[Skypalace]

Saying that the fastest way around a course is the highest average speed is a bit of a tautology - average speed is of course course distance divided by time, the shorter the time the higher the speed. Course distance is (relatively) fixed.

Any of these fine suggestions above are rules of thumb, but are never perfectly correct. The fastest way around a course depends on a lot of variables, and is extremely car-dependent, as well as course dependent.

The best line through a particular corner very much depends on what's after the corner. If there's a long straight, then the best line will likely have a very late apex so that you can get on full throttle as quickly as possible. But if there's another corner immediately thereafter turning the same direction, you might do a very early apex to be able to get you to the outside edge of the track as quickly as possible to set up for the following corner, which might be more important.

Then there are corners with a broad range of lines, turn 7 at the (repaved) Portland International Raceway is an example, it's 6 or 8 cars wide, but leads onto a the long back straight. Sure you want a late apex in order to maximize exit speed onto the straight, but it's SO wide on entry that I see many people enter too wide, adding material track distance at the slowest corner on the course, adding time that they don't need to add as it's not actually improving their exit speed. Also, as with many corners, the line through there is very different on a momentum car (944 or Spec Miata) than in a car with significant power, as you normally need a later apex in a higher-powered car so that you can be on full throttle without rear tires getting loose (ie. you have less available grip for lateral acceleration). With a lower powered car, there's less traction required for acceleration on full throttle, and thus more available for lateral grip (turning), so you can apex earlier and continue turning but still have full throttle through the exit.


And of course, on almost any corner a 911 wants a later apex than a Vette with equivalent power, because of the rearward weight bias, and a late turnin works because trail braking allows you to rotate that big pendulum behind you, whereas a car with 50/50 weight distribution won't react to trailbraking the same way.

I completely agree with Nick on sensitivity. It's one reason I've been having braking problems with the new Cup - the brakes are rock-hard, like pushing against a brick wall, and have very little travel, and virtually zero feedback through the pedal. I'm basically having to calibrate how hard my quads are pushing the pedal, rather than feeling the pedal travel through the side of my foot through my driving shoes, like in the 996 that has a softer pedal and much more pdeal travel during braking. I'm developing the feel, but I'm not fully there yet (or I wouldn't have spun under braking twice in my last race weekend :-))

I like to drive with my bottom seat cushion removed, as it gives me better feel of the car - literally improving the feedback through the seat of my pants. The biggest difference in being able to sense yaw (esp. rear of the car moving out) as quickly as possible. With hard engine and tranny mounts, there's enough engine & tranny noise in the car (plus I have earpieces in) that it's difficult to hear the tires, which are a ton easier to hear in a street car.

I also have a seat with very high side (thigh) bolsters, in a seat that most people would probably consider to be one size too small, but I like to have zero movement in the seat so that I can feel the car as much as possible. I like to think it's one reason I'm much better in a real car than in video games. As Nick says above, loose is fast, I definitely prefer a car a little on the loose side, but it requires high sensitivity (but it's SOOO fun when it's working!). I spin fairly often in video games because I can't sense the yaw movement of rear getting loose, in a game the only way to sense yaw is through the front windshield, but that's definitely not how I sense it in a real car. Or so I like to think :-)

I have full Motec data in my cars, and spend a very bit of time analyzing it, to see exactly what works and what doesn't. Often it's counter-intuitive. I've seen many times where I've changed a line to make up .3 sec on corner entry, only to lose .5 second through to the next corner. I've also seen the other way, lost .3 sec on corner entry, and made up .5 second along the next straight. Data is very helpful, but it takes some practice to figure out how to use it properly, and to actually help yourself rather than confuse yourself with so much information.

I find video even more useful, I will never pass up an opportunity to have someone look at my video and make comments. I've had many pros and fellow competitors (most helpful if they're in a different class so not directly competitive :-)) willing to look at video of a lap and make comments. Usually it works, occasionally it doesn't - I had one pro insist that I was taking a particular corner in too short a gear (I should roll through in 3rd rather than downshift to 2nd). I must have done a dozen laps in my next session, and it never felt right. My data showed that I was losing launch speed out of the corner every single time, and lost more time on the next straight than I gained. It was probably gearing/grip difference (I was in the 996 and he had been in a 2010 Cup, which has both slightly different gearing, as well as quite a bit more mechanical grip esp. in the back), but could have been me trying his gearing advice, on perhaps the wrong line.