(user is banned)
Don't think! feeeeeal
So for some reason demos aren't required for HL runs anymore and new runs are being hosted without public demos. Also Nate believes they haven't even been used to verify for some runs.
I'm proposing that demos be required for all PC games in which it takes little effort on the runner (ie. doesn't lag them to shit) and little effort for someone to use a demo to verify authenticity (ie. it works properly) AND ESPECIALLY FOR SS runs.
I think having one authentic working demo for a whole run is the best way to verify it as a SS and that the runner didn't use in-game methods for cheating like slow-down (host_framerate in HL). Otherwise I think a webcam video should be required to prove SS on games with quick save which could be segmented and forged video made easily, showing keystrokes and mouse movement well enough you can tell the runner isn't acting.
Now I particularly most care about this for HL speedruns. So far every HL speedrun has had demo proof (or at least all mine did, and so did groobo's) There is one HL speedrun however which no one ever got to see a demo for and nate has told me through PM that there probably never was one. I am referring to 'Best Hazard course time with scripts: 0:02:54 by Daniel 'rayvex' Babik on 2008-10-24.', the only speedrun to beat one of my speedruns. Now it's very possible the run is totally legit, those hazard course runs are by far my worst speedruns and I only made them because I didn't like someone else's name on the HL page when all the did was make a terrible 3 minute run, but beating it by 11 seconds if fairly impressive. Also Daniel could have segmented his run and never thought anything of it because for some reason it never said my hazard course runs were SS. I think Daniel is an honest person and I give him the benefit of the doubt. The problem here was not created by Daniel but is merely a result of poor management at SDA.
For this particular video Daniel sumbitted, there are three extremely easy ways he COULD have cheated (not saying he did) in the run and made the very video that is hosted on SDA. First would be to use a different mod all together; there is a version if hlsp bunny hosted on the creator's ftp which has easy quake-style jumping for perfect jumps. I could never get Mike to approve this. This I think is most likely because for some reason the hlsp bunny site went down and if Daniel did a Google search to try and find it, this version would be the mostly likely one he’d find. Second he could have played the game with a slower host_framerate, then played the demo back at normal speed to make the video. Again, this is extremely easy to do. Third he could have segmented his run and then put it together into one video. This is very easy to do with HL and you could make the change between segments effectively undetectable in a video.
Needless to say I could host a forged "video" beating that speedrun by tomorrow morning and it'd look even more legit than his.
Now I know there's always ways to cheat and you can't always take every means of verification possible. However I think how easy it is to cheat, and how easy it is to require a means of verification which prevents it, or makes it many times harder, is very relevant to this debate. In the case of HL speedruns, if only a video is required, there's three very easy ways to cheat described above. But fortunately for HL there is also a very easy way to help verify that speedruns did not use these means to cheat. That is demos of course. Demos don't lag the runner, are easy to make and submit, run perfectly and are easy to verify with; hell that's all quake runners even submit, then someone else makes the videos if they want.
So if there's no demo verification for this HL speedrun I would kindly ask that it be removed if the runner can't provide it. I would also request that in the future that all HL speedruns require demos and demos hosted publicly. I really don't think it's fair to have a speedrun obsolete another without being held to the same verification standards as the run it obsoletes.
Since someone said I should point out where the video is apocryphal, I decided to. Here is my case:
Exhibit one, the runner does not start in the area where the map first loads, when the run starts he is already going through the double doors which are closed when you start, yet in his run he going through them as they are open. He most likely just made his video start too late, regardless this lowers the total time of the video and does not follow the standards all the other speedruns before it used.
Exhibit two, at 0:58 the runner enters the first new level load. Notice how the video has a large jump, he teleports immensely. If the video was made with one demo this wouldn’t happen, compare with my single segment runs, or hell even my segmented runs look much better than this. This is a strong indication of discontinuity.
Exhibit three, at 1:50 the runner enters the 3rd map, again a large teleport is noticed. This is much more noticeable than even in my segmented runs. Again, this is a strong indication of discontinuity.
Exhibit four, at 1:50 when the runners enters the 3rd map, the white text displaying “ACTIVATE WHEELS AND DIALS BY HOLDING DOWN THE USE KEY HOLD USE KEY UNTIL BRIDGE IS IN POSISTION” disappears for an instant then comes back, this is very noticeable. Compare to my hazard course run; the text stays visible through-out the whole level transition.
Exhibit four, at 2:06 the runner enters the 4th map, again a large teleport is noticed. Again, this is an indication of discontinuity.
Exhibit five, at 2:06 the white text displaying “PRESS FLASHLIGHT KEY AGAIN TO TURN FLASHLIGHT OFF” disappears for an instant then comes back again. Compare
this to my hazard course run; the text stays visible the whole time.
Exhibit six, at 2:06 you can hear a large discontinuity and pause in the female vox wav being played. Compare with my hazard course run; the discontinuity is much
smaller and much less noticeable.
Exhibit seven, at 2:16 when the runner enters the 5th map the wav file which says “morphine administered” stops at morphine and then nothing else is heard. Compare with my hazard course run where you hear a small pause after morphine, then hear “ministered”
That's seven indications of discontinuity. How could such a video possibly obsolete a proven SS run?
Also, I find it out right comical that Mike could so adamantly disapprove of a mod yet then do ABSOLUTELY NOTHING to make sure it isn't used in hosted videos approved by SDA and himself.
I'm proposing that demos be required for all PC games in which it takes little effort on the runner (ie. doesn't lag them to shit) and little effort for someone to use a demo to verify authenticity (ie. it works properly) AND ESPECIALLY FOR SS runs.
I think having one authentic working demo for a whole run is the best way to verify it as a SS and that the runner didn't use in-game methods for cheating like slow-down (host_framerate in HL). Otherwise I think a webcam video should be required to prove SS on games with quick save which could be segmented and forged video made easily, showing keystrokes and mouse movement well enough you can tell the runner isn't acting.
Now I particularly most care about this for HL speedruns. So far every HL speedrun has had demo proof (or at least all mine did, and so did groobo's) There is one HL speedrun however which no one ever got to see a demo for and nate has told me through PM that there probably never was one. I am referring to 'Best Hazard course time with scripts: 0:02:54 by Daniel 'rayvex' Babik on 2008-10-24.', the only speedrun to beat one of my speedruns. Now it's very possible the run is totally legit, those hazard course runs are by far my worst speedruns and I only made them because I didn't like someone else's name on the HL page when all the did was make a terrible 3 minute run, but beating it by 11 seconds if fairly impressive. Also Daniel could have segmented his run and never thought anything of it because for some reason it never said my hazard course runs were SS. I think Daniel is an honest person and I give him the benefit of the doubt. The problem here was not created by Daniel but is merely a result of poor management at SDA.
For this particular video Daniel sumbitted, there are three extremely easy ways he COULD have cheated (not saying he did) in the run and made the very video that is hosted on SDA. First would be to use a different mod all together; there is a version if hlsp bunny hosted on the creator's ftp which has easy quake-style jumping for perfect jumps. I could never get Mike to approve this. This I think is most likely because for some reason the hlsp bunny site went down and if Daniel did a Google search to try and find it, this version would be the mostly likely one he’d find. Second he could have played the game with a slower host_framerate, then played the demo back at normal speed to make the video. Again, this is extremely easy to do. Third he could have segmented his run and then put it together into one video. This is very easy to do with HL and you could make the change between segments effectively undetectable in a video.
Needless to say I could host a forged "video" beating that speedrun by tomorrow morning and it'd look even more legit than his.
Now I know there's always ways to cheat and you can't always take every means of verification possible. However I think how easy it is to cheat, and how easy it is to require a means of verification which prevents it, or makes it many times harder, is very relevant to this debate. In the case of HL speedruns, if only a video is required, there's three very easy ways to cheat described above. But fortunately for HL there is also a very easy way to help verify that speedruns did not use these means to cheat. That is demos of course. Demos don't lag the runner, are easy to make and submit, run perfectly and are easy to verify with; hell that's all quake runners even submit, then someone else makes the videos if they want.
So if there's no demo verification for this HL speedrun I would kindly ask that it be removed if the runner can't provide it. I would also request that in the future that all HL speedruns require demos and demos hosted publicly. I really don't think it's fair to have a speedrun obsolete another without being held to the same verification standards as the run it obsoletes.
Since someone said I should point out where the video is apocryphal, I decided to. Here is my case:
Exhibit one, the runner does not start in the area where the map first loads, when the run starts he is already going through the double doors which are closed when you start, yet in his run he going through them as they are open. He most likely just made his video start too late, regardless this lowers the total time of the video and does not follow the standards all the other speedruns before it used.
Exhibit two, at 0:58 the runner enters the first new level load. Notice how the video has a large jump, he teleports immensely. If the video was made with one demo this wouldn’t happen, compare with my single segment runs, or hell even my segmented runs look much better than this. This is a strong indication of discontinuity.
Exhibit three, at 1:50 the runner enters the 3rd map, again a large teleport is noticed. This is much more noticeable than even in my segmented runs. Again, this is a strong indication of discontinuity.
Exhibit four, at 1:50 when the runners enters the 3rd map, the white text displaying “ACTIVATE WHEELS AND DIALS BY HOLDING DOWN THE USE KEY HOLD USE KEY UNTIL BRIDGE IS IN POSISTION” disappears for an instant then comes back, this is very noticeable. Compare to my hazard course run; the text stays visible through-out the whole level transition.
Exhibit four, at 2:06 the runner enters the 4th map, again a large teleport is noticed. Again, this is an indication of discontinuity.
Exhibit five, at 2:06 the white text displaying “PRESS FLASHLIGHT KEY AGAIN TO TURN FLASHLIGHT OFF” disappears for an instant then comes back again. Compare
this to my hazard course run; the text stays visible the whole time.
Exhibit six, at 2:06 you can hear a large discontinuity and pause in the female vox wav being played. Compare with my hazard course run; the discontinuity is much
smaller and much less noticeable.
Exhibit seven, at 2:16 when the runner enters the 5th map the wav file which says “morphine administered” stops at morphine and then nothing else is heard. Compare with my hazard course run where you hear a small pause after morphine, then hear “ministered”
That's seven indications of discontinuity. How could such a video possibly obsolete a proven SS run?
Also, I find it out right comical that Mike could so adamantly disapprove of a mod yet then do ABSOLUTELY NOTHING to make sure it isn't used in hosted videos approved by SDA and himself.
Thread title: