Skip to content

Instantly share code, notes, and snippets.

@kevinburke
Last active April 7, 2024 05:13
Show Gist options
  • Star 0 You must be signed in to star a gist
  • Fork 0 You must be signed in to fork a gist
  • Save kevinburke/190b4c7fedfae12bc8e115519f4a9541 to your computer and use it in GitHub Desktop.
Save kevinburke/190b4c7fedfae12bc8e115519f4a9541 to your computer and use it in GitHub Desktop.

F1 market resolution criteria

Last edited: April 6, 2024

The results of a race can change after the finish for several reasons.

  1. Cheating writ large, for example, stealing from another team, deceiving the race officials, or running an out-of-spec car, where the penalty is not applied until days or weeks after the race. This type of incident, while unfortunate, will not change the outcome of a market.

  2. Penalties applied after the race, for example, for exceeding track limits, impeding other drivers or speeding in the pit lane. Typically, by the end of the race it will be public knowledge that a driver is "under investigation" or suspected of a penalty for certain issues. Any time a market can be affected by an investigation that is announced within 5 minutes of the race conclusion, the market will not resolve until the penalty outcome is known.

  3. Car did not meet the post race technical specifications - for example, the FIA must be able to extract 1 liter of fuel from each car, each car must meet minimum weight requirements, and a wood plank under the car must be undamaged. It may not be clear an issue exists here for several hours after the race (Vettel 2021, Hamilton/Leclerc 2023). Usually the offending driver is disqualified.

While (3) obviously can happen, it is rare, and it would be inconvenient to wait several hours for the official classification to resolve all markets - everyone would have their money tied up until the official results come out.

For penalties in category (3) to affect the outcome of a market, one of these criteria must be met:

  • A market participant leaves a comment within 5 minutes of race end asking me to wait for official classification because they suspect that a driver may be disqualified.
  • It was possible to know or suspect a penalty was coming during the race, for example, team asking driver to save fuel or stop the car immediately after the finish line, or a reporter posts a Tweet saying it may be questionable, AND someone in the market takes a position that could be perceived to be a bet on that outcome (for example, buying NO at 99.9%).

If a person (not a general-purpose bot) takes a position that could be perceived as a bet on post-race disqualification, but there was no public evidence a disqualification was coming (ie none of the reporters, TV, F1 Reddit, or other teams commented on the potential for it), market will resolve on a case-by-case basis.

News comes in after a race while the market is still open

I try to resolve all markets about 5 mins after a race ends. However, if for some reason I can't resolve a market, and new news comes in that could affect the outcome - for example, an announcement of an investigation into a driver - feel free to bet on it and leave a comment explaining your bet. I will take into account betting activity when deciding whether to resolve a market in the face of new news.

Season long or multiple race bets

If a bet applies to multiple races over multiple weeks for example, "Will Red Bull win every race in 2023," then category 3 disqualifications will absolutely affect the market's resolution.

Drivers must compete in the race weekend

If a driver does not participate in a race weekend - defined as, a replacement driver has been announced before the Thursday press conferences - markets involving them will resolve N/A.

In general

I'm going to resolve markets using the spirit of the question and, as best as possible, the spirit of the bets in the market, to decide how to resolve them. For example, if I incorrectly say a "Who will win" market "Resolves based whoever is leading after 55 laps," and the race is actually 56 laps long, don't try to argue that the description matters more than the question title.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment