Quality over Quantity (GT7)

[Griffith500]

It was the Toyota I noticed was missing.

No, the real number is around 40 and the same in practice. The modeling effort that went into the standard to semi-premium conversion doesn't add any new premiums into the game and that's what I'm talking about, not a measure of "total modeling effort". And since we can't quantify it in any way because we don't know how long it takes to do a conversion it's meaningless. It might only take a week to brush up a car, meaning 26 conversions = 1 premium car, almost inconsequential. Even the VGT's I wouldn't count as a full car because they have no interior. They are pretty on the outside but without a cockpit they are just semi-premiums and not up to the full "premium" status of other cars or other sims.

It's important to know in case the strategy shifts. Big picture.

How many more Premiums could they have made if they didn't make the semi Premiums? How many fewer will we get if they convert all Standards to semi Premiums? That's what I'm interested in.

Remember you claimed that the rate of production has stayed the same, but this is despite the team growing dramatically (and the semi Premiums appearing) - plus your estimates double PD's once ridiculed claim of six months per car. How does all of that fit in?

I find it amusing that you state that you can't possibly know how many of the various retouching tasks take place and what they're worth, when their relative proportion may change from one game to the next. This is no different from my picking a duplication rate of, say, 25%, and applying that to the inflated car list (by dividing by 1.25) - only there's no pretence of accounting for uniqueness, just a general trend in the split of visible effort that contributes to that uniqueness.

I'm just trying to figure out what PD are up to, what possible approaches PD could take, and what the likely outcome would be in each case. In stating you don't know how much the smaller tasks are worth, how can you state with any authority how many of the larger ones will get through, without making assumptions?

 

[Johnnypenso]

I don't know how long it takes to make a livery and make it work on the PS3 but I do know that the Mazda 787B was released on Assetto Corsa as a mod 2 days ago and already one player alone has created and made available for download, 3 liveries for the car:

Maybe it's harder for the Playstation?

 

[05XR8]

Man, I'm wondering for GT7, if we'll have about 800 Premiums ready to use. Then, each month PD start FREE monthly DLC packs.

 

[SSQES]

Man, I'm wondering for GT7, if we'll have about 800 Premiums ready to use. Then, each month PD start FREE monthly DLC packs.

I am expecting 600 premiums at launch, with like 1-2 new cars each months (like GT6)

 

[05XR8]

I am expecting 600 premiums at launch, with like 1-2 new cars each months (like GT6)

Do you feel the DLC will be cars we already have that are being released as full premium, new cars to the series, "new cars to GT7"(like the new to GT6 cars) or mixed with new cars(2014, 2015, 2016) & cars from GT6?

 

[SSQES]

Do you feel the DLC will be cars we already have that are being released as full premium, new cars to the series, "new cars to GT7"(like the new to GT6 cars) or mixed with new cars(2014, 2015, 2016) & cars from GT6?

Probably like GT6 : New cars to the serie. I think the cars from GT6 will be in GT7 at launch.
@Johnnypenso : I don't know this guy, but it's an impressive work.

 

[Griffith500]

I don't know how long it takes to make a livery and make it work on the PS3 but I do know that the Mazda 787B was released on Assetto Corsa as a mod 2 days ago and already one player alone has created and made available for download, 3 liveries for the car:

Maybe it's harder for the Playstation?

Because that's the only thing I mentioned.

 

[FiftyBelowZero]

you're having us vote for 400 premiums when we already have more than 400 premiums

 

[Johnnypenso]

Probably like GT6 : New cars to the serie. I think the cars from GT6 will be in GT7 at launch.
@Johnnypenso : I don't know this guy, but it's an impressive work.

There's quite a lot of impressive work going on in the AC camp. You should see some of the stuff I have! I was wondering though, if a guy working alone on a PC without professional resources presumably, can do this kind of thing in 2 days, how long does it take to create liveries on the PS3? When they release 3 Nascar cars, but with 11 liveries and add 11 cars to the car count, how long did it take to make those liveries?

 

[SSQES]

There's quite a lot of impressive work going on in the AC camp. You should see some of the stuff I have! I was wondering though, if a guy working alone on a PC without professional resources presumably, can do this kind of thing in 2 days, how long does it take to create liveries on the PS3? When they release 3 Nascar cars, but with 11 liveries and add 11 cars to the car count, how long did it take to make those liveries?

I don't think the problem is the livery, but the car itself.

 

[Johnnypenso]

I don't think the problem is the livery, but the car itself.

It was suggested above that we don't know how much the "little things" occupy the modelers at PD. PD has said a car takes a modeler 6 months. That's 130 days. If they can produce 1 livery per day as those 787B liveries suggest they might be able to, we can equate 130 liveries to one premium car. I'm sure there's more to it and it might even be easier for PD since they have professional resources, rather than having to grab stuff off the internet but it is an indication I would think, of how adding liveries fits into the overall scheme of things.

 

[Bloodmasters]

we need a livery editor so you can create all the liveries you want in little time

 

[Griffith500]

It was suggested above that we don't know how much the "little things" occupy the modelers at PD. PD has said a car takes a modeler 6 months. That's 130 days. If they can produce 1 livery per day as those 787B liveries suggest they might be able to, we can equate 130 liveries to one premium car. I'm sure there's more to it and it might even be easier for PD since they have professional resources, rather than having to grab stuff off the internet but it is an indication I would think, of how adding liveries fits into the overall scheme of things.

What about interiors?
What about external parts?
What about the semi Premiums?

There's an argument to be had that GT's liveries aren't the typical full coverage templates seen in PC sims, so the optimisation of texel usage, UV mapping, transparency etc. takes a bit longer, but it's not the lion's share of the effort.

Still, do enough of them and they'll add up.

By the way, your 40 cars per year figure is 12 real months per modeler per car for GT5 (more for GT6), not 6. Make your mind up. Unless you're implying it's man months, but I'll let you work that one out.

We want to know how many Premiums could be in GT7, not how many liveries a hobbyist can produce using pre-made templates (that an artist took the time to make for them).

 

[Johnnypenso]

What about interiors?
What about external parts?
What about the semi Premiums?

What about them?

There's an argument to be had that GT's liveries aren't the typical full coverage templates seen in PC sims, so the optimisation of texel usage, UV mapping, transparency etc. takes a bit longer, but it's not the lion's share of the effort.

Still, do enough of them and they'll add up.

And neither one of us can quantify it. However, as I said, I do know that you can make 3 liveries in 2 days for a PC game. If you have some other data points feel free to include them.

By the way, your 40 cars per year figure is 12 real months per modeler per car for GT5 (more for GT6), not 6. Make your mind up. Unless you're implying it's man months, but I'll let you work that one out.

I imply nothing and there's nothing for me to work out. 40 unique premiums/year is their historical output for 10 years, that's not in dispute AFAIC. It's PD that said 6 months not me, so if the figures don't jibe it's not my problem to work it out.

We want to know how many Premiums could be in GT7, not how many liveries a hobbyist can produce using pre-made templates (that an artist took the time to make for them)

Of course. But isn't it also possible that on cars with liveries, a pre-made template is also made during the "6 month" period it takes to make the average car, meaning that adding additional liveries should take no more time than it does a hobbyist? Just throwing that out there.

 

[FiftyBelowZero]

we need a livery editor so you can create all the liveries you want in little time

Please don't go off-topic.

 

[05XR8]

I dont know how many Premiums can be made for GT7. I do know they made the extra base models in about 4years time. White ones and matte black on some. And that's if they didn't get a head start while GT5 was being completed. If GT7 comes out December 2015, halve the number of base models for a start.

They've got lots to do. The GT Awards cars, deleting standard cars and replacing them with the premium version with a year model difference

I think I waa better off just waiting for each game to drop. First reading about what cars would be in the game via gaming magazines. Oh well.

 

[Griffith500]

What about them?

They take time, too, but I don't understand why you've been so reluctant to address that and reconcile it against your 40 car figure. It is only coincidence that it was 40 for GT5 and for GT6 - for it to remain 40 for GT7, PD would have to create more additional work (outside of unique Premiums) proportional to the extra modelers they have.

You seem to only be considering the evidence that reinforces your pre-existing belief. You're taking an empirical view and extrapolating it when the basis for that empiricism has been changed, and so it is no longer valid. At worst, it's superstition: PD have always made 40, so they'll continue to make 40.

And neither one of us can quantify it. However, as I said, I do know that you can make 3 liveries in 2 days for a PC game. If you have some other data points feel free to include them.

The difference in livery format between games means you have no relevant data point. It's not an order of magnitude difference, but it adds up.

As above, there were more modelers working on GT6, but they produced the same number of unique Premiums. That's probably a good indication of the time taken to make the semi Premiums. "I only look at what I can measure."

I imply nothing and there's nothing for me to work out. 40 unique premiums/year is their historical output for 10 years, that's not in dispute AFAIC. It's PD that said 6 months not me, so if the figures don't jibe it's not my problem to work it out.

So you're not actually interested in contributing to the discussion, other than simply "being right" on your own terms? You did mention 6 months per car, and you equated it to 130 liveries. Why deny that?

We're not interested in historical output, except for how it may inform the future, but you seem reluctant to consider the holes in your historical accuracy (false empiricism) in order to ensure an accurate prediction.

Of course. But isn't it also possible that on cars with liveries, a pre-made template is also made during the "6 month" period it takes to make the average car, meaning that adding additional liveries should take no more time than it does a hobbyist? Just throwing that out there.

No. Have you seen how PD makes liveries? There will be some overlap with the fixed details, but where the liveries differ, PD need unique UV mapping, transparency and texel density optimisation. Not 6 months' work, but it's not like this constrained view you have on PC sims, either. Extrapolate that lack of understanding to the rest of your assumptions, won't you.


The texturing is a small task compared to the modeling, including the unique interiors on the NASCAR and AJGT cars, or extra external parts for existing models. Or, you know, the semi Premiums, of which PD produced an equal number to the Premiums. All of that needs to be accounted for, too.

Your 40 car prediction (which I was originally responding to) is therefore unreliable; it's not even a reliable minimum, for reasons already explained.

 

[mister dog]

All this premium/standard/semi premium nonsense. Let's face it they just port their entire library over each time to brag with the car count on the back of the box (or for the love of cars like the fans like to state).

You are driving around with 800 PS2 era models in a 2013 game (and the possibility it will be the same for a 2016 game when GT7 is released is really mind boggling). Add to that this VGT project which is a nice publicity stunt but they don't have to model any interiors and it's basically the only DLC they release...

Let's compare racing games:

- Raceroom = quality over quantity
- Game stock car = quality over quantity
- AC = quality over quantity
- PCARS = quality over quantity
- Forza = quality over quantity
- Forza horizon = quality over quantity
- Grid Autosport = quality over quantity
- Mario Kart = quality over quantity

Gran Turismo is practically the only game that = quantity over quality. See where i'm getting at?

 

[Griffith500]

All this premium/standard/semi premium nonsense. Let's face it they just port their entire library over each time to brag with the car count on the back of the box (or for the love of cars like the fans like to state).

You are driving around with 800 PS2 era models in a 2013 game (and the possibility it will be the same for a 2016 game when GT7 is released is really mind boggling). Add to that this VGT project which is a nice publicity stunt but they don't have to model any interiors and it's basically the only DLC they release...

Let's compare racing games:

- Raceroom = quality over quantity
- Game stock car = quality over quantity
- AC = quality over quantity
- PCARS = quality over quantity
- Forza = quality over quantity
- Forza horizon = quality over quantity
- Grid Autosport = quality over quantity
- Mario Kart = quality over quantity

Gran Turismo is practically the only game that = quantity over quality. See where i'm getting at?

That's great, thanks.

Can I presume you don't think the semi Premiums are an improvement over the Standards? Would that mean you'd be more or less disappointed with 400 Premiums and the rest semi Premiums, or (however many more Premiums they can make) and all the other cars carried over again?

I'm sure many would want only Premiums, but it's hard to say how many of them we'll get if it's the only thing PD works on (and it's technically the same as the carrying over existing assets situation above, in terms of modeling alone, since they don't really cost anything).

 

[mister dog]

That's great, thanks.

Can I presume you don't think the semi Premiums are an improvement over the Standards? Would that mean you'd be more or less disappointed with 400 Premiums and the rest semi Premiums, or (however many more Premiums they can make) and all the other cars carried over again?

I'm sure many would want only Premiums, but it's hard to say how many of them we'll get if it's the only thing PD works on (and it's technically the same as the carrying over existing assets situation above, in terms of modeling alone, since they don't really cost anything).

They might look like an improvement as they got rid of those jaggies (which makes it more acceptable i guess), but it's exactly the same PS2 era model. I can understand that people that play bumper cam or third person don't mind as in those cases it's not noticeable, but for someone that only plays with the cockpit cam it renders 70% of the ingame cars useless as they don't want to be driving from inside a letterbox.

In any case i do not find it acceptable for a developer to port over simple car models like that and have such great quality differences. We know how it works but average Joe that buys the game to have a racer in his collection sees that "there are more than 1200 cars in GT6!", whilst only later he finds out the majority are standards, so one could even say it is misleading advertisement in a way.

Porting over premiums from GT6 to 7 would be more acceptable as they are already very detailed and have interiors, but porting PS2 era cars over yet again is just laughable. And for the small team no time excuse; either hire more staff, outsource or get out of the development business.

 

[Johnnypenso]

They take time, too, but I don't understand why you've been so reluctant to address that and reconcile it against your 40 car figure. It is only coincidence that it was 40 for GT5 and for GT6 - for it to remain 40 for GT7, PD would have to create more additional work (outside of unique Premiums) proportional to the extra modelers they have.

It is only a coincidence that it's been 40 for ten years. Okay.

You seem to only be considering the evidence that reinforces your pre-existing belief. You're taking an empirical view and extrapolating it when the basis for that empiricism has been changed, and so it is no longer valid. At worst, it's superstition: PD have always made 40, so they'll continue to make 40.

I am considering the only evidence that I have, which is 40/year. Consistent like clockwork too. When it changes we'll have new evidence.

The difference in livery format between games means you have no relevant data point. It's not an order of magnitude difference, but it adds up.

Perhaps, which is why I said, "maybe it's different on the PS3".

As above, there were more modelers working on GT6, but they produced the same number of unique Premiums. That's probably a good indication of the time taken to make the semi Premiums. "I only look at what I can measure."

That's a good guess. How do you know what they are working on? You don't. Maybe they new premiums have full undercarriages and full engine bays and you can pop open the boot and see a spare tire and now they take 9 months to complete. Maybe in true PD spirit they've gone to 1M polygons per car. Maybe the new modelers aren't working on cars they are working on tracks. Maybe Kaz thinks 40 has been fine for 10 years, after all look how many games we sold and they are putting their modeling effort into something to do with the FIA, or it takes 10 guys a year to model Isle of Man. T10 did say it took them 13000 hours for Nurburgring and IOM is longer than that and you know PD's attention to detail.

So you're not actually interested in contributing to the discussion, other than simply "being right" on your own terms? You did mention 6 months per car, and you equated it to 130 liveries. Why deny that?

There is a quote function:

First:

I don't know how long it takes to make a livery and make it work on the PS3

Then:

how long does it take to create liveries on the PS3? When they release 3 Nascar cars, but with 11 liveries and add 11 cars to the car count, how long did it take to make those liveries?

Then:

If they can produce 1 livery per day as those 787B liveries suggest they might be able to, we can equate 130 liveries to one premium car. I'm sure there's more to it and it might even be easier for PD since they have professional resources

If that's equating, then I guess I equated. To me it looks more like pure guesswork, but that's just me.

We're not interested in historical output, except for how it may inform the future, but you seem reluctant to consider the holes in your historical accuracy (false empiricism) in order to ensure an accurate prediction.
No. Have you seen how PD makes liveries? There will be some overlap with the fixed details, but where the liveries differ, PD need unique UV mapping, transparency and texel density optimisation. Not 6 months' work, but it's not like this constrained view you have on PC sims, either. Extrapolate that lack of understanding to the rest of your assumptions, won't you.

What I'm reluctant to do is extrapolate based on nothing.

The texturing is a small task compared to the modeling, including the unique interiors on the NASCAR and AJGT cars, or extra external parts for existing models. Or, you know, the semi Premiums, of which PD produced an equal number to the Premiums. All of that needs to be accounted for, too.

Your 40 car prediction (which I was originally responding to) is therefore unreliable; it's not even a reliable minimum, for reasons already explained.

You haven't explained anything, all you've done is throw out a bunch of what ifs, maybes, and could be/might be/should bes. When you know how long it takes to model a car currently, how long it takes to model a livery, how many people are working on cars as opposed to tracks and other things, whether the new recruits are replacing people that got fired or quit etc. etc. etc. then you let us know. Until then I'll stick with what I do know.

 

[Tornado]

They might look like an improvement as they got rid of those jaggies (which makes it more acceptable i guess), but it's exactly the same PS2 era model. I can understand that people that play bumper cam or third person don't mind as in those cases it's not noticeable, but for someone that only plays with the cockpit cam it renders 70% of the ingame cars useless as they don't want to be driving from inside a letterbox.

There's a difference between what they did for the RUF models, which are essentially scratch built to modern design sensibilities; and the Toyota Celica where they smoothed out some sloppy texture work and modelled door panels.



Attaching the semi-Premium moniker to the latter is kind of a joke, yes. And, unfortunately, when I was presented with this list it was done in a highly misleading fashion. I really should have checked it before commenting on it.

 

[mister dog]

There's a difference between what they did for the RUF models, which are essentially scratch built to modern design sensibilities; and the Toyota Celica where they smoothed out some sloppy texture work and modelled door panels.

I didn't know there was a difference but if they bothered to put in extra work and time on those RUF's according to modern design standards, why didn't they finish the job and make it a premium model like they did with the A110?

ps. About that A110 being represented as both a standard and a premium, i can't think of any other reason why they did that but to add another one to the car count. At least take out the freaking standard then

 

[Tornado]

I didn't know there was a difference but if they bothered to put in extra work and time on those RUF's according to modern design standards, why didn't they finish the job and make it a premium model like they did with the A110?

I don't know. Probably difficulty with sourcing an actual model to do the interior. They didn't technically make the A110 a Premium, either.

They take time, too

The one's that they actually converted to something that could be called Semi-Premium certainly did. Now that I'm actually looking at the "Semi-Premium" list and comparing it to the pictures in the comparison thread, the amount of work actually done to some of those cars was sometimes so minute that it is practically irrelevant to the discussion. Some of them have notable changes, but they are still just touched up versions of Standard cars that already looked better than most of the other Standard cars. Some of them are just Premium cars with the interior removed. Some of them are entirely rebuilt multi-piece models that are so low quality that they barely look any better than the Standard they replaced. Some of them are the same mesh but retextured.

 

[Griffith500]

Regardless of the level of detail, it's still effort not spent on Premiums. It's interesting that the semi Premiums paint a spectrum of "completeness".

Even the Rufs, barring the obvious omission of interiors (which applies to some supposed full Premiums also), are incomplete; I refer mostly to the head- and tail-lights.

From the frequency of touched up elements, it may be possible to estimate the relative complexity of each task.

It is only a coincidence that it's been 40 for ten years. Okay.

I am considering the only evidence that I have, which is 40/year. Consistent like clockwork too. When it changes we'll have new evidence.

Perhaps, which is why I said, "maybe it's different on the PS3".

That's a good guess. How do you know what they are working on? You don't. Maybe they new premiums have full undercarriages and full engine bays and you can pop open the boot and see a spare tire and now they take 9 months to complete. Maybe in true PD spirit they've gone to 1M polygons per car. Maybe the new modelers aren't working on cars they are working on tracks. Maybe Kaz thinks 40 has been fine for 10 years, after all look how many games we sold and they are putting their modeling effort into something to do with the FIA, or it takes 10 guys a year to model Isle of Man. T10 did say it took them 13000 hours for Nurburgring and IOM is longer than that and you know PD's attention to detail.


There is a quote function:

First:


Then:


Then:

If that's equating, then I guess I equated. To me it looks more like pure guesswork, but that's just me.



What I'm reluctant to do is extrapolate based on nothing.


You haven't explained anything, all you've done is throw out a bunch of what ifs, maybes, and could be/might be/should bes. When you know how long it takes to model a car currently, how long it takes to model a livery, how many people are working on cars as opposed to tracks and other things, whether the new recruits are replacing people that got fired or quit etc. etc. etc. then you let us know. Until then I'll stick with what I do know.

You really don't know any different. Your 40 car figure is useless for prediction: superstition.

If you think getting the same output from double the staff is to be expected, then fine. I find it troubling. And that is something we can measure, by the way. I don't like arbitrarily dismissing evidence, even if it means not being able to draw conclusions.

I just think we're missing something. You think you have it covered with a false empiricism based on an incomplete metric.

It'll be funny if PD focuses on the semi Premiums for GT7. In a way, we should almost expect it...

EDIT: The environment modelers are credited separately. Extra polygons would be apparent in GT6, because progressive meshes. Once again, check your level of understanding.

 

[badAZbuddha]

What Polyphony needs to do is do what they did when going from gt2 too gt3. Put in as much high quality cars as possible for gt7 and add the other fully modeled cars for future updates or save them for gt8 so there is no low quality cars on the track.

 

[karelpipa]

Semipremiums instead of standards is a good deal for me, i mean, look at those Rufs.

Can the Semis have additional parts like front bumper or at least their own rear wing not the hideous standard rear wing?

 

[Harsk100]

There are 447 premium cars, 642 standard cars and 152 Semi-Premium cars in GT6. 1241 cars in total.

So, if we take the 642 standard cars out and if we keep the 152 semi-premiums and 447 premiums, then we'll have an total of 599 cars without standard cars...

So, as I said, don't throw away the standard cars, please. But PD: try to improve their quality, please.

 

[Imari]

So, if we take the 642 standard cars out and if we keep the 152 semi-premiums and 447 premiums, then we'll have an total of 599 cars without standard cars...

Oh no, how would we ever manage to cope with only 600 cars in the game? If one were to drive each car for five minutes, that's only 50 hours of gameplay! Oh, the humanity! Won't someone think of the children!

I don't know how people managed to cope with GT1 and GT3. That must have been like some sort of masochistic torture going on, to play games with so few cars.

[/horrendoussarcasm]

 

[Johnnypenso]

Oh no, how would we ever manage to cope with only 600 cars in the game? If one were to drive each car for five minutes, that's only 50 hours of gameplay! Oh, the humanity! Won't someone think of the children!

I don't know how people managed to cope with GT1 and GT3. That must have been like some sort of masochistic torture going on, to play games with so few cars.

[/horrendoussarcasm]

Each car for 5 minutes on each track is only 5000 hours. A full time job for 2.5 years