Hope my theorizing was entertaining is all...
Wow. I don't even know how that happened. Fixed.
No.
When I was rethinking the wall, AKA premeditating it's death, this was one of my big questions. At some point I thought of dropping the ratings on the extremes and having the top 5/bottom 5 be in each. Top 5 would have made sense, bottom 5 would have had to skip a full rank for a long time.
Or that it was nominated before GTP became quite so divided. No way in hell does it make it if gets polled tomorrow.
Wait...what? Could you explain so I don't assume what you meant?
The standard deviation is, simply put, the average difference between a data point and the mean in a set of data points. More relevant to the cool wall, it would take into account not just the average vote, but the distribution. Currently, a car that gets 50 SZ votes and 50 uncool votes scores the same as a car that gets 50 cool votes, even though the first was divisive and the second was unanimous. I aim to design a system that would change that. This way, cars that receive a large number of votes in one category will get a more extreme score than cars that don't have one dominant voting category.
Thank you for the explanation. Two things; First, don't expect people to get behind it....at all. Second, I wish you the best of luck in your endeavor.
I agree, however, I feel I can make it more democratic. The current system is very mathematically simple, when in fact there is a lot more information contained in the data set generated by each poll that could be tapped into to give a fairer and more accurate score for each car.
How easy/difficult would it be for someone to input results in Python? I know when @GranTurismo916 took over from me, I gave him a set of instructions for how to put the results into the Formula to get the result, since he said he didn't have that much experience with Excel. Ease of entry for the curator would have to be something taken into account.
It was actually your comments which inspired me to try and come up with an improved system.
If the code is well written, it should be as simple as entering the five numbers corresponding to the number of votes (the programme would show "How many Sub Zero votes were there?" and you'd type the number and hit enter) and then the programme would give you a few outputs, including the original score, standard deviation, adjusted score, and total votes. Even then, though, the rawness of it can be intimidating.
The one in the OP I just checked goes to the 2011 Porsche 911 GT3 RS 4.0 (997)It was actually your comments which inspired me to try and come up with an improved system.
If the code is well written, it should be as simple as entering the five numbers corresponding to the number of votes (the programme would show "How many Sub Zero votes were there?" and you'd type the number and hit enter) and then the programme would give you a few outputs, including the original score, standard deviation, adjusted score, and total votes. Even then, though, the rawness of it can be intimidating.
To be honest I'll probably do it in LabVIEW once I've got the basic ideas sorted since LabVIEW is much better at processing data and it's also much more user-friendly. Unfortunately it's much less common since it's not free, so if I do it that way I'll upload the VI but I doubt anyone else will be able to use it. I'd also consider doing it in MATLAB but that's not free either.
Anyway, I'll keep you all posted on how it goes. @GranTurismo916 Do you have all the raw data from the cars already in an Excel spreadsheet or similar format? If so, I may ask to use it at some point. If not, well, a raw data sheet might be useful... I'm going to spend way too much time on this.
D'oh. Thanks, I'll use that for now. I don't think I'll need an up to date version until the programme is finished.
I have the updated version, I just haven't updated it on the actual threads. I should do that now, actually.
Thanks. I try my best.
