I had to do a regressioin project for my Business stats II class. I did a linear and exponential regression analysis for 60' times and 1/4 mile times. See if how accurate this equation is for you guys. Plug in your 60' time for x and y should be close to your 1/4 mile time. Of course, all this info is done with time slips from my car so it won't be so accurate for everyone.

Oh, and it is only accurate for 60' time from like 1.8-2.6....so if you plug in 1.4 your not going to get a reasonable anwser. Please post if this is equation is accurate for you.

Linear Y= 7.8906353 + 2.9121395(X)

exponential Y= 9.2259399*1.2203618^X

 

hmm...

with a 2.096 60'...

linear Y = 13.994
exp Y = 14.005



--Paul

 

of course it's not gonna be accurate for your boosted times, but is that close to you NA times? I guess this is only good for me and those with the VI and JWT ECU or guys like ceaser.

I do have the excel program to calculate all of this. If someone wants it I can send it to you, but you need excel to use it. I used 30 data points for mine.

 

linear: 14.55
exponential: 14.55

Close. course i'm in a 2k2 auto, vs a stick (i assume)

 

If I could just pull a 0.00 60' I could run 9's!!!

 

Most of ours were only off by 0.05 - cool formula !!

 

With my data ranging from a 2.14 - 2.4 60' and the 1/4 ET from 14.1-15.2 the regression shows that for every .1 decrease in 60' I see an average of .29 reduction in ET at a 95% confidence level. Of course that seems a bit high, but the error factor was 1.4x IIRC and the R squared value was 84%.

 

Well I finally found a calculator to do the exponential Y formula, and I used some old time slips with 2.4 60's and they were off a little more.

It seems like, (just with the data used I used), the more you move away from the 2.1 - 2.3 60's the less accurate the end figure becomes, but Ceasar never gets above a 2.2 anyways most times.

I thought that was a really cool formula... it'll certainly give some measure of certainty predicting a time if we go to some of these country tracks that don't print of the 1/4 mile times etc.

 

The most accurate way to do this would be to use the excel program I used and enter in Ceaser's data. 1/8 mile times will not work with this. Would would have to create a new equation for that using 60' times and 1/8 mile times. I actually did that test as part of the project and I was able to predict 1/8 mile times more accurately than 1/4 mile times.

There is a couple reasons that the equations will be inaccurate. The 60' times I have only go as low as 2.14. If I had some times on slicks the lower 60' would be more accurate. The more data the better.

I guess you have to understand how this works in order to see that really low or high 60's won't give accurate predictions. A linear curve will end up going negative at some point and a exponential curve will give have too much of an upward slope the higher the X variable is.

The purpose of regression this project I did was to see if there was a correlation between 60' (X) and 1/4 mile ET (Y), which of course there is. I also was trying to find out which curve was better, the linear or exponential. They proved to be both about the same. The last part is to use a prediction table.

All of this info is based on a 95% confidence intervel so there is error factor involved. The error factor for the slope is +/- 1.45 and the prediction has a 5% chance of being outside the upper and lower ends of the error factor. The problem is that the error factor is a very large percentage the B value in the y=a+bx equation.

I plan to do some more work with this excel program. I hope to build up enough data to give better predictions. I also want to do a regression to see if there is a correlation between 60' time and trap speed.

 

hey, that sounds great, I don't have excel on this PC but I know I have office around here somewhere - I have been wanting to make some graphs of Ceasars times for ages, now it just gives me more of a reason to get on it.

When we get the timeslips database up, there will certainly be a bunch of info floating around for a range of times..... might could even make a javascript program where racers could plug in their 60' times and let it splurt out some predictions etc.... personally, I really like those racing calculators that work out all your relative horsepower, and other info..

Nice work Mike !