GT5's Game-Breaking Online Flaw (OP Updated: 11 Feb)

[carracerptp]

I'm running shuffle races as they appear to be only races that are not badly affected.

Why is that? I have done many shuffle races in Destinkeys rooms so I would think a massive "Slow down" and "FPS drop" would have had people talking sooner?

 

[Speedy6543]

Great work MGR! 👍
Very usefull.

 

[Famine]

Great work MGR! 👍
Very usefull.

Indeed.

Since we're not going to get a Blog post out in the next hour, I'll just advise everyone keeps their lobbies under 12 - remember that the more stuff going on, the more pronounced the effect, so wheels, USB mics, heavily modified cars and traffic will all add to it but seemingly so long as you're less than 12 drivers per lobby you shouldn't see anything more severe than usual P2P-caused lag.

 

[Filip_Ovik]

Here is a video showing how it feels to drive affected of the issue.

Please click on this link: http://www.youtube.com/watch?v=lir3cop1Cgg

 

[dsgerbc]

I'll just advise everyone keeps their lobbies under 12
...
seemingly so long as you're less than 12 drivers per lobby you shouldn't see anything more severe than usual P2P-caused lag.

I'd say 'nothing more severe than usual P2P-caused lag' applies to lobbies with under 10 players. Bump it to 11-12 drivers and it will be bearable/livable, but still more laggy then what I used to consider normal.

 

[Famine]

We managed a 20,000 mile lobby last night with 12 (and one watching non-participant) just fine. The only person who was apparently lagging was me - and of course I didn't see it. A slightly more geographically-discerning lobby should be fine.

 

[Rdp616]

I think this affects time trials too. Same way with that David Coulthard challenge not too long ago. Some of the times were just insane. We're all on the same tires, same power, yet some people are 10-15 seconds quicker than others? I consider my self a good racer and i couldn't get no where near the top, and i use a Fanatec wheel..

Watching some of the replays it looked like some guys were on racing hard tries. Not everyones internet is the same speed, ping is also different from household to household. And i think this affects speeds, grip, and power in big ways.

 

[carracerptp]

I think this affects time trials too. Same way with that David Coulthard challenge not too long ago. Some of the times were just insane. We're all on the same tires, same power, yet some people are 10-15 seconds quicker than others? I consider my self a good racer and i couldn't get no where near the top, and i use a Fanatec wheel..

Watching some of the replays it looked like some guys were on racing hard tries. Not everyones internet is the same speed, ping is also different from household to household. And i think this affects speeds, grip, and power in big ways.

I think some people were cutting the track, they later fixed it.

I don't think the "David Coulthard challenge" was "Online" it was Online. Offline would not have this problem (in this way).

 

[Hydro]

We ran our regular Wed. WRS Online race with 11 drivers and had no issuses other than the usual (minor) lap 1 lag when the field was all together.
Settings were:
Standard room quality.
Low Mic quality.
Visual damage was set to 'off'.
EDIT: Premium Cars, Most drivers using wheels, 8 mics.
Replay lap times were consistent with stopwatch times (+/- <1 sec. - my slow / fast stopwatch finger could account for that discrepancy)

I think this affects time trials too. Same way with that David Coulthard challenge not too long ago. Some of the times were just insane. We're all on the same tires, same power, yet some people are 10-15 seconds quicker than others? I consider my self a good racer and i couldn't get no where near the top, and i use a Fanatec wheel..

Was that before or after the event was 'reset' due to a massive corner cut (chicago IIRC) that wouldn't disqualify the lap, and allowed for significantly faster lap times?

 

[carracerptp]

Here is a video showing how it feels to drive affected of the issue.

Please click on this link: http://www.youtube.com/watch?v=lir3cop1Cgg

Great video, the case (problem) can easily be seen.

 

[Furinkazen]

Why is that? I have done many shuffle races in Destinkeys rooms so I would think a massive "Slow down" and "FPS drop" would have had people talking sooner?

Mostly standard cars.

 

[jon-]

We just officially delayed the start of the champ.

Started as 15 person room, all but 1 person was about 6 seconds off their quali time. By the time it had got to 13 people things were a bit more stable with most of us only a few seconds down, though it seemed to vary from user to user.

At 13 cars me and a friend were switching between pad and wheel and both noticing a second or two gain over the wheel.

At 12 users thing stabilised for a lot of us, though not every one.

This was at trail mountain, all running stock Merc SLSs.

 

[Rdp616]

Was that before or after the event was 'reset' due to a massive corner cut (chicago IIRC) that wouldn't disqualify the lap, and allowed for significantly faster lap times?

After, i'm talking about just lap times in general, and watching different replays. Some people were sliding all over the place, some people seemed like they were on racing tires.

I watched throttle control the most. Some guys would be full throttle out of the hammer head and would get no wheel spin at all. While others, (like myself) would get insame amounts of wheels spin, even at half throttle coming out of the same corner.

I was always trying to point out this "flaw" but could never really find a way to put it into an explanatory form. This thread will hopefully bring these problems to PD's attention. IMO this thread should be stickied.

 

[carracerptp]

After, i'm talking about just lap times in general, and watching different replays. Some people were sliding all over the place, some people seemed like they were on racing tires.

I watched throttle control the most. Some guys would bu full throttle out of the hammer head and would get no wheel spin at all. While others, (like myself) would get insame amounts of wheels spin, even at half throttle coming out of the same corner.

I was always trying to point out this "flaw" but could never really find a way to put it into an explanatory form. This thread will hopefully bring these problems to PD's attention. IMO this thread should be stickied.

I don't think it was a glitch, more than likely line choice. I did that challenge and found that if you get the car off "balance" it would madly spin the wheels. Keeping it level was hard but when you did, grip was in good supply.
Also, the really fast guys can use just off of 100% WOT, keeping the car just under the power band and keeping the grip.

I would really be shocked if the Offline was messed up as there is no lag, or other factors outside of the games control.

Then, there was that Go-kart TT.... That was really odd....

 

[jon-]

We just officially delayed the start of the champ.

Started as 15 person room, all but 1 person was about 6 seconds off their quali time. By the time it had got to 13 people things were a bit more stable with most of us only a few seconds down, though it seemed to vary from user to user.

At 13 cars me and a friend were switching between pad and wheel and both noticing a second or two gain over the wheel.

At 12 users thing stabilised for a lot of us, though not every one.

This was at trail mountain, all running stock Merc SLSs.

Race 2 we started with 10 people, everyone had the right amount of grip.

 

[Rdp616]

I don't think it was a glitch, more than likely line choice. I did that challenge and found that if you get the car off "balance" it would madly spin the wheels. Keeping it level was hard but when you did, grip was in good supply.
Also, the really fast guys can use just off of 100% WOT, keeping the car just under the power band and keeping the grip.

I would really be shocked if the Offline was messed up as there is no lag, other factors outside of the games control.

Then, there was that Go-kart TT.... That was really odd....

Who knows. Watching different replays tho definitely made it look fishy. Like i said i use a wheel, and it's very easy to keep the balance, and i'd still get wheel spin. Even in a straight line.

 

[Jordan]

I've co-authored this post on the blog with Famine. Thanks for your additional contributions, MGR and Filip. 👍

 

[EDK]

Thanks to the dedicated staff and members who showed up for our test yesterday.

The next moaner gets a special prize.

 

[amar212]

I've co-authored this post on the blog with Famine. Thanks for your additional contributions, MGR and Filip. 👍

Great article, great job done by all involved, hope it will bring additional attention and hopes it can actually be solved. Thank you guys for the effort in the mutual behalf 👍

 

[Christian_C]

Isn't the problem that the GT clock runs faster than real life? the blog post makes it out like the GT clock runs slower than RL but to me it seems the other way round?

 

[jon-]

Good article, fair play.

 

[carracerptp]

I wounder if the "USB data" could have anything to do with it?
Many DS3 users seem to have a lower "glitch amount". I always have it (the controller) connected to the PS3 via the USB cord. I'm sure I have seen the laps just as bad as most Wheel users.

 

[EDK]

Isn't the problem that the GT clock runs faster than real life? the blog post makes it out like the GT clock runs slower than RL but to me it seems the other way round?

The problem is that when the right set of circumstances are present, lap times are much slower than normal. The clock runs at the correct rate in real time. We have live, in race video from our test that shows this.

When watching the replay, the game clock from some drivers runs much more slowly than a real time clock.

Affected drivers are in heavier traffic, and are using a wheel and/or headset.

 

[jon-]

I wounder if the "USB data" could have anything to do with it?
Many DS3 users seem to have a lower "glitch amount". I always have it (the controller) connected to the PS3 via the USB cord. I'm sure I have seen the laps just as bad as most Wheel users.

I USB my controller and it's fine when my wheel isn't.

People are also getting improved laptimes by turning their bluetooth mics off, which would point to an overall system load issue rather than specifically USB.

 

[dudeondacouch]

Nice factual and thorough article in the News section. 👍

Hopefully this will spur awareness, and in turn, bring us closer to a proper online racing environment.

 

[nealcropper]

Glad it's been brought up in the blog, great post gtp! All you guys who have done all the testing should be the ones to thank though, let's hope it finally gets noticed and fixed, or at least word from PD.

 

[dsgerbc]

When watching the replay, the game clock from some drivers runs much more slowly than a real time clock.

Game clock shows replay lap lasting 2:30.
Real time clock shows replay lasing 2:15. I think this suggests that "the GT clock runs faster than real life" is more correct.

I'm not sure 'clock runs faster/slower' is a proper characterization though. De-synchronized clock is just an easily observable symptom. Dropped frames/driver inputs seems like a better description.

I imagine if I watch the replays on youtube frame-by-frame I'll find frames where the clock jumps forward by more than it does between 'normal' frames.

 

[cicua]

Isn't the problem that the GT clock runs faster than real life? the blog post makes it out like the GT clock runs slower than RL but to me it seems the other way round?

Neither. The game clock runs at correct realtime speed *when playing*.

It is the replay that plays too fast, and when it does it is a sure sign that you were hit by the bug.

The article is somewhat unclear on this.

 

[EDK]

Game clock shows replay lap lasting 2:30.
Real time clock shows replay lasing 2:15. I think this suggests that "the GT clock runs faster than real life" is more correct.

How do you arrive at that conclusion?

If 1 minute takes 1 minute and 10 seconds, is the clock in the game not running more slowly than a real one, taking longer to arrive at the same point in time?

 

[dsgerbc]

If 1 minute takes 1 minute and 10 seconds, is the clock in the game not running more slowly than a real one, taking longer to arrive at the same point in time?

During one minute of real time, the replay clock does more than one minute. I think this means that replay clock is "faster".