Formula Onigami - The case for Scorigamis in Formula 1

If you happen to be both a sports fan and slightly too addicted to youtube, you might have come across the term Scorigami - a concept invented by Jon Bois to denote “the act, and art, of producing a final score in a football game that has never happened before”. These Scorigamis were originally just funny statistical oddities, but after Secret Base’s viral video essay on Scorigami, these unique scoring events have gained a life of their own:

Nowadays it’s a massive deal¹ when a Scorigami occurs in an NFL game, to the point it might even be mentioned on the official broadcast! But as more of a Formula 1 fan myself I kept wondering: how cool would it be if we had Scorigamis in F1?

The hunt for a good Scorigami statistic for Formula 1 Link to heading

Formula 1 is very different from NFL, and that means that we need to put some more thought into what a F1 Scorigami should look like.

We have to consider how often these Scorigamis happen - if they happen too often the pool of available Scorigamis will be exhausted too quickly, but if they don’t happen often enough people lose interest in them. The goal we should strive towards is a rate of single-digit Scorigamis per year, so that the threat of a Scorigami occurring in any given Grand Prix is ever-present.

Other factors to keep in mind: these Scorigamis should be meaningful and easy to understand by the average F1 fan, and they should be able to happen in every Grand Prix². The possibility of “cursed” Scorigamis should also exist: things that will either never happen again, or would take divine intervention to happen again.

Taking all of this into consideration, I have an idea:

My Scorigami proposal: Points scored over a Grand Prix weekend Link to heading

I have been thinking about this on and off for a couple of months, and I believe this is the right foundation for us to build our Scorigami castle on:

• We can track Scorigamis per team, which I will do in this blog post
• The Scorigamis can happen in every Grand Prix weekend
• Easy to calculate: just add the points scored by the team’s drivers across the weekend
• Decent combinatorial potential: in a sprint weekend, there are 4 point-scoring events (2 drivers * 2 races). This leads to the existence of rare point scores that can only be achieved by finishing in very specific positions in both the sprint and the race
• Potential for weirdness

What teams to include? Link to heading

It’s obvious that the teams currently in the grid should have their Scorigamis tracked. What is less obvious is how far back through a team’s history we should go. Take Mercedes, for example: this team can trace its roots back to Tyrrell which started competing in F1 under its own name in 1970³, which would predate Williams’ presence in Formula 1 by ~5 years. That would mean that Mercedes would score Scorigamis before Williams!

You might consider this to be absolutely nonsensical, and argue that a team’s “Scorigami history” should only begin from its last ownership change: well, now you have wiped away the legacies of Team Enstone (Alpine) and Team Silverstone (Aston Martin) in Formula 1.

This is a thorny subject with no right answers. I’ve decided to be generous and include as much of the teams’ genealogy as possible:

OLD_TEAM_NAMES = {
    # Teams with stable branding
    "McLaren" :  [(0, "McLaren")],
    "Ferrari" :  [(0, "Ferrari")],
    "Williams" : [(0, "Williams")],
    "Haas F1 Team" : [(2016, "Haas F1 Team")],

    "Red Bull" : [
        (1997, "Stewart", 1999),
        (2000, "Jaguar", 2004),
        (2005, "Red Bull")
    ],
    "Mercedes" : [
        (1968, "Tyrrell", 1998),
        (1999, "BAR", 2005),
        (2006, "Honda", 2008),
        (2009, "Brawn", 2009),
        (2010, "Mercedes")
    ],
    "Aston Martin" : [
        (1991, "Jordan", 2005),
        (2006, "MF1", 2006),
        (2006, "Spyker MF1", 2006),
        (2007, "Spyker", 2007),
        (2008, "Force India", 2018),
        (2019, "Racing Point", 2020),
        (2021, "Aston Martin")
    ],
    "RB F1 Team" : [
        (1985, "Minardi", 2005),
        (2006, "Toro Rosso", 2019),
        (2020, "AlphaTauri", 2023),
        (2024, "RB F1 Team")
    ],
    "Sauber" : [
        (1993, "Sauber", 2005),
        (2006, "BMW Sauber", 2009),
        (2010, "Sauber", 2018),
        (2019, "Alfa Romeo", 2023),
        (2024, "Sauber"),
    ],
    "Alpine F1 Team" : [
        (1981, "Toleman", 1985),
        (1986, "Benetton", 2001),
        (2002, "Renault", 2011),
        (2012, "Lotus F1", 2015),
        (2016, "Renault", 2020),
        (2021, "Alpine F1 Team")
    ]
}

Calculating Scorigamis for 70 years of Formula 1 Link to heading

Now that the finer details are figured out, we need to figure out how we’re going to process 70 years of Grand Prix racing. I won’t go through the technical details today, but I’ll give you the gist of it: I downloaded a data dump from the Jolpica API and then developed a python script that takes the information from this data dump, and creates a table that keeps track of the points scored by each each team for every Grand Prix, and the number of times that score has occurred for that team.

Here is the table with the Grand Prix scores of every team that currently exists on the grid, and the number of times that score has occurred previously (html, csv):