General Discussion

Cars with limited top speed

OzzmanOzzman Member Posts: 1,257 ✭✭✭✭✭
edited August 23 in General Discussion
Just re-posting this from the Car corrections sub-forum, to get more attention. Also tagging our most enthusiastic staff members @Hutch_Tim and @Hutch_Katie.

I'd like to provide an example of this issue, by comparing two cars.
As you can see the 718 Cayman and the XE S has pretty similar stats. Similar 0-60, similar MRA, the main difference being that the Jag has it's top speed limited at 155MPH.
Now let's see some times, I happen to have both cars stock (Porsche vs Jaguar):
1/4 mile: 13:33 - 13:40
1/2 mile: 20:86 - 21:03
1 mile: 33:70 - 34:33
0-120MPH: 17:03 - 17:90
75-125MPH: 12:20 - 13:43
They are similar, the Porsche has a slight advantage everywhere, but nothing to see here. However:
0-150MPH
: 35:76 - 51:46
As the Jag is getting close to it's top speed, the acceleration curve gets too flat, and it takes ages for it to get to 150. And also, here's a source claiming it's 0-150MHP time to be 36:50, which seems much more realistic, as it's close to the Porsche.
There are many-many cars in TD with a limited top speed, so this issue is a big one, although I'm not sure how it would affect these cars' RQ, as the >120MPH range of cars is not that relevant on most tracks.
I tested some other limited cars as well for 0-150MPH time (Stuttgart 2), and though some don't seem to slow down so much as they get closer to the limit, many 3-400 HP cars seem to have trouble reaching their top speed. Feel free to post some times here if you have stock, limited top speed cars.

Comments

  • David_FookDavid_Fook Member Posts: 2,373 ✭✭✭✭✭
    Like this a lot, it’s always been an obvious bug but good work for getting some numbers behind it.

    Hutch will claim that they won’t be able to get accurate numbers for delimited top speeds to properly define the curves, but I don’t think that would be necessary. The top speed they program in just has to be an estimate to bump the curve up, it will still be limited to 155. Even poor guesses will end up with a lower error than the one detailed above
  • AndreasSimmerAndreasSimmer Member Posts: 829 ✭✭✭✭✭
    Oh boy, the next major error in the engine...
  • BoxtaSBoxtaS Member Posts: 1,492 ✭✭✭✭✭
    Great posts, @O__VER   @samagoichi ;

    really highlights this bizarre flaw in the way the game works 
  • Ivo_KamburovIvo_Kamburov Member Posts: 777 ✭✭✭✭
    Oh boy, the next major error in the engine...
    We've actually been pointing this out since the beginning 
  • AndreasSimmerAndreasSimmer Member Posts: 829 ✭✭✭✭✭
    Can we please have a comment on this?! Obvs this is a huge issue. @Hutch_Tim , @Hutch_Katie
  • mauro07mauro07 Member Posts: 2,344 ✭✭✭✭✭
    it's a shame.. ok limited top speed but the engine can reach a speed (150 mph) in less seconds in real life but in this game we have MRA!! What the Hell!!! :/
  • GT47LMGT47LM Member Posts: 372 ✭✭✭
    Didn't Hutch say they get all their data from Evo? If so, how come all the information is wrong? Or does Hutch decide to only implement imaginary stats they care about? 
  • O__VERO__VER Member Posts: 1,501 ✭✭✭✭✭
    GT47LM said:
    Didn't Hutch say they get all their data from Evo? If so, how come all the information is wrong? Or does Hutch decide to only implement imaginary stats they care about? 
    I think one of the Hutch guys mentioned that they expected the Evo data to be more accurate than it was. Think maybe a few regrets with that partnership.
  • 0171801718 Member Posts: 103 ✭✭✭
    GT47LM said:
    Didn't Hutch say they get all their data from Evo? If so, how come all the information is wrong? Or does Hutch decide to only implement imaginary stats they care about? 
    Well if they get a 0-60 and a 0-100 and a top speed but no 0-150 time, Hutch will model all cars according to those known values and that means limited cars just get a very 'flat slope' after the last known time for a given speed (e.g. 0-100)

    Now if Hutch changes the cars they need to find values for higher speeds (like 0-150) and also might need to change the model, otherwise post 150 the cars would rapidly fall off (maybe on test Bowl or a 1 mile race). another way would be to find an alternative model for cars known to be limited which would result in artificial race times again, however the error would likely be smaller.

    Also, if the cars change, they will all be improved. According to the system Hutch employs to determine RQ, those cars will most likely gain RQ. Because the RQ system ist relative, other cars might then lose RQ or even lose class (e.g. go from B to C). I'm not even sure I want that.




  • RuvlaRuvla Member Posts: 47
    edited August 29
    01718 said:
    Also, if the cars change, they will all be improved. According to the system Hutch employs to determine RQ, those cars will most likely gain RQ. Because the RQ system ist relative, other cars might then lose RQ or even lose class (e.g. go from B to C). I'm not even sure I want that.
    But that is the idea of the initial post. :)

    I guess the problem lies in a little bit different area: We all here are car enthusiasts (I guess that is the majority of the player base).

    We play this game, because we are interested how different cars will behave relatively in different conditions. We try to predict the result looking at car stats on the card and based on our car knowledge and experience.

    While the game uses it's engine (rules of which are hidden from us), and stat on the card is interpreted differently by the engine (that human would do). So it creates those dissonance in our expectations and the results.

    I see 2 possible solutions:
    - fix the engine to be close to real world (this definitely requires a lot of effort and is questionably doable)
    - make the engine transparent to the players, so we know how the game works
  • Hutch_KatieHutch_Katie Member, Administrator, Wrenchmen, Hutch Staff, Senior Wrenchmen Posts: 65 admin
    GT47LM said:
    Didn't Hutch say they get all their data from Evo? If so, how come all the information is wrong? Or does Hutch decide to only implement imaginary stats they care about? 
    We get the majority of car data form Evo, but occasionally some data comes from the manufacturer too.
  • 0171801718 Member Posts: 103 ✭✭✭
    @BoxtaS I think this part about the model is fairly well understood. However, if you would feed this very simple model with more data closer to the max speed, there would be less of a problem. So it is a weak model fed with insufficient data. You could fix either one and would get a somewhat better result or you could fix both to get a very realistic result.
  • AndreasSimmerAndreasSimmer Member Posts: 829 ✭✭✭✭✭
    Can we please have a comment on this?! Obvs this is a huge issue. @Hutch_Tim , @Hutch_Katie
    Thanks for successfully ignoring this whole topic dear devs.
  • BoxtaSBoxtaS Member Posts: 1,492 ✭✭✭✭✭
    01718 said:
    @BoxtaS I think this part about the model is fairly well understood. However, if you would feed this very simple model with more data closer to the max speed, there would be less of a problem. So it is a weak model fed with insufficient data. You could fix either one and would get a somewhat better result or you could fix both to get a very realistic result.
    Actually quite a good point - If the model used the theoretical (de-limited) top speed instead of 155 as the v-max, then acceleration between 120-150 would be as expected. Might be quite labour intensive to implement tho. 
  • HeissRodHeissRod Member Posts: 4,886 ✭✭✭✭✭
    I'd like to know where EVO got their "time to top speed" figures from.  As you said, that's not a commonly measured stat.
Sign In or Register to comment.