Course Dependency vs. Driving Lines

I was at an autocross event this past weekend with a relatively small group I run with only two or three times a year. On my last run the starter (who I didn’t know) leaned into my car and remarked something about how I show up in different cars but I’m always fast.

It may be true that I’ve raced in three different cars (not all mine) in my last three visits.

Still, I had to think about what he said because I wasn’t sure why he thought that would be in any way remarkable. I mean, ok, thanks for the complement!, but if you know how to be fast why should the car matter too much? Of course, it takes me a while to get the hang of any particular car and I don’t have nearly the multi-car experience like some of my car-slut friends.

I think we need to make the following distinction:

Course dependency between cars is a real thing. One car can be highly favored over another because of course differences. This is mostly influenced by power-to-weight ratio which varies widely between classes and preparation levels and to a lesser extent is influenced by car width and transient response. On the other hand, cornering and braking performance is very similar for all cars on Street tires and all cars on race compound tires, respectively.

But driving line differences between cars, while real, are relatively smaller than course dependency. The principles that produce speed on the autocross course are constant and apply quite generally to all cars and all classes.

Move The Corner To Suit 2

Based on comments received I think I need to more explicitly state the moral of the story told in Move The Corner To Suit.

To review, I decided it best to move the “standard” line on the approach to B, as shown in Figure 1, to the right in order to not be required to lift on the approach to C. This allowed me to maximally accelerate coming out of B all the way to D. I decided it was better to give up some time early to gain more later. Compared to my run when I didn’t do this and compared to all my co-drivers’ runs who never did this I Saved Time.

Figure 1: Standard Line vs. Intended Line With The Corner Moved To The Right

A friend (a better driver than me) was at the event and commented that there was no way he did not have to partially lift on the approach to C. I have no reason to doubt that he is 100% correct. But, he was in a very different car with more power and worse tire-to-weight ratio and Street tires while I was on R-comps.

So, to be clear, I’m not advocating that every driver in every car should slow down enough at B (which is one way to change the path shape) or move the corner enough at B, to allow a no-lift approach to C. Anyone in most any car could slow down to 5mph at B and not have to lift approaching C, right? Of course, this would not be optimum. Slowing that much would be… no, I won’t say it! You get the idea.

The moral of the story is this: For every corner you meet ask yourself: Is the proper apex point, is the proper shape of the path, completely obvious?

We have so many connected corners in autocross that the answer is almost always that it is not obvious to us non-aliens.

I was walking the Bristol Tour course a few weeks ago with the friend mentioned above and I recall that we disagreed by a few feet about the location of the apex at the final super-sharp corner on day 2. Even now I have no idea who was right or wrong. I’m not even sure if I executed the corner as I intended during the course-walk. Again, we were driving very different cars. We could have both been right, but only for our particular car. (Ok, probably I was wrong!)

Here is my advice: look at every corner or corner complex and decide, for you and your car, the location of the apex and the best path shape, remembering that often you have the option to move the corner earlier or later and change the path shape with the steering wheel, brakes and throttle. There may be a cone there that 99 out of 100 people will say, “That’s obviously the apex.” Decide for yourself and play with the possibilities in your mind: What happens if I move the corner later? What happens if I move the corner earlier? Do I lose and then gain more? Do I gain and then lose more?

As someone once said, “Highest average speed wins every time.”

Move The Corner To Suit

One of the great freedoms (and complications) of autocross is that often we must decide the location and shape of the “corner.” Sometimes in a major way, but more often in minor ways that create small, but crucial, advantages. Let’s explore a real-life example from a recent autocross event.

Figure 1 shows the relevant section of the course. We enter at A, accelerating strongly, must brake hard and turn 90 degrees to the right at B, then sweep through a gate at C, followed by a long acceleration zone through and even beyond the gate at D.

Figure 1: Section Of A Recent Autocross Course

The corner at B contains the key questions: Where is the entry, apex and exit? As I’ve drawn it in Figure 1 it may seem straightforward. It’s basically a symmetrical entry and exit, right? The cone at B is the apex, i.e. the slowest point. As we approach C there will be either a braking or lift point allowing us to apex at C and then accelerate out.

After a run I realized that this was not optimum.

Figure 2 shows the same cones with two actual car paths from GPS data. Red is me, Blue is my co-driver.

Figure 2: Actual Paths

Neither one of these paths is perfect, by any means. For instance, I think my Red path should have gone out wider on entry similar to Blue. (I think I did go wider on other runs.) But, after cone B, Blue has a problem that allows Red to be Saving major Time. Do you see the problem?

Maybe I should ask instead, do you see the advantage of the red path? Or at least suspect what it might be? When I show you the data I think it will become clear.

Let me tell you my thought process. What I told myself was that I needed to move the corner at B earlier. I wanted to make the cone at B more of an exit cone, not an apex cone. Even though this would require me to brake earlier as I approached cone B the advantage would be that, partially because B was a relatively slow corner, I could accelerate beyond B without ever lifting for the gate at C. I hoped that this would increase my average speed from B to D and beyond, allow me to traverse that section in less time and make up for the initial loss caused by braking earlier. I didn’t intend to change the shape of the corner at all. I only wanted to move it earlier. I wanted to move it to the right as compared to what’s shown in Figure 1.

Figure 3 shows the data with reference numbered locations and Figure 4 shows the numbered locations on the paths.

Figure 3: Annotated Data
Figure 4: Numbered Path Locations

At 1 both cars are traveling at exactly the same speed. Blue has correctly gone a little wider. Red is braking harder and will slow more than blue.

At 2 Red reaches his minimum speed point. This is his apex. Blue reaches his minimum speed point at 3, slightly further along and definitely after cone B. Therefore Red gets back on the gas earliest, but this is not very important in this case and not the reason for the large gain Red will eventually get. Red is slower than Blue by 3mph (mostly because he didn’t enter wide enough) and losing time until 4, at which point Blue has taken a 0.10s lead.

The problem Blue has is even before 4 when he must lift off the throttle in order to negotiate C gate. The drop in Longitudinal Acceleration is down to zero at 5 while Red maintains strong acceleration. Blue has been forced to approach the gate at C at a very different angle than Red because he chose to carry more speed around B. After 4 Red starts to gain as seen by Blue’s Delta-T trace angling sharply upward.

[Note: please ignore the cursor marks between 6 and 7 in Figure 4. They are not at the cursor location in Figure 3.]

Mostly by virtue of Blue having to lift at 4, Red is 8mph faster by 6. He maintains a speed advantage all the way until the next braking point at 7. At 7 Red has turned the earlier 0.10s deficit into a 0.33s gain. So, Red gave up 0.10s to gain 0.43s by moving the corner.

Did Red succeed in making the cone at B an exit cone? Not really. The data is clear that I was only able to make it the apex cone while Blue (and many others) created an apex further beyond cone B. What I intended is shown as the dashed path in Figure 5, below.

Figure 5: Intended Line

I’m not completely clear on why we as autocrossers do this so often, i.e. our slowest point is beyond where the theoretical apex location should be. I wanted to move the apex earlier than B, but I didn’t do it! At least not on this run. I did succeed in achieving the main goal, which was to not have to lift to get through the gate at C, in fact I was at full throttle in 2nd gear, but I had to slow too much rounding B, which probably hurt the total gain.

I see this phenomenon in myself and others over and over again, event after event, local and national. Do many of us just lack discipline? What’s our problem? If you have any thoughts on why this should be, or if you think I’m not seeing it correctly, please let me know in the comments.

By the way: We each got 4 runs that afternoon. My co-driver did this section with a lift on the approach to C all four runs. The blue run shown was by far the fastest of the four. Of my four runs, I did a partial lift once and stayed in the gas three times. The red run shown was the fastest but very similar to the two other no-lift runs. The lifted run was by far the slowest.

Key Differentiators in SCCA Solo® (Autocross)

What makes SCCA Autocross different from other forms of motorsport? Six items come to mind.

Getting It Done In Three All National Tour events have new courses which racers may walk, but not practice. Each of only three runs is for time and only the best run counts. Regional racers are often dismayed at their first national event when encountering the difference between learning a course in six to eight runs, as is common at regional events, and reaching your potential in three runs at a national event. Some regions run a “Pro” class where only the first three runs count in order to foster the capacity needed to compete at the national level.

Lack of Corner and Edge Definition Modern autocross courses have a unique characteristic in having ill-defined “track edge” limits. (This may not have always been the case.) Multiple line choices become available for different classes of cars while drivers are required to decide where a corner begins and ends, its proper radius and where to create apexes where only some (or none) of these features are rigidly defined by cones. This situation requires a skill set not normally developed by racing on traditional “ribbon” race tracks.

Extreme Feature Connectivity Courses often have multiply-connected features at a level not typically found on fixed race tracks. Such connectivity requires making complicated, multi-variable decisions to obtain the highest possible average speed from the start to the finish.

Urgent and Intense Mental Discipline    An intense mental planning discipline in the limited time during and after the course walk and between runs is required based on imagining how the course will drive on the first run and then analyzing mistakes and considering alternatives to get faster on subsequent runs. To win you usually must be fast on the first run and faster yet on each run thereafter.

Both Intuitive and Theoretical Drivers Can Win   Intuitive racers who drive what they feel and see based on experience are on an equal footing with more analytical types, allowing room for different approaches, as long as the first differentiator in this list is obeyed.

The Class Structure Supports Both Drivers and Builders The sport provides success paths for both drivers and builders and all mixtures in between by classing regular production automobiles at various price and performance levels as well as special-purpose racing machines at various levels of modification, performance and build uniqueness.