Is there a best number of split times when looking at motorsports data analysis?
If you are looking at motorsports logged data then you maybe “split” the lap into sections. These sectors or split times enable you to understand more and start to explain the overall lap time.
What you want is to know is how well your driver performed in order to focus on opportunities to improve. Unfortunately, a lap time alone is too crude. You want a way to help you isolate, analyse and prioritise where your driver can get better. Too much happens around a whole lap to give you much information.
In fact, I have often had situations where overall lap times are similar but they were actually driven very differently.
Why Split Times?
If you are to improve your race driving you need more information about what went into your lap times. Split times are a great way to start that process.
Most motorsports data analysis software enables you to create sectors. Some software even tries to do this for you (eg AIM Studio)
If you are looking to use split times to help your analysis, where should you start? Are the automated sectors any good? What else should you consider?
This article aims to help and guide you through choosing the best type and best number of split times.
Less Is More
It might seem like more is better with sectors. After all, if you take smaller and smaller sections then surely you are going to be able to see more and more detail? More detail is better right?
The AIM studio takes this approach creates a lot of sectors – each straight, each corner gets its own split times.
The issue with this approach is that the smaller the segments the less reliably you can use them.
The reason is that you can easily, say, overshoot your braking one lap and run off the track. Your small segment will show you completing that split in record breaking time (yay!) The issue being that you cannot do a segment in that time and still keep the car on the track!
The result is that the small segment, whilst your fastest time through that portion of the track, is completely useless for your driver analysis.
You need information that can help you put together the puzzle of completing the best actually possible lap time.
You do not need to know a completely imaginary time that would be impossible to achieve in reality.
When you are trying to determine the best number of split times, less is definitely more.
But how many?
Recommended Best Number of Split Times
In my experience, the best number of split times is 3 to 5 sectors of approximately 20 to 30 seconds in duration.
Furthermore, to make them more reliably comparable, you should try to cut them on the straights and just before your braking zone.
The reason for this is that your driver will not be involved in any significant skill based driving activity on the straights (compared to braking zones and corners.) Therefore they are likely to be similar each lap.
You want to completely avoid cutting splits mid corner or in braking zones.
If you do not have any straights then consider “meaningful chunks” of the track. Chunks that take around 20 to 30 seconds to complete.
Choosing Split Times – Worked Example
If you’ve never done this before then here is an example. It is part of a more detailed worked example I take students through on my Complete Beginners Guide To Motorsports Data Analysis. 👈 Worth a look if you are new to data analysis.
1- Start with the Track Map:
2- Look at the track and identify the straights.
3- Pick a point towards the end of the straight before you typically begin to brake. Pro Tip: You can use the speed trace to help you in your analysis software for this.
4- Below you can see the 5 sections I have chosen for this track.
Walkthrough Of Why These Splits Are Chosen
In more detail, here are how the five sectors end up:
So split 1 is from the start / finish line and around the first corner. This is a key corner at this track. The drivers performance through here defines a lot of how the next half of the lap will go. The split is not technically on a straight. It is in an acceleration zone though and one the driver will consistently drive the same line through.
Split 2 is similar, in so far as it is ends not on a straight but in an acceleration zone. This section though is a logical and linked driving event. The end of the split is the end of that event. For the driver this point is a slight mental pause before the next section so makes a good split point.
Split 3 takes in a key fast left into right right corner combination. This corner is off camber on the exit and is another one where there is a lot of time to be gained or lost. Knowing the drivers performance through this section can help you identify a lot of consistency issues and also potential car handling issues. Therefore that is what I mean by a “meaningful chunk” of the lap.
Split 4 starts just before the braking zone for this key right hander. This corner leads down the long back straight. Knowing how the driver navigated this section is key to understanding where they might be able to gain more time. If the driver was consistent here but on one lap went much faster, you might check to see if they had a slipstream before assuming a better drive. Again the aim is to make sure you are always making fair comparisons in your data analysis.
Finally Split 5, starts just before the braking zone for the chicane. The split nicely encapsulates the drivers performance through this event and see them over the line.
Knowing the best number of split times is not a straight forward decision. What you want is the ability to understand where the driver has opportunities to improve. Split times are a good place to start in your data analysis. The splits however have to be meaningful and relevant. Too many splits and you will end up getting mislead by what you might think is possible. Too few and you are no better off than just the lap time on its own.
My suggestion is to:
- pick between 3 to 5 split times,
- avoid mid corner or braking zones for your sectors splits.
- segmented along straights (or fairly straight acceleration zones)
- aim for segments that are around 20 seconds long.
I hope this helps you gain more confidence when considering how to use split times and sectors for your data analysis?
If you use AIM studio or another software that auto generates segments, just make sure that the results you get follow reasonably close to these recommendations or else you are potentially going to get confused.
Check out the further reading suggestions below. Also consider signing-up to my weekly(-ish) newsletter for the latest articles delivered direct to your inbox.
Where do you go after looking at split times? Try the delta-t channel. Learn about it here.
Are you using all your racecars grip? Know how to get this info from your logger? Here’s the article.
No clue where to start with motorsports data analysis? Could this be the course for you?
Run an AiM logger? Know how to create custom sectors in AiM Race Studio? Full guide here.