znod

I had a tad of time today so I tried ordinary least squares (OLS) regression for zero to 60's. My results look promising.

Unadjusted 1/4 Times: 13.629 13.744 13.702 13.742 13.720 13.764 13.686 13.742 13.759 13.774 13.772
Unadjusted 0-60 Times: 5.149 5.270 5.231 5.273 5.206 5.301 5.209 5.237 5.286 5.299 5.282

Regression Coefficients: a = -8.808; b = 1.024

Applying the coefficients to the unadjusted 1/4 times yeilds the following values and differences:

Actual 0 to 60:................5.149...5.270...5.231...5.273.. .5.206...5.301...5.209...5.237...5.286...5.299...5 .282
Less calculated 0 to 60:..5.146...5.263...5.220...5.261...5.239...5.284 ...5.204...5.261...5.279...5.294...5.292
Difference:......................0.003...0.007...0 .011...0.012..-0.033...0.017...0.005..-0.024...0.007...0.005..-0.010

I could easily live with predictions of adjusted zero to 60's being off only by such amounts.

Here are my related weather/altitude adjusted 1/4 times:

Adjusted 1/4 Times: 13.451 13.565 13.523 13.562 13.593 13.637 13.559 13.61 13.627 13.642 13.550

Applying the regression coeffients gives:

Predicted Adjusted 0 to 60: 4.963 5.080 5.037 5.077 5.109 5.154 5.074 5.126 5.144 5.159 5.065

I like the way the above values look. Maybe I had a sub 5 second zero to 60 too. I am going to see what happens using OLS regression on other values.

grogan545

Very interesting Zman.Without going into great depth it looks like this method is reasonable in predicting corrected 0-60 times.I have to review some of my previous conversions but I believe they will show that similar 1/4 et's will give similar 0-60 times.The actual times seem to mirror the corrected times.Example:
et=13.45 actual,0-60=5.00(grogans formula for 1/4 time)
et=13.45 corrected,0-60=4.961 corrected (Z's best run).
It would be interesting to see how the magazines get their time to speed data since MT & CD do use weather corrections for their acceleration data.There must be a formula based in fact to do this conversion.

znod

Right, I thought about that too. What else to do you know about MT's procedures. I am accumulating data on the mags procedures.

I had 3 very, very good runs this morning and will be reporting on them soon. New best--at least under 20.01.00. Persistence and patience with car learning.

grogan545

Not very much Znod.Occasionally the magazines make mention of their testing methods but do no go into detail.It is usually in response to a readers question under "letters to the editor".This is where I saw R & T describe their test method.They indicated that they use raw data with no corrections because this is the way the average person would test.I observed this sometime in the last 6 months but I no longer have the magazine(probably no articles on BMW's).I think I incorrectly noted in the past that MT does not use a "rollout".I don't remember why I said that but it is not correct they do use a "rollout".Maybe a letter to the editor of the 3 major magazines would give us the answers.When I find the time I will do so,probably can do it through email.

I am waiting for your new results.I still find it amazing that your car continues to learn over a long period of time.I will not have more testing until after next week.I am hoping that in May the weather conditions will be close to the "standard" conditions so that very little correction is necessary.Last year when I began testing in May I got very good results and I know the temp was between 55 & 65 F.I know the corrections are necassary for comparison but I still need to be convinced that at the extremes(very poor,very good)the results are accurate.

znod

I have the published data on how RT and CD test, although the data are incomplete. I am glad to have MT data you had. I haven't studied your issue with over correction for good weather, but I will start thinking about it.

I wanted to go out badly this morning because I had some good runs yesterday, but I had too much wheel spin. Nevertheless, I though my car was running good and decided that I had over brake torqued. Today I used very minimal brake torquing and had three essentially zero wheel spin runs in a row.

Below are the three new sets of data from today, along with two older sets. Together, these five runs yield my top five sets of adjusted data (with one exception from yesterday--a run with noticeably worse unadjusted data, but slightly better adjusted data than one of those below). I used the regression apporach in getting my adjusted zero to 60's. I am starting to believe that my car has mostly overcome 20.01.00. Soon, I am going to do a comparison of the type I usually do using my unadjusted and adjusted data from today's best run--to see if all of the relationships, given the use of regression, look reasonable.

Unadjusted:
.......................2/18..........2/20.......2/25..........2/25.......2/25
1/4 Speed.....102.710...103.240..104.050...103.640.103.710
1/4 Time.........13.629.....13.686..13.564.....13.663.....13.631
1/8 Speed.......82.450.....82.170....82.880.....82.530 ...82.490
1/8 Time..........8.853........8.894.....8.810.......8 .859.....8.858
0 to 60.............5.149.......5.206....5.079.......5.145......5.161

Adjusted:
.........................2/18........2/20........2/25........2/25........2/25
1/4 Speed.....104.083...104.218.104.862.104.449....104.519
1/4 Time.........13.451.....13.559...13.461....13.559......13.527
1/8 Speed.......82.584.....82.644......82.928.....82.7 46....82.776
1/8 Time..........8.725.......8.802........8.732...... ..8.802.....8.779
0 to 60............4.959.......5.071......4.970.......5.071.......5.038

Bokke

Rain today Leaving on a business trip tomorrow, so no testing until next weekend!

znod

Have a great trip Bokke.

EBMCS03

NICE so your car learned to be fast again. I havent been following since you guys posted waaay too much

can someone give me a re-cap really fast.

Znod: Have you just been beating the heck out of your car to make it re-learn?

What did you, or did you "un-mod" something to make your car fast again?

znod

Thank you for your interest Eric. Well, I can say at least that I have used my car to its fullest extent a significant number of times since 20.01.00. My car started learning, but then seemed to cease after 20.01.00. I decided that the problem might be my minimal mods and had them removed. The car quickly started learning again and soon moved beyond its mods-on level under 20.01.00. It appears to have had another recent learning spurt. And, I am producing some very good test results now. I think that's it in a nut shell. Let me know if I can provide more info.

grogan545's car immediately took a liking to 20.01.00. And, its performance seems better after than before.

znod

There is an article in the current issue of Drag Racing on weather stations. It gives the same (essentially) standard conditions--60 F, 29.92 PSI; and 0% humidity. Do you know how to get humidity from temp, barometric pressure, and dew point?