Is Your Driving Actually Improving? The Metrics Lap Times Hide

If you think that simply looking at your lap time is the way to track your progress and your improvement — I think you're probably wrong. And that's the mistake I made myself at the beginning of my karting career. I would always look at the lap time, and I would always decide whether I'd had a good session or not based on the lap time.

But the reality is that your lap time only tells one part of the story, because conditions on the track change massively from session to session and from day to day. You can't really compare your lap time today against your lap time from yesterday, or two days ago, or two months ago, or six months ago.

Why lap time gives you false positives and false negatives

The track is always going to develop, either in a faster way or a slower way. You may find yourself on a track that's rubbered in with very high-quality rubber, perhaps very soft rubber, and you'll set really quick lap times — but maybe you're not driving well. Whereas another day you're driving much better in terms of your consistency, your gap to your theoretical, and your lap times are slower, because that day the track conditions were much worse.

So you can't rely entirely on lap time — which unfortunately 90% of drivers and parents do. I did too. But after all these years at the racetrack I've realized that merely looking at lap time isn't the solution, and it will actually give you false positives and false negatives.

It can trick you into thinking you had a really good session: wow, this lap time is so fast. When the reality is the track was just super quick, and if you'd put somebody else in that kart they'd have been half a second quicker than you.

Conversely, it can trick you into thinking you had a bad session, because your lap time is four tenths off your personal best at that track. But your gap to your theoretical is absolutely zero, your consistency is extremely good, you repeated the laps like crazy, and you were already on the pace on lap two — which is impressive. That may have felt like a bad session because of the four tenths. The reality is it was an outstanding session.

Relying on lap time is the losing strategy.

And there's a second problem: when you're driving, you always know your personal best on that track. Everybody does. So if you drive constantly thinking about it — I just did a 55.6 but normally I do a 55.1 — is that a good thought to have while driving, or a negative one? Of course it's negative, because you'll always be comparing against your personal best. And as I said, you can't really do that. One day you'll have new tires and the track will be super fast. Another day you won't have new tires and the track will be slow — you may be driving at your best ever and still not get within six tenths of that personal best.

Sections beat sectors

That's why at PURPL we built the dashboard so you can see how good your session was through several features, independently of the raw lap time.

The first is your gap to your theoretical. To make that meaningful you need proper mini sectors — I call them sections. In PURPL there are deliberately many more of them than the three you get from a MyChron, which are split evenly across the track by GPS. I wanted more, because that's how you can tell whether you did some really good corners on one lap and some really bad corners on another.

On a track like Lonato you'll find six sections instead of three. Same at Franciacorta — seven splits instead of three. Why is that useful? If there are 14 corners on the track, roughly every two corners becomes a mini section. In an AiM sector you may have four corners: you might be doing really well in two of them and really badly in the other two, and it can still be your fastest sector. That's not useful, because what you want is to see which corners were genuinely your best, so they can be combined into a real theoretical best lap.

Most of the time three sectors aren't accurate enough to tell the right picture. The logger is doing its job — it's given three sectors per track from the GPS and it follows that reliably — but the funny thing is that the start and end of those sectors usually land right in the middle of a corner. That's exactly what we're trying to avoid: we don't want to cut a sector and start a new one in the middle of a corner or in the middle of a chicane. And don't worry, the algorithm doesn't create unrealistic splits either — it won't cut a chicane in half. I took care of that. If you want the underlying idea, it's the same argument as breaking the lap into pieces, just with pieces that respect the corners.

Chase the meta lap, then close the gap

Once you can see all the sections across the lap, you know: on this lap I did this section really well, on that lap I did it badly. The goal is to put all the best mini sections together, and those combine into your perfect lap.

I call it the meta lap — where you're really putting everything together, using all the grip the track and the kart had available.

And you want to combine that with your optimal peak grip window, which maybe comes on lap six or seven, because the tires are a bit cold at the beginning. Then you have three or four laps at peak grip — those are the laps to go for your perfect lap in.

The goal is to close the gap to that perfect lap as much as possible. It's going to be very hard to match it exactly; the closer you are, the better. And on top of that you rely on your consistency.

Consistency, measured properly

How do we value and calculate consistency? We use the standard deviation. It's a statistics value that measures how much variance there is between your fastest lap and all your other laps — how many tenths you're swinging by. The lower the standard deviation, the higher your consistency, so the closer to zero the better.

• 0.01 — you're doing an absolutely phenomenal job.
• 0.5 — a huge difference, because on average you're five tenths off your own fastest lap. You are not consistent enough.

It's almost impossible to bring that to zero — in the first laps the tires won't be at peak performance, so it's hard to get there. Nobody really reaches zero. But that should be your target, and you should track the number over time. Six months ago my standard deviation averaged 0.55 across the day; now it's 0.25. That's massive improvement — and it's visible no matter what the track was doing. This is the number behind measuring and fixing lap-to-lap consistency.

And in a race, you don't just want a good fastest lap. Nobody wins on fastest lap alone. You need consistently fast laps.

I'd rather have a fastest lap a tenth worse and repeat it every single lap, than an unbelievable fastest lap I'm three or four tenths away from the rest of the time.

It's what I say to my drivers at Babyrace: when you do a lap and you can't replicate it, it's almost like you had it by luck, by chance. That's not how it's supposed to be. It's supposed to be repeatable.

Now, I know there are effects from slipstream, from traffic, from tire degradation and tire temperature. I understand — I've driven many years myself and I still drive. But those variables don't have as much effect as you think, so be real with yourself: yes, there are variables, but they're not the main factor. In some categories they matter more — in 60 Mini or Juniors the slipstream is quite effective, so on a track like Zuera with very long straights you can see two, three, even four tenths. That's pretty wild. Keep it in consideration and live with it.

The real signature of a fast driver

The real skill a successful racing driver has is this: you look at their lap times in the data and you can't quite tell one lap from another. The laps are so similar that you almost have to zoom in massively to see the differences in speed or in the lines.

We have some really cool ways to see that repeatability. There's braking repeatability, and there's racing line repeatability — how close you were to your fastest lap's racing line, how many laps you managed to repeat it, and on average how many meters away from that line you are. If one lap you're perfect, the next you're two meters away, the next three meters, then you're averaging 2.5 meters from what you did on your fastest lap. Obviously you want to bring that down. For braking, once you've found the fastest lap, the goal is to repeat that braking point as closely as possible — and the corners costing you most are where to start.

Now — we all know that as karting drivers we drive by feeling. Don't focus too much on reference points. I totally agree with that, and you'll never hear me telling you to find a reference point and always brake there. That's for formula cars and GT cars, because you arrive at much higher speeds — 300 km/h on the straight — and there you must use reference points for braking. In karting you arrive at much lower speeds, you don't really have clear reference points, and everything happens so fast because every corner is right next to the next one. It's almost impossible to use them. So don't focus on those. Aim at getting braking repeatability and line consistency as close to zero as possible, because those are the two things drivers at their peak are able to do.

Adaptability: how fast do you get up to pace?

The other thing is adaptability. The best drivers adapt massively from session to session, day to day, and track to track — so they get up to pace much sooner than when they were starting out. They're not going to take 15 laps to reach a quick lap; they might take five laps to adapt to the track conditions and get on the pace.

Because when you go out, you always have to discover the grip. I have this theory of grip discovery: drivers are discovering the grip as soon as they go out, and then they have to adapt. If you feel there's a lot of grip, you adapt and carry more speed into the corners, brake later, release the brakes, get on the throttle earlier. If you discover there's no grip compared to the session before, you have to adapt to that too. That's grip discovery adaptability — but in general you have to adapt to everything: your setup, the track conditions, how the grip evolves through the day.

So if it normally took you 15 laps to adapt to the track and the kart, you need to take that down to two, three, four laps. Drivers like Verstappen, Leclerc, Schumacher, Hamilton are the kind who adapt super fast. And I'm sure they weren't like that on day one — they developed that adaptability.

It matters even more in the rain. The conditions are never the same in the wet; they're different all the time. The grip level in one wet session differs massively from another wet day — a bit more rain or a bit less, more standing water or less, more rubber on the track which becomes soap, or less rubber which is actually better in the wet because there's more grip available. You have to adapt massively, which is why wet sessions demand different data habits.

Race metrics, not just session metrics

And it's not just about your individual lap time or an individual driving session. You should also trace and measure your race metrics. For example: how many overtakes can you make in a race? How good are you at holding position on lap one?

Some drivers are really fast but on lap one they lose five positions on average. Others aren't as fast, but they always gain two or three. Those are measurable metrics you can track, and it's really cool, because it shows you the drivers who may not be quickest but who always recover positions.

Obviously you want drivers who are really fast and don't have to recover anything, because they're starting P1 — that's the nirvana, the ultimate holy grail. But it's unlikely, and everybody needs to make up positions at some point. Whether you have the pace or not, you've got to wrestle your way through and grab the positions, even without maximum pace. So it's also: how good are you at overtaking, how good at defending, how many laps can you do without making mistakes, how many laps can you drive at the front of the pack without losing your focus?

We already track some race metrics at PURPL — in the Race Intel feature you can drop in the PDF from your race and see things like how many positions you gained and lost. And there are a few more features to build, which is what we're working on in the next few weeks, to make it even more understandable and more valuable. The broader case for this is one I've made before: consistency matters more than speed.

So stay tuned, keep analyzing your sessions on PURPL, keep finding lap time. Go purple, and remember to keep sending it. Ciao ciao.