F1 2022 Mid Season Review and Driver Rankings

As we are now in the Summer Break, I thought it’d be worthwhile to see how the Mathematical Model has ranked drivers for 2022 so far, as well as the evolving trends over the 2022 season. For any new readers (welcome!), the model looks at results from the entire history of F1 to assess how strong each driver is based on their results versus their teammates. By doing this it can rate and rank drivers regardless of the car they drive.

Predictions vs Reality

Prior to the start of the season I posted some predictions of how teammates would perform against each other. So far, the predictions have been a bit of a mixed bag. Ferrari is actually totally accurate. Others are not exactly right but in the right direction relative to their 2021 results (Red Bull, AlphaTauri, Aston Martin). There a couple of dud predictions too, for various reasons (McLaren, Mercedes, Alpine and Williams). We’ll review again at the end of the year to see if anything’s changed.

2022 Predicted numbers vs actual results.
2022 Predicted numbers vs actual results.

There is a general trend that the more points a team has scored, the more accurate the prediction appears. This can be taken in a few different ways. Being generous to the model, more points means that freak results have less of an impact on the overall percentages, meaning the points scored are likely to be more reflective of their relative abilities.

A less generous interpretation is that the amount of variation is going to be a lot less in competitive teams (e.g. Albon has scored 100% of Williams’ points, but such a scenario would be basically impossible at Ferrari regardless of who the drivers were.)

I thought it’d be useful to look at a couple of examples of how a driver’s career rankings have been adjusted due to 2022, how factors outside of their control (such as the performances of previous teammates) can cause adjustments, and some more general information of how the model works in practice.

Fall of Red Bull Academy Drivers.

Verstappen and Ricciardo have both underperformed (in terms of results relative to their teammates) compared to the model’s expectations, whilst Sainz’s rating this year has varied between achieving expectations and underperforming. As these 3 drivers are all linked via previous teammates, the lowering of one ranking is further fueling the lowering of the others too.

By comparison, the 2021 season did little be dampen Ricciardo’s overall ranking, as his poor results were somewhat insulated by Verstappen and Sainz exceeding expectations. (Although Ricciardo never partnered Sainz, the link to Norris important here).

The slight lowering of Verstappen and Ricciardo’s rankings has knock on effects for other Red Bull drivers such as Gasly and Albon. Whilst Albon has actually exceeded expectations in his comeback against Latifi, he has seen his career rating as a whole has actually decrease.

The changes in all of these driver’s overall ranking is not dramatic, given that half a season of additional data is not a lot. For example, the 2021 rankings are unchanged for Verstappen, Sainz and Tsunoda, with Ricciardo falling one place (to Russell) and Gasly being leapfrogged by Vettel.

Sainz’s 2022 Rating

This one is not much a trend but a good example of how the model works and how it corrects itself. There’s no doubt that Sainz’s 2022 season has been a mixed bag, and I know several people that have had changing opinions of the driver as the year has evolved. The model has also had changing opinions, with his results after 4 rounds being considered one of the least impressive on the grid (due to a couple of DNFs) before a remarkable turnaround. It’s surprising that Ferrari’s scores are the ones that (currently) match the 2022 prediction given how much their relative points have fluctuated due to unreliability and other factors.

The graph below shows how Sainz’s and Leclerc’s ratings are dependent of each other, with Sainz taking larger swings downwards/upwards than Leclerc due to the fact that the most likely explanation was considered to be Sainz underperforming relative to his ability. The idea that one driver’s score goes down as their teammate’s goes up is a core part of how the model works, and although the swing is quite extreme here the general principle is as expected. However, the model doesn’t always follow this simple rule of teammate dependency due to other factors (as we shall soon see).

Graph showing Leclerc and Sainz's ratings. Leclerc's ratings increase as Sainz's decrease, and vice versa.
Graph calibrated so that Leclerc’s current 2022 rating is given a score of 1.

Russell’s Growth

One of the big questions going into the year was how Russell would compare to Hamilton at Mercedes. There’s no doubt that Russell has fared remarkably well, and the model’s rating of him has correspondingly shot up due to how little data the model previously had.

The graph of Russell’s 2022 scores relative to Hamilton’s (calibrated so that Russell’s current score is equal to 1) conveys a lot if information.

Graph showing Hamilton and Russell's ratings. For a period of time both their scores are improving.
Graph calibrated so that Russell’s current 2022 rating is given a score of 1.

The first surprise is that for many races both driver’s scores are increasing. How is this possible? Remember that the model has an overall rating for a driver’s total career that is then adjusted for a given season. Although the situation is complex, a simple explanation is that Hamilton’s score is increasing as his percentage of Mercedes’ points also increases.

Meanwhile Russell’s score is increasing due to the fact that the model’s overall perception of him as a driver is improving. You can also see that Russell’s score plateaued and then began to fall. This is partly due to Hamilton scoring the majority of Mercedes’ points in the last few rounds, and partly because the model’s correction of Russell’s abilities is probably mostly complete.

Another consequence of Russell’s strong results is that Hamilton’s career rankings have decreased slightly relative to the rest of the grid. As the scores are calibrated relative to Hamilton’s career (with a mean year for Hamilton equal to a score of 100%), this has the effect of raising almost every other driver’s ranking slightly ( the difference is less than 1 percentage point increase for most drivers).

Finally, let’s look at how Russell’s rise has affected Latifi, a driver that the model has always struggled to place due to a combination of a lack of experience, inexperienced teammates and unrepresentative results. Prior to this year, Latifi’s ranking was based solely on Russell’s, so Russell’s improving rating has seen Latifi rise as a driver too. This is despite a string of relatively poor results vs Albon in 2022, and Albon’s own ranking being lowered due to the aforementioned fall of Red Bull drivers.

Latifi’s career therefore provides some contradicting inputs, with his results across 2020-21 being surprisingly close to Russell’s. The model has (probably correctly) concluded that Latifi overperformed vs Russell (or to flip the argument, Russell failed to get the points at Williams that he should have done given his talent), but the exact extent of this is still being assessed.

Notes on Rookies and their Teammates

As Alfa Romeo are fielding a rookie driver, Zhou’s rated is dependent solely on this years comparison to Bottas, whilst Bottas’ ranking is unfortunately only based on his previous career.

A similar situation exists for Haas, whereby Schumacher has no comparison linked to the rest of the grid except for Magnussen. The general perception seems to be that both Bottas and Magnussen are performing above expectations. If this is true, it means that all 4 drivers will be underrated slightly by the model.

2022 Driver Rankings

C Tier

20) Zhou Guanyu, 42%

Zhou has shown a lot of potential, but has been lacking in terms of points. Despite this Alfa Romeo have been impressed by the Chinese driver, particularly in his ability to go through the steep learning curve of F1 whilst not having major crashes or mistakes that one that expect from a rookie.

One positive sign for him is that the 3 rookies of 2021 were the bottom 3 ranked drivers that year, emphasising the difficulty of adapting to Formula One. Once experience is taken into account, the model currently believes that Zhou has a slightly higher potential than all 3 of them.

Given all this, is the model being too harsh on Zhou by placing him last? Perhaps, but it lives and breathes by results, and he’s scored fewer than 10% of Alfa Romeo’s points. Whilst teammate Bottas has been performing well, few would consider him a top line driver and the model would expect any other current driver to have scored more points in the car.

19) Mick Schumacher, 54%

Schumacher had a lot going for him at the start of the season, after gaining experience, easily outpacing his teammate in 2021 and finding himself with a much more competitive car. However, the start of the 2022 season was poor, with several mistakes that are particularly costly during the current cost cap era. His lack of points was also holding Haas back in the constructors championship.

Much improved form and confidence in the last few races have ended questions about whether he is F1 material, and delivered an obvious jump in his rating too (see graph below), even if it’s yet to translate into a boosted position. Like Zhou, his current scoresheet vs a non-top tier driver is disappointing on paper, but the trend for both of them is probably upwards.

B Tier

18) Kevin Magnussen, 62%

Drafted in at the last minute for due to global geopolitics, Magnussen instantly impressed, scoring points in 3 of the first 4 races. As he’s facing a driver with no connection to the rest of the grid, there is little room for the model to fully assess how impressive his results are. Unfortunately the model does rate his previous efforts as particularly spectacular, which fixes him stuck in the bottom half for the remainder of the year.

17) Yuki Tsunoda, 63%

After a tough first year, the Japanese driver has looked more sure footed in 2022. Tsunoda has been delivering more consistent results for AlphaTauri and has been noticably closer to teammate Gasly than in 2021. He’s earnt his place within F1, but withdrawal of Honda and the renewal of Perez’s contract at Red Bull raise questions about his long term career path, even if he can keep delivering results.

Graph showing driver ratings over time for Tsunoda, Magnussen, Schumacher and Zhou

One thought on “F1 2022 Mid Season Review and Driver Rankings

  1. Hey! I really like your content!

    Just a heads up: you need to do something about the posts layout. The “pages” links at the bottom are barely visible and didn’t find out they exist until just now (I used to just click on the index links to see the rest of the content).

    Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s