Cheating everywhere, GMs???

[Samoht]

Originally Posted by titanshub [You must be logged in to view images. Log in or Register.]

if you have a valid argument against the methodology used to come to these conclusions then lets see it.

Yeah, it's nonsense. There is no correlation between ping and FPS.

Ping is the response time between your client and the server.

FPS is the speed that it takes pixels to render from your video card to your monitor.

OP is trying to measure reaction time in ping from different computers with unknown relation to the geographic location of the hosted server when all he's got is FPS.

OP dismisses obvious flaws as "rubber banding" when in reality almost everything you see from the movement of other characters not your own fall under rubber banding. It's not precise.

The server does not monitor the position of your character. That is 100% client side (which is why warping in MQ works). Your client sends updates to the server, and the server shares those updates with other players, so the locations are no longer accurate by the time the other players receive them.

You ever sent a tell in guild chat at the same exact time of somebody else? It shows up first on your screen, but it shows up in a different order from someone else. Why? Because your chat was displayed client side without interaction from the server and chat from everybody else had to go from your client to the server to their client.

OP spent 2 months producing a 30 minute video based on the flawed assumption that he could measure reaction time frame by frame when there's literally nothing to keep the players locations synchronized in real time.

When measured against an admitted cheater, Stunningly's reaction time lost. Therefore Stunningly must be a cheater, right?

[DeathsSilkyMist]

Originally Posted by Samoht [You must be logged in to view images. Log in or Register.]

Yeah, it's nonsense. There is no correlation between ping and FPS.

Ping is the response time between your client and the server.

FPS is the speed that it takes pixels to render from your video card to your monitor.

OP is trying to measure reaction time in ping from different computers with unknown relation to the geographic location of the hosted server when all he's got is FPS.

OP dismisses obvious flaws as "rubber banding" when in reality almost everything you see from the movement of other characters not your own fall under rubber banding. It's not precise.

The server does not monitor the position of your character. That is 100% client side (which is why warping in MQ works). Your client sends updates to the server, and the server shares those updates with other players, so the locations are no longer accurate by the time the other players receive them.

You ever sent a tell in guild chat at the same exact time of somebody else? It shows up first on your screen, but it shows up in a different order from someone else. Why? Because your chat was displayed client side without interaction from the server and chat from everybody else had to go from your client to the server to their client.

OP spent 2 months producing a 30 minute video based on the flawed assumption that he could measure reaction time frame by frame when there's literally nothing to keep the players locations synchronized in real time.

When measured against an admitted cheater, Stunningly's reaction time lost. Therefore Stunningly must be a cheater, right?

Yup, this is correct.

Online games are not nearly as accurate as they may appear to be. Generally we use what is called interpolation for the location of any player that is not your own. This means another player's movement on your screen is actually calculated purely on your own client in between movement packet updates. We basically just take the last know speed and direction of that player, and assume they are still going in that direction and at that speed until we get the next update.

When everybody's latency is low, you don't really notice interpolation happening, because the time it takes to receive the next movement packet is so low. Any minor discrepancy between your client's interpolation calculation, and the actual location of the player is going to be unnoticeable.

"Rubber Banding" occurs when there is a considerably long time between movement packet updates. Lets say it took 1 full second between movement packet updates. That means for a full second, all of the other players were simply moving in the last known direction they were going, and at the last know speed. Lots of changes in speed and direction can occur within one second, which is why you will see the "rubber banding" player teleport to a different location.

Interpolation will never be a magic bullet to solve this issue, because we can't actually see into the future to know what you are going to do next. All we can do is predict, and hope you have a good internet connection[You must be logged in to view images. Log in or Register.]

[titanshub]

Originally Posted by Samoht [You must be logged in to view images. Log in or Register.]

Yeah, it's nonsense. There is no correlation between ping and FPS.

Ping is the response time between your client and the server.

FPS is the speed that it takes pixels to render from your video card to your monitor.

OP is trying to measure reaction time in ping from different computers with unknown relation to the geographic location of the hosted server when all he's got is FPS.

OP dismisses obvious flaws as "rubber banding" when in reality almost everything you see from the movement of other characters not your own fall under rubber banding. It's not precise.

The server does not monitor the position of your character. That is 100% client side (which is why warping in MQ works). Your client sends updates to the server, and the server shares those updates with other players, so the locations are no longer accurate by the time the other players receive them.

You ever sent a tell in guild chat at the same exact time of somebody else? It shows up first on your screen, but it shows up in a different order from someone else. Why? Because your chat was displayed client side without interaction from the server and chat from everybody else had to go from your client to the server to their client.

OP spent 2 months producing a 30 minute video based on the flawed assumption that he could measure reaction time frame by frame when there's literally nothing to keep the players locations synchronized in real time.

When measured against an admitted cheater, Stunningly's reaction time lost. Therefore Stunningly must be a cheater, right?

I think that there is an error in this logic. The idea that anything in the real world is completely provable is false. If we assume for the moment that the video in question contains no methodological or calculation errors and we compared that to a hypothetical admission of guilt from the accused parties then the statistical argument presented in the video is still more valid than the admission of guilt. This is because human testimony is a terrible source of information that is generally avoided in favor of empirical data.

All we can ever have is evidence not proof and assuming no methodological or calculation errors in that presentation then the author has moved the bar from something like 99.9% to 99.999%. If you set the bar at 100% then you have asked for the bar of proof to be set so high nobody can ever achieve it. This is bad for a server that promises a fair raid scene and impartial staff. (I'm not saying that staff are not these things just pointing it out) Everyone here including Vanquish should have a vested interest in the server maintaining its integrity.

As stated in the video, this kind of analysis in speed running is done with direct input logs of when buttons were pressed. The devs of the server presumably don't have access to that level of information about the users of p99. For this server to have any integrity in the raid the burden of reasonable doubt not absolute proof needs to be used.

If we use the burden of reasonable doubt then the accused look guilty af.

That being said, I'm bias so if you want to prove it wrong go poke actual objective holes in his analysis instead of hand waving lag from the monitor to the whatever whatever whatever. Provide a counter example where you use his methodology and produce provably honest reaction times that contradict his claims.

[Samoht]

Originally Posted by titanshub [You must be logged in to view images. Log in or Register.]

I think that there is an error in this logic. The idea that anything in the real world is completely provable is false. If we assume for the moment that the video in question contains no methodological or calculation errors and we compared that to a hypothetical admission of guilt from the accused parties then the statistical argument presented in the video is still more valid than the admission of guilt. This is because human testimony is a terrible source of information that is generally avoided in favor of empirical data.

All we can ever have is evidence not proof and assuming no methodological or calculation errors in that presentation then the author has moved the bar from something like 99.9% to 99.999%. If you set the bar at 100% then you have asked for the bar of proof to be set so high nobody can ever achieve it. This is bad for a server that promises a fair raid scene and impartial staff. (I'm not saying that staff are not these things just pointing it out) Everyone here including Vanquish should have a vested interest in the server maintaining its integrity.

As stated in the video, this kind of analysis in speed running is done with direct input logs of when buttons were pressed. The devs of the server presumably don't have access to that level of information about the users of p99. For this server to have any integrity in the raid the burden of reasonable doubt not absolute proof needs to be used.

If we use the burden of reasonable doubt then the accused look guilty af.

That being said, I'm bias so if you want to prove it wrong go poke actual objective holes in his analysis instead of hand waving lag from the monitor to the whatever whatever whatever. Provide a counter example where you use his methodology and produce provably honest reaction times that contradict his claims.

wut.

Isn't that all OP provided here? Some false correlations and hand waving?

The only thing supporting his point of view that what he's saying is true is that he wants it to be true.

I don't have to prove anything false. His argument is fucking bullshit, and the fact that GMs haven't made any changes based on the video indicates to me that they agree.

Originally Posted by titanshub [You must be logged in to view images. Log in or Register.]

The racers with the lowest latency has the biggest advantage. If I had to guess other factors are at play here as well such as read write speeds on the I/O. The game is writing that 900+ random to a log file that then needs to be read by the 3rd party software so delays here might be significant or not. None of that negates the use of cheat software. The claim being made by the OP that they are cheating and I'm pointing out that extra sources of lag on his computer alone does not exonerate them from that accusation. The other racers all fell within that statistical distribution dataset from humanbenchmark.com.

The argument presented here is looking at statistical distributions of the reaction times of racers (not just the accused). The OP has controlled for the major and obvious sources of latency and mechanical lag to the user as best as he can. When we do this over multiple races we can see a pattern of inhuman reaction times emerging.

If this was about 1 time then it would be an anomaly. It's the fact that it repeatedly happens time and time again is the accusation. Statistically he should fall further back in that distribution across multiple races. It's always possible that the OP cherrypicked the data which is why a more formalized accusation would be nice to see.

However, even if the OP cherrypicked the data then we are meant to believe that multiple people who have world class Olympian level reaction times all play classic everquest, are very unusually much older than normal for people with these sorts of reaction times, and all play in vanquish. (I assume nobody in riot has been caught doing this but if they are caught everyone should be banned).

I'm skeptical af about that being true. However, I'm not an expert in these things and this is just my understanding of what was presented.

It can't be that he's better than them, nah. NAH.

[titanshub]

Originally Posted by Samoht [You must be logged in to view images. Log in or Register.]

wut.

Isn't that all OP provided here? Some false correlations and hand waving?

The only thing supporting his point of view that what he's saying is true is that he wants it to be true.

I don't have to prove anything false. His argument is fucking bullshit, and the fact that GMs haven't made any changes based on the video indicates to me that they agree.



It can't be that he's better than them, nah. NAH.

OP made a reasonable argument to show something many have suspected for a long time. If he has made an error then the burden is on the people who disagree with the assertion to refute it. It's how science works. Now this isn't real science as its just a youtube video and not a very rigorous presentation of evidence but its enough to convince a LOT of people who play here. Do you care if people think Vanquish race winners are cheating or not? I guess I don't really care but it will stand as evidence until refuted. If you do care I think that refuting it should be relatively simple with a larger randomized dataset of races.

[Profyx]

I support LAN Party racing!

[Trexller]

anyone remember back when like 90% of the blue server was using MQ2, so many people that they couldn't ban us all, so they stripped us of all coin on our accounts and removed 10% of our xp?

Good times.

[getsome]

Originally Posted by Trexller [You must be logged in to view images. Log in or Register.]

anyone remember back when like 90% of the blue server was using MQ2, so many people that they couldn't ban us all, so they stripped us of all coin on our accounts and removed 10% of our xp?

Good times.

It was showeq. It was 600~ accounts. We had one guy in guild who tagged seven different accounts. Most were just making maps of skyfire.

[bobadilx]

Guy tries to use simple data, people either say wow get a life or why can't you calculate the airspeed velocity of an unladen swallow? Either way when all that kael races boil down to is time off the line it's really exciting and skill based stuff.

[Profyx]

Originally Posted by bobadilx [You must be logged in to view images. Log in or Register.]

Either way when all that kael races boil down to is time off the line it's really exciting and skill based stuff.

If you think kael race boils down to time off the line you've never raced it. While I disagree with the OPs conclusions, I wont dismiss people having opinions or questions regarding reaction time off the line. But the truth is that the line time is irrelevant. People who win the race are running better lines than people who lose the race. This is a 91-95 second race, and People win it by more than 50ms every week, sometimes 20x that. There are few people who know the race well enough to compete, period.