1

Topic: Craven's Latest

Craven's latest replays are so obviously created at an artificial speed it is hilarious.
(Personally, I believe there are at least 8 people doing this).

It will be interesting to see what the communities does about it.

Could we PLEASE PLEASE do something to make this more difficult?

@tonnes: as the new official nevertable moderator, could you weigh
in here?


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

2

Re: Craven's Latest

@ protonspring:

I invite you to present me with detailed analysis or evidence to support your accusation. I watched all of Craven's recently submitted replays and did not find them suspicious.

I have noted your tendency to fling about assertions and accusations that are poorly conceived, highly subjective, and not infrequently erroneous. This, in combination with the fact that you are the only confirmed cheater on the history of the table means the following:

Unsubstantiated claims or accusations will not find any traction with me. They are unwelcome and a waste of my time.

Discussion regarding the need for cheat detection measures is ongoing and I believe you have made your opinions and ideas known in that area.

Let me re-iterate: I am always open to consider detailed analysis or actual evidence, but I won't be bothered with blanket accusations or blunt declarations.

3

Re: Craven's Latest

My motivations are aimed at making the nevertable and neverball replays more legitimate.

I've seen and made replays at all kinds of speeds, and I know what it looks like when they aren't made at full speed.  Unfortunately, there is no "evidence" because there is insufficient information in the replay file.  I'm trying to address this in a different thread (could we please do something?)

If everyone thinks these replays look perfectly normal, then there is nothing more I can do other than to be on the record asserting that they weren't done at full speed.  Someday I will be vindicated, but it doesn't appear as though it will happen for quite some time.

4

Re: Craven's Latest

http://table.nevercorner.net/record.php?id=1485

This is a decent example.  If you watch this replay, you'll see lots of little tiny and VERY SHORT ticks to get the ball in the perfect position.  You'll also see similar super short ticks to rotate the camera.  It is simply implausible to do this to exact at full speed.  I'll make an example and post it so you can all see what I'm talking about.

5

Re: Craven's Latest

http://table.nevercorner.net/record.php?id=1507

I made this replay at 1/2 speed to demonstrate the point. 
You'll see the same short ticks to correct the ball position.  Isn't my reaction time god-like? smile

This IS evidence that it can be easily done (and by only adding two characters to the source code)
and no one can tell if this replay was done at full speed or not.  Anyone?
At the same time, there is NO evidence AT ALL that people aren't submitting these kinds of replays.
(there is not sufficient information in the replay file be be able to tell. . hence. . no evidence can be provided).
Shouldn't a rational mind conclude that my position deserves legitimate consideration?

Seriously. . . just think about it.  Perhaps my assertions aren't that poorly conceived.

6

Re: Craven's Latest

protonspring wrote:

I've seen and made replays at all kinds of speeds

The only bogus replays you can legitimately say you've "seen" are your own. Oh, and neverballaddict, oh! and let's not forget monarch...

I know what it looks like when they aren't made at full speed.

This seems to be the substance of your argument. It's not adequate.

Unfortunately, there is no "evidence" because there is insufficient information in the replay file.

This is true. And this is why we don't make accusations based on nothing more than our own assumptions.

If you watch this replay, you'll see lots of little tiny and VERY SHORT ticks to get the ball in the perfect position.

Tiny and "very short" ticks are easily created at normal speed when using the keyboard (do I need to demonstrate this?). I'm not clear what you are talking about when you say "perfect position". Do you mean that he gets the ball to stand still on a moving platform? There are many imperfections in these replays, and I honestly don't see anything "God-like".

I made this replay at 1/2 speed to demonstrate the point.
This IS evidence that it can be easily done.

This fact is patently obvious and does not require evidence.

there is NO evidence AT ALL that people aren't submitting these kinds of replays.

You are assuming that anyone who submits a replay is guilty of cheating until proven innocent. Sour grapes?

Shouldn't a rational mind conclude that my position deserves legitimate consideration?

Are you implying that I am irrational? I have legitimately considered it and I think you have been treated fairly and have had thoughtful responses to your complaints and suggestions. You seem to be lacking in patience and you are quick in jumping to conclusions. I am not disagreeing with you on the potential benefit of replay verification. I am taking issue with inappropriate accusations and divisive, unproductive whining.

Perhaps my assertions aren't that poorly conceived.

Instead of congratulating yourself, I would like to suggest that you refrain from pointing fingers and make a meaningful contribution to this project. As I said to you before, I am glad you like the game and that you want to help make it better. If you can't stay focused on that, but instead feel driven to defame other people because you are convinced they are cheaters like yourself, then I'm afraid you are wasting your time. Work on a viable solution. It won't happen overnight and I won't be doing it for you. Welcome to community development!

Please upload your example replay to this thread as a file attachment. I trust you will agree with me that it doesn't belong on the table.

7

Re: Craven's Latest

I'm so sick of battling this.  Perhaps my perception is completely skewed.  It's so hard to believe that I'm the only one that seems to see this.  Either way. . . I resign.

Here's the file.  It was done at 1/2 speed.
mym can just delete it from the table.

All I'm after is a general agreement that we (the community) should do something to make it more difficult to make these kinds of replays (as attached).  Do we agree here?

Actually, I've put in a fair amount of time trying to add features to neverball.  Parasti seems a bit adverse to most changes, but I understand his motivations, and all-in-all the product is pretty good, so no complaints here
(Although I would like a little less delay for code reviews).

I still don't know the code well, but I'm getting there.  Hopefully, at least the multi-stream code can get committed someday.  It also makes a pretty good base for multi-play.

8

Re: Craven's Latest

The delay is inflated on purpose.  It seems that helping you is only counterproductive and unhealthy for this little community, which is why I don't.

9

Re: Craven's Latest

@parasti:

Hmm.  I didn't think your personal feelings would get in the way, but I guess I expected too much.

Since you are the primary approver of changes, I see no point spending any time trying to contribute to this community.

I won't bother you guys anymore.  I'm sure there are more productive things I can do with my time.

10

Re: Craven's Latest

Here's protonspring's 1/2 speed replay demonstration mentioned above for completeness.

Post's attachments

mym2_20_protonspring_01507.nbr 865.9 kb, 226 downloads since 2010-12-23 

11 (edited by elfakyn 2011-01-03 03:15:21)

Re: Craven's Latest

I know it's a sensitive matter, but just my two cents here. As a full-time keyboard player, I believe that the "short ticks" in Craven's replay are in fact normal consequences of keyboard use. However, protonspring's 1/2 speed example does show a noticeable shortening in the length of those "ticks", as well as an overall acceleration of phenomena.

Even though I have almost nonexistent coding experience, and having seen neverball's source only in passing, my intuition says that the duration of the keyboard-driven "ticks" (i.e. the time in which the floor tilts from zero to maximum when a key is pressed (or vice-versa when one is depressed)) is indeed dependent on the playing speed. Therefore, the duration of these "ticks" should be indicative on the playing speed. Based on these assumptions and my untrained eye, I feel that Craven's replay was done at full speed (implying legitimate), whereas protonspring's half speed replay is evidently forged.

These are extremely subtle signs, and in order to make it easier to understand what I mean, I urge you to compare my legitimate (and weak) mym2-20 replay against protonspring's.

Okay, now to figure out how to upload files using the quick reply.

Edit: Replay attached.

Edit 2: It goes without saying that these signs can only be interpreted if the person is using the keyboard. I'm afraid that mouse users will probably show no signs of this.

Edit 3: Goofing around in hard-9 to compare known "normal" ticks with Craven's. I cannot see any difference in tick speed.

Edit 4: Fixed typo; I should really go to bed now. Edit 5: Okay, in edit 4 I forgot to actually fix the typo (I think it's fixed now); good night.

Post's attachments

hard-09_04_ticks.nbr 645.47 kb, 195 downloads since 2011-01-03 

mym2-20_01.nbr 1.37 mb, 216 downloads since 2011-01-03 

12

Re: Craven's Latest

elfakyn wrote:

I know it's a sensitive matter, but just my two cents here. As a full-time keyboard player, I believe that the "short ticks" in Craven's replay are in fact normal consequences of keyboard use. However, protonspring's 1/2 speed example does show a noticeable shortening in the length of those "ticks", as well as an overall acceleration of phenomena.

Even though I have almost nonexistent coding experience, and having seen neverball's source only in passing, my intuition says that the duration of the keyboard-driven "ticks" (i.e. the time in which the floor tilts from zero to maximum when a key is pressed (or vice-versa when one is depressed)) is indeed dependent on the playing speed. Therefore, the duration of these "ticks" should be indicative on the playing speed. Based on these assumptions and my untrained eye, I feel that Craven's replay was done at full speed (implying legitimate), whereas protonspring's half speed replay is evidently forged.

These are extremely subtle signs, and in order to make it easier to understand what I mean, I urge you to compare my legitimate (and weak) mym2-20 replay against protonspring's.

Okay, now to figure out how to upload files using the quick reply.

Edit: Replay attached.

Edit 2: It goes without saying that these signs can only be interpreted if the person is using the keyboard. I'm afraid that mouse users will probably show no signs of this.

Edit 3: Goofing around in hard-9 to compare known "normal" ticks with Craven's. I cannot see any difference in tick speed.

Edit 4: Fixed typo; I should really go to bed now. Edit 5: Okay, in edit 4 I forgot to actually fix the typo (I think it's fixed now); good night.

Discontinuous press keys cause this situation.

13

Re: Craven's Latest

balius07 wrote:
elfakyn wrote:

I know it's a sensitive matter, but just my two cents here. As a full-time keyboard player, I believe that the "short ticks" in Craven's replay are in fact normal consequences of keyboard use. However, protonspring's 1/2 speed example does show a noticeable shortening in the length of those "ticks", as well as an overall acceleration of phenomena.

Even though I have almost nonexistent coding experience, and having seen neverball's source only in passing, my intuition says that the duration of the keyboard-driven "ticks" (i.e. the time in which the floor tilts from zero to maximum when a key is pressed (or vice-versa when one is depressed)) is indeed dependent on the playing speed. Therefore, the duration of these "ticks" should be indicative on the playing speed. Based on these assumptions and my untrained eye, I feel that Craven's replay was done at full speed (implying legitimate), whereas protonspring's half speed replay is evidently forged.

These are extremely subtle signs, and in order to make it easier to understand what I mean, I urge you to compare my legitimate (and weak) mym2-20 replay against protonspring's.

Okay, now to figure out how to upload files using the quick reply.

Edit: Replay attached.

Edit 2: It goes without saying that these signs can only be interpreted if the person is using the keyboard. I'm afraid that mouse users will probably show no signs of this.

Edit 3: Goofing around in hard-9 to compare known "normal" ticks with Craven's. I cannot see any difference in tick speed.

Edit 4: Fixed typo; I should really go to bed now. Edit 5: Okay, in edit 4 I forgot to actually fix the typo (I think it's fixed now); good night.

Discontinuous press keys cause this situation.

Thanks for the info. I'm now going to make time to analyze the observable differences between normal speed and half speed / quarter speed etc.

14

Re: Craven's Latest

I would like to know if you discover anything.  I can make replays at any speed you like if this this helps you.

My findings led me to the conclusion that it is possible to to a one-time-frame-tick at full speed.  Therefore, it's not possible to "prove" someone was playing at a non-normal speed just because some of their "ticks" were really fast.

15

Re: Craven's Latest

protonspring wrote:

I would like to know if you discover anything.  I can make replays at any speed you like if this this helps you.

My findings led me to the conclusion that it is possible to to a one-time-frame-tick at full speed.  Therefore, it's not possible to "prove" someone was playing at a non-normal speed just because some of their "ticks" were really fast.

Indeed, after some experimentation of my own with different speeds, I haven't found any mismatches between different speed replays (and come to think of it, it's too easy to 'bend time' in neverball). It seems the only options are:

1) implementing some sort of timestamping, which may yield false positives (already discussed)
2) reprogramming the real (i.e. machine time) minimum tilting time (the time it takes to move from a level surface to a maximum angle one) to be independent of playing speed (i.e. if one plays slower they should make extra effort in order not to tilt the platform too fast and therefore uncover themselves (this is because, in my experiments, I have seen that when playing extremely slowly, the platform is also slow to react on a keypress)). I'm an (almost) complete programming n00b, but I believe this would require a relatively large rewrite. (just came up with the idea).
3) ???

16

Re: Craven's Latest

After some time thinking about this, I would recommend a few simple things which I could probably code up without much work:
1.  Increasing the updates per second from 90 to something more. . . 180?
  - this will help discover small ticks and abnormal reaction times
  - would double the size of the replays, but I'm not sure this would cause much problems.
2.  time stamp each command/record
  - not sure, but this may help discover some irregularities.
  - very easy to code
  - makes replay files a little bit bigger, but not much.

I'm not saying this is going to catch all of the cheaters, but I do believe it would be a pretty decent step in the right direction and wouldn't take much work.

just my 2 cents.

17

Re: Craven's Latest

fwiw: It is soooo nice to hear that at least one person agree with at least something I've been harping on for way too long.  So. . thank you elfakyn for that!

18

Re: Craven's Latest

The other I made for this thread was to demonstrate intentional ticks.

I made this one at 1/2 speed and tried to minimize the ticks so it "looks" normal.

No one can prove it wasn't done at full speed and this is my issue.

Post's attachments

easy-slow.nbr 88.71 kb, 203 downloads since 2011-01-21 

19

Re: Craven's Latest

This one is especially bad.

http://table.nevercorner.net/record.php?id=1491

20

Re: Craven's Latest

Sorry for reviving this old thread, but I completely missed this discussion and just have to add my two cents. It's just too funny!

First of all, it is kind of ironic that protonspring is accusing me of cheating, because I'm a guy who tries to beat every game in the vanilla version. I am also an Icy Tower veteran. (Cute little game, but the community died some years back.) There was a huge scandal there about Slowdown cheats before I even joined the community. And afterwards some players still kept using personalized mods to improve their records. I tried this mod too, but got bored after awhile. After a new version of the game came out, I switched back to vanilla, because the mod no longer worked. I still managed to set a few records and to win a few tournaments.

Second, and more on topic, I have recently taken up playing Neverball again. The old records on the Nevertable were all done with the keyboard, but recently I've switched to the mouse. Watching old replays, I also noticed all these little ticks. Mouse replays are much smoother. If you've only tried the mouse, I can therefore understand if a keyboard replay seems strange.

But there is also a gameplay issue. I believe that the mouse is actually superior to the keyboard in many ways. And that's probably also how the developers play the game. But some things are better done with the keyboard, like moving in a straight line or changing the camera mode or angle. I used the keyboard in other games (Icy Tower), so this was the natural way for me to start. But the mouse gives you way more control over the ball (or is it the table?). So, without the mouse it is probably impossible to finish all levels (especially the Nevermania ones) or to reach all record goals.

I guess that what is really important to be good at this game (as with many things) is practice, practice, practice. I still haven't managed to reproduce all stunts that uau, Elviz, and also protonspring had posted on the Nevertable. So there are obviously still better players than me.

A final word: I do not waste hours upon hours to improve a record by a fraction of a second. If you doubt that, you should look up my Enigma highscores (google "enigma nongnu"). But if I make a record while trying to finish the levels and collect all coins, why not share it with other people? Isn't that an encouragement for them to reproduce the result?

21

Re: Craven's Latest

Yeah, it's pretty obvious that most of protonspring's replays are tool-assisted. I generally prefer the mouse over the keyboard for most levels, though I think the best way of controlling is with a joystick. You've basically got the precision of a mouse, but you can easily go back to level by releasing the stick. I found it very easy using the GameCube controller on my Wii port of Neverball (minus the frame rate issues).

22

Re: Craven's Latest

camthesaxman wrote:

I found it very easy using the GameCube controller on my Wii port of Neverball

Hell fffffff yes! why didn’t I think of this.

Leveling out is what makes neverball tricky for me.  I’m a pretty good judge of how far to move the mouse depending on how sharp of a turn I need to make, but I have some stupid issue with re-centering. I end up wobbling back and forth over-correcting myself until I fall off the side.

Unlike a full sized joystick for flying games, the small analog stick on a gamepad springs back to center position almost instantaneously.  xboxdrv simulates mouse and keyboard events in linux flawlessly.  Pretty sure I have a few 360 controllers buried in my closet.

23

Re: Craven's Latest

Joystick or gamepad is a good idea!

However, it can be a disadvantage if the table immediately levels out when you stop pressing a button. Like, if you need to hug a wall to keep the ball from falling off a narrow ledge. Or in high-speed levels where you constantly accelerate. The mouse "remembers" the tilt position if you let go, until you move it in a different direction.

Just for fun:
You can test this nicely in a fenced-in level (e.g. Easy 3). Just move the mouse once in any direction, then let go and watch.

Another thing about gamepads: With mine I would need to switch hands and tilt the table with the left instead of the right. Also, needing two hands for the gamepad, I would be unable to use the keyboard to change the camera.


Warning: count(): Parameter must be an array or an object that implements Countable in /home/customer/www/neverforum.com/public_html/fmpbo/include/parser.php on line 814

24

Re: Craven's Latest

On some levels I have found it helpful to use the mouse and arrow keys, mostly to level the table and to zero the ball speed. Do do this, I first slowly move the mouse to about level the table. Then I very briefly and immediately press one opposing arrow key after doing the same with another. I often have to press two opposite keys in this manner one or two more times before the ball speed is almost zero. I am not so sure if pressing two keys simultaneously would either ignore one key or better zero the ball speed.

Assuming I am not interested in the lazy cam, I have immediate access to all desired controls if there is a middle mouse button to toggle between chase cam and manual cam.

It the it's level set: http://neverforum.com/fmpbo/viewtopic.php?id=2922
6/19/17: a bit of help needed here, plz: http://neverforum.com/fmpbo/viewtopic.php?id=2988
avatar drawn with ANVIL-1000MD (c) and scaled down with GIMP
にほんご べんきょう する。 (i am studying japanese)