Official GT5 discussion thread

Audi TTS Coupe ‘09

I have little need for super fast cars at the moment since hardly any of the events allow me to use them.
 
What about online ?

I am afraid to go online now. What with people like tha_con with racing experience. I'll be road kill all the time. XD
 
I've not gone online for the same reasons. Not to mention I returned the game today. My roommate will get it for Christmas so I'll be back at it soon enough :p
 
Silly? Surely a driving simulator should be trying to get as close to real life as possible. Ergo an improvement in the simulation should result in a decrease in the delta between real-world lap times and in-game lap times. Thus if GT1 had Car X getting around the Ring in 4 minutes where in real life Car Z would take 8 minutes, and GT2 shows the car taking 5 minutes, GT3 6 minutes, and GT4 7 minutes, you'd see the times getting closer as a result of the simulation moving from a very weak arcade racer to a more accurate simulation (yes, very coarse, unrealistic figures just to illustrate the trends!). If GT5's simulation engine is more sophisiticated that predecessoers, those time advantages caused by not factoring in real-life limitations should be reduced, meaning closer-to-real results.

That's all of course, no brainer, taking all things as equal and accounting for varables. A change in variables like tyres or weight of the driver will add variation to the gam eresults and real-world result. But as others say, these variables aren't going to have a huge difference. the same car on the same track with the same driver isn't going to see a 10% improvement from chaning frmo one brand of slicks to another, unless the intial brand are extremely sucky tyres! A windy day isn't going to add a minute to your lap times on the Ring. So we'd just be looking at same ball-park race times for an accurate simulation, and not exactly the same times. Where in one test of GT4, a Ford GT managed Leguna Seca in 1:38 in game best time, and 1:54 in real life, that difference of 16 seconds (16%) isn't going to be explained by a different brand of tyres and a little more headwind, but a difference between the game and real life. The question is whether the core simulation isn't very accurate, allowing cars to perform better in the game than in RL, or if it's a matter of the drivers.

The ultimate simulation game would have players getting about the same best times, give or take a few percent difference, because the cars' handling and the racing conditions reproduce the physical state of the cars and the mental state of the driver. We should be seeing some more realistic times in GT5 versus GT4 for the same car+track combinations if the simulation has improved.

Well, I think GT5 among others are slowly reaching the point of dimishing returns where adding in more factors will not necessarely result in a smaller difference between real life and in-game lap times. It's pretty much as good as it gets - the only way to get an even better simulation would be to include and match the variables Robert listed to the simulation you want to run. Be able to set ambient temperature, air pressure, humidity, track temperature, tyre pressure (!) etc. yourself. As long as you can't do this, you just have to assume some or perhaps "best-case" scenario for everything in the game.

On a small track, badly set up tyres can easily amount to a few seconds. What is a few seconds on a small track, is easily 10 fold on a track like the Nordschleife, simply due to it's length.

If you want a real representative test of in-game laps vs. real-life - I'd chose a very easy track and a car (not a FordGT or anything too high powered) that is easy to drive and can do consistent lap times lap after lap that aren't too dependant on the factors we have no idea about. The more difficult the car or the track, the higher the chance of trial/error come into play - or as tha_con sais, the driver becomes his own factor.
 
Last edited by a moderator:
If you want a real representative test of in-game laps vs. real-life - I'd chose a very easy track and a car.
At which point, how can they miss out all three of the Top Gear reasonably priced cars?! Every TG fan willl want to see where they'd come on the leader board, but there's not one valid car for testing. PD should have included a car set up for that purpose.
 
What about online ?

I am afraid to go online now. What with people like tha_con with racing experience. I'll be road kill all the time. XD

You should definitely just hop online in a social lounge. We can do free laps and just chit chat :p It'd be fun! Me and Phil played yesterday for maybe an hour and had a great time (especially when we smashed up a few cars, lol). A B3D casual room would be really fun :)
 
How does it compare to the latest Nissan GTR ? and Mercedes SLS ?

I :love: the SL300 before playing GT5. Like it even more after the AMG challenge. The above 2 cars are the only ones I'm interested to try in the game.

The RUF BTR '86 isn't as fast as the Nissan GTR. And it very easily starts to slide when going into corners and when you try to accelerate out of corners too fast. Doing that won't be fast, but it's a lot of fun.

I tried the Nissan GTR in Arcade mode. I'm definitely going to buy it in GT mode at some point.

I you want scary, take the Zonda R down the Nürburgring. It's incredibly fast, has loads of grip, and make one mistake and I'll send you flying.


Does anyone know when this better AI is supposed to be unlocked? I'm at level 19 and 43% complete. So far it's a disappointment. Outside of occasionally deliberately blocking my way to keep from overtaking and not crashing into me when I do the same, they'll more often act like a bunch of bumpercars. And always break way too early before the same corners.
 
What's a good car for Historic Cup in Expert Mode? This event is much harder than the other ones in Expert. It's even much harder than some of the events in Extreme.
 
I tried the Nissan GTR in Arcade mode. I'm definitely going to buy it in GT mode at some point.

Yes, I like the Nissan GTR *a lot* in real life.

EDIT:
Great movie showing AI in GT 5
http://www.youtube.com/watch?v=r_HJSJLL8_I

There are many situation i've been talking about weeks ago and some quite new. Really good video. All reviewers should watch it if they claimed that AI is bad in this game.

I think it's because the Special Events have hardcoded path to train people like me. The reviewers saw them, and use the video conveniently to illustrate their points. Perhaps, PD can introduce AI mode to these Special Events (after the initial training period).

FWIW, I really really like the License Tests and Special Events. Now I understand what Ostepop meant when he said Gran Turismo taught him how to drive. ^_^

The game has enough variety to challenge and "train" the player under different racing or road conditions.


Oh yes, thanks for clarifying the AI for players. ^_^
 
I think it's because the Special Events have hardcoded path to train people like me
Licence Tests and Special Events like Top Gear dont have normal GT 5 AI, rather something totally scripted.

What's a good car for Historic Cup in Expert Mode?
Toyota 7 from 'Like The Wind' event [extreme]
 
The Special Events actually have AI that will behave exactly the same almost 99% of the time. In a sense (and I don't really like this) each race against other competitors in the Special Events is essentially a puzzle. A game of finding very very tight and precise opportunities in which you can pass safely whilst not sacrificing your own line or speed (or getting disqualified).
 
The Special Events actually have AI that will behave exactly the same almost 99% of the time. In a sense (and I don't really like this) each race against other competitors in the Special Events is essentially a puzzle. A game of finding very very tight and precise opportunities in which you can pass safely whilst not sacrificing your own line or speed (or getting disqualified).

That's exactly how I treated the first two Top Gear challenges. I can tell you exactly which car you will pass on each turn, what position you will be in, etc. I could probably do them in my sleep now lol.
 
Yap. But the reviewers should have noticed that GT5 has to have bigger scope for AI drivers simply because of B-Spec. In B-Spec, besides the enemy AI, Polyphony also needs to implement a parameterized AI driver for us to control. Judging from forum posts, it seems that they also implement some sort of emotion variables into our drivers.

If they want, I think all PD needs to do is to implement "more fear" or "less aggressiveness" into certain A-Spec levels, so that the enemy AI won't run into us.

These issues should not be confused with Forza comparisons, which is largely a PR problem that Sony/gamers should clarify separately.
 
They're quoting Sony's report, which is shipped numbers. It sold 440k in the US alone in the 3 days of NPD's reporting period for November. It probably sold another 2-300k (or maybe a little more) since then in the US, plus whatever it sold in Canada and Latin America.
 
Last edited by a moderator:
Back
Top