There’s nothing more fun in the world than messing around with statistics, is a fun way of admitting that I have a limited social life.

Regardless, I do enjoy the occasional statistical model, so I made my own, the Teammate Performance Index. This measures ten different statistics between teammates, weights each one to 10% and finds an overall percentage performance between the drivers. Whilst the categories are completely made up based on personal preference and are weighted towards successful drivers (as wins, poles and championships are included), but my hope is that when I collect a wide range of drivers, it will create a clearer picture as to driver performance. Whilst the statistics will not be treated as gospel, as they are made up for fun, it will help to broaden opinions if used as a tool to compare driver performances.

With no ado, here is a template of the index.

	Driver One	Driver Two
Wins
Poles
Podiums
Points (Excl. Sprints&FL)
Point Positions
Qualifying H2H
Race H2H
DNF/S/Q’s
Fastest Laps
Championships

Wins:

This is self-explanatory, compare the number of wins each driver has over their time as teammates.

Poles:

Compare the number of poles each driver has over their time as teammates.

Podiums:

Compare the number of podiums each driver has over their time as teammates.

Points (Excluding Sprints and Fastest Laps):

Compare the points each driver has over their time as teammates. I excluded the sprints as they have only been in F1 for a few years, so are not a fair comparison to make between eras. I also exluded the fastest laps, both for this reason and because when fastest laps are later included as a category. I will use the modern points system between eras, as it creates a consistent value for each position and a consistent percentage gap between 1^st and 2^nd place.

Point Positions:

The amount of times a driver finished in the points paying positions. I am not using the modern points system for this category, as there is limited value applied to the position itself and the points have expanded as the reliability of the cars have expanded. Additionally, it also allows for both the modern points system and past points systems to be represented in the data of the model.

Qualifying Head-to-Head:

How many times each driver beat the other in qualifying before grid penalties are applied.

Race Head-to-Head:

How many times each driver beat the other in the races. I am including one driver finishing and the other not in the statistics, both due to the frequency of retirements before the modern era of F1 leading to the data set becoming incredibly small to compare drivers to and because many times a retirement is in part or in full that drivers fault, so should be considered as part of performance.

Did Not Finish, Did Not Start, Did Not Qualify, Disqualified:

The amount of times each driver met one of these conditions during their time as teammates. Unlike the other categories, the percentage is flipped when measuring it, so if the result was 2-1, the driver with 2 would receive 33.3% and the driver with 1 would receive 66.6%

Fastest Laps:

Compare the number of fastest laps each driver got during their time as teammates. For the classic era, when multiple drivers may have received fastest laps due to inaccurate timing data, all drivers will be considered to have had one fastest lap.

Championships:

How many championships each driver has. To note, if drivers receive 0-0 on any of the categories, they will both receive 50% as default, despite the fact that is not actually how maths works when dividing with zero. However, it is the most useful way of expressing the data in the model. Of course, this balances the model towards successful teams and drivers, as there are more complete data sets to work with, but I want to put my focus on comparing the more successful drivers, at the present time.

So, this is my new model. Its flaws will likely become apparent as I collect more data for it, but should be a fun way of looking through an comparing the statistical record. The next article I am writing will use this model, hence my introduction of it today.