|
I've put the scores in a spreadsheet. This is a preliminary analysis, which does not include data still on the raw sheets in the bus, which will provide a finer granularity.
The analysis concerns only SU2 vehicles. There were not enough cars in SO4 to have good data, although they appear to not be affected significantly by course degradation.
Cars that had a DNF were not counted in these statistics, as the 90 second time applied was not indiciative of their speed through the course.
Cone penalties were also ignored. I'm interested in how much slogging the cars were doing, and taking the absolute times seem more indicative of how much the car/engine would bog.
First course, in the morning, where SU2 cars ran after the SO4 cars:
Average amount slower on the second run than the first: -2.0 seconds
Average amount slower on the third run than the first: -1.8 seconds
Yes, on average, all the SU2 cars went faster on their second run, then lost 0.2 on their third run compared to the second, which was still faster than the first run.
Second course, in the afternoon, where SU2 cars ran first on a fresh course:
Average amount slower on the second run than the first: 1.8 seconds
Average amount slower on the third run than the first: 3.0 seconds
So, due to course degradation, the SU2 cars on average lost 1.8 seconds on their second run, then an additional 1.2 seconds on their third run.
First thoughts
First Course: I'm going to bet that the very first run of the day being slower was due to people gearing up, learning the course, and getting used to driving on dirt. Then, the minimal 0.2 seconds degradation to the next run is because the 4wd vehicles ran first, so the course was "pre-torn-up" and the additional wear caused by the SU2 cars was not very significant.
Second Course: For SU2 cars, running first on a fresh course exposed them to a larger variety of conditions: instead of going from a 5 to a 4 as in the morning, the "course goodness" rating went from 10 to 5. This larger variation caused larger time differences per run.
----------------------------------------------------
What amount of time is "significant"?
For the upper half of the field (again, on raw times), which I will assume to be (at least on this day) relatively consistant drivers, the average difference between consecutive places was 2.4 seconds. However, that includes first place, which was over 11 seconds ahead of second place. Removing that outlier, the average was reduced to 1.2 seconds. And looking at the trophy spots (top 5, minus outlier of first-second delta) the average difference was 0.5 seconds.
HOWEVER, for this new discussion of "signifcance" and how it affects winning, I think we need to consider times including cones. In that case, top half of the field had an average of 2.5 seconds between places, (reduced to 1.9 seconds between places if you ignore the outlier of first to second, which is now only 5 seconds, not eleven) and the places 2-5 are spread, on average, over 1.6 seconds.
----------------------------------------------------
Summary Thoughts
1) First, let's imagine that all drivers run their first run as a group, then their second, etc., and that they always ran in the same order. These runs are on a course already, uh, "prepared", by the 4wd cars. Over six runs, the first driver would have a 0.2*6 = 1.2 second advantage over the last driver. On average, this is less than the 1.6 seconds between final trophy places.
2) Same scenario, but on a fresh course. The first driver would have 1.9*6 = 11 second advantage over the last driver. This is certainly plenty of trophy positions.
Both scenarios imagine a 3 runs on two courses setup. The course would not deteriortate as much between runs 4-6 as it would 1-3, if running 6 runs on the same course. That is, like Four Oaks, with six runs on the same course, the effect would not be as large.
----------------------------------------------------
Ideas for the future
1) As Carl mentioned some time ago, it seems that the SU2 cars should always run after the SO4 group. Yes, the times will be slower, but the competition within the SU2 class will be closer.
2) A scheduled run order, that gets inverted each heat, may even out the discrepencies. First driver of heat one runs last on heat two. This works much better if we are having an even number of runs, of course. An alternate method, for a three run setup, would be to "slide" the start around by a third each run: Of 21 cars, car 1 runs first the first time around, car 14 runs first the second time, and car 7 the third. This would reduce the "going first" advantage to 1/3 of the 0.2 seconds per run (assuming running after SO4), giving the overall luckiest slot of the day only 0.4 second advantage (total, over six runs), well under the 2.3 to 1.6 second average delta between the top five spots. (2.3 including first place, 1.6 not)
----------------------------------------------------
Further study
Looking at the raw sheets and finding out where in the order people ran may reveal more data. As I said, this finer granularity will likely reveal, for example, that the first 10 cars on a fresh course have a larger advantage that then tapers off to the average advantage. On a non-fresh course, there may be only a very small difference. We'll see. When I get that data, I'll see if I can get anything out of it.
Cheers,
Anders
_________________ Lina Racing: As Seen On Radio
Last edited by AndersGreen on Thu Aug 12, 2004 1:01 pm, edited 1 time in total.
|
Login
Register
FAQ
Search
