Will the relatively poor sales of GT6 affect the development of GTSport?

Will the abysmal sales of GT6 have an effect on how GT7 is designed and developed?

  • Definitely. I think they will take this as a sign that they need a major overhaul of the franchise.

    Votes: 34 16.4%
  • Somewhat. Much of the game will remain "GT", but some parts will be overhauled completely.

    Votes: 111 53.6%
  • Not at all. Business as usual. A familiar game targeted towards a more casual audience ala GT6.

    Votes: 62 30.0%

  • Total voters
    207
Status
Not open for further replies.
Because the AI (in GT4) actually races you, as in does not deliberately let you win? How about a competent GT-Mode? You can say what you like about online this or Academy that. It doesn't make the game better.
Actually, they exhibited almost no signs of intelligence at all, and they barely did any racing. This was the game which people began to complain about Gran Turismo's bots being mindless road rockets, with PC sims beginning to show competent artificial intelligence. What they did do was set a good pace and took turns pretty well - and was when bot ramming began to be complained about a great deal, as some players were actually slower in turns than the bots! And strangely, I think this was the game in which the uncatchable "rabbits" first appeared... :P

Online does make a game better, potentially, and it would be suicide to offer a racing game without online capability. Heck, even fighting games have online play, something that makes me nervous. Still, everyone's remarks about the pitfalls and minuses of online racing are true too. For some it's a godsend, for others, a headache, so offline play is just as important.

The bots in GT6 can race, but you have to find the series. I found halfway competitive bots in the I-A Class Super GT series. Instead of first to last being spread out to an appalling average of a minute and a half, the difference in those races is 15 to 20 seconds, and the fields are much tighter as a consequence. Why this series and very few others, I haven't a clue. But being vocal and insistent that PD juice up the bots is the only way to insure this is addressed by PD, along with everything else we want improved in GT7.

While I haven't been all that impressed by the A.I. in most racing games so far, one did: RaceRoom. They don't exactly race like people do, they struggle around a lot of turns for instance. But they can set a blistering pace and be uncatchable, and they can be tuned to suit your racing style to where the feeling is remarkably similar to racing online with competent people close to your skill level. Sector 3 and I think their one A.I. programmer show that it can be done reasonably well. They can sure run you off the road like people, though!

GT3 and GT4 are so much better than GT6, the absolute worst GT game ever made.
Yeah... I think this is a highly personal opinion. Those boards are fairly quiet these days, and I sincerely doubt many people at all have played them for many years. I think of them fondly, GT4 at least which I still regard with awe, but it's collecting dust along with GT5. And when 7 hits, GT6 will be joining them.
 
Last edited:
Yeah... I think this is a highly personal opinion. Those boards are fairly quiet these days, and I sincerely doubt many people at all have played them for many years. I think of them fondly, GT4 at least which I still regard with awe, but it's collecting dust along with GT5. And when 7 hits, GT6 will be joining them.
Then in about 10 or 15 years. GT1 will probably be more popular than GT2,3,4,5,6 and 7. :lol:
 
AI competency isn't about their pace, even though GT's AI drive pretty damn slowly. My biggest gripe with the GT6 AI is they are overly submissive. They hit the brakes whenever you are within about 2 feet of them, which makes them impossible to actually race against. As soon as you catch them, they slam the brakes on and you sail away to an instant 2 or 3 second lead.

I've found it very frustrating having AI cars slam the brakes on when I'm directly behind them in the slipstream, simply because I'm close to their bumper, in the middle of a straight.
You also notice it when getting a good exit onto a straight. Instead of slowly getting alongside the AI, and having to br brave into the braking zone to pass, you get alongside, and they instantly hit the brakes mid straight, and there goes any chance of an actual race.

It can also be used as a tactic to prevent AI in faster cars from overtaking. When they are pulling up alongside you, just move over close to them, you don't even need to contact them, and they'll slam on the brakes. Even mid straight.

PD seem to have attempted to make the AI good at avoiding accidents with the player and each other. The problem with this is they are so paranoid about contact that they rarely attempt a pass, often needing 5 car widths of room to get by, and they'll panic and slam on the brakes whenever they get close to another car.

The Nascar events seem to be the only exception to the rule, but in those the AI act like you're not there, and will turn right into you and punt you off. They've got a lot of work to do to get the AI in good shape.
 
If we go by the past figures it would seem Sony has some worrying to do:

In the millions:

Gran Turismo 10.850
Gran Turismo 2 9.370
Gran Turismo 3 A-spec 14.890
Gran Turismo Concept Series 1.560
Gran Turismo 4 “Prologue” 1.400
Gran Turismo 4 11.730
Gran Turismo 5 “Prologue” 5.350
Gran Turismo PSP 4.220
Gran Turismo 5 10.890
Gran Turismo 6 2.370
That GT 3! Man those were the days, after that its been meh.

Look at that drop off GT5 to GT6. Holy hell.:banghead:
 
Actually, they exhibited almost no signs of intelligence at all, and they barely did any racing. This was the game which people began to complain about Gran Turismo's bots being mindless road rockets, with PC sims beginning to show competent artificial intelligence. What they did do was set a good pace and took turns pretty well - and was when bot ramming began to be complained about a great deal, as some players were actually slower in turns than the bots! And strangely, I think this was the game in which the uncatchable "rabbits" first appeared... :P

Online does make a game better, potentially, and it would be suicide to offer a racing game without online capability. Heck, even fighting games have online play, something that makes me nervous. Still, everyone's remarks about the pitfalls and minuses of online racing are true too. For some it's a godsend, for others, a headache, so offline play is just as important.

The bots in GT6 can race, but you have to find the series. I found halfway competitive bots in the I-A Class Super GT series. Instead of first to last being spread out to an appalling average of a minute and a half, the difference in those races is 15 to 20 seconds, and the fields are much tighter as a consequence. Why this series and very few others, I haven't a clue. But being vocal and insistent that PD juice up the bots is the only way to insure this is addressed by PD, along with everything else we want improved in GT7.

While I haven't been all that impressed by the A.I. in most racing games so far, one did: RaceRoom. They don't exactly race like people do, they struggle around a lot of turns for instance. But they can set a blistering pace and be uncatchable, and they can be tuned to suit your racing style to where the feeling is remarkably similar to racing online with competent people close to your skill level. Sector 3 and I think their one A.I. programmer show that it can be done reasonably well. They can sure run you off the road like people, though!


Yeah... I think this is a highly personal opinion. Those boards are fairly quiet these days, and I sincerely doubt many people at all have played them for many years. I think of them fondly, GT4 at least which I still regard with awe, but it's collecting dust along with GT5. And when 7 hits, GT6 will be joining them.
I rather an A.I that followed a straight line over ones that desperately try to let you win.

Though GT4 did start rabbit chasing it was a passible thing back then. The cars had a decent margin towards each other, there was less cars and on Championship Events, they allowed you to Qualify.

I wouldn't say Online = Instantly, better game. but it does open games up to people who prefer to race with other people instead of A.I. Though PD seems to only focus on Online for GT5 and GT6 instead of focusing on both which isn't any better than no Online mode.

I think the reason for this is because we have already done everything in GT4, so there is no need to look back at it, though I still do for my single player GT Fix, same goes for GT5 and for some now GT6 and overall people found GT4 the better game to go through the entire game because they enjoyed that one the most.
 
@Mike_grpA
I've read a few similar complaints here, and it's got me to thinking that there might be some sort of adaptive system that tunes the A.I. to the player. Possibly. I say this because mine rarely acts like that.

The potential reason mine don't is I've been racing in GT6 "wrong," in that I use a car around 75% or so of the PP rating for the races. This gives me a challenging race, and also one I can lose. I've been doing races over and over in the I-A Class events, especially the Super GT 300 and 500 which provide the closest fields of bot drivers. They still give up if you get more than a three-quarter second lead on them, but when they're close, they race pretty competitively. Drafting, rubbing fenders, nudging, bumping, passing, the kinds of behavior we want to see in these bots.

A couple of us suggested this on GT Planet, mostly me. Some people tried it and thought it was great. Some thought it stunk. But those that liked it had experiences similar to mine. This is one possibility.

Another is it's just possible that the PS3 is so complex, the game plays slightly different for some people. I have noticed that some games have bugs and other issues that don't manifest for most people. I had a bug in Killzone that was pretty rare in one level. This is another possibility.

Just throwing things out there, because it does seem that a few people will remark about little details every once in a while that I don't experience, and this is one of them.
 
Last edited:
I'm not sure it's that complex. I have found some events in the career where the AI seems to have been "tuned" so to speak to not care quite so much. Perhaps the Super GT events are one where the tuning is much better, and as I said earlier, the Nascar events seem to be tuned to the AI have no problems trading paint and bumping, but the side effect is they tend to turn into you like they're not aware of you at all.

This isn't a GT6-specific problem either. I remember it being the same in GT5 too. I reckon it was probably an attempt from PD to avoid the same complaints from GT3 and 4 about the AI acting like the player isn't there.

If you park in the middle of the racing line and watch the AI reacting to you as they come past in GT5 and 6, you'll see their avoidance is actually quite good, but I think the AI as a whole are too simple, and perhaps that's why this great avoidance comes at the cost of being way too conservative and careful in other conditions, like running side by side, and also why they are far too scared to attempt an overtake.

I don't believe they are adaptive to the player at all, as I used to use cars with lower than recommended pp when I played through the career in GT6, and am quite an aggressive driver (very much a late braker, and tune for oversteer, but not a dirty racer), and I didn't find any of the races hard to beat the AI, and they never adapted to my style, they just always plodded along in their little Trulli trains waiting for me to pass them, and never once did I see one attempt to defend a position, they'd stick religiously to their line and just allow me to pass them on any corner.

I'm positive PD will work on it for GT7. If they can keep their avoidance of accidents close to as good as GT6, but make them more aggressive and capable of close racing, through adding complexity to them, I'm confident they could end up with very good AI. This is what I'm hoping anyway.
 
I'm not sure it's that complex. I have found some events in the career where the AI seems to have been "tuned" so to speak to not care quite so much. Perhaps the Super GT events are one where the tuning is much better, and as I said earlier, the Nascar events seem to be tuned to the AI have no problems trading paint and bumping, but the side effect is they tend to turn into you like they're not aware of you at all.

This isn't a GT6-specific problem either. I remember it being the same in GT5 too. I reckon it was probably an attempt from PD to avoid the same complaints from GT3 and 4 about the AI acting like the player isn't there.

If you park in the middle of the racing line and watch the AI reacting to you as they come past in GT5 and 6, you'll see their avoidance is actually quite good, but I think the AI as a whole are too simple, and perhaps that's why this great avoidance comes at the cost of being way too conservative and careful in other conditions, like running side by side, and also why they are far too scared to attempt an overtake.

I don't believe they are adaptive to the player at all, as I used to use cars with lower than recommended pp when I played through the career in GT6, and am quite an aggressive driver (very much a late braker, and tune for oversteer, but not a dirty racer), and I didn't find any of the races hard to beat the AI, and they never adapted to my style, they just always plodded along in their little Trulli trains waiting for me to pass them, and never once did I see one attempt to defend a position, they'd stick religiously to their line and just allow me to pass them on any corner.

I'm positive PD will work on it for GT7. If they can keep their avoidance of accidents close to as good as GT6, but make them more aggressive and capable of close racing, through adding complexity to them, I'm confident they could end up with very good AI. This is what I'm hoping anyway.
I like the idea behind B spec in GT5, where "hot drivers" were more likely to crash, but were much faster than the "cold drivers" it didn't work very well. But the idea behind it is great. I gain time when Im underpressure too... but sometimes I go wide too. And sometimes being too hot leads to bad lines... if the AI in GT7 do this right. It will be a fantastic feature.
 
I like the idea behind B spec in GT5, where "hot drivers" were more likely to crash, but were much faster than the "cold drivers" it didn't work very well. But the idea behind it is great. I gain time when Im underpressure too... but sometimes I go wide too. And sometimes being too hot leads to bad lines... if the AI in GT7 do this right. It will be a fantastic feature.

Yeah that illustrates what I was talking about well. I had forgotten about that in GT5 B-Spec. So by my theory, the 'hot' drivers will be less scared of a crash, so will drive closer to others and attempt passes more often, and won't back off so much in the corners, but they risk crashing in numerous ways, but the "cool" drivers are more like the standard AI you find in most A-Spec races, where they're mega careful, rarely attempt overtakes unless their car is way faster and they've got a big wide straight bit of track to do so, back off when you get near them, and have that annoying tendency to jab the brakes mid turn lol, but they'll pretty much never crash.

I agree they could make the B-Spec pretty damn good if they can get those kinds of things balanced well, but with more complex AI, with less emphasis on the 'A' and more on the 'I' :lol:
 
GT5 Prologue was revolutionary, it took another 3 years for the final game to come out which had a graphical downgrade and was full of PS2 standard cars. GT5 was when we saw that PD can't keep up with the competition and have well and truly lost the title of being the "king"of racing games.

GT5 was great, a slight improvement but lacked some detail to Prologue. A example is the blocky shadows. The biggest problem I figured was the Standard cars. I expected at least a improvement on most of them in GT5 but 6 years from prologue we had not many improved. Now this is carried onto GT7. That is 9 or 10 years. Thats way more than enough time to improve them.
 
I agree they could make the B-Spec pretty damn good if they can get those kinds of things balanced well, but with more complex AI, with less emphasis on the 'A' and more on the 'I' :lol:

I suspect that as long as B Spec remains "push button to go faster" it will be a weak game mode.

On the other hand, if it were to progress into something where you had to "build" your own AI out of parts in some fashion, that might be interesting as a game. One could set braking points and a racing line for one's AI fairly easily, but then constructing a set of reactions as to how it reacts to other cars would be the real art. Making an interface that allowed the player to alter those reactions in a simple yet intuitive way would be tough, but probably doable.

And then once you've built your AI driver, you take him to the track and let him loose. And then shout commands at him from the virtual pit wall, like our current B Spec mode.

The other advantage to a "build your own AI" mode is that it allows for online play, either normally or asynchronously. Pitting your own AI against your friends could be good fun.

B Spec is a decent basic concept, it just needs a lot of work on how to gamify it into something that is generally more engaging.
 
Really. If all AI is like the BSpec driver, itll beat any racing game! Miles better than current AI. What a missed opportunity.
 
The GT6 A spec AI is just slow as turtle :lol:

Agree with @FrzGT the Bspec driver is better than the race opponent AI. GT5 bobs were varied in aggression, pace etc, if these parameter are tuned to give good competitive AI, it would be great indeed.

Right now, to get decent pace AI, I have to get them better cars to drive, this means I have to bump my car power, enter the race and bring it back to original power, use lower grip tire and if not enough, put 20%-30% water on track to make the race interesting :lol:
 
Not exactly sure of where to put this : was hoping to see GT7 in the list - was surprised to see no racing sim at all.
 
On the other hand, if it were to progress into something where you had to "build" your own AI out of parts in some fashion, that might be interesting as a game.
Along with your notion of drawing/racing out a racing line for B-Spec Bob and setting brake points/regions, how about building your Bob's character with development points, a la Fallout 3? I discussed several character traits that defined a human race car driver way back in 2006, and how this would be good factors in defining GT's bots. Let me see if I can remember all seven...
  • Skill - this should be self evident, overall ability
  • Aggression/tenacity - desire and ability to achieve better position, or hold onto one
  • Strength/endurance - how well the driver can perform consistently over the course of a race
  • Intelligence/savvy - how clever he is, how able to deal with split second decisions like avoiding a wreck, take advantage of brief opportunities like a missed braking point
  • Senses - ability to handle sun glare, rain, fog, night racing etc
  • Courage - ability to shake off incidents and get right back at it, take chances
  • Luck - possible wild card ability to deal with and avoid trouble
It could be argued that what separates drivers from each other are the other six characteristics, but I thought I'd include Skill in there too. Maybe you could have seven development points and have a well rounded bot driver. Or a strong smart driver, or a clever, lucky risk taker.

I'm more partial to having something like Forza's Drivatar that you train by going through skill tests, but this is another way to approach "building" a bot.
 
Along with your notion of drawing/racing out a racing line for B-Spec Bob and setting brake points/regions, how about building your Bob's character with development points, a la Fallout 3? I discussed several character traits that defined a human race car driver way back in 2006, and how this would be good factors in defining GT's bots. Let me see if I can remember all seven...
  • Skill - this should be self evident, overall ability
  • Aggression/tenacity - desire and ability to achieve better position, or hold onto one
  • Strength/endurance - how well the driver can perform consistently over the course of a race
  • Intelligence/savvy - how clever he is, how able to deal with split second decisions like avoiding a wreck, take advantage of brief opportunities like a missed braking point
  • Senses - ability to handle sun glare, rain, fog, night racing etc
  • Courage - ability to shake off incidents and get right back at it, take chances
  • Luck - possible wild card ability to deal with and avoid trouble
It could be argued that what separates drivers from each other are the other six characteristics, but I thought I'd include Skill in there too. Maybe you could have seven development points and have a well rounded bot driver. Or a strong smart driver, or a clever, lucky risk taker.

I'm more partial to having something like Forza's Drivatar that you train by going through skill tests, but this is another way to approach "building" a bot.
To add to that, it would be interesting to be able to put them through certain trials that will develop a selected trait somehow. It could be limited to, lets say 25 sessions, where you would have to pick which traits you'd want the character to portray, so that way you can't just max out everything and each one would be individual to the creators liking.
 
To add to that, it would be interesting to be able to put them through certain trials that will develop a selected trait somehow. It could be limited to, lets say 25 sessions, where you would have to pick which traits you'd want the character to portray, so that way you can't just max out everything and each one would be individual to the creators liking.

Tenacious D would make an excellent dungeon master

d20-giant-brushed-steel.jpg
 
That GT 3! Man those were the days, after that its been meh.

Look at that drop off GT5 to GT6. Holy hell.:banghead:

Once again those sales numbers for GT6 are incorrect. Its best to wait for whenever PD decides to release info.
 
I hope so. I genuinely hope so. My fear is that if GT7 doesn't deliver the goods, it'll kill the franchise off altogether, at the very least in its current format. That would be an absolute tragedy for the former market leader in racing realism, and what was, once upon a time, the greatest racing game around. It would be an absolute tragedy for a charming, nuanced, largely enjoyable series, that shows off its utter passion and love for the motor car in all its forms. It would be an absolute tragedy for the wonderfully talented staff of PDI. It would be an absolute tragedy for a series a great many of us grew up with, a series that gave us fond memories, and best of all, gave us this wonderful community. For better or for worse, there is nothing quite like Gran Turismo, and there may very well never be again.

Please Kaz, don't mess it up. For me, for you, for PDI, for Sony, for all of us.

Please.

They have one last chance with GT7. We can put GT6's poor sales performance down to being released after the PS4, and after the casual players moved on, with no real desire to go back to a PS3 game. We can put the issues with GT5 and GT6 down to the hardware and the difficulty PDI have had with development for it.

There are no excuses like that when it comes to GT7 on PS4. The development team has never been in a better place to turn out the best GT game ever, hopefully having learned a lot from GT on the PS3. Yet, at the same time, if they haven't, the series has never been more likely to fail altogether. I know Kaz and co. have the potential, and I truly hope it all comes together for them.
 
Last edited:
Along with your notion of drawing/racing out a racing line for B-Spec Bob and setting brake points/regions, how about building your Bob's character with development points, a la Fallout 3? I discussed several character traits that defined a human race car driver way back in 2006, and how this would be good factors in defining GT's bots. Let me see if I can remember all seven...
  • Skill - this should be self evident, overall ability
  • Aggression/tenacity - desire and ability to achieve better position, or hold onto one
  • Strength/endurance - how well the driver can perform consistently over the course of a race
  • Intelligence/savvy - how clever he is, how able to deal with split second decisions like avoiding a wreck, take advantage of brief opportunities like a missed braking point
  • Senses - ability to handle sun glare, rain, fog, night racing etc
  • Courage - ability to shake off incidents and get right back at it, take chances
  • Luck - possible wild card ability to deal with and avoid trouble
It could be argued that what separates drivers from each other are the other six characteristics, but I thought I'd include Skill in there too. Maybe you could have seven development points and have a well rounded bot driver. Or a strong smart driver, or a clever, lucky risk taker.

I'm more partial to having something like Forza's Drivatar that you train by going through skill tests, but this is another way to approach "building" a bot.

That's one way to do it, and probably a good one. But I suspect in the age of the internet such would be min/maxed really fast and everyone would just Google the "best" setup. I look at things like WoW talent trees and how hard Blizzard had to work to drive people away from min/maxing and allow them to be able to express some creativity and personality.

I think you could do a good job with such a system though, certainly better than current B Spec. There's a lot of design space in the traits you choose, the effects they have and how you allow the player to distribute points.



What I had thought of (that probably would never work) is something more like FF12s gambit system, if you've ever played that. The player chooses sets of conditional actions, puts them in a ranked order and lets that describe which actions the AI takes.

Having a limited set of conditional slots, and potentially a limited set of conditional actions as well if the actions had to be unlocked or purchased (because Gran Turismo is a game that encourages collecting, after all) might make it so that the optimal setup changes from track to track and car to car. Perhaps "powerful" actions are exclusive, you can only one "legendary action" active at a time. Perhaps strong actions have cooldowns on how often they can trigger each lap, or each race.

Hell, maybe you don't customise your set for the driver, maybe you do it for each track. Maybe you assign custom offensive and defensive maneuvres to corners or sections of the track. How do you want your driver to attack the first chicane at Monza? How about the Lesmos? Parabolica? How about defense? That could have a real advantage of making players think about how each section of the track is approached, and thinking about potential racecraft and strategy.

I feel with this sort of system, a designer can build in a lot more depth, and give the player a lot more decisions that are not easily broken down into numerics. It also has room for finding and building "combo" sets, where the actions work together particularly well and are greater than the sum of their parts. Or "wacky" actions for those who like a bit of humour ("when within 100m of the finish line in 1st place, do a handbrake turn and reverse over it!").

Of course, it would be tough to design and very tough to balance, but that's why designers get paid the big bucks. It's certainly a system that could start small and be expanded through later patches, or modified if need be. I'd be interested to see how far a small, basic set of instructions could be pushed, to start with.

There's probably dozens of viable ways to build an interesting B Spec system, and I don't claim that's the best, or even better than a traits based system like the one above. Mine isn't even original, it's just a mash-up of B Spec + FF12 combat. But maybe someone else will get a giggle out of reading it. I think it's interesting, and it's certainly an approach that will probably feel fresh to most players.
 
@Imari tl; dr... sorry. :embarrassed:

However, you got me at "FF12...Gambit". One of the better AI management scheme I've seen in a console game.
 
people seem to forget the drop off which happened on GT5's release. It wasn't a good game when it came out and it took a near entire rewrite/upgrade with the 2.0 release patch for it to pick up.

GT6 is currently on 1.20 patch. I'm guessing that by 2.0 patch which will almost certainly arrive as the lead times are long for GT games that this will too pick up more players/Sales.

However the thing which took most of my mates off to other games was the launch of the game came over a month after the launch of the PS4 most of the guys I raced with had upgraded so were hoping this would be a dual platform release.

Had they done (if they do) a dual platform release then this would have impacted sales positively.
 
people seem to forget the drop off which happened on GT5's release. It wasn't a good game when it came out and it took a near entire rewrite/upgrade with the 2.0 release patch for it to pick up.

GT6 is currently on 1.20 patch. I'm guessing that by 2.0 patch which will almost certainly arrive as the lead times are long for GT games that this will too pick up more players/Sales.

However the thing which took most of my mates off to other games was the launch of the game came over a month after the launch of the PS4 most of the guys I raced with had upgraded so were hoping this would be a dual platform release.

Had they done (if they do) a dual platform release then this would have impacted sales positively.
The logic of your post escapes me. If most of your mates have graduated to PS4, how is a 2.0 update going to pick up more players/sales? Do you think they'll drag their dusty PS3's up from the basement to play GT6 when they could just use their PS4 to play Drive Club or Project Cars?
 
For me its just a sign that customers want something more than same stuff and features promissed delivered. So if they planned doing same with GT7 and then rethinked then ok. But GT6 has hurt PD, low sales and features that seem hard to fix still cost them in development time.
 
Status
Not open for further replies.
Back