Gran Tursimo Brain Droppings: Random Thoughts, Theories, Ramblings and Inquiries

  • Thread starter Mafio
  • 30 comments
  • 2,829 views
327
United States
United States
Mafiodaproducer
This is a thread for all the weird, goofy and/or random thoughts that come across the mind while playing GT7.

THIS IS NOT A GRIEVANCE OR COMPLAINT THREAD
There are PLENTY of other threads for that.

I recently realized that the driver changes size according to the car their driving. This lead me to the conclusion that we don't control one driver, but any given driver at any given track at any given time. Its like the movie Ghost. We see our avatar, Patrick Swayze...but it's really Whoopi Goldberg.


The infamous Castrol Celica GT-4 Rally Car. The game has it rated at ~300 hp....

Fry GIFs - Find & Share on GIPHY


....is it REALLY tho?? Are they doing the thing????? Has anyone tested it out? Does GT account for manufacturer sandbagging or do the cars just perform at whatever power the manufacturer says the car does?

I maybe misremembering this, but in the old PS1 days, one of the draws into GT was the fact that you could race with "regular" cars. Looking about at it now, it was really just sports cars. I think ours dumb, childish brains didn't qualify a Nissan 240sx as a "sports cars". They were just "litte Japanese cars".
 
Driver's head position in PSVR is hilarious. I have tons of headroom in the Audi R18D, feel like I want to crank the seat up an inch or two.

In others, especially old classics, I pop into VR view and get a steering wheel in my face, like I'm an old man (which I am) who has forgotten his glasses (which I do) and is peering through the wheel trying to read the instruments (only done this in GT7 so far)

I believe I saw somewhere that the driving position can be adjusted by popping out of VR, adjusting the view in Cockpit View, then going back to VR. Still pretty funny though. I never knew a Countach was so spacious!

I also feel like the cars are about 0.9 scale. They all feel just a little smaller and narrower than I'd expect.

None of that stops me thinking that VR is the best thing to happen to computer driving games since the DFP. I love it! And the weird quirks that PD always manage to drop into their game give Gran Turismo a unique charm.

Yesterday I pulled in to the Sardegna pit flat out, behind two AI cars who had at least made some effort to slow down to the pit limit speed. They both vanished just before I got there, of course, which was a good thing because otherwise I'd have piled into the back of them at about 3 times their speed.

But then, it cuts to a scene from inside the garage, looking out at the pitlane and a pit crew waiting for their car to arrive, and instead of watching a car pull into the pit, my mind imagined three cars tumbling down the pitlane, past the crew, in a huge fireball.

If I had the skillz I'd make a video clip of that..
 
The infamous Castrol Celica GT-4 Rally Car. The game has it rated at ~300 hp....

Fry GIFs - Find & Share on GIPHY


....is it REALLY tho?? Are they doing the thing????? Has anyone tested it out? Does GT account for manufacturer sandbagging or do the cars just perform at whatever power the manufacturer says the car does?

Thinking on this some more, it would be funny if it is saying 300 but really has 350, performs like it has 350 and it wins every race you enter it in, except every time you race it you get disqualified. Also, if you replace the turbo with an "upgrade" turbo, the power inexplicably drops, but all you notice is that the car got slower.

It then gets reported as a glitch.
 
Last edited:
Driver's head position in PSVR is hilarious. I have tons of headroom in the Audi R18D, feel like I want to crank the seat up an inch or two.

In others, especially old classics, I pop into VR view and get a steering wheel in my face, like I'm an old man (which I am) who has forgotten his glasses (which I do) and is peering through the wheel trying to read the instruments (only done this in GT7 so far)

I believe I saw somewhere that the driving position can be adjusted by popping out of VR, adjusting the view in Cockpit View, then going back to VR. Still pretty funny though. I never knew a Countach was so spacious!

I also feel like the cars are about 0.9 scale. They all feel just a little smaller and narrower than I'd expect.

None of that stops me thinking that VR is the best thing to happen to computer driving games since the DFP. I love it! And the weird quirks that PD always manage to drop into their game give Gran Turismo a unique charm.

Yesterday I pulled in to the Sardegna pit flat out, behind two AI cars who had at least made some effort to slow down to the pit limit speed. They both vanished just before I got there, of course, which was a good thing because otherwise I'd have piled into the back of them at about 3 times their speed.

But then, it cuts to a scene from inside the garage, looking out at the pitlane and a pit crew waiting for their car to arrive, and instead of watching a car pull into the pit, my mind imagined three cars tumbling down the pitlane, past the crew, in a huge fireball.

If I had the skillz I'd make a video clip of that..
You can solve the problem of the VR view being too close to the steering wheel by looking slightly downwards before entering the VR view. If you are using a real steering wheel, I suggest looking just below the real steering wheel.
 
You guys think Kaz used to street race?

Im a RELIGIOUS user of Praiano’s tunes, but sometimes I wonder if having the rear of the car sitting up so high really necessary or does he just use it as a calling card of sorts? So when he’s racing online and sees a car pass by with its *** all in the air he’s like…

IMG_8609.gif
 
Im a RELIGIOUS user of Praiano’s tunes, but sometimes I wonder if having the rear of the car sitting up so high really necessary or does he just use it as a calling card of sorts?
I believe it's a game-physics exploit. Someone mentioned on the GT Sport board that the game applies down-force perpendicular to the car floor, so by increasing the rake you also increase the horizontal component of the down-force vector pointing backwards, giving additional forward thrust.

You could try asking @praiano63.
 
They both vanished just before I got there, of course, which was a good thing because otherwise I'd have piled into the back of them at about 3 times their speed.
I did that yesterday, except I got there before they disappeared. Both of us went flying, and then it cut to that ever so peaceful cut scene...
 
This issue has been solved "long ago" in GTT7 to my knowledge, alongside putting cars all high because it was used to reduce PP much more than it was affecting handling while at the same time increasing top speed.
You are correct in that it's no longer an exploit.
The high rear is to force weight onto the front wheels to overcome the understeer due to the stiffness of the suspension settings chosen in order to reduce weight transfer, it's just a fast way to get a lot of safe tunes out instead of understanding how the car behaves and tuning the suspension settings for the car's characteristics.
if this approach doesn't work then the car get's labelled as overpowered or horrible or in need of a chassis swap... which I find comical
 
I finally watch the movie yesterday. Perfect choice for Father's Day movie actually. Me during the police chase

Hands On Hips Sassy GIF - Hands On Hips Sassy Posing - Discover & Share GIFs
 
Sometimes I keep theorizing the Paddock should feature "open layouts" for all the three Tokyo Expressway tracks. Removing the barriers to merge the inner and outer loops of each track would give a nice illusion of a free roam mode, since they run mostly side by side.

  • Tokyo Expressway - Central Open Layout
  • Tokyo Expressway - East Open Layout
  • Tokyo Expressway - South Open Layout
 
... PSVR ...

I also feel like the cars are about 0.9 scale. They all feel just a little smaller and narrower than I'd expect.

...
I had the same feel as well (in PSVR1). Not that I ever was inside a 996 GT3, but it seemed too "small".
 
Ok here is a thought to ponder....
Do you think it would be possible / feasible to include a safety car or full course yellow to the online racing as a way to control the dirty driving that occurs?
 
Ok here is a thought to ponder....
Do you think it would be possible / feasible to include a safety car or full course yellow to the online racing as a way to control the dirty driving that occurs?
Wouldnt that just make it even easier for the dirty driver to pull bad moves as the field stacks up again?
 
You are correct in that it's no longer an exploit.
The high rear is to force weight onto the front wheels to overcome the understeer due to the stiffness of the suspension settings chosen in order to reduce weight transfer, it's just a fast way to get a lot of safe tunes out instead of understanding how the car behaves and tuning the suspension settings for the car's characteristics.
if this approach doesn't work then the car get's labelled as overpowered or horrible or in need of a chassis swap... which I find comical
In that case, can you recommend some guys in the tuning threads with cars that don't show their behinds that much? I never go to the tuning section of the forum these days. I only apply suspension numbers close to what i saw from praiano years ago (it's typically upping roll bar and natural frequency an downing toe and camber a bit) without touching the height of it.
 
In that case, can you recommend some guys in the tuning threads with cars that don't show their behinds that much? I never go to the tuning section of the forum these days. I only apply suspension numbers close to what i saw from praiano years ago (it's typically upping roll bar and natural frequency an downing toe and camber a bit) without touching the height of it.

I think that gets it covered
 

I think that gets it covered
Thanks, i already favourited that post long ago.

I knew you would assume i do feel ANYTHING helpful while driving. :D
 
Thanks, i already favourited that post long ago.

I knew you would assume i do feel ANYTHING helpful while driving. :D
If there is no "feeling wrong" then the car seems to be in an alright position, no? ;)

If you think it can be improved, then
1. go lower until the car isnt stable anymore OR has the rubbing issues.
2. if rubbing, up again. if unstable refer to the thread to fix it.
3. if lowered as much as possible and stable, try reducing toe in and out
4. it still stable try reducing camber angle
5. if still stable congrats, the only thing left is finding the gears to match the track which then is either a) sprint race with no regard to fuel or b) eco race IF you only run short of fuel over the whole event for about a max of 1 lap, this can propably be fixed, otherwise back to a) and sprint + pit
 
Ok, maybe a low powered car isn't the right thing to learn tuning, i don't know.

I bought a Toyota GR86 '15 with LSD and racing suspension. First thing: i didn't even notice a real difference between standard suspension and racing suspension. When i tuned the suspension to my kind "praiano settings" i also didn't notice much. When i set the LSD to extreme highs and lows a also didn't feel much.

Yes, with some setting i felt a bit more power oversteer mid-to-end-corner and yes, times at Tsukuba varied but I'm not constant enough to say which one is better/faster.

So my upper post stays: I don't feel it and I'm useless at tuning myself.

I'm was a DR:B driver when i did sports mode 2 years ago, my record at the online Time Trials this year is 15 gold, 8 silver, no bronze. So I'm not the best driver but also not that inferior.

But i will stay doing what I'm doing: Go with standard suspension when cars are weak and using "about praianos known direction" when i want to tune suspension and LSD.

I can't go full praiano because mostly i don't like widebodys and big, custom spoilers in the rear and also the rake looks terrible mostly.

i tried tuning more than once. It's enough.
 
In that case, can you recommend some guys in the tuning threads with cars that don't show their behinds that much? I never go to the tuning section of the forum these days. I only apply suspension numbers close to what i saw from praiano years ago (it's typically upping roll bar and natural frequency an downing toe and camber a bit) without touching the height of it.
Praiano will have a tune for pretty any car thats not a FF and it will be safe to use but not the best/fastest tune,
I have some tunes dotted around,
There are some good tuners out and about in Discord servers to be honest.
 
Ok, maybe a low powered car isn't the right thing to learn tuning, i don't know.

[...]

i tried tuning more than once. It's enough.
If you can't get a good idea of it, you can try some form of A/B/etc and using some form of statistical method to back it up. I'm thinking possibly a Kruskal-Wallis test (don't worry, I need to review this, too).

I remember using it for a homework assignment a few years ago and realized that it would be really useful for what I am trying to do now, in absence of a test driver.

Essentially, you can collect your lap times and rank them...
Screenshot from 2024-06-27 14-17-33.png

Then do multiple comparisons between the sets...
Screenshot from 2024-06-27 14-21-54.png

Do a bunch of math and come up with conclusions based on the results...
Screenshot from 2024-06-27 14-41-54.png


The problems are always going to be the 1) sample size of each category, 2) the amount of hours it would take to obtain and enter all the lap time data and 3) variance caused by the skill level of the driver. Technically, you can just take the lap times from each category, create an average of each sample and compare them that way, but the problem is that you can't verify the conclusions at all.

Also going to take some time out to throw this at @Tidgney since it may be relevant for a certain series of his, even if it's just food for thought. I know that optimals were used to help deal with variance and save time.

Maybe someone in the future someone could create some sort of analysis tool that uses the game's API to collect the laptime data during the session and rank the cars/setups. A machine learning model could also be used to predict an optimal setup given a large enough dataset, but that is way beyond my paygrade at the moment. Some other person's project, ha.
 
If you can't get a good idea of it, you can try some form of A/B/etc and using some form of statistical method to back it up. I'm thinking possibly a Kruskal-Wallis test (don't worry, I need to review this, too).

I remember using it for a homework assignment a few years ago and realized that it would be really useful for what I am trying to do now, in absence of a test driver.

Essentially, you can collect your lap times and rank them...
View attachment 1368158
Then do multiple comparisons between the sets...
View attachment 1368160
Do a bunch of math and come up with conclusions based on the results...
View attachment 1368171

The problems are always going to be the 1) sample size of each category, 2) the amount of hours it would take to obtain and enter all the lap time data and 3) variance caused by the skill level of the driver. Technically, you can just take the lap times from each category, create an average of each sample and compare them that way, but the problem is that you can't verify the conclusions at all.

Also going to take some time out to throw this at @Tidgney since it may be relevant for a certain series of his, even if it's just food for thought. I know that optimals were used to help deal with variance and save time.

Maybe someone in the future someone could create some sort of analysis tool that uses the game's API to collect the laptime data during the session and rank the cars/setups. A machine learning model could also be used to predict an optimal setup given a large enough dataset, but that is way beyond my paygrade at the moment. Some other person's project, ha.
It's more of a time issue (Which is why I use optimum), each car profile test takes 1 hour 30 minutes with everything I test so if I ended up doing multiple tests on the same track (Which I think this is? Apologies if I've mis read) I'd be there all day. I do suspect Sophy will have some impact on BoP testing in the future though.
 
It's more of a time issue (Which is why I use optimum), each car profile test takes 1 hour 30 minutes with everything I test so if I ended up doing multiple tests on the same track (Which I think this is? Apologies if I've mis read) I'd be there all day. I do suspect Sophy will have some impact on BoP testing in the future though.
I spent a bit more time looking into ways to implement it and it seems very simple. There's a Python library built to handle statistics (SciPy), that can automate the Kruskal-Wallis test. Just need to feed it data (repeatedly). I think you have most of the data stored somewhere already, and the test sessions wouldn't need to get any longer.

It could be outside of the scope of what you want to do with that series, but it could be useful to anyone who needs to constantly test setups or cars and can't easily determine what is consistently faster or needs objectivity when ranking them.

Where it might get interesting is if you have, say, 3-5 people per DR rank collecting data. You could probably sort out the fastest picks really fast, and it would be a nice, inclusive team effort.

There's just so many cars to get through, though.
 
Last edited:
After month of waiting I finally got another Ferrari invite. The FXX-K is the only Ferrari I don't have yet. I can finally buy it!🙂 The only problem is that I just can't decide what color to buy. Thinking about this for 3 days now😂
 
Back