Makeshift Shuffle Club - Time Trials & Testing for club car lists - all welcomeOpen

[amarynceos]

1 and 2, ideally 3, but not as fussed on the latter. :3 The Coupe part is important, though. No sports sedans!

 

[watermelon punch]

1:35.761 -- Lotus Evora '09 (old time 1:36.846)
1:37.308 -- Hyundai Genesis Coupe 3.8 Track '13
1:37.712 -- Nissan SKYLINE Coupe 370GT Type SP '07
1:38.650 -- Audi TTS Coupe '09


The Audi is a bit too slow, perhaps, but the Skyline and the Hyundai are going on the list of potentials.
👍

A 6 second spread at Laguna Seca between fastest & slowest of 20+ cars at 450+pp is nothing.
The issue would be if the next faster car was relatively different than others. - the gaps between the cars should be fairly equal.. with any larger gaps occurring lower in the list.
IE: you have gaps, you have the smaller gaps between #1, 2, 3... and if the gap increases at all, it should be at #9, #10... the cars that come into play when there's a lot of cars on track.

I know everyone's really worried about there being too much difference between cars in the ilst, but in application in the races, we all know that the difference isn't anywhere near as big as you believe it would be on most tracks.

I'm more worried about the differences in top speeds in the fast cars. Which will be raced more often on the big tracks.
Much like with the Low Power list, which every track is a big track to the low power cars.
So top speeds should be integrated into the data for the faster car lists. (Like I did with the low power cars and traffic.)

 

[snowgt]

A 6 second spread at Laguna Seca between fastest & slowest of 20+ cars at 450+pp is nothing.
The issue would be if the next faster car was relatively different than others. - the gaps between the cars should be fairly equal.. with any larger gaps occurring lower in the list.
IE: you have gaps, you have the smaller gaps between #1, 2, 3... and if the gap increases at all, it should be at #9, #10... the cars that come into play when there's a lot of cars on track.

I know everyone's really worried about there being too much difference between cars in the ilst, but in application in the races, we all know that the difference isn't anywhere near as big as you believe it would be on most tracks.

I'm more worried about the differences in top speeds in the fast cars. Which will be raced more often on the big tracks.
Much like with the Low Power list, which every track is a big track to the low power cars.
So top speeds should be integrated into the data for the faster car lists. (Like I did with the low power cars and traffic.)

In the old days, I used to disagree about the spread of cars, but in the meantime I think you're right. The lists can surely have a spread of 10 seconds or so on a medium-long circuit (lap time maybe 1:45 or so). That's because the lists have to work in 2 scenarios:

Scenario #1: Room is half full. That essentially reduces the spread to half of what it is from car #1 to car #16. If 16 cars are too close, then the handicap-system doesn't work anymore for smaller fields.

Scenario #2: Room close to being full. The previous race's winner might not be able to compete for a win or even a podium in the next race because of the big spread, but he might end up in the midfield, and in the next race after that, will have a car to compete with again. It evens out over a couple of races.

It's quite a good system actually.

ps: I agree that top speeds should play a part in choosing the cars. For list order, the lap time has to be the deciding factor though. But if a car's performance in a straight line is massively different than on a whole lap, some cars might just have to be cut.

 

[tarnheld]

ps: I agree that top speeds should play a part in choosing the cars. For list order, the lap time has to be the deciding factor though. But if a car's performance in a straight line is massively different than on a whole lap, some cars might just have to be cut.

With this in mind i would like to call all test drivers (@snowgt has done this already 👍) to note the top speed on a long straight, preferably on a high speed track. You can also test the top speed on SSRX. Or if anyone has a list of top speeds recorded from 1.09 on, please share. Top speeds have changed considerably, i don't know if they changed relative to one another.


A different thing:

As is noticed we don't have the Shelby GT350 '65 in any list, please share your views on that car, as i would really like to add it to a list. From my view it might fit in the Classic List, but might be too slow. It's really a great car, premium, maybe too tame for the Legends list.

 

[Patrick8308]

As is noticed we don't have the Shelby GT350 '65 in any list, please share your views on that car, as i would really like to add it to a list. From my view it might fit in the Classic List, but might be too slow. It's really a great car, premium, maybe too tame for the Legends list.

Actually this refreshes my idea of a Classic Muscle Car List.💡

I have seen in any of those timetable txt files that someone had tested them in the past already.

Maybe the idea was cut due to some reasons. I don't know about this.

But this would actually be the potential list:

Pontiac Firebird Trans Am '78
PP 415 FR 220 HP 1640 kg 435 Nm 30,000 Cr
Ford Mustang Mach 1 '71
PP 453 FR 300 HP 1615 kg ------ 50,000 Cr
Dodge Charger Super Bee 426 Hemi '71
PP 473 FR 425 HP 1841 kg 650 Nm 80,000 Cr
Plymouth Cuda 440 Six Pack '71
PP 479 FR 385 HP 1576 kg 651 Nm 58,360 Cr
Chevrolet Chevelle SS 454 '70
PP 479 FR 450 HP 1762 kg 677 Nm 70,000 Cr
Dodge Charger 440 R/T '70
PP 466 FR 375 HP 1650 kg 651 Nm 75,500 Cr
Plymouth AAR Cuda 340 Six Barrel '70
PP 435 FR 290 HP 1594 kg 467 Nm 110,000 Cr
Chevrolet Nova SS '70
PP 467 FR 375 HP 1590 kg 563 Nm 65,000 Cr
Dodge Challenger R/T '70
PP 490 FR 425 HP 1724 kg 664 Nm 75,000 Cr
Plymouth Superbird '70
PP 480 FR 425 HP 1755 kg 664 Nm 42,440 Cr
Chevrolet Camaro SS '69
PP 451 FR 300 HP 1401 kg 515 Nm 60,000 Cr
Chevrolet Camaro Z28 '69
PP 451 FR 290 HP 1415 kg 393 Nm 65,000 Cr
Chevrolet Corvette Stingray Convertible (C3) '69
PP 456 FR 300 HP 1552 kg 515 Nm 50,000 Cr
Chevrolet Corvette Stingray L46 350 (C3) '69
PP 464 FR 350 HP 1490 kg 515 Nm 71,500 Cr
Chevrolet El Camino SS 396 '67
PP 453 FR 325 HP 1468 kg 555 Nm 30,000 Cr
Mercury Cougar XR-7 '67
PP 462 FR 310 HP 1418 kg 581 Nm 47,280 Cr
Shelby GT350R '65
PP 455 FR 306 HP 1268 kg 446 Nm 106,110 Cr
ShelbyGT350 '65
PP 454 FR 306 HP 1270 kg 446 Nm 50,000 Cr
Pontiac Tempest Le Mans GTO '64
PP 461 FR 325 HP 1420 kg 580 Nm 52,600 Cr
Chevrolet Corvette Coupe (C2) '63
PP 430 FR 249 HP 1370 kg 474 Nm 84,890 Cr

With the Shelby in it of course!

 

[LongbowX]

@Patrick8308 I actually did a test with all those muscle cars way back. The only problem with them is that they would be restricted to certain tracks because their top speed is so limited. There really are rather few tracks that are even suitable. And the spread of times wasn't really enough as well.

 

[Patrick8308]

Oh, I wasn't aware of that.
When I have been driven muscle cars I probably had installed adjustable transmission, so never recognized that issue.
But with what you explaining, all the limits, this list might not be worth investing time in the testing process.
So the idea is cut again!

 

[snowgt]

With this in mind i would like to call all test drivers (@snowgt has done this already 👍) to note the top speed on a long straight, preferably on a high speed track. You can also test the top speed on SSRX. Or if anyone has a list of top speeds recorded from 1.09 on, please share. Top speeds have changed considerably, i don't know if they changed relative to one another.

I would like to reiterate that SSRX top speeds are not representative for 2 reasons. 1) because apparently for some reason that only PD knows there's decidedly less or no drag on that track, so the top speeds are way higher than on any other track, and 2) you will not reach that top speed on any "normal" track. Better to note down the top speed achieved on a normal circuit. (The Ferrari GTO might top out at 300+km/h, but you'll usually never get there)

A different thing:

As is noticed we don't have the Shelby GT350 '65 in any list, please share your views on that car, as i would really like to add it to a list. From my view it might fit in the Classic List, but might be too slow. It's really a great car, premium, maybe too tame for the Legends list.

Actually this refreshes my idea of a Classic Muscle Car List.💡

I have seen in any of those timetable txt files that someone had tested them in the past already.

Maybe the idea was cut due to some reasons. I don't know about this.

But this would actually be the potential list:

(...)

With the Shelby in it of course!

The general problem with muscle cars is that most of them will run into the limiter on the majority of tracks, whereby they are rendered useless when tuning (installing a custom transmission) is prohibited.

The Shelby GT350 might just have a high enough top speed before running into the limiter, I only know the GT350R tops out at 210km/h. But this could already be too low when using it on tracks with longer straights.

I've started my testing of the potential Classics-cars yesterday, and the nominally slowest of the bunch I'm looking at currently is the Ferrari Dino. The Shelby should be roughly in the same territory, maybe a tiny bit slower. I'll make sure to test it before I'm through.

 

[watermelon punch]

RE: SSRX & top speeds
They should ONLY be judged against top speeds at SSRX.
IE: no mixing "straight at GVS" and "SSRX top speed".
This avoids the issues of SSRX being relevant. Because the cars are merely compared against each other.

FYI: The top speeds at SSRX are mainly to find potential limiter issues in slipstream draft scenarios.
IE: DO not schedule a 1-make race with this car at Indy SS or SSR7!
(So this data is MAINLY for the host, and for people picking combos.)

Unfortunately... there's no other easy track to test top speeds, because it all involves getting out of the turn before the straight properly. :/
Meaning that ONE person would have to do ALL the top speeds on ONE track for EVERY car in the list.... for that to be relevant.
You could not compare S4 at GVS with me driving, with GTO MR at GVS with snowgt driving. :/ Because there would be error... major room for error.

But don't worry too much about the relevancy...

RE: CAR LIST ORDERING & factors

The new spreadsheet system seems to be working VERY similar to the old way with my comparisons.

Except nice that we don't need all cars on one track by only one driver.

I would ALSO say that the car list order it spits out is not "written in stone".
It will of course also require the keen eye to suss out discrepancies.
The spreadsheet system is not capable of determining the "gaps". This needs familiarity with the cars, and looking at the sets of times.

I don't know of a way the spreadsheet can be used to suss out the gaps between cars. Maybe there is a way, but that would require @tarnheld because I just don't understand the mathcrobatics well enough to come up with something like that. And besides... I don't think it's necessary anyhow, as the person who comes up with the list is usually aware of the gaps of performance anyway.
(Example: I'm most familiar with this with the Vintage & Legends... Longbowx is most familiar with the 90s... And amar will recognize the differences in the Sport Coupes... and this is because it's your "baby". When you pick cars for a list, you have a good sense of where they stand in the first place or you don't consider them... even before test results are accumulated.)

I'm also hoping that MAYBE @tarnheld can find a way in the spreadsheet (NO PRESSURE THOUGH! ) ...
To give more or less weight to some sets of times than others.💡
I don't know if this is possible... or too crazy!
It can also be done manually though... so it's not really pressing that the spreadsheet should be able to do this.
Right now I'm really working hard to bring out the new Vintage list.
And I'm using my own method to give more weight to certain sets of times. (By just doubling those sets of times in the data).

Example: Top speeds being worth less than a full set of times on one track by one driver.
Example: Individual tests between specific sets of cars on various tracks being given more weight than all other times
OR... whatever makes sense.

RE: VARIABLE DIFFERENCES WITH SMALLER GRIDS

I see this issue, and this is why I think it's okay to have less difference at the 1,2,3 between the cars.
And to have larger differences as the list goes down.

There is some problem having the lists work with different amounts of drivers.

This is because with a small grid... even when fastest drivers start out in back, in slightly slower cars... there is less traffic to pass and maintain ahead... So a little less challenge for the aliens.

But generally, I think the list works best when the average drivers outnumber the fastest drivers.
For example, for every very fast driver present, there should be 1.5 average drivers.
This shouldn't be too hard to achieve because average drivers (by the very nature of being average) far outnumber aliens!!

So the answer to this is... invite your friends to join us!!
Any friends... whether faster or slower... because the more people, the more average drivers!


RE: Muscle cars & American cars :/

Yeah, it stinks. Nearly all of them have such horrible transmission issues with tuning prohibited. That's why American cars are severely underrepresented in our shuffle lists.

 

[Chiochan]

Have we considered a Rally collection? I noticed the increased interest in snow tracks, and a rally group might be fun for all to throw around once in a while. Any interest and I could get a list together.

 

[watermelon punch]

@Chiochan
Certainly!! 💡

Another note about spreads/gaps...

In the old days, I used to disagree about the spread of cars, but in the meantime I think you're right. The lists can surely have a spread of 10 seconds or so on a medium-long circuit (lap time maybe 1:45 or so).

Yes... I think a 10 second gap on a track with a 1:45 ballpark lap time is EASILY proper. 👍
My rule of thumb with the lower power cars has been to stay under 10 seconds spread between 15 cars on Deep Forest.

But for higher power cars, I think on a longer track, you can go as high as 12 seconds difference between 15 cars without creating a problem.

I say 15 cars because I think 15 should be the amount everyone should shoot for on a list.

It's okay to have 16 cars... though we have rarely had more than 13 drivers in a race.

And I think the minimum should be 12 cars.

If there are less than 12 cars appropriate for a list (and this could happen) it should be set up something like this:

10 cars:

1. Car A
2. Car B
3. Car C
4. Car D
5. Car D
   
6. Car E
7. Car E
   
8. Car F
9. Car G
10. Car H
11. Car I
12. Car I
    
13. Car J
14. Car J
15. Car J

 

[snowgt]

FYI: The top speeds at SSRX are mainly to find potential limiter issues in slipstream draft scenarios.
IE: DO not schedule a 1-make race with this car at Indy SS or SSR7!
(So this data is MAINLY for the host, and for people picking combos.)

Ok, for that purpose only it works perfectly fine.

Unfortunately... there's no other easy track to test top speeds, because it all involves getting out of the turn before the straight properly. :/
Meaning that ONE person would have to do ALL the top speeds on ONE track for EVERY car in the list.... for that to be relevant.

Which is what I do. And as far as I saw, the top speeds are quite robust, as long as you get a halfway decent exit out of a turn - it doesn't need to be perfect. By the end of the straight, the speeds are asymptotically converging as the rate of acceleration slows down. They are just a helper to see, if there are oddities with a car maybe though.

I don't know of a way the spreadsheet can be used to suss out the gaps between cars. Maybe there is a way, but that would require @tarnheld because I just don't understand the mathcrobatics well enough to come up with something like that. And besides... I don't think it's necessary anyhow, as the person who comes up with the list is usually aware of the gaps of performance anyway.

For mere ordering of a list, I'd say you don't need to take the gaps into account. As long as there is no cluster of cars at some point in the order, it should be fine. I think everyone knows that not all cars will have the same gaps between each other. Sometimes the handicap is a little bigger and sometimes less. Changes with track anyway and the drivers don't have even gaps between them either.

 

[snowgt]

Car List Ordering Mathcrobatics®

As i was involved in the crunching of the TT times of the 90's list and the Hot Hatch reordering, i caught a glimpse on the difficulty making these lists. The experience i got from these exercises was:

• It's easy to order by one tester/track combo
• It's hard to get a meaningful order by multiple testers/tracks (think of possible contradicting car orders of two testers)
• It's hard to extract any info more than the order of the cars, for example time differences depend on tester,track,etc.

The ranking method of @watermelon punch takes all of this into account, but has the flaw that you need a full list of TT times for one driver/track combo, as the ranking for a subset of cars cannot be compared to another ranking of a different subset.

With the new TTs in the club there was good data, but the old method didn't work cleanly with this, so we had to massage the data, and try to figure out how to fit this into the ordering. Various ways were tried, but all brought some form of bias into the ordering. So the task was to fight the bias and come up with a method that just works and can incorporate even just two extra TT times from two cars by one driver without blowing up.

So what is the most basic info we can extract from the TT times? It's which car is faster than another. For this you only need the two TT times of two cars. So if we go through all car pairings, and extract this info, we can give each car pair A,B a number of 1 if A is faster than B, and 0 if A is not faster than B. If we sum up this numbers for all driver/track combos, we get a table like this:

__A B C D ...
A 0 5 7 6
B 2 0 4 7
C 0 5 0 8
D 0 0 0 0
...

This is just a compact way to store the info "A beats B 5 times", "A beats C 7 times", "A was beaten by B 2 times" and so on.

Now we need the info how many times a comparison was possible for a car pair, because not every car pair has been tested by each driver on each track. So we need another table like this:

__A B C D ...
A 7 7 7 6
B 7 7 7 8
C 7 7 8 8
D 6 8 8 6
...

This table is a compact way to store the info "We have 7 TT times for car pair A,B, 6 TT times for car pair A,D, 8 TT times for car pair B,D" and so on.

Now we can divide each entry of table one by table two, to get a percentage value that says for each car pair "A beats B 5 times out of 7, that's 71%"

__A B C D ...
A 0/7 5/7 7/7 6/6
B 2/7 0/7 7/7 8/8
C 0/7 5/7 0/8 8/8
D 0/6 0/8 0/8 0/6
...

Now check out the rows: If you sum it up, the number tells you how many times a car was faster than all others.
For example car A: 2.71 times it was faster than all others.
And the columns: If you sum it up, the number tells you how many times a car was slower than all others.
For example car A: 0.29 times it was slower than all others.

______A B C D ...
______0.29 1.42 2.0 3.0
A 2.71 0/7 5/7 7/7 6/6
B 2.29 2/7 0/7 7/7 8/8
C 1.71 0/7 5/7 0/8 8/8
D 0.00 0/6 0/8 0/8 0/6
...

Subtract those two numbers of the car, and you get a measure for the fastness of the car against all others.

A: 2.71-0.29 = 2.42
B: 2.29-1.42 = 0.87
C: 1.71-2 =-0.29
D: 0-3 =-3

These numbers tell you the fastness of a car relative to the others, 0 means faster than half of the other cars, anything higher is faster than half of the other cars, negative is slower than half of the other cars.

So that is the new (world) order, i will stop here and let you digest that, hope somebody get down here!
Just some advantages:

• As there is only the counting of 'faster than' points, there is no bias of track length, number of cars done by driver, etc.
  
• It's easy to extract problematic cars, were more testing is needed.
• We can easily add other 'faster than' tests, for example top speed.

I have finally thought this through and I didn't find any major flaw with it, so it's a good method.

There's one mistake in your example though. The B/C values are wrong from the second matrix onwards. That's what threw me off at first, because the results then violate a couple of principles that the method should sustain. E.g....

The sum of element a(ij)+a(ji) = 1 always. (3rd matrix onwards, B&C violate that rule)
The elements a(ii) = 0 always.
The maximum possible sum of 1 row or column is always N-1, where N is the number of cars tested.
The sum of Row X + Column X = N-1.

You basically wouldn't need to subtract column from row anymore, but it moves the figures to a different interval, i.e. from [0,N-1] to [-N+1,N-1], thereby doubling the gap between each cars "index figure".

Good work though! 👍

 

[watermelon punch]

Which is what I do. And as far as I saw, the top speeds are quite robust, as long as you get a halfway decent exit out of a turn - it doesn't need to be perfect. By the end of the straight, the speeds are asymptotically converging as the rate of acceleration slows down. They are just a helper to see, if there are oddities with a car maybe though.

I was thinking mainly this is an issue with lower power cars.
IE: GVER in low power... might not be the same for 2 different drivers.

 

[tarnheld]

There's one mistake in your example though. The B/C values are wrong from the second matrix onwards. That's what threw me off at first, because the results then violate a couple of principles that the method should sustain. E.g....

Yeah, just thought this up for illustration, this was not derived from real TT times.

The sum of element a(ij)+a(ji) = 1 always. (3rd matrix onwards, B&C violate that rule)

Hey that's a good point! 👍 Right now there's a subtle error in my calculation, because i don't count a tie in TT times, but add that comparison to the total number of comparisons. That means a tie is actually counted as "slower than"!

And with that invariants, we don't need the final subtraction anymore. 👍

Thanks!

 

[snowgt]

Yeah, just thought this up for illustration, this was not derived from real TT times.



Hey that's a good point! 👍 Right now there's a subtle error in my calculation, because i don't count a tie in TT times, but add that comparison to the total number of comparisons. That means a tie is actually counted as "slower than"!

And with that invariants, we don't need the final subtraction anymore. 👍

Thanks!

Actually, if all testers arrive at the same order of cars, you would get a result-series for the cars with increments of 1 (or 2 with subtractions). "0" for the slowest car, and "N-1" for the fastest. Any divergence from that series is an indicator that the result was not unambiguous and might need some study.

It's very rare that you would get the same time to 1/1000s, so there will always be a "faster than". Even if they turn out to be "the wrong way round", results from others should even that out.

 

[tarnheld]

Actually, if all testers arrive at the same order of cars, you would get a result-series for the cars with increments of 1 (or 2 with subtractions). "0" for the slowest car, and "N-1" for the fastest. Any divergence from that series is an indicator that the result was not unambiguous and might need some study.

It's very rare that you would get the same time to 1/1000s, so there will always be a "faster than". Even if they turn out to be "the wrong way round", results from others should even that out.

Yeah, if there were not those STIG times only available up to the seconds...

Here is a link to an example Spreadsheet with the new ideas incorporated:

https://docs.google.com/spreadsheets/d/1KlsNAnoVR202MHBQwjgvauclwxKqO-fW_x5FunoZuzc/edit?usp=sharing

It contains TT times for 33 cars from various anonymous sources. The List is on page "PreliminaryList" (you can switch between pages at the bottom). There is an "FTCheck page" with various checks if the ordering is valid, and to see where more tests are needed. The other pages contain the calculation steps.

 

[watermelon punch]

Vintage (revamped)

1. Honda Civic 1500 3door CX '79
2. Nissan Skyline 2000GT-B (S54B) '67
3. Alfa Romeo Spider 1600 Duetto '66
4. Mini Marcos GT '70
5. Isuzu Bellett 1600 GT-R '69
6. Isuzu 117 Coupé '68
7. Alfa Romeo Giulia Sprint Speciale '63
8. Honda 1300 Coupe 9 S '70
9. Lotus Europa S.2 '68
10. Toyota Celica 1600GT (TA22) '70
11. BMW 507 '57
12. Mazda Cosmo Sport (L10A) '67
13. Mitsubishi Lancer 1600 GSR '74
14. Toyota Sprinter Trueno GT-Apex (AE86) '83
15. Nissan Fairlady 2000 (SR311) '68

This is how it looks now.
And I think we might have enough tests at this point actually.


90s Sports (revision)

1. Toyota CELICA GT-FOUR (ST205) '98
2. Nissan Silvia K's Aero (S14) '96
3. Mitsubishi Lancer Evolution II GSR '94
4. Nissan Fairlady Z 300ZX Version S TwinTurbo 2seater (Z32) '98
5. Audi S4 '98
6. Mitsubishi GTO Twin Turbo MR '98
7. Chevrolet Camaro Z28 Coupe '97
8. Acura NSX '91
9. TVR V8S '91
10. Mazda éfini RX-7 Type R (FD) '91
11. Toyota SUPRA RZ '97
12. Nissan SKYLINE GT-R V・spec II (R32) '94
13. Subaru IMPREZA Sport Wagon WRX STi Version VI '99
14. Chevrolet Corvette GRAND SPORT (C4) '96
15. Lotus Elise Sport 190 '98

This is how it looks if we keep both replacements for MR2.
I think it looks pretty fair.

Retro List

1. Honda Accord Coupe '88
2. Honda Civic 1500 3door 25i '83
3. Isuzu Piazza XE '81
   
4. Nissan Exa Canopy L.A. Version Type S '88
5. Honda City Turbo II '83
6. Toyota MR2 1600 G '86
7. Honda Ballade Sports CR-X 1.5i '83
8. Volkswagen Golf I GTI '76
9. Nissan Silvia Q's (S13) '88
10. Alpine A310 1600VE '73 <-------- Potentially on Chopping Block (but would be a shame)
    
11. Nissan Fairlady Z 280Z-L 2seater (S130) '78
12. Toyota Corolla Levin GT-Apex (AE86) '83
13. Nissan Skyline Hard Top 2000 GT-R (KPGC10) '70
14. Mitsubishi Lancer EX 1800GSR IC Turbo '83
15. Toyota Celica 2000GT-Four (ST165) '86
16. Honda Prelude Type S '98 <-------- Potentially on Chopping Block (due to hostility lol)
    
17. Ford Taurus SHO '98

This list might benefit from
@snowgt running
Volkswagen Golf I GTI '76
(comfort soft) on Motegi East
Because I don't have a time from snow from VW Golf 76
(It has a tie with the Ballade right now.)
(I also can complete another set of times. I am 4 cars short now of having 2 tracks completed with all cars.)


Legends (revamped)

1. Renault R5 Turbo '80
2. Mercedes-Benz 300 SL Coupé '54
3. Alpine A110 1600S '73
4. Nissan Skyline Hard Top 2000 Turbo RS (R30) '83
5. Lotus Europa Special '72
6. Jaguar E-Type Coupe '61
7. Lancia Stratos '73
8. Nissan Fairlady Z 300ZX (Z31) '83
9. Pontiac Firebird Trans Am '78
10. DMC DeLorean S2 '04
11. Chevrolet Camaro IROC-Z Concept '88
12. Dome Zero '78
13. Peugeot 205 Turbo 16 '85
14. Nissan Fairlady Z 300ZX TT 2seater (Z32) '89
15. Nissan Silvia 240RS (S110) '83
16. Toyota Sprinter Trueno GT-Apex (AE86 Shuichi Shigeno Version) '00
17. Ford RS200 '84 <------------ Maybe too fast for this list, even as fastest car

The bold cars need further comparison testing.


Historic

1. Toyota MR2 1600 G '86 <------- This might be too slow for #1 but I don't know
   
2. Toyota Celica XX 2800GT '81
3. Alfa Romeo Giulia Sprint GTA 1600 '65
4. Nissan Skyline 2000GT-R (KPGC110) '73
5. Honda CR-X del Sol SiR '92
6. Nissan Fairlady 240ZG (HS30) '71
7. Toyota MR2 1600 G-Limited Super Charger '86 <------- Only ONE MR2 would be kept (if any)
8. Toyota 2000GT '67
9. Mazda MX-5 (NA) '89 <------- This needs to be tested against the BMW 2002 Turbo 73
10. BMW 2002 Turbo '73 <------- This needs to be tested against the Miata MX-5 '89
11. Audi quattro '82
12. Mitsubishi Galant 2.0 DOHC Turbo VR-4 '89
13. Mazda Savanna RX-7 GT-Limited (FC) '85
14. Nissan Silvia K's (S13) '88
15. Chevrolet Corvette Coupe (C2) '63 <------- Several more tests needed
16. Honda Prelude Si VTEC '91 <------- Several more tests needed
17. Buick GNX '87 <------- Several more tests needed
18. Toyota Supra 3.0GT Turbo A '88

The cars in bold need more testing.


Premium Sports (initial preliminary view)

1. Audi TTS Coupe '09
2. Nissan SKYLINE Coupe 370GT Type SP '07
3. Lotus Evora '09
4. BMW Z4 M Coupé '08
5. Art Morrison Corvette '60
6. Lotus Esprit V8 '02
7. Jaguar XKR Coupe '10
8. Mazda RX-7 Spirit R Type A (FD) '02
9. Cadillac CTS-V Coupé '11
10. Nissan Fairlady Z (Z34) '08
11. Nissan SKYLINE GT-R V-Spec II Nur (R34) '02
12. Chevrolet Corvette Z06 (C5) '04
13. Aston Martin V12 Vantage '10
14. Amuse NISMO 380 RS Super Leggera
15. Toyota FT-86 G SPORTS Concept '10
16. Hyundai Genesis Coupe 3.8 Track '13
17. TVR Tamora ’02
18. Honda NSX Type R '02
19. SRT Viper GTS '02
20. Alfa Romeo 8C Competizione '08
21. Ferrari California '08
22. Chevrolet Corvette Stingray (C7) '14

This is just with the preliminary data.
This is up to @amarynceos to whittle this list down a bit, after some more initial testing. ^^ 👍

So I will leave you to it, and just give just my initial opinions.
But I have NOT yet tested any of these cars yet... so my opinions are based on looking over the specs, looking at initial sets of lap times, and "optics".

I don't think the spread between #1 & #22 is a problem at all.

I think more cars at the top are "cuttable" because there's more bunching of times (the cars are closer)... but also because the cars in the bottom part of the list are more exciting. (But maybe that's just me, I don't know about their popularity in general.)

But my big opinion is that the Ferrari California should be UNCUTTABLE!!
I think this list would be incomplete without the Ferrari!!!


So....
Comments?
Questions?
Criticisms?
Complaints?
Whining & crying & gnashing of teeth?
Excitement & enthusiasm maybe? ^^

 

[amarynceos]

Liking the revamped Vintage list (yay for Europas!) and the 90s list looks good. Good to see another Europa on the Legends list now, too. 👍 Not sure about that IROC-Z, but that's only because my immediate reaction to those cars is 'kill it with fire.' Pity the 2000GT had such issues, it was a fun car to drive but there was always the 'ah maaaan' feeling when you got stuck with it. Good for it to go.

I'm excited about new lists!

----

I wasn't clear I guess about my concerns with spread on my list... I was worried that it wasn't enough, actually, hence my desire for a few more slower cars. I've pruned it down to 16 at the moment, with a 6 second spread at Laguna (same spread as the Luxury list). I'm still weighing the reintroduction of the 430 and GT-R at the fast end of the list, which would enhance the spread to 7s or so, and require further cuts (I have a group in there that's running too close to each other, which needs to be pruned anyway) -- the 430 and GT-R mark the bottom end of supercar territory, though.

Anyway, current working list:

Premium Sports Coupes (initial preliminary view)

1. Hyundai Genesis Coupe 3.8 Track '13
   
2. Nissan SKYLINE Coupe 370GT Type SP '07
3. Lotus Evora '09
4. BMW Z4 M Coupé '08
5. Lotus Esprit V8 '02
6. Jaguar XKR Coupe '10
7. Cadillac CTS-V Coupé '11
8. Nissan Fairlady Z (Z34) '08
9. Chevrolet Corvette Z06 (C5) '04
10. Aston Martin V12 Vantage '10
11. TVR Tamora ’02
12. Honda NSX Type R '02
13. SRT Viper GTS '02
14. Alfa Romeo 8C Competizione '08
15. Ferrari California '08
16. Chevrolet Corvette Stingray (C7) '14

The Ferrari is uncuttable, btw. And yes, there are still two Corvettes, but I'd only part with one reluctantly.

Even though it's not on the working list, I'm still going to run the Audi TTS at a few tracks to evaluate its relative gaps to the other slow fellers before I toss it completely.

 

[watermelon punch]

Well the Toyota 2000GT has migrated to the Historic List.
I think it will have a better opportunity to really shine there anyway!!

Premium SPorts...
RE: faster cars... yes, I think you can go faster... we want to get the Ferrari into play more!
It should be at #10.

 

[tarnheld]

If anybody wondered about the midfield of the 90's list, there were some TT times registered wrong.

I fixed this and here is the new list (MR2 axed) with the new calculation measure discussed above:

0.00 Toyota CELICA GT-FOUR (ST205) '98
1.00 Nissan Silvia K's Aero (S14) '96
2.42 Mitsubishi Lancer Evolution II GSR '94
2.75 Nissan Fairlady Z 300ZX Version S TwinTurbo 2seater (Z32) '98
4.00 Audi S4 '98
6.00 Acura NSX '91
6.33 Mitsubishi GTO Twin Turbo MR '98
6.45 Chevrolet Camaro Z28 Coupe '97
7.95 TVR V8S '91
8.10 Mazda éfini RX-7 Type R (FD) '91
10.53 Toyota SUPRA RZ '97
11.63 Subaru IMPREZA Sport Wagon WRX STi Version VI '99
12.17 Nissan SKYLINE GT-R V・spec II (R32) '94
12.60 Chevrolet Corvette GRAND SPORT (C4) '96
13.07 Lotus Elise Sport 190 '98

You see that the midfield 6-8 is pretty close together. But we really need some more data for the A4, as there are only 2 tests against the Acura and the Fairlady.


The Premium Sports list is very volatile now, after i found the same misregistering bug there, it now looks like this:

0.00 Nissan SKYLINE Coupe 370GT Type SP '07
1.00 Hyundai Genesis Coupe 3.8 Track '13
3.42 Lotus Evora '09
4.00 Nissan Fairlady Z (Z34) '08
4.00 BMW Z4 M Coupé '08
6.08 Cadillac CTS-V Coupé '11
6.33 Jaguar XKR Coupe '10
6.50 Lotus Esprit V8 '02
9.17 TVR Tamora ’02
9.17 Chevrolet Corvette Z06 (C5) '04
9.67 Honda NSX Type R '02
10.25 Alfa Romeo 8C Competizione '08
10.25 Aston Martin V12 Vantage '10
11.92 SRT Viper GTS '02
13.58 Ferrari California '08
14.67 Chevrolet Corvette Stingray (C7) '14

 

[amarynceos]

There's so little data thus far for the Premium Sports Coupes, no need to worry about ordering for some time yet. 👍

 

[watermelon punch]

There's so little data thus far for the Premium Sports Coupes, no need to worry about ordering for some time yet. 👍

Yeah I was just hoping to "put it out there" to whet the appetite at this point, because we had at least a little bit of info.
I haven't even done test one yet in that list.
Of course, I figure my testing will come in most useful for helping iron out cluster ties.

Also, I'm saving up for that 20million cr. Ferrari now.
So I was hoping the Premium Sports list would be cut down a tad before I went shopping.



ALSO>>> we have the GErman list... Scwarzwald League
But that has 26 cars on it (and lots of same model) still... so I think we'll see that list (more trimmed) when @Patrick8308 finishes his initial evaluations of the cars. 👍

 

[amarynceos]

1.7 million plus the '07 Skyline and the Hyundai. The Aston, Alfa and Ferrari are the most expensive at over 200k, and also three I'm not planning on cutting. But hey, at least I ditched the half-million FT-86 Concept!

 

[tarnheld]

Also, I'm saving up for that 20million cr. Ferrari now.

In my humble opinion, and fully aware of the backlash from GTO lovers, i have to say the driving experience is not worth the 20mil. It's a gorgeous car, but if you have 20 millions to spend, the 330 P4 is way cooler on Nordschleife. And for 1.25mil, the other GTO is a real bargain -- fun-wise. The 250 GTO was the most-'meh' car from my 20mil investments.

 

[watermelon punch]

Not sure what to do... but I think I won't be satisfied now until I have at least the option to buy another 20million cr. car.
So let me get 20million and then worry about where to spend it.

I only think the Premium Sports list is expensive because
a) I hadn't seen the cut down list yet
b) as a whole, they cumulatively cost a bit
c) they would seem cheap to me, but I've been slumming it in QMs.

Edit:
d) I don't have any already

I only had 7million last week. Now I have 17million... so it's not so bad.
At least somebody posted a good tune for the Megane QM. So we're making 350k every race.

 

[watermelon punch]

Just an FYI to @Patrick8308
I've collected all your times from the GVER time trial. 👍
So you don't need to type them out.

 

[snowgt]

Premium Sports, huh? Guess I know what I'm doing after testing the Classics list.

 

[tarnheld]

Actually, if all testers arrive at the same order of cars, you would get a result-series for the cars with increments of 1 (or 2 with subtractions). "0" for the slowest car, and "N-1" for the fastest. Any divergence from that series is an indicator that the result was not unambiguous and might need some study.

I think most of the time the data is ambiguous, that's the nature of the game. The TT times and the resulting ordering are what mathematicians (i guess you are in this league too ) call nontransitive, that is if car A is faster than B for tester X on track Y, and car B is faster than C for another tester/track, one can not say that A is faster than C. The ordering is futile mathematically, just as you cannot order rock,paper,scissors in a 'better than' ordering.

But the TT times ordering is only mildly nontransitive, it's mostly problematic for direct consecutive cars in an ordering. For example take @amarynceos times in his Premium Sport Coupes post:

Laguna Seca, SM Tyres

1:32.955 -- Chevrolet Corvette Stingray (C7) '14
1:33.540 -- Ferrari California '08
1:33.781 -- Amuse NISMO 380 RS Super Leggera
1:34.056 -- TVR Tamora ’02
1:34.111 -- SRT Viper GTS '02
1:34.138 -- Alfa Romeo 8C Competizione '08
1:34.188 -- Toyota FT-86 G SPORTS Concept '10
1:34.199 -- Aston Martin V12 Vantage '10
1:34.455 -- Honda NSX Type R '02
1:34.598 -- Lotus Esprit V8 '02
1:35.301 -- Jaguar XKR Coupe '10
1:35.343 -- Mazda RX-7 Spirit R Type A (FD) '02
1:35.663 -- Chevrolet Corvette Z06 (C5) '04
1:36.054 -- Nissan Fairlady Z (Z34) '08
1:36.579 -- Nissan SKYLINE GT-R V-Spec II Nur (R34) '02
1:36.672 -- Cadillac CTS-V Coupé '11
1:36.857 -- BMW Z4 M Coupe '08
1:36.846 -- Lotus Evora '09

Deep Forest, SM Tyres

1:21.268 -- Ferrari California '08
1:21.754 -- Toyota FT-86 G SPORTS Concept '10
1:21.867 -- Chevrolet Corvette Stingray (C7) '14
1:22.032 -- Aston Martin V12 Vantage '10
1:22.297 -- Honda NSX Type R '02
1:22.416 -- SRT Viper GTS '02
1:22.436 -- Alfa Romeo 8C Competizione '08
1:22.500 -- TVR Tamora ’02
1:22.801 -- Amuse NISMO 380 RS Super Leggera
1:23.091 -- Art Morrison Corvette '60
1:23.230 -- Chevrolet Corvette Z06 (C5) '04
1:23.664 -- Mazda RX-7 Spirit R Type A (FD) '02
1:23.895 -- Lotus Esprit V8 '02
1:23.951 -- Nissan SKYLINE GT-R V-Spec II Nur (R34) '02
1:24.058 -- Jaguar XKR Coupe '10
1:24.607 -- Nissan Fairlady Z (Z34) '08
1:25.308 -- Cadillac CTS-V Coupé '11
1:25.384 -- BMW Z4 M Coupé '08
1:25.821 -- Lotus Evora '09

If we bring it into this sensible order (first is slowest):

BMW Z4 M Coupé '08
Lotus Evora '09
Cadillac CTS-V Coupé '11
Nissan Fairlady Z (Z34) '08
Nissan SKYLINE GT-R V-Spec II Nur (R34) '02
Jaguar XKR Coupe '10
Mazda RX-7 Spirit R Type A (FD) '02
Chevrolet Corvette Z06 (C5) '04
Lotus Esprit V8 '02
Art Morrison Corvette '60
Honda NSX Type R '02
Alfa Romeo 8C Competizione '08
Amuse NISMO 380 RS Super Leggera
TVR Tamora ’02
Aston Martin V12 Vantage '10
SRT Viper GTS '02
Toyota FT-86 G SPORTS Concept '10
Chevrolet Corvette Stingray (C7) '14
Ferrari California '08

Then all bold consecutive car pairs have contradicting TT times in the TT data. For example the Z4 is faster than the Evora on Deep Forest but slower on Laguna Seca. @amarynceos is not to blame here, it's just illustrating the nature of these tests. There will always be a driver/track for whom the order is wrong, but then it will be only for consecutive cars.

We all agree that there is a 'big picture' ordering, the small discrepancies can be checked and weeded out by

• more tests
• deciding by good taste if a discrepancy is a car/track issue

For example the Luxury list has this 2ton monsters with top speed a lot higher than some other cars, but they are weak on a twisty track. If we have TT data for both types of tracks we will get discrepancies for some cars, so we have to do good judging for the final ordering.

 

[amarynceos]

ANYONE:
a track with longer straight parts and/or hills (not a tight track)
Nissan Skyline 2000GT-B (S54B) '67
Honda Civic 1500 3door CX '79
Alfa Romeo Spider 1600 Duetto '66
Marcos Mini Marcos GT '70
(COMFORT SOFT)

ANYONE:
ANY TRACK
Isuzu Bellett 1600 GT-R '69
Alfa Romeo Giulia Sprint Speciale '63
(COMFORT SOFT)

ANYONE:
a bigger track with longer straight parts (not a tight track)
Renault R5 Turbo '80
Mercedes-Benz 300 SL Coupé '54
(COMFORT SOFT)

ANYONE:
ANY TRACK
DMC DeLorean S2 '04
Dome Zero '78
Chevrolet Camaro IROC-Z Concept '88
(COMFORT SOFT)

I did these all on Spa, Offline TT.

Time -- Car -- Top Speed going into Les Combes and into the Bus Stop (mph)

3:19.587 -- Nissan Skyline 2000GT-B (S54B) '67 --105/104
3:21.889 -- Honda Civic 1500 3door CX '79 -- 97/99
3:19.835 -- Alfa Romeo Spider 1600 Duetto '66 -- 101/101
3:19.485 -- Marcos Mini Marcos GT '70 -- 100/101
3:16.678 -- Isuzu Bellett 1600 GT-R '69 -- 108/107
3:16.339 -- Alfa Romeo Giulia Sprint Speciale '63 -- 107/106

3:03.345 -- Renault R5 Turbo '80 -- 117/116
3:03.830 -- Mercedes-Benz 300 SL Coupé '54 -- 125/118

2:59.037 -- DMC DeLorean S2 '04 -- 126/121
2:56.982 -- Dome Zero '78 -- 130/124
2:59.293 -- Chevrolet Camaro IROC-Z Concept '88 -- 126/124

Note the 300SL, half a second slower than the R5 despite an 8 mph advantage down Kemmel.