Deprecated: Function create_function() is deprecated in /home/customer/www/neverforum.com/public_html/fmpbo/include/parser.php on line 811

1

Topic: Timer deltas in Nevertable records

I've checked the deltas for the values of timer commands in adjacent updates for all current contest records in the Nevertable.

In all cases they fell between 0.011108s (~90.025 UPS) and 0.011112s (~89.993 UPS), except for the replays listed below where they were in the range from 0.010893s (~91.802 UPS) to 0.010895s (~91.785 UPS). Obviously, the accumulated time differences have an effect on the value of the timer command at the time of entering the goal.

I would appreciate it if someone could crosscheck these results and/or come up with an explanation.

2

Re: Timer deltas in Nevertable records

I can't crosscheck the entire Nevertable database with you, but my ad hoc tests seem to confirm there's a discrepancy in these replays.  I don't believe this can happen unless the code has been tampered with.

3

Re: Timer deltas in Nevertable records

I hope shino won't consider my wget frenzy a DoS attack.  (Sorry!)

I checked all the replays in the contest, the list of flagged replays matches exactly with the above.

4

Re: Timer deltas in Nevertable records

You're just banned for 48 hours. See you !

I'm joking ;-)

5

Re: Timer deltas in Nevertable records

what does this mean guys?

6

Re: Timer deltas in Nevertable records

If it is verified, this means someone is cheating.

7

Re: Timer deltas in Nevertable records

This appears to be the same person who was accusing Elviz of playing at 1/2 speed.
hmmmm...

8 (edited by nue 2010-09-24 07:22:35)

Re: Timer deltas in Nevertable records

It's intriguing that they're all from the same player. What fraction of total replays / of protonspring replays show this behavior? If the code would have been tampered with, would all replays be broken or just some fraction?

9 (edited by parasti 2010-09-24 08:40:20)

Re: Timer deltas in Nevertable records

These are relatively old replays of protonspring that have this peculiarity.  I can only assume his later replays use a different cheat.

BTW, this investigation appears to have been prompted by protonspring accusing Elviz once again.  Funny thing, he describes the detection method, which only triggers on his own replays.

10

Re: Timer deltas in Nevertable records

This is quite serious ... Do the nevertablers have a suggestion on how to resolve this? Protonspring sounds extremely frustrated in the "thread" you link to. Does anyone know how other open source games handle this kind of cheating? (Must be a known problem, no?)

11

Re: Timer deltas in Nevertable records

The flagged contest records cannot be allowed to stand, as the resulting time would have been different (slower) had the timer been advanced at the same rate as the simulation. Unless further tests show otherwise, protonspring's other records may be legitimate, though I no longer have trust in their complete authenticity.

Thanks to parasti for taking a look at the replays.

12 (edited by parasti 2010-09-25 14:09:22)

Re: Timer deltas in Nevertable records

I think the immediate action should be to move all of his replays out of the contest (edit: as well as invalidate any new submissions). Their legitimacy can be determined over time.  In the long term I wouldn't mind accumulating more such methods of detecting possible cheats.

13

Re: Timer deltas in Nevertable records

This is quite a serious matter. As a Nevertable moderator, I will stop validating his replays for now.
I will contact him and ask him to explain himself in this thread.

14

Re: Timer deltas in Nevertable records

I checked those records.  They are all broken.  Back when I was messing
with code to keep up with Elviz.  They should all be removed from the table
as they are invalid.

Subsequent replays are all just like Elviz's (and valid. . right?)

btw. . very interesting things happen with you find the GetTicks() and divide
them all by a common integer.  Let the games begin.

15

Re: Timer deltas in Nevertable records

btw.  It's nice to see some action here.  This web site will be
WAY cooler, when more people other than a few hackers post
replays.

16

Re: Timer deltas in Nevertable records

parasti wrote:

I can't crosscheck the entire Nevertable database with you, but my ad hoc tests seem to confirm there's a discrepancy in these replays.  I don't believe this can happen unless the code has been tampered with.

I'll go check the rest of my replays, and let everyone know if there
are others.

17

Re: Timer deltas in Nevertable records

parasti wrote:

These are relatively old replays of protonspring that have this peculiarity.  I can only assume his later replays use a different cheat.

BTW, this investigation appears to have been prompted by protonspring accusing Elviz once again.  Funny thing, he describes the detection method, which only triggers on his own replays.

I have no interest in posting cheat replays.  If so, I would have eliminated every one of Elviz's records.
It's pretty easy to do, but no fun when cheating.  (Admittedly. . many of them are REALLY good).

There are more than just mine that have this "peculiarity."

18

Re: Timer deltas in Nevertable records

tonesfrommars wrote:

This appears to be the same person who was accusing Elviz of playing at 1/2 speed.
hmmmm...

It's quite easy to cheat by increasing the gravitational constant, increasing the tilt angles, or
just slowing time down, but that wouldn't be any fun.  I didn't intend to post the cheat ones.

Playing at a lower speed is easily done by dividing the GetTicks() in ball/main.c by 2 or another
integer.  This doesn't violate any physical rules of the game and the replays should be just fine.

19

Re: Timer deltas in Nevertable records

nue wrote:

It's intriguing that they're all from the same player. What fraction of total replays / of protonspring replays show this behavior? If the code would have been tampered with, would all replays be broken or just some fraction?

I've only seen this behavior in my broken version, and some of Elviz's replays.

20

Re: Timer deltas in Nevertable records

FYI: I have a replay parser which analyzes time passage, angles, min/max accelleration, player's average reaction time.

Unfortunately, if someone is very careful with the mouse (and moves slowly) there is no way to tell if someone is just running the game at a slower speed (like dividing the GetTicks()), and when a particular individual (who also knows the code really well), just happens to have the reflexes of a praying mantis.

21

Re: Timer deltas in Nevertable records

I reviewed all of my replays.  Here are the ones with the time "bug:"
easy_05_02
easy_05_03
easy_08_02
fwp-12_01
fwp-13_02
medium-02_02
medium-05_04
medium-14_01
mym-14_02
mym2-15_01
mym2-19_01

I have deleted these replays and they can be removed from the table.

22 (edited by parasti 2010-09-27 18:58:54)

Re: Timer deltas in Nevertable records

I can only admire the patience displayed by the people here, including Elviz.  I'm not that patient, so I have to vent my anger on IRC.

21:38 <parasti> my biggest concern really is the fact that he just doesn't get what has been said to him
21:38 <parasti> to him, Elviz is a cheater, and he's just playing catch-up
21:38 <parasti> even though contributions-wise Elviz is on the same level as me, the only difference is that I have a "developer" tag in the forum and he doesn't
21:41 <parasti> hell, even that is an understatement, Elviz has done MORE than I have, as I only work on code, he works on code as well as assets
21:41 <parasti> it's just beyond ridiculous

23

Re: Timer deltas in Nevertable records

protonspring wrote:

There are more than just mine that have this "peculiarity."

protonspring: I would like for you to substantiate this statement, can you provide us with examples of such other replays?

Also, re: your belief that Elviz is producing replays at slower speed, I'm sure that he is willing to be transparent, as obviously we would take cheating on the table seriously no matter who is doing it.

In spite of calling your early replays "broken" or "bugs" you've got to own up to the admission that you were (and for all I know may still be) cheating. That is the core issue here.

If your allegations about others cheating are true, I would be interested in seeing some proof of it. It sounds like you have done some analysis of replays, maybe you could share your data with us instead of just throwing accusations out there.

I agree that the more players participate, the better the contest is.

I am reminded of the film "King of Kong" where the players are sometimes required to make their record scores in public. Also reminds me of pro cycling and doping allegations.

24

Re: Timer deltas in Nevertable records

parasti wrote:

I can only admire the patience displayed by the people here, including Elviz.  I'm not that patient, so I have to vent my anger on IRC.

21:38 <parasti> my biggest concern really is the fact that he just doesn't get what has been said to him
21:38 <parasti> to him, Elviz is a cheater, and he's just playing catch-up
21:38 <parasti> even though contributions-wise Elviz is on the same level as me, the only difference is that I have a "developer" tag in the forum and he doesn't
21:41 <parasti> hell, even that is an understatement, Elviz has done MORE than I have, as I only work on code, he works on code as well as assets
21:41 <parasti> it's just beyond ridiculous

I have no doubt that Elviz has contributed LOTS to neverball (and say more than I ever will), but I don't see how that has any bearing on whether or not his replays were done at full speed or not.

25

Re: Timer deltas in Nevertable records

OK. . come on people. . .

You all seriously believe that all of Elviz's replays were done at full speed
(especially when it is so easy to slow the game down)?
Why doesn't Elviz deny it, here on the forum?

Take his record on TDF12 for example.  His times are achieved by hitting the
ramp on the very edge many times (which is very hard to do even at 1/4 speed).

There are many many other examples which do these kinds of things and throw in sub
frame reaction times.

You all are so gullible that you seriously believe Elviz can do that at full speed
and that consistently?  Do you have any idea how hard that is?

OK. . here's the deal. . Is there no one out there that will agree with me that
many of Elviz's replays are at least suspicious?  If not, then I must be way out
there, and I will concede and go away. . but seriously. . .is there no one?