Gran Turismo 7 Custom Race thread

[Guffaluff]

Was planning to post this here, but I rambled too much and it became too long for a comment 😅

I set up a custom race championship using the VGT Gr.1 cars, and it starts off quite fun! Just made a post about it here if you want to follow along

VISION GT LMP SERIES

 

[A_Higher_Place]

Some group A DTM fun. I have seriously detuned this skyline and the AI simply just plod along. Even with boost set to weak, it's a bit of a snoozefest. Looks amazing though!

 

[Stefan Clift]

Currently adding some fictional but realistic 2016 Super GT liveries for the GT500 class. Feel free to use them, the idea is to bulk out a whole grid whilst making the whole thing believable.

Honda TAKATA MUGEN NSX CONCEPT-GT ‘16

gran-turismo.com

Gran Turismo Official Website

www.gran-turismo.com

Nissan MINE’S ADVAN GT-R ‘16

gran-turismo.com

Gran Turismo Official Website

www.gran-turismo.com

 

[05XR8]

Some group A DTM fun. I have seriously detuned this skyline and the AI simply just plod along. Even with boost set to weak, it's a bit of a snoozefest. Looks amazing though!

That’s weird. Now that you have the V8 Quattro maybe pop in a few R32 Nismos to carry the PC1/PC2 theme around the circuit.

Edit: I was trying to figure out why the cars are going so slow. I remembered, I think, all cars run on treaded tyres.

 

[A_Higher_Place]

That’s weird. Now that you have the V8 Quattro maybe pop in a few R32 Nismos to carry the PC1/PC2 theme around the circuit.

Edit: I was trying to figure out why the cars are going so slow. I remembered, I think, all cars run on treaded tyres.

Yeah, they're all pretty much stock but I've seriously detuned the 'Audi'. It has a power limiter set to 80 here and it still pulls time on them. The laps vary by 8 seconds the AI slows so much. Most of what you see if me feathering the throttle. If I go full throttle then I'd be in first by the end of lap 3.

 

[05XR8]

Yeah, they're all pretty much stock but I've seriously detuned the 'Audi'. It has a power limiter set to 80 here and it still pulls time on them. The laps vary by 8 seconds the AI slows so much. Most of what you see if me feathering the throttle. If I go full throttle then I'd be in first by the end of lap 3.

It does mess up the grid. When I sit on the side of the circuit for a spectating video at Brands Hatch, the AI are full of life. Racing them is a bit up and down with the programming behaviour. They'll take themselves out for fear they're holding me up.

 

[24HuresdeLoL]

Back trying to get a few more competitive races done for my British GT-Esque GT3/GT4 ‘GTWS Pre-Season’ series. Hoping to turn this into a league someday​

 

[KAlex122]

First time I'm trying custom races and i have some questions. For now it's only one:

- would "boost of" show me an objective performance of all AI cars? So for example i set up 20 cars and let the other 19 go on for 20 laps without intervening (all set to 520pp), could that show me which are good 520pp-setups and which are not performant? Or is AI performance not consistent enough throughout the 19 drivers?

 

[suomi1]

First time I'm trying custom races and i have some questions. For now it's only one:

- would "boost of" show me an objective performance of all AI cars? So for example i set up 20 cars and let the other 19 go on for 20 laps without intervening (all set to 520pp), could that show me which are good 520pp-setups and which are not performant? Or is AI performance not consistent enough throughout the 19 drivers?

You will never be able to objectively know what the real performance of cars are in the hands of an AI. Regardless of boost level.

 

[KAlex122]

Okay. So to determine the car's performances i should drive all 20 by myself first.

Any suggestions for a track that is medium speed and relatively easy to be consistent at? I want to take this grid to many different circuits in the future. Planing a championship with names and designs and so on. First thing i noticed is that @525pp the FF cars are the slowest (2 Civics, an Integra and a Renault RS Trophy in the Field).

I may have to get some cars above 525pp to have a compact field.

Any suggestions on how to approach a championship concerning performance differences? Reverse starting order, like winner (or championship leader?) start last on the next course?

 

[Batschisane52]

Okay. So to determine the car's performances i should drive all 20 by myself first.

Any suggestions for a track that is medium speed and relatively easy to be consistent at? I want to take this grid to many different circuits in the future. Planing a championship with names and designs and so on. First thing i noticed is that @525pp the FF cars are the slowest (2 Civics, an Integra and a Renault RS Trophy in the Field).

I may have to get some cars above 525pp to have a compact field.

Any suggestions on how to approach a championship concerning performance differences? Reverse starting order, like winner (or championship leader?) start last on the next course?

Reverse starting order worked very good for my championships...

And don't forget to restrict the power of the car you are using by 5-10% because there is this full-throttle glitch for the AI cars in custom races...

Put the boost on "weak", that usually yields the best results when it comes to AI-performance...

 

[05XR8]

Okay. So to determine the car's performances i should drive all 20 by myself first.

Any suggestions for a track that is medium speed and relatively easy to be consistent at? I want to take this grid to many different circuits in the future. Planing a championship with names and designs and so on. First thing i noticed is that @525pp the FF cars are the slowest (2 Civics, an Integra and a Renault RS Trophy in the Field).

I may have to get some cars above 525pp to have a compact field.

Any suggestions on how to approach a championship concerning performance differences? Reverse starting order, like winner (or championship leader?) start last on the next course?

The programming will sort out the performance of the cars. All aids are turned on so the Boost will just do what it does to keep the grid tight.
Now, the main things are…. everything else. AI programming is based on the cars used, the track layout, race distance and wear multipliers. You can have an excellent two lap race and an average one hour race. An awesome one hour race and a last to first victory in a few laps.
Many of us have different ways we set up cars and settings. Not to mention different skills to our own driving. It’s not a chore for me, but it’s pretty much constant testing.

With cars, it’s all testing. I have good races with certain cars and horrible experiences with others. I thought one grid works well at one track and tried a different grid of cars with failing results. One such grid was an Abarth 500 One Make like the real series. Cars had the Racing Hards and all set up nicely. The AI programming was soo slow, at every track I tested. Yet, when I built an all FF grid with mixed cars, the programming provided competitive racing.

Just ran two 10 lap test races with a GT3/GT4 grid at Lago and Sardegna. Took a long time to pass cars on these short laps. At Tsukuba, I’d pass cars much quicker. So far so good.