forums.marpirc.net

JG: YES, RONCLI's mspacman-fast recording looks VERY suspicious. Just watch the first screen, obviously few show off close calls with ghosts, but too many. Also he seems to getaway later on so f*king often from so many close situations, too often, too easy. Not really wiggles back and forth to buy time in close situations. Not really those occasions when you stop after a wrong move, stunned you got away with it, planning what to do next. Then again he just might have a perfect controller for the game, but I believe this might be the case of some slowdown. If he is not willing to submit with TGmame, hes mspacfast recordings should be DELETED. There is not really any other reason for him not to submit. If you haven't played the game in long time -it takes only couple of games to get your skills back imo.

Negative's gameplay looks ok to me, there are those fine little tricks and maneuvers which a good player should have. Impressive, positive.

The TJT wrote: Negative's gameplay looks ok to me, there are those fine little tricks and maneuvers which a good player should have. Impressive, positive.

If you are talking about his Super Breakout recordings, I have to disagree. Have you ever played Super Breakout more than one game? I don't think it's humanly possible (without slowdown) to react the same way as his recordings demonstrate.

What makes his Super Breakout recordings even more questionable to me is that two of them are submitted less than 1 hour apart from each other..

sbrkout-cav 15 Oct 02 04:21:12
sbrkout-dbl 15 Oct 02 05:04:33

If you were talking about his Ms-Pacman recordings, I can't comment about those as I haven't watched them.

Niinpä, tarkoitin tietenkin negativen Mspacfast inppejä.
Kyllä, olen pelannut superbreakouttia, olisiko jonkun verran alle tonni ennätys.


Retroterveisin TJT

J.K. Voihan v näitä pakkasia, rupee jo ketuttaa

My opinion is that negative1´s recordings should all be deleted. He has admited that he uses pause in his recordings, at least once (Somewhere in one of his long posts and talk about mspacman) Also he has way too many recordings deleted due to slowdown.

About the TG/Alphamame: I agree with using only Alphamame. As long as someone is making a mame32 gui version. (And yes Barry not your front end this time If that will not happen I see no reason for me to change and use the cli version with front end or without. My point is just that if you must use alphamame there must be an option for both mame32 cli, gui (and maybe dos) versions. Just like TG mame 35 has dos and gui right now.

Just my opinion

The TJT wrote:Niinpä, tarkoitin tietenkin negativen Mspacfast inppejä.
Kyllä, olen pelannut superbreakouttia, olisiko jonkun verran alle tonni ennätys.

Joo, epäilinkin että noita tarkoitit, vastasin kun en ollut aivan varma. Sori, jos kirjoitukseni vaikutti vittuilulta, englanniksi on joskus vaikea ilmaista itseään...

The TJT wrote:Retroterveisin TJT

J.K. Voihan v näitä pakkasia, rupee jo ketuttaa

What doesn't kill you makes you stronger Onpahan ainakin hyvä syy pysyä sisätiloissa ja pelata mainioita MAME-pelejä.

I am in complete and utter agreeance with Anders. If there's a MAME32 GUI version of AlphaMame, and it's able to be updated at the same time or a reasonable time after a new release, I will use that and ONLY that... even if it means game restrictions... because there's over 3500 roms... frankly that's alot that CAN be done.

Maybe some kind of hybrid of FanMame and AlphaMame [best of all worlds... stability and every game to date ] but then I know that's nothing but a dream. Till then, Mame32 will be what I use [save for the TG Versions where applicable]

mahlemiut wrote:If the game intends for there to be autofire, then where's the problem? What we check for is autofiring inputs, on such things as joysticks, or whatever. The only difficulty would be those games where autofire is an option. But DIP switches can be checked, and NVRAM must be the default (the game should use default settings when the data in the NVRAM is invalid).

Maybe I'm wrong, but it seem your slowdown detection is based on the length time between 2 hit on the buttons. So, if the game had an inner autofire system, you can't detect this time and detect a possible slowdown. I'm not speak of the autofire detection.

Francois

QRS wrote:
About the TG/Alphamame: I agree with using only Alphamame. As long as someone is making a mame32 gui version. (And yes Barry not your front end this time If that will not happen I see no reason for me to change and use the cli version with front end or without. My point is just that if you must use alphamame there must be an option for both mame32 cli, gui (and maybe dos) versions. Just like TG mame 35 has dos and gui right now.

Yes, I agree too.

Francois

[quote="Francois DanielMaybe I'm wrong, but it seem your slowdown detection is based on the length time between 2 hit on the buttons. So, if the game had an inner autofire system, you can't detect this time and detect a possible slowdown. I'm not speak of the autofire detection.

Francois[/quote]

Not sure i follow you on this. We've already mentioned we cannot detect an "inner autofire" system because inner autofire systems are legal. Also inner autofire would NOT be detected as slow down because the buttons joystick movements would be pressed slower than normal rather than faster than normal when slowdown is used. But for many games slowdown is easily detectable by checking the press length distributions, slowdown recordings reveal a "faster" pattern than regular recording (provided the game is using a known frames per second for comparison which is usually 60.)

Chad wrote: Not sure i follow you on this. We've already mentioned we cannot detect an "inner autofire" system because inner autofire systems are legal. Also inner autofire would NOT be detected as slow down because the buttons joystick movements would be pressed slower than normal rather than faster than normal when slowdown is used. But for many games slowdown is easily detectable by checking the press length distributions, slowdown recordings reveal a "faster" pattern than regular recording (provided the game is using a known frames per second for comparison which is usually 60.)

Exactly, but with the other method (the Willou's one), it's possible to detect also this sort of slowdown.

Francois

It can If you believe in the logic of Willou's method. Willou's method makes a lot of assumptions, that the frame sizes are equivalent (which they are certainly not for analouge games) and that the times are not faked (which has been done) and i'm not sure if the theory behind the math is correct (I could be wrong.) :)

Chad wrote:It can If you believe in the logic of Willou's method. Willou's method makes a lot of assumptions, that the frame sizes are equivalent (which they are certainly not for analouge games) and that the times are not faked (which has been done) and i'm not sure if the theory behind the math is correct (I could be wrong.)

We had tested the methode and it appezrs it work well, but maybe some games fooled the method. We haven't tested all games

Francois

Chad wrote:I'm pretty sure you can't do it that way LN2, because you don't know the exact time length of the recording, you can only tell the exact time of the playback.

hmm, I sort of see your point. It fails if the user had a forced frameskip 0 on...cuz then the true gametime would be a certain #frames regardless if that ran at 100% or 25% speed. The info you have on playing it back etc. would result in just concluding the user had all frames saved as input...no frameskip was used.

This means that method for confirming speed is bogus at best cuz all it does is verify a frameskip of 0 was used....not that the game actually was played and recorded at 100% speed. If you assume they had auto-frameskip on then that would mean the game played at 100% speed but they easily could have just had a forced frameskip 0.

As far as creation and modified date/time of the input file. That should be legit if a new input file was created at the time of recording. However, I'm frequently replacing the previous inp file for a game which results in the new input having the original creation date of the file from previous runs. Even given that, it's too easy to fake a creation or modified date.

Breakout is a really simple game though. I can't see even a 200MHz P2 type system having problems playing that at 100% speed. negative1 did have issues with arkanoid playing at full speed though and he asked for his arkanoid scores to even be removed I think in a previous thread.

The TJT wrote:JG: YES, RONCLI's mspacman-fast recording looks VERY suspicious. Just watch the first screen, obviously few show off close calls with ghosts, but too many. Also he seems to getaway later on so f*king often from so many close situations, too often, too easy. Not really wiggles back and forth to buy time in close situations.

Gimme a break. I had watched one of roncli's mspac-fast recordings way back when and it seemed fine to me. Sure he has mastered the controls, but I didn't see anything that would make me think he wasn't playing at 100% speed. A game like that is all tempo anyway....if his speed wasn't 100% that means it would be varying which overall I think would make the game even harder to play as timings wouldn't be consistent.

I played those a lot a few months ago and have decent scores for it but not where I was going to submit them cuz for regular mspac-fast at least, I want to reach the kill screen. I'm sure if you viewed mine you would see several close-calls also. I know when I was playing it at a friend's house it was freaking him out. He was like how can you control it at all. A couple months later he is nearly matching me in terms of control.

I haven't played those for a couple months now though...but will return soon enough. I would hope Ron's scores aren't suspected. I accept them.

JoustGod wrote: I'm frankly quite tired of giving negative 1 and roncli the benefit of the doubt when both players have repeatedly refused and/or ignored requests to do at least a few recordings on a more secure version of MAME.

Rick(ln2), I admire your pacxpertise, but if they have had trouble with speeds and REFUSE making TGmame recordings....The one question remains, why.

Btw what do you use for controls?