Gran Turismo 7 Custom Race thread

[Theufcveteran]

Ready to post my Sophy findings.

Boost Weak is the fastest apart from Strong but strong has considerable rubberbanding.

Weak will have rubberbanding ONLY if the car that is behind is way behind. I couldn't specify how many seconds (anyone have an Idea?). For this I was using a Gallardo vs Civic SiR (sophy car), both stock.

At tokyo, the Civic will reach the tunnel at 200km if driven by me, at the end of the straight, 218 km/h. With Sophy, after she is way behind, it can reach around 228 km/h at the end of the straight. So in boost weak it gets a 10 kilometer boost, ONLY if way behind.

When its in racing situations, i.e., has cars around it, even on boost weak It will be stock and have stock speed, doesn't matter if there's like 2 or 3 cars way ahead, If one is close to It, it will have regular speed. This means either my car or another sophy car.

So I tried the other way around, SOPHY with the Gallardo and me on the Civic with boost weak to see if anyone of us is affected: nope.

Sophy will Max out at 311km/h, like me when I tried the Gallardo. My Civic, even if way behind SOPHY, will not gain 10 kilometers per hour like she does, was maxing out at the same 218km/h as before.

So, If SOPHY on boost weak rubberbands ONLY if the next car is waaaay ahead, its a great compromise to have boost weak turned on. When in racing situations its running on regular speed. It does not rubberband.

If anyone can disagree and show I am wrong I am waiting to discuss this. So Far, boost weak sophy is great.

Thanks for sharing all your findings! This stuff is great for figuring out the ins and outs, what the best options are. It's good to know that suspicions about minimal rubberbanding on boost weak seem to be confirmed here.

Now I hope to god PD don't change it. It is possible that the single races previously available with Sophy were the same AI but with boost set to off.

With any luck either PD will only improve it (actually improve, not slow it down/add more rubberbanding).

And boost off needs to actually be no rubberbanding, but to be honest I don't have much faith on that.

I know PD probably want to balance things for the widest audience, if they don't feel they can change anything then perhaps they could hard lock boost to off if the difficulty is lowered? Assuming they can't genuinely slow it down on a lower difficulty.

I don't know how much they can tweak given the AI is an external agent, but clearly their hardcoded rubberbanding that appears with boost off can affect the AI.

 

[appu]

Your 1) is very intriguing
Does it really mean that a mixed field of gr3 and gr4 behave properly, no matter if the human driver is on the gr3 or gr4?

I have only tried in the slower class but it was flawless for me. The slower class has no slowdown, and Sophy is fantastic at managing traffic. Can't recommend giving it a go enough.

Same answer for me. I was in Gr3.

A few more points I noticed:
1. Sophy stays within real life track limits which results in me being way faster than Sophy at some turns which was annoying. But I like that it is realistic. The current track limits in GT7 is unrealistic. See the picture below:

This kind of riding on edge of the track limits from me is plain comical. Having two tires off the track never dirties them as long as the other two tires are technically on track. This wouldn't float in real life racing and in many other sim/simcade games. I have been retraining myself to respect real life track limits. This has cut down on my ridiculous gains over Sophy at many corners (which I prefer).
2. I can confirm that unfortunately both boost weak and off have rubber-banding, which is annoying.
3. Some of my favorite Sophy moments are close pack-racing at night. This is where the unfortunate headlight bug kicks in. It is pitch-black ahead because the light from my headlights have cast shadows on the headlight illumination area for the car in front of me. But the day-night cycles are still fantastic and I love it every time. I keep weather random so I love surprises like glorious sunset and foggy morning.
4. Multiclass racing unfortunately is not that fun because of the lowly 20 car limit.
5. I am continuing to figure out how to make a race where strategy plays a good role. The inability to turn off boost is screwing up with any potential to make strategies work. Any ideas are welcome on how to add diverse strategy while keeping BoP (I am lazy, you see).

 

[pritchyTim]

Same answer for me. I was in Gr3.

A few more points I noticed:
1. Sophy stays within real life track limits which results in me being way faster than Sophy at some turns which was annoying. But I like that it is realistic. The current track limits in GT7 is unrealistic. See the picture below:
View attachment 1440285
This kind of riding on edge of the track limits from me is plain comical. Having two tires off the track never dirties them as long as the other two tires are technically on track. This wouldn't float in real life racing and in many other sim/simcade games. I have been retraining myself to respect real life track limits. This has cut down on my ridiculous gains over Sophy at many corners (which I prefer).
2. I can confirm that unfortunately both boost weak and off have rubber-banding, which is annoying.
3. Some of my favorite Sophy moments are close pack-racing at night. This is where the unfortunate headlight bug kicks in. It is pitch-black ahead because the light from my headlights have cast shadows on the headlight illumination area for the car in front of me. But the day-night cycles are still fantastic and I love it every time. I keep weather random so I love surprises like glorious sunset and foggy morning.
4. Multiclass racing unfortunately is not that fun because of the lowly 20 car limit.
5. I am continuing to figure out how to make a race where strategy plays a good role. The inability to turn off boost is screwing up with any potential to make strategies work. Any ideas are welcome on how to add diverse strategy while keeping BoP (I am lazy, you see).

I think road cars (would require you BoP them yourself, but same PP or power/weight is probably good enough) with a mixture of default tires and the appropriate tire wear multiplier/race length would be a decent recipe for ‘strategy’ - this spreadsheet has each car’s default tire on the fifth tab.

 

[manou]

I'd love to watch some of the custom races you guys race in, would be entertaining and would free my mind off my daily life and stress.

And I would love to test some of your custom races. I suck at creating a good custom race.
If therewas a list of great custom races with sophy somewhere, that'd be a great + for gtplanet users.

 

[05XR8]

Probably could be a “GT7 Custom Race” forum like the Livery, Tuning and Photo forums. Create, search, share, discuss in a location easy to find player set ups.

 

[bliprunner]

Did a little test to see how Sophy 'Weak' rubberbanding behaves differently when you are P1 or P2.

5 lap races at Interlagos with two identical GT3 cars. Race one has an extra GT2 to keep hold of P1.

Race 1:
GT2 car races off and wins by 40 seconds. I start behind GT3 Sophy and despite being right behind don't manage a move that sticks until lap 2. Once in P2 I pull away gradually and beat them by 5 seconds. This felt pretty natural.

Race 2:
No GT2 car but Sophy GT3's pace feels the same at the start, I'm on their bumper and this time get the move done lap one. But now as the P1 car I can't pull more than a second away from Sophy, even though I'm lapping slightly quicker than race 1. This felt... less than natural, obviously.

I know these aren't findings that will shock anyone, but might be useful info (maybe for multiclass if you are in the slower class). Screenshots from the end of both races attached.

 

[pritchyTim]

Did a little test to see how Sophy 'Weak' rubberbanding behaves differently when you are P1 or P2.

5 lap races at Interlagos with two identical GT3 cars. Race one has an extra GT2 to keep hold of P1.

Race 1:
GT2 car races off and wins by 40 seconds. I start behind GT3 Sophy and despite being right behind don't manage a move that sticks until lap 2. Once in P2 I pull away gradually and beat them by 5 seconds. This felt pretty natural.

Race 2:
No GT2 car but Sophy GT3's pace feels the same at the start, I'm on their bumper and this time get the move done lap one. But now as the P1 car I can't pull more than a second away from Sophy, even though I'm lapping slightly quicker than race 1. This felt... less than natural, obviously.

I know these aren't findings that will shock anyone, but might be useful info (maybe for multiclass if you are in the slower class). Screenshots from the end of both races attached.

Love any and all tests in here! Thank you for sharing. It’s funny, it makes it seem like we went from avoiding multiclass to ENSURING it! I know some others (suomi?) have found that having a runaway lead rabbit can be useful for ‘regulating’ how the rest of the grid behaves. I have not tested that much, myself.

—

As a suggestion to someone feeling testier than I, I’m curious how number of opponents might affect the AI performance. I’ve heard a lot of “Sophy can’t do this/that” since 2.0, referring to its scope and relating to its inability to scale due to resource demand. Is it BETTER when we don’t stretch it across 19 competitors? I kind of doubt it, but think it’s a worthy topic to explore, if only briefly.

—

Did some Intrgrale series last night and the bouncing between Sophy and Reggy (thanks, 05XR8… do we like “-ie” or “-y” better?) is AWESOME! Funnily, I got 1st on two Sophy races and then third at Fuji in GT-Rs. The added shuffling of having different ‘drivers’ behind the cars on my custom grids across a several-race series is really welcome, as even in a one-make series I saw a lot of different finishing orders. There’s a slight hurdle in needing to save your Custom Race twice, but it’s hardly added time going from event to event once you jump that hurdle once.

 

[KAlex122]

I just realised it would be fun to let Sophy race against Reggie. How much of a advantage can Sophy gibe Reggie and still win? How annoyed will Sophy be by Reggie's corner entrance brake checks?

 

[suomi1]

removed

 

[suomi1]

Love any and all tests in here! Thank you for sharing. It’s funny, it makes it seem like we went from avoiding multiclass to ENSURING it! I know some others (suomi?) have found that having a runaway lead rabbit can be useful for ‘regulating’ how the rest of the grid behaves. I have not tested that much, myself.

[...]

In a rather funny way, as I was in a pain against Sophy on Road Atlanta / Boost WEAK even on the lower difficulty, I found (I think) a way to silence rubberbanding with boost STRONG. In a 3 categories multiclass racing.

Whatever the category I was racing in, AI kept the same pace every time in each class. I tried both of them.
Or the way categories are spread in term of performances is the key, or the fact that one car (namely, the C4 Corvette) was by far dead last because Sophy wasn't able to drive it correctly saved the day.

 

[05XR8]

In a rather funny way, as I was in a pain against Sophy on Road Atlanta / Boost WEAK even on the lower difficulty, I found (I think) a way to silence rubberbanding with boost STRONG. In a 3 categories multiclass racing.

Whatever the category I was racing in, AI kept the same pace every time in each class. I tried both of them.
Or the way categories are spread in term of performances is the key, or the fact that one car (namely, the C4 Corvette) was by far dead last because Sophy wasn't able to drive it correctly saved the day.

I found that before we got SOPHY, the better the stability, the easier AI can drive the car, the faster it’ll drive. Just fitting RH to the car isn’t enough. Adding fully tuned suspension isn’t enough. Once the brake package is added, Reggie would be able to adjust the car better over a lap. Reason I noticed was having a car that got lapped in one of my races.

I thought the car had been involved in collision early in the race and being damage was Heavy, took to the pits and just couldn’t catch the rest of the field. When I looked at the replay with display on, the car just out raked itself into the barrier.
I went back to the garage to check the setting. Maybe I forgot to tune the suspension on that duplicate. Checked the parts and it had everything the other cars had, but not the Slotted racing rotors(I don’t use race pads- found there’s no difference in AI braking). Once I equipped those, that car was on pace with the pack as normal.

Reggy (thanks, 05XR8… do we like “-ie” or “-y” better?)

I’m used to “ie” like Reggie Jackson or Eddie Murphy. Some prefer the y. One more letter or one less letter. All good.

 

[TurismoValor]

I have a request: After about 3 weeks of playing I'm about to complete menu 39, I'm CL ~35 and I've got about half of the CE completed at gold. What's a good way to use custom races to grind money? Of course, it's easy enough to find the headline game breaking races on YouTube but they always get patched it seems. What can/should I be doing with custom races to make a pretty penny, even if its not a min/max setup?

 

[pritchyTim]

I have a request: After about 3 weeks of playing I'm about to complete menu 39, I'm CL ~35 and I've got about half of the CE completed at gold. What's a good way to use custom races to grind money? Of course, it's easy enough to find the headline game breaking races on YouTube but they always get patched it seems. What can/should I be doing with custom races to make a pretty penny, even if its not a min/max setup?

1. the 4 money maker events seem pretty stable at this point, and are the best way to make money if you’re driving the car

2. custom races offer a more fun way to make less money per event, but for many of us the scenic route is preferred to the expressway

3. you can set up a (less profitable than #1) race that lets you AFK and make money, by tuning an otherwise-bad car (have you rolled any good engine swaps yet? you can buy one at lvl50) to beat a custom grid filled with otherwise-good cars (tomahawk w a crappy tune) while it wallrides around Daytona. If you go looking into this method, be aware most of the videos/posts are from a brief pre-patch period where it was more profitable than it is now.

 

[TurismoValor]

1. the 4 money maker events seem pretty stable at this point, and are the best way to make money if you’re driving the car

2. custom races offer a more fun way to make less money per event, but for many of us the scenic route is preferred to the expressway

3. you can set up a (less profitable than #1) race that lets you AFK and make money, by tuning an otherwise-bad car (have you rolled any good engine swaps yet? you can buy one at lvl50) to beat a custom grid filled with otherwise-good cars (tomahawk w a crappy tune) while it wallrides around Daytona. If you go looking into this method, be aware most of the videos/posts are from a brief pre-patch period where it was more profitable than it is now.

I hate to ask, but what are those 4 money maker events? I see people reference some things a lot, but never so clearly that I understand what to do.

 

[pritchyTim]

I hate to ask, but what are those 4 money maker events? I see people reference some things a lot, but never so clearly that I understand what to do.

This is probably not the best forum for it, check for these in your World Circuits map:

Sardegna, 800pp, 15 laps / 25min
La Sarthe, 700pp, 30min
Tokyo, 600pp, 12 laps / 25min
Spa, 800pp, 60min

They unlock after a certain amount of Café progression - you can surely find more details elsewhere!

 

[2cam16]

I hate to ask, but what are those 4 money maker events? I see people reference some things a lot, but never so clearly that I understand what to do.

One of them is WTC800 at Sardegna. That's my go-to.

 

[05XR8]

@Jordan @Famine would it be worth making a GT7 offline Forum the same as the GT7 Online Forum and Sub-forums? The activity in this thread is ramping up(from what I'm seeing, unless gtplanet staff have different data) and would be helpful to current members, future members and gtplanet visitors to search and share Custom Race set ups and information.
At the least similar to the photo and livery threads where players can select from ideas and get inspiration in creating their own.

 

[Dayzura]

This is my take on the IMSA MX-5 Cup series, based loosely on their car specs.

Car: Mazda Roadster NR-A (ND) '22
Engine capacity : 181hp
Minimum weight : 2474lbs (this is as close as GT7 allows to the real world 2475lbs)
Minimum ride height : 102mm
Tires : Comfort Soft (as limited by what Sophy will use)

Each car costs ~$158,750 to buy and build, plus any expense to apply desired Style to each car (user VIKINGMIKE13 has a great collection of 2025 Whelen MX-5 Styles in Showcase, as shown in build video). After applying a base setup to all cars (as per build video) I adjust each AI car individually, with small changes to things like anti roll, camber, toe to create some slight variance between car performance - some cars also have a very small ballast increase or power reduction. On track performance (lap time) is less than real world, mostly due to the tires. They should really be on Sports, but Sophy uses the MX-5's default Comfort Soft, which are actually quite fun to race on (heat is more of an issue than wear). For reference, my Driver rating is AS and these guys are giving me some decently challenging racing.

Build and Base Setup


MX-5 Cup (GT Sophy Custom Race) at Road Atlanta - Onboard (and race config)


MX-5 Cup (GT Sophy Custom Race) at Road Atlanta - Replay

 

[Vitesh Patel]

Apologies if I'm posting old info to the regulars to this thread. Just looking to share my GT Sophy findings after some testing.

I tend to run with Boost Off and Slipstream set to real. I'm looking to turn the rubber banding off, so I thought these would be the best choices.

Today, I ran a 40 lap event at Monza in the F3500. All the Sophy agents decided to pit twice for fuel, where as one stop would've done the trick like I did. Their second stop came with 1 lap to go.

I love that they race mostly cleanly and put up a decent challenge. However, the rubber banding is still woeful. I made my one and only stop at about 75% distance, and came out P11. I was able to retake 1st on my outlap, such was the pace drop off of all the Sophy agents. Then while holding the lead, a driver in 5th posted a laptime 3 seconds faster than my best lap.

It's so disappointing after being excited that I'd be able to run the endurance events that PD never brought to the game. But it seems it needs a lot more tweaking yet.

Yesterday I ran the Super Formula around Lago Maggiore and they actually undercut me by pitting earlier for fresh tyres. So that was awesome. I guess if you try to dictate the strategy, that's when the proverbial wheels fall off...

 

[Dayzura]

Apologies if I'm posting old info to the regulars to this thread. Just looking to share my GT Sophy findings after some testing.

I tend to run with Boost Off and Slipstream set to real. I'm looking to turn the rubber banding off, so I thought these would be the best choices.

Today, I ran a 40 lap event at Monza in the F3500. All the Sophy agents decided to pit twice for fuel, where as one stop would've done the trick like I did. Their second stop came with 1 lap to go.

I love that they race mostly cleanly and put up a decent challenge. However, the rubber banding is still woeful. I made my one and only stop at about 75% distance, and came out P11. I was able to retake 1st on my outlap, such was the pace drop off of all the Sophy agents. Then while holding the lead, a driver in 5th posted a laptime 3 seconds faster than my best lap.

It's so disappointing after being excited that I'd be able to run the endurance events that PD never brought to the game. But it seems it needs a lot more tweaking yet.

Yesterday I ran the Super Formula around Lago Maggiore and they actually undercut me by pitting earlier for fresh tyres. So that was awesome. I guess if you try to dictate the strategy, that's when the proverbial wheels fall off...

Set Boost to weak, it makes a world of difference.

 

[Hasnain282]

Tried my own multiclass race with Sophy and it did pretty well for the most part, it nearly botched it in the high speed S but it didn't crash or anything (I'm the driver behind in the Peugeot)
I was using the early 2010s LMP1s, unfortunately they didn't turn out that balanced, the Toyota was quickest by far, the Peugeot and Audi had similar pace but as the Audi needs to be short shifted and Sophy doesn't do that, it ended up using way more fuel than the others

 

[2cam16]

Been playing around with Sophy on different tracks. All Miata TC spec race. Just little 5-6 lap sprints. Had to turn off boost as I was not getting anywhere. Pretty fun to battle the AI. Like others have said, no sudden braking by the AI in the middle of a turn is so much more fun. One of these days I'll have to try the multi-class races like you all.

 

[Dayzura]

To follow on from my MX-5 Cup post, I've been impressed with how the well the Sophy cars are doing with battling each other (at least with my config). If you watch further up the road, or in the rear view mirror, you can regularly see them two wide, sometimes 3, looking for a pass or defending - and it's all totally separate to racing me. Really adds to the immersion and is the best single-player racing I've had in any Gran Turismo game - it's saved GT7 for me. The Dragon Trail Seaside race below was particularly fun.

I really think the key is buying the full grid of cars, applying a base setup to them all, then slightly modifying each cars setup, and adding some variety with small ballast adjustments and power restrictions - think of each car setup as a performance file for individual cars/drivers. Then randomize your grid and enjoy! I think I am going to create myself a little MX-5 Cup championship, with 45 minute races, and get ChatGPT to simulate a qualifying order for each round. If they would just add qualifying to custom races, like we have for lobby racing, I would be completely happy!

Brands Hatch


Dragon Trail - Seaside

 

[Aphelion]

To follow on from my MX-5 Cup post, I've been impressed with how the well the Sophy cars are doing with battling each other (at least with my config). If you watch further up the road, or in the rear view mirror, you can regularly see them two wide, sometimes 3, looking for a pass or defending - and it's all totally separate to racing me. Really adds to the immersion and is the best single-player racing I've had in any Gran Turismo game - it's saved GT7 for me. The Dragon Trail Seaside race below was particularly fun.

I really think the key is buying the full grid of cars, applying a base setup to them all, then slightly modifying each cars setup, and adding some variety with small ballast adjustments and power restrictions - think of each car setup as a performance file for individual cars/drivers. Then randomize your grid and enjoy! I think I am going to create myself a little MX-5 Cup championship, with 45 minute races, and get ChatGPT to simulate a qualifying order for each round. If they would just add qualifying to custom races, like we have for lobby racing, I would be completely happy!

Brands Hatch


Dragon Trail - Seaside

Granted, this takes an awful lot of effort, but the fact that this is possible in GT right now is something to behold. So cool!

 

[Goldin]

I am really having fun with Sophy custom races, but as I am mostly doing some races with road cars, I would love that they add tyre selection beside stock or racing hard.
Mostly driving road cars on stock compound or one better, CS, RH or max.RM and when using some older cars that would fit as opponent with power/weight they finish handiccaped because of stock tyre usually being CM (like Civic '84), I really hope that they would add that option for sophy cars in next updates.

 

[Cleveland8]

Thanks for sharing all the Sophy testing. Has anyone tested the variances between Pro and Intermediate on the custom race setting? I began doing this as I was having a tough time making progress in positions on Gr3 pro. I did some 5 lap races on Tokyo at both Pro and Intermediate and found that lap times for both were about the same with Sophy fastest at sub 2:03. My fastest was 2:04.3. I then jumped into the Quick race and ran beginner and intermediate using same gr3 bop. They were very different - I won on beginner by 7.9 secs and won by .224 on intermediate. In custom race, has anyone done testing for the difference between Intermediate and Pro?

 

[pritchyTim]

Lucked into a neat spectator angle at Interlagos!


Clearly, a couple of the cars need work. Shout if you can tune the '92 NSX!