Forza 3 vs GT5P Physics: Objective Analysis
- Thread starter MowTin
- 71 comments
- 19,412 views
The Outlaw
Premium
- 3,042
- Bay Area, California
- GTP_timeattack
Without a doubt T10 talked the physics of the game up well beyond reason 👎 But what's new with salesmen today
I'm probably one of the least biased individual/drivers you will find on this forum, and you simply can't compare the realism of the physics between GT5P and FM3. Simply put, if the physics in Forza 3 even came close to the physics of GT5P I wouldn't have been back to Prologue a day after picking up FM3
And I think a lot of us know how worn out Prologue is...that just shows how disappointing Forza's physics are for anyone who is into the challenge of balancing and driving a car to it's absolute limit.
- 1,620
... if the physics in Forza 3 even came close to the physics of GT5P I wouldn't have been back to Prologue a day after picking up FM3
I think that pretty much sums it up - as I don't believe anyone would want to buy a 360 and FM3 just to bash it, but want to like it - esp considering how conservative PD are with new features etc. Its a pity FM3 isn't better with the basic stuff despite having lots of appealing features.
I blame MS for whipping their devs too hard, but maybe PD aren't quite whipped hard enough...
- 1,029
- Uruguay
Thank you for an excellent and well thought out comparison.
I think FM3 tries to appeal to a broader audience. Their physics seem like it's simplified for drifting. I often find myself automatically drifting when I make a mistake in a corner.
What Turn 10 did was make a fun realistic racing game that could be played by anyone. Some probably will just play FM3 to pimp their rides and tune them. Taking a Ford Focus and turning it into some 500hp monster is a game in itself.
FM3 may win by simply appealing to more people.
I dont agree with this at all. Suzuka is far more to real life in GT5P than FM3. I was watching Trullis onboard lap on the official F1 site and used his lines to improve my lap time in GT5P, thats how on the money they are with this track in GT. I played FM3 last night and its not even close to the real Suzuka, the corner profiles are all off, cambers, gradients.
If you are talking about graphics detail then maybe you have a point but in terms of real track mapping GT5P is far superior.
Seismica
Premium
- 4,346
- Guisborough
- GTP_Seismica
- 23,800
- Philippines
Full throttle at second gear... in what cars? I find it hard to use more than part throttle in cars like the M3 and F430 in anything below fourth gear at full attack in a corner.
And that's in standard trim... on N tires, which faithfully represent stock levels of grip. Yes, I'm a time attack junkie.
If you're complaining that you can be full throttle in second gear on high level S or R tires, then that's another level of grip, entirely... that's the reason why they're called sports tires. (and no, S-tires aren't equivalent to stock on any in-game cars... Spec III increased the grip of S-tires, so they no longer faithfully represent stock levels of grip for the the sports cars in game)
And that's in standard trim... on N tires, which faithfully represent stock levels of grip. Yes, I'm a time attack junkie.
If you're complaining that you can be full throttle in second gear on high level S or R tires, then that's another level of grip, entirely... that's the reason why they're called sports tires. (and no, S-tires aren't equivalent to stock on any in-game cars... Spec III increased the grip of S-tires, so they no longer faithfully represent stock levels of grip for the the sports cars in game)
Seismica
Premium
- 4,346
- Guisborough
- GTP_Seismica
- 567
- aGrenadeShark
- aGrenadeShark
My x360 just bricked the day i bought my MS wheel. I will also be doing a comparison of the 2 games very soon both with wheels. DFGT and MS wheel. Both which are the low cost options for the games.
First thing i notice is that I have to relearn everything moving to GT5P. I instantly spun out on almost every turn. When i left GT5P, i was competing for top lap times.
Playing forza inherently makes you bad at GT5P. The games are just so different as to the way they handle the physics. GT5P is like ice skating while forza is like driving on rails. I actually did a comparison a while back between the two and I still feel like GT5P's low speed physics need some real help as they are overly penalizing compared to real life.
A perfect example is a mazda rx8 coming out of a corner (stock) at full speed. It is almost impossible in real life to spin out a rx8. Its balanced so well that if you are on the racing line, you can mash the accelerator with no problem (i have driving time in rx8 on a track). In gt5P, I can spin a stock rx8 out no problem on almost any turn using stock tires.
In forza, the rx8 reacts more realistic at low speeds IMO.
Again, I will be doing a comparison review in a few weeks when I get my 360 back. I am sure lots others will as well. It's just good to see multiple opinions on the matter.
First thing i notice is that I have to relearn everything moving to GT5P. I instantly spun out on almost every turn. When i left GT5P, i was competing for top lap times.
Playing forza inherently makes you bad at GT5P. The games are just so different as to the way they handle the physics. GT5P is like ice skating while forza is like driving on rails. I actually did a comparison a while back between the two and I still feel like GT5P's low speed physics need some real help as they are overly penalizing compared to real life.
A perfect example is a mazda rx8 coming out of a corner (stock) at full speed. It is almost impossible in real life to spin out a rx8. Its balanced so well that if you are on the racing line, you can mash the accelerator with no problem (i have driving time in rx8 on a track). In gt5P, I can spin a stock rx8 out no problem on almost any turn using stock tires.
In forza, the rx8 reacts more realistic at low speeds IMO.
Again, I will be doing a comparison review in a few weeks when I get my 360 back. I am sure lots others will as well. It's just good to see multiple opinions on the matter.
- 567
- aGrenadeShark
- aGrenadeShark
- 20,145
- Australia
- VIPERGTS01
That list was actually released the time of GT5P spec I (original) then updated with the new cars in Spec II (PAL, NTSC release), later (August 1st 2008) during spec II we got a update which increased grip in S rated tyres only, this furthered the gap between N and S tyres BUT the list was never updated or modified at all reflecting the changes. So suddenly the cars that nudged into S tyres (as new tyres) were so much quicker. Just on Eiger alone the cars gained a good 1-2 seconds over pre patch.
For this reason I don't pay attention to that old list for accuracy anymore.
- 567
- aGrenadeShark
- aGrenadeShark
that sucks. Also, there's apparently no GT-R.
- 738
- RaLpH89 / GTP_RaLpH89
GT5 prologue's take off oversteer and N3 tires grip needs "a lot" of improvement, because.
1. Too many rookie in GT5p online just take an 4WD/AWD to get ahead in front.
2. Driving F1 seems effortless in some corners. (it's still hard to drive sometimes)
3. Time attack seems way off in real stock cars counter part (maybe PD should have picked N3 tires to get the almost identical time, 2:15: in a)
P.S. Someone should make a race about stock cars vs stock cars, like Best Motoring style. That would make GT5 prologue fun. It's kinda boring seing the same cars in all types of PP races.
1. Too many rookie in GT5p online just take an 4WD/AWD to get ahead in front.
2. Driving F1 seems effortless in some corners. (it's still hard to drive sometimes)
3. Time attack seems way off in real stock cars counter part (maybe PD should have picked N3 tires to get the almost identical time, 2:15: in a)
P.S. Someone should make a race about stock cars vs stock cars, like Best Motoring style. That would make GT5 prologue fun. It's kinda boring seing the same cars in all types of PP races.
AWD cars naturally blast off the line (when you don't care about mechanical damage, anyway) faster than their 2WD competitors, so while it may be considered a 'n00b' thing to do, it's spot on.
As for the F1, anyone here can bumble around Suzuka at what might initially seem like a good time (because it's a billion times faster than everything else in the game), but I'd say only 5% would be able to get it within a second of what it can do. There's a huge difference between stumling around and managing a 37 or something and pushing harder and nailing a 32 and then breaking into the 20's, it starts of ridiculously easy and gets a LOT harder VERY quickly.
Odds are, if it seems easy, you're not going fast enough. Push anything to the edge of it's limits and I guarantee it's a)a LOT harder and b) a LOT faster.
- 20,145
- Australia
- VIPERGTS01
Thats pretty spot on TBR 427, I will add aswell the reason why the F2007 sometimes seems effortless in corners is because F1 cars mostly rely on aero grip which makes high speed long corners seem quite easy even when accelerating.
Standard cars racing is fantastic, sometimes myself and Jackington (a GTP member) join events with standard cars and have a blast racing each other. At one stage there was a couple months with F430vsGTRvZ06 (standard tune) events which were pretty good since everything has been PP based. When GT5 comes we will be able to make our own races and rules.
Standard cars racing is fantastic, sometimes myself and Jackington (a GTP member) join events with standard cars and have a blast racing each other. At one stage there was a couple months with F430vsGTRvZ06 (standard tune) events which were pretty good since everything has been PP based. When GT5 comes we will be able to make our own races and rules.
Its pretty hard to objectively analyse these games. Very soon its going to be subjective...
Objective would be getting a copy of a detailed magazine test (Sport Auto (german)) and then find a way to measure the stopping distance, cornering skidpad G force, 0..200km acceleration times, plus compare the car in bone stock form on the Nurburgring, driven by someone who's really good at the games.
In the pc sim rFactor for example you can look at all the telemetry and when I create cars, I make sure the real tests are matched with what my cars do on a skidpan.
For gamepad orientated semi sims such as GT5 and Forza, I bet most cars in stock form, when driven by a good driver, are quite a bit faster than the real cars at the Ring for example; indicating exagerated grip for the sake of playability.
Objective would be getting a copy of a detailed magazine test (Sport Auto (german)) and then find a way to measure the stopping distance, cornering skidpad G force, 0..200km acceleration times, plus compare the car in bone stock form on the Nurburgring, driven by someone who's really good at the games.
In the pc sim rFactor for example you can look at all the telemetry and when I create cars, I make sure the real tests are matched with what my cars do on a skidpan.
For gamepad orientated semi sims such as GT5 and Forza, I bet most cars in stock form, when driven by a good driver, are quite a bit faster than the real cars at the Ring for example; indicating exagerated grip for the sake of playability.
- 23,800
- Philippines
Actually... a number of us did objective tests for GT4... and found the numbers reasonably accurate in terms of braking, cornering and speed.
Laptimes are something you can never accurately compare... there are too many variables... and the fact that you can drive just over the edge in a videogame without fear of death will tilt the odds in favor of the game in some cases... and the lack of "feel" of a videogame, compared to real life, will tilt the odds in favor of real life, in others.
Laptimes are something you can never accurately compare... there are too many variables... and the fact that you can drive just over the edge in a videogame without fear of death will tilt the odds in favor of the game in some cases... and the lack of "feel" of a videogame, compared to real life, will tilt the odds in favor of real life, in others.
Yeah there are always differences, no matter how good the game / sim is. But people say GTR Evolution is a 'sim' yet the Koenigsegg there laps the Ring about 1 minute faster than the real car does, which is clearly only possible by a hefty grip / power / ninja brakes overdose.
I'm actually convinced a good sim will give similar laptimes to real life. Sure yeah at the Ring, some of the faster twisty bits a real driver might take a bit easier than we can do on a LCD screen, but overall..
The Ring has actually been converted from GTR Evolution to rFactor (they're identical platforms really) and I create physics for cars in rFactor. The track is off in places, but fairly good. Same goes, hopefully, for my cars which match the Sport Auto test data down to the aerodynamic lift / drag.
If you can sort of focus 'on the middle' in this video you'll see just how close a sim can be even when the car and track are obviously not perfect: http://www.youtube.com/watch?v=FoB9EzL-KCM (poor video / audio sync!) The essential thing being of course that it wouldn't count if I copied his lap just by matching speeds by studying the video. Instead this is just the result of trying to do a fast lap.
The reason I threw in the gamepad argument is that I always see GT4 times on the Ring being much faster than the real cars do, but that could all be down to the tuning people do to their cars in the game.
I'm actually convinced a good sim will give similar laptimes to real life. Sure yeah at the Ring, some of the faster twisty bits a real driver might take a bit easier than we can do on a LCD screen, but overall..
The Ring has actually been converted from GTR Evolution to rFactor (they're identical platforms really) and I create physics for cars in rFactor. The track is off in places, but fairly good. Same goes, hopefully, for my cars which match the Sport Auto test data down to the aerodynamic lift / drag.
If you can sort of focus 'on the middle' in this video you'll see just how close a sim can be even when the car and track are obviously not perfect: http://www.youtube.com/watch?v=FoB9EzL-KCM (poor video / audio sync!) The essential thing being of course that it wouldn't count if I copied his lap just by matching speeds by studying the video. Instead this is just the result of trying to do a fast lap.
The reason I threw in the gamepad argument is that I always see GT4 times on the Ring being much faster than the real cars do, but that could all be down to the tuning people do to their cars in the game.
- 23,800
- Philippines
It's just down to how far over the limit you can drive in a game... and the fact that in GT4, since fuel temperature, intake temperature and heat-soak are not modelled, you can pretty much thrash it all you want with no performance drawbacks.
Also note: there's no tire-wear in free-lapping mode, so that's another factor to consider.
You can look at real life times and see how big the variance is for a track like the Nurb. The GT-R... Porsche did it in 7:54, in the dry.
Chris Harris did it in 7:55... in the wet. (stupid German test drivers... )
SportAuto did it in 7:38... in the dry... validating the original Nissan times that had everyone in an uproar... (consequently... he did it in a ZR1 in exactly 7:38, also... more on this later).
Now... with tremendous factory efforts and a zillion laps... Nissan had the GT-R down to around 7:26.
People cried "foul" over that... especially Porsche, who went on to demonstrate that if you put a Porsche driver in a Nissan car, it will automatically make 1/2 the horsepower and limp its way around the track.
but GM had the Corvette ZR1 around the Ring in 7:24 or thereabouts with their own factory effort.
So... with factory backing, a zillion laps of tweaking, and a near robotic driver, you can knock about 12 seconds off of Horst von Saurma's time.
If your driver has absolutely no sense of self-preservation... it's entirely conceivable that you could shave another dozen seconds off of that.
And... if your car experienced no tire wear whatsoever, no overheating of oil, water, brakes, etcetera... another dozen seconds...
This is where Virtual Reality and Reality part ways... but long before that, the driver pucker factor gives you a huge difference in lap times.
Also note: there's no tire-wear in free-lapping mode, so that's another factor to consider.
You can look at real life times and see how big the variance is for a track like the Nurb. The GT-R... Porsche did it in 7:54, in the dry.
Chris Harris did it in 7:55... in the wet. (stupid German test drivers...
)SportAuto did it in 7:38... in the dry... validating the original Nissan times that had everyone in an uproar... (consequently... he did it in a ZR1 in exactly 7:38, also... more on this later).
Now... with tremendous factory efforts and a zillion laps... Nissan had the GT-R down to around 7:26.
People cried "foul" over that... especially Porsche, who went on to demonstrate that if you put a Porsche driver in a Nissan car, it will automatically make 1/2 the horsepower and limp its way around the track.
but GM had the Corvette ZR1 around the Ring in 7:24 or thereabouts with their own factory effort.
So... with factory backing, a zillion laps of tweaking, and a near robotic driver, you can knock about 12 seconds off of Horst von Saurma's time.
If your driver has absolutely no sense of self-preservation... it's entirely conceivable that you could shave another dozen seconds off of that.
And... if your car experienced no tire wear whatsoever, no overheating of oil, water, brakes, etcetera... another dozen seconds...
This is where Virtual Reality and Reality part ways... but long before that, the driver pucker factor gives you a huge difference in lap times.
- 601
- eggs2see
My thoughts on FM3 V GT5P Physics are pretty much the same as the OP's. With one exception. I Actually find the car's hook up better off launch in FM3 than GT5P, and they also do step out or "torque steer" as you called it, but im pretty sure that's actually what happens to front wheel drives and not rear wheel drives, as the torque is actually steering the car as opposed to making the rear end step out.
Yeah there are always differences, no matter how good the game / sim is. But people say GTR Evolution is a 'sim' yet the Koenigsegg there laps the Ring about 1 minute faster than the real car does, which is clearly only possible by a hefty grip / power / ninja brakes overdose.
I'm actually convinced a good sim will give similar laptimes to real life. Sure yeah at the Ring, some of the faster twisty bits a real driver might take a bit easier than we can do on a LCD screen, but overall..
The Ring has actually been converted from GTR Evolution to rFactor (they're identical platforms really) and I create physics for cars in rFactor. The track is off in places, but fairly good. Same goes, hopefully, for my cars which match the Sport Auto test data down to the aerodynamic lift / drag.
If you can sort of focus 'on the middle' in this video you'll see just how close a sim can be even when the car and track are obviously not perfect: http://www.youtube.com/watch?v=FoB9EzL-KCM (poor video / audio sync!) The essential thing being of course that it wouldn't count if I copied his lap just by matching speeds by studying the video. Instead this is just the result of trying to do a fast lap.
The reason I threw in the gamepad argument is that I always see GT4 times on the Ring being much faster than the real cars do, but that could all be down to the tuning people do to their cars in the game.
2 things should immediately be ringing bells in your head:
Firstly, tyres. Even a small difference in softness will shave dizens od seconds of a lap time at the ring.
Secondly, in GTR the driver doesn't afraid of anything. Crash? Meh, restart. Not my car, don't have to pay for repairs, don't care if I bin it. NOT SO in real life. The driver still has to make sure the car makes it back in one piece.
Given those two differences, I reckon everyone here could knock at least 30 seconds off any time IRL.
There is absolutely a big difference in what a factory driver can do after loads of laps and what a 'regular expert' (Sport Auto) can do with limited laps. The 7:26 times of the factory driven Nissan and Corvette vs the 7:38 Sport Auto laps show that quite well. And of course in a sim / game, we can do loads and loads of laps so the 7:26s in these cars would be a more realistic target.
I doubt though that there is more than a few seconds left in those real 7:26 laps. These factory drivers really get close to the limit. It is just physically impossible for those laptimes to drop much more without cheating with slick tires. A few seconds sure, but never 30 seconds. Tire wear over one lap isn't an issue and tightly driven, tire temperature shouldn't be significant either. There is only so much extra time to shave off with tires that make the car pull certain G forces.
Its pretty interesting that you need quite a bit more grip to really gain 10 seconds on the ring. If I'd tune 'my corvette' so that I can do a 7:11 lap, it would pull 1.15G on a flat skidpan or there abouts, which it just won't ever do in real life. If I stick to realish 1.07 numbers though and have the matching brake G force, I could probably just get a 7:23 out of it during a Ninja Lap (tm).
So yeah absolutely conditions vary, driver skill varies, but comparing one of the faster real laps to a sim, I don't think you can put it all down to these variables.
I doubt though that there is more than a few seconds left in those real 7:26 laps. These factory drivers really get close to the limit. It is just physically impossible for those laptimes to drop much more without cheating with slick tires. A few seconds sure, but never 30 seconds. Tire wear over one lap isn't an issue and tightly driven, tire temperature shouldn't be significant either. There is only so much extra time to shave off with tires that make the car pull certain G forces.
Its pretty interesting that you need quite a bit more grip to really gain 10 seconds on the ring. If I'd tune 'my corvette' so that I can do a 7:11 lap, it would pull 1.15G on a flat skidpan or there abouts, which it just won't ever do in real life. If I stick to realish 1.07 numbers though and have the matching brake G force, I could probably just get a 7:23 out of it during a Ninja Lap (tm).
So yeah absolutely conditions vary, driver skill varies, but comparing one of the faster real laps to a sim, I don't think you can put it all down to these variables.
Think of what the drivers are doing...then think if they had a several more years to practice, and didn't have to worry about the cars, and could crash them again and again and it wouldn't matter.
I do not believe that they will ever push as hard as they could, because it's a long track and there's too many places to stick it into the fence or armco, and they wouldn't be getting a pay rise for the next few years if they bend a $400,000 prototype.
I'm not sure where you got the 1.15g bit from, but are you considering grip not only from a latitude standpoint, but also the longitudinal factor? A lot of places on te ring aren't corners; just slowish straights and braking zones; and that difference in grip would stack up VERY quickly.
I do not believe that they will ever push as hard as they could, because it's a long track and there's too many places to stick it into the fence or armco, and they wouldn't be getting a pay rise for the next few years if they bend a $400,000 prototype.
I'm not sure where you got the 1.15g bit from, but are you considering grip not only from a latitude standpoint, but also the longitudinal factor? A lot of places on te ring aren't corners; just slowish straights and braking zones; and that difference in grip would stack up VERY quickly.
Similar threads
