How's your driving? The "Stormer error correction", thread.
#76
The only thing you can hope for with the youngsters is that they have some fun with Dad at the track. Because now that my son is 18, I think I've become about 80 IQ points stupider. Same thing happened to my dad. Oddly enough, the old man smartened up a LOT by the time I had my own place and was paying some bills and had bricks and plywood for living room furniture.
#77
Tech Elite
iTrader: (51)
Originally Posted by rcnewb2004
trilerian: (just curious) why do you "know" it won't be normally distributed? (sorry, i am not the brightest... not even close). Thank you for correcting me, yes it should be 3 deviations away. Also the Stormer Error correction is much easier to apply at the track than to do some funky standard deviation.
#78
As for chassis I did run the main with a different chassis that I've been trying to get dialed as well as my other car, the new one was close in the 3rd Q run so I decided to give it a try and leave the older one I ran the 1st 2 Q's with sitting.
At least this approach gives you a very tangilbe comparison track to track where purely laptimes and # of laps leaves way to much variable
#79
Tech Master
I do not mean to hijack this thread and I do not mean to turn this into some stats thread, so this will be my last irrelevant question. I just wish to learn a few more things, and perhaps find other meanings for the Stormer-Error correction.
trilerian: Could you plot over 20 laps? maybe 40-100 laps (over a month if the track layout doesn't change) and see if the shape of the graph approaches normal? (I just got curious now, if it is possible to predict one's stormer-error by just observing the average time of a lap.) Or perhaps even try and correlate the stormer-error to a number of standard deviations.
Ok... for those who have found me annoying... i am sorry... i be quiet now.
Happy Racing.
trilerian: Could you plot over 20 laps? maybe 40-100 laps (over a month if the track layout doesn't change) and see if the shape of the graph approaches normal? (I just got curious now, if it is possible to predict one's stormer-error by just observing the average time of a lap.) Or perhaps even try and correlate the stormer-error to a number of standard deviations.
Ok... for those who have found me annoying... i am sorry... i be quiet now.
Happy Racing.
#80
Tech Master
Bob - I heard a rumor that Montana Choc. Chip cookies will help your driving consitancy and descrease your error..
#81
Can I get these in quantity
#85
Although it is theoretically possible for individual laps times to be normally distributed, it is highly unlikely (at least for one individual). A normal distribution requires the mean (average) and the median (the middle-most value) to be close or identical. This means that 50% of your laps are above and 50% are below your median.
The more likely statistical representation for individual lap times is a skewed distribution. Skewness occurs when there is some type of boundary (like the fastest lap) where data congregates towards that boundary value but not past it.
Great topic Bob! What is the date of the Magic City Classic this year? I'm not sure if business travel will allow me to sneak that one in again, but I sure enjoyed the last time I was up there...if it wasn't for Blackstock out-qualifying me! (Like THAT was a surprise!)
Cheers,
John
The more likely statistical representation for individual lap times is a skewed distribution. Skewness occurs when there is some type of boundary (like the fastest lap) where data congregates towards that boundary value but not past it.
Great topic Bob! What is the date of the Magic City Classic this year? I'm not sure if business travel will allow me to sneak that one in again, but I sure enjoyed the last time I was up there...if it wasn't for Blackstock out-qualifying me! (Like THAT was a surprise!)
Cheers,
John
#86
Although it is theoretically possible for individual laps times to be normally distributed, it is highly unlikely (at least for one individual). A normal distribution requires the mean (average) and the median (the middle-most value) to be close or identical. This means that 50% of your laps are above and 50% are below your median.
The more likely statistical representation for individual lap times is a skewed distribution. Skewness occurs when there is some type of boundary (like the fastest lap) where data congregates towards that boundary value but not past it.
Great topic Bob! What is the date of the Magic City Classic this year? I'm not sure if business travel will allow me to sneak that one in again, but I sure enjoyed the last time I was up there...if it wasn't for Blackstock out-qualifying me! (Like THAT was a surprise!)
Cheers,
John
The more likely statistical representation for individual lap times is a skewed distribution. Skewness occurs when there is some type of boundary (like the fastest lap) where data congregates towards that boundary value but not past it.
Great topic Bob! What is the date of the Magic City Classic this year? I'm not sure if business travel will allow me to sneak that one in again, but I sure enjoyed the last time I was up there...if it wasn't for Blackstock out-qualifying me! (Like THAT was a surprise!)
Cheers,
John
http://www.rctech.net/forum/showthread.php?t=195146
#87
Tech Champion
iTrader: (38)
Bob, hope you dont mind. But I wrote a program to parse Rc Scoring Pro results and made some graphs for our local drivers in Ohio based on the results using your calculation. I also expressed it in a different way. % of capability. Seems to be a hit!
Here are some screen shots. I used Eli's results since they are probably the best
Here are some screen shots. I used Eli's results since they are probably the best
#88
Bob, hope you dont mind. But I wrote a program to parse Rc Scoring Pro results and made some graphs for our local drivers in Ohio based on the results using your calculation. I also expressed it in a different way. % of capability. Seems to be a hit!
Here are some screen shots. I used Eli's results since they are probably the best
Here are some screen shots. I used Eli's results since they are probably the best
Good job with the software, very nice... HOWEVER, stop helping Eli go faster... that'll be enough of THAT!!!
There are a few things I've noticed can slightly mess with the results. If you miss when you should have hit your best lap (somewhere in the first 10 laps), and hit it later in the run, your accuracy will be slightly higher than it should have been. Because your fastest lap won't be quite as fast as it could have been.
All in all though, a simple way to show what your doing. And how close you are to driving your package to it's potential.
#89
Tech Master
iTrader: (1)
Cool, I had the same idea. I wrote an Excel macro that analyzes RC Scoring Pro's RoundX.txt file and pumps out the Stormer Factors.
Unfortunately, the usability is severely hurt by Scoring Pro's lack of a good, simple format for storing results. I did the best I could.
To use this, open the excel file attached. You'll probably have to turn down Macro security settings and restart Excel for it to work. Then, open the roundX.txt file that you want Stormer numbers for. In the Import window, use "Fixed Width", and set a column break every 8 steps. So, there will be a column line at 8, 16, 24, 32, 40, 48, 56, 64, 72, and 80. If that works, do a control+a (select all), and paste into Stormer.xls. Then, go Tools --> Macro --> Macros and press Run. That should pump out the stormer factors just to the right of the results (column K).
It's a bit of work and probably requires some Excel experience to use. We'll have to bug the Scoring Pro guy to have his program output an easier to use format, like csv or tab-delimited. If the program would put out results in one of those formats, I would write a .net app where you simply use a windows dialog box to open the results file. It would automatically add Stormer factors to the results.
The really strange thing is that my results directory has a RoundXImport.txt and a RoundXImport-NoLaps.txt, but unfortunately both have no laps. It's like there is a bug in the program, because it seems like RoundXImport.txt should have laps.
Standard deviation is still probably the all-around best marker for consistent driving.
-Adam
Unfortunately, the usability is severely hurt by Scoring Pro's lack of a good, simple format for storing results. I did the best I could.
To use this, open the excel file attached. You'll probably have to turn down Macro security settings and restart Excel for it to work. Then, open the roundX.txt file that you want Stormer numbers for. In the Import window, use "Fixed Width", and set a column break every 8 steps. So, there will be a column line at 8, 16, 24, 32, 40, 48, 56, 64, 72, and 80. If that works, do a control+a (select all), and paste into Stormer.xls. Then, go Tools --> Macro --> Macros and press Run. That should pump out the stormer factors just to the right of the results (column K).
It's a bit of work and probably requires some Excel experience to use. We'll have to bug the Scoring Pro guy to have his program output an easier to use format, like csv or tab-delimited. If the program would put out results in one of those formats, I would write a .net app where you simply use a windows dialog box to open the results file. It would automatically add Stormer factors to the results.
The really strange thing is that my results directory has a RoundXImport.txt and a RoundXImport-NoLaps.txt, but unfortunately both have no laps. It's like there is a bug in the program, because it seems like RoundXImport.txt should have laps.
Standard deviation is still probably the all-around best marker for consistent driving.
-Adam
#90
Tech Champion
iTrader: (38)
Dont even get me started on the quirks in the RC Scoring Pro output. I used the same output file. The code took me about an hour to write and then 6 hours to account for all the weird things it can output. Good thing we had a blizzard and I had nothing else to do. Otherwise I would have just given up.