MS
- Nov 7, 2020
- 3 min read

Comparative Lap Time Distribution for 2020 Top Race Finishers (as of Eifel Grand Prix)

This work displays a representative race lap time distribution comparison for the top 5 finishes in each race though to 2020 Eifel GP. In having to present this data some “outliers” have to be taken off from the raw lap times that are recorded. A data treatment process was first used and then the filtered data is plotted as shown in the ridgeline plots below.

Outliers in this context are laps that are vastly different from the average lap time per individual driver. These include but are not limited to the very 1st lap, safety car laps, laps going into and out of pit stops, etc. While generally the higher variance is limited to these factors, there also is a possibility driver/car errors resulting in longer lap times. To best accommodate these outliers, I’ve used a modified IQR (Interquartile Range) method based off the average lap time (50th percentile) of each driver. For the upper outlier cutoff limit, I’ve used the 85th percentile (anything above 35% of the mean lap time is relegated as an outlier) as opposed to the standard 75% to possibly include any naturally slow laps without skewing the data too much. The number of outliers are marked in the notes of each plot along with the IQR scale (1.7) used to obtain the current filtered data. As for fast laps, no such cutoff limit was set as this would practically be improbable. This takes to the fact that a fast lap, however much faster than the mean lap time, cannot be discarded form the representative data as such data would indicate a significantly better lap from a driver.

If we look at Tuscan GP, the variation in Alex Albon’s race time is stretched out quite a bit (101sec cutoff limit) compared to the rest (86s, 94s, 96s, 95s), this results in a more lumped up and skewed plot compared to the others races presented in this work. In analyzing data, a low standard deviation indicates that the values tend to be close to the mean of the set whereas a high standard deviation indicates that the values are spread out over a wider range. This is seen to be true in the distribution of Albon's lap times where his average is lap time is lowest almost the top 5 and has the highest standard deviation which is double the average of the rest. Also with the Tuscan GP, the multiple restarts probably didn't help the cause either for consistency.

σ_ham = 1.314s AvgLap_ham = 82.178s
σ_bot = 1.633s AvgLap_bot = 82.493s
σ_alb = 3.306s AvgLap_alb = 83.414s
σ_ric = 1.619s AvgLap_ric = 83.029s
σ_per = 1.675s AvgLap_per = 83.391s

As for the fastest lap it is mostly shared between Hamilton, Bottas and Verstappen. Bottas has some fastest-laps over Lewis, say for example in Spain, that probably is attributed to that extra pit stop towards the end of the race to gain the extra point when the win was outside of the packet (pit stop: L64/65, fastest lap: L66/66).