What kind of top speeds are people seeing on the track? (Page 3 of 3)

The teams at CMP (and MSR) have gotten a lot faster as they've piled on the seat time, no doubt about it.

What I've learned from all these races is that you don't have to be fast to get in the top 10 of any given race; a reliable car that turns endless black-flag-free laps— even fairly slow ones— will usually accomplish that. You do have to be fast to get in the top 5, however. Not the quickest thing on the track, but within a few seconds of the quickest car.

Alright, CMP fans:
Here you go.

All CMP races since the start of 2010:

Southern Discomfort 2010:

South Spring 2010:

South Fall 2010:

Southern Discomfort 2011:

Note that there is one outlier that's WAY offscreen: What appears to be a transponder glitch on Student Drivers at Southern Discomfort 2011.

It is interesting, however, that if you look at the individual races, you'll see that it DOES appear to matter how fast you are in the top 10-15% of the field - but only in comparison with each other, not the field at large.

As in x=bestSpeed / laptimeInSeconds ? That seems a bit off, since the speed value is already computed from laptime and the track length.

At any rate, if you do that you end up with the uncorrelated column-looking thing you do above on a per-race basis, a bunch of seperate (occasionally partially overlapping) columns for each event.

As for the data, it's copy-paste from Mylaps - which gives you something similar to a tab-separated values file, except with really mangled data fields for some races. Here's the raw data:
http://www.phyrefile.com/file/view/X1xReTd8XMO5NiEv

If you'd like, I can massage that into XML, or a proper CSV, or whatever pretty easily, since the parser I wrote uses (almost) all the available data (It ignores most DNS cars)

There are two races missing from the 2010 season - the season opener because it's missing timing data on mylaps, and one other race because it seems to be completely missing.

Nope, you basically did what I was asking when you broke the races out individually... 

The track isn't constant throughout the race, much less over the course of multiple races.  You can see this by the differences in the average speeds in the different races.  The racetrack itself is getting slower each race as we smooth over the aggregate  in the asphalt.  That difference needed to be accounted for and it wasn’t. 

What I do see are a couple of trends.  One is actually the one we say doesn't exist - but it stastically does.  The faster cars finish higher.  This trend is strongest within the top 10% of cars.

That's why I get paid the big bucks...  :roll:

A cleaner graph might be if you normalize the best average speed against the winners best average speed.

that sure looks like half a bell curve to me.

It just warmed the cockles of my heart watching the wagon go round and round at CMP. Plus the Honda cycle with direct injection was pretty damn cool too.

What all this tells me is what most of us already knew: reliability, clean driving, and pit strategy puts you in the top 10% - from there, speed is key.

We saw this when the Audi ran a perfect race - clean, quick driver changes, no black flags, and long stints.  They finished 7th out of about 100 cars.