validating the index on a known style change (Klopp’s arrival) is what made me believe the rest.
• Both of us were partly right. Again. Data doesn’t always pick a winner — sometimes it just tells you which half of your claim to keep.

---

What’s your definition of sexy football? I bet your weights would be different. Would love to hear them.

Best, Olek