OPINION: F1 2021's Driver Ratings need to change

OPINION: F1 2021’s Driver Ratings need to change

Jul 9, 2021 8 Min Read Posted by Jacob Hancox

F1 2021‘s driver ratings system lacks precision and consistency, and it needs a change.

Photo credit: Codemasters / EA

At present, it seems very likely that George Russell will replace Valterri Bottas at Mercedes-AMG Petronas Formula One Team for the 2022 Formula 1 season. Various media outlets are abuzz with the notion, and the paddock rumour mill is certainly churning. But why would Mercedes do this? Why would Mercedes replace the third best driver on the grid, according to F1 2021, with Russell, who ranks a lowly 14th best on the grid? The answer: F1 2021’s driver ratings are deeply flawed.

How the Ratings Work

In F1 2021, drivers are given ratings in four criteria, which then have a weighted average applied to find out the overall rating of that driver. The criteria are experience, awareness, racecraft, and pace. Issues with nomenclature and the murky difference between awareness and racecraft aside, this doesn’t seem to be a bad system on the face of it. A little simplistic, sure, but not outright bad.

Let’s look at how these ratings are determined. Supposedly, Codemasters uses cold, hard maths to ascertain the scores. For instance, experience is simply a measure of how many races a driver has participated in. This sounds fair enough, until one examines the actual ratings given to each driver.

Flaws in the System

George Russell, who has 46 race starts, earns an experience score of 60. Antonio Giovinazzi, who has 49 race starts, is only given an experience score of 54. There are multiple other examples of this same peculiar error. Pierre Gasly, who has an entire season and a half of race starts over Russell with a total of 73, has only got an experience score of 59. Esteban Ocon has nine more race starts than Charles Leclerc, yet his experience rating is two points below the Ferrari driver. If these scores are based only on real-world facts, how do they come out incorrect?

Awareness is the next category. Codemasters call this ‘A measure of the driver’s ability to avoid incidents and penalties in a race’. A little more arbitrary here, and open to human judgement and interference a little, but still mainly based on facts. At the time of writing, Lando Norris has the highest current penalty point total on the grid, he is just two penalty points away from a race ban. So he should get one of the lowest awareness ratings on the grid, right? Not a bit of it. Lando Norris has a perfect 99 in awareness.

Next up is racecraft, which is ascertained by looking at how many positions a driver gained or lost during a race, compared to the average for their grid position. This, at least as it’s advertised, is way too simplistic. It fails to take into account retirements or a lack thereof, bad strategies from the team or unlucky safety car timings which are also not mentioned. It’s possible that these are all taken into account, and if they are racecraft is probably the most accurate rating in the system, but if they are that information has not been made public.

Finally we have pace, by far the most important category for a racing driver to be proficient in. If it’s so important, then surely a complex algorithm is used to get accurate results that manage to separate the driver’s performance from their car’s, right? Wrong. According to Codemasters, pace is determined by measuring “a driver’s best lap times – the closer they are to the fastest lap in a race, the higher the score.” Oh dear.

Not only are drivers in a Haas F1 Team compared to drivers in a Aston Martin Red Bull Racing or Mercedes to measure their pace, apparently without consideration of their machinery, but a number of other potential variables are swept under the rug. What if a driver runs a strategy which seems them pitting onto soft tyres late in the race – do they therefore deserve a higher pace rating compared to their teammate who went from softs to hards on lap 16 and slogged it out from there? And what of qualifying, arguably the purest indicator of a driver’s pace that exists? According to the system, Yuki Tsunoda has a higher pace rating than both Carlos Sainz Jr. and George Russell. I doubt there are many people out there who would agree with that outcome.

What needs to be done?

Short answer: I don’t know. One of the beautiful things about Formula 1 is that it is so hard to analyse a driver’s quality as a separate entity from that of their car. Even more confusing is the fact that some cars suit drivers better than others – look at Daniel Ricciardo this season, or Gasly in the Red Bull back in 2019. These factors make rating drivers on their own merits an incredibly difficult task.

I’m neither a mathematician nor a statistician, so I’m not the person to come up with the formula for how to rate drivers in the F1 game series. The human hours necessary to truly scout and pick out a driver’s qualities and flaws are likely very high, and they are hopelessly tangled in subjectivity, so a mathematical solution does seem preferable, even if it occasionally gets something a little wrong. However, the equations and algorithms that Codemasters currently use are just too simple, and they fail to take a great deal of factors into account. It’s not game breaking, it’s certainly not the end of the world, but it could be better.

Do you have any ideas for how to improve the driver ratings system in F1 2021? Tell us on Twitter at @OverTake_gg or in the comments below!

Tags:
Jacob Hancox
View More Posts
My name is Jacob and I have been writing for OverTake since November of 2020. I come from the UK, but I'm now living in Berlin. I love to watch, write about and sometimes shout about all forms of racing.