Loading...

Page 1/1
  • DanielHall
  • avatar of DanielHall
  • Member
  • 2 messages
  • Gender
  • Location
  • Offline  
2019-11-07 16:11:59  
Hello! I was pointed onto this site a while back, and I have some queries/suggestions on some specific rules to when it comes to submission.

I am a Sonic the Hedgehog (Mega Drive) runner and I had initially planned on submitting my times on here, but I couldn't because of the emulation rules. I've came back here because I'm trying to archive as many records as possible for a future video series, but this is still one that confuses me. Basically, I don't exactly see what the stigma is against unofficial emulation. As far as emulation accuracy goes, the 'official' emulation versions are some of the absolute worst in terms of accuracy, and the only noticeable emulation differences for the Mega Drive have absolutely no effect on timed records, and even less on score and rings records. While I am aware that this is quite a broad spectrum as there are nearly 5,000 games on here for different consoles, where the emulators for a lot of newer consoles can be wildly inaccurate, could a happy medium be found in 'acceptable' emulators for each game? A lot of websites usually have proof moderators to handle this for their specific game/console, and they would generally determine which emulators would be suitable for competition. Just remember that there will be a point in time where the hardware will completely degrade and the only way to play some of these games will be through emulation, so it's not a good idea to treat unofficial emulation with contempt.

This would be a good time to segue into my next criticism: There are proof moderation issues, with a multitude of pages having incorrect or flat out fake times that are currently proven, and I'm not even sure if there's a report button. Currently, we use a picture proof system on this site, which while video proof would be a better way to detect false times, it can work. Back when I joined, 1:11 had been listed as the world record on Final Zone. The tool-assisted speedrun (TAS) for this level is 1:11 but it is two frames off of 1:12, and is debatable as to whether it could even be performed on console due to the controller differences. This means that if it is possible, it would have to be a literally-perfect run. I initially thought it was a mistake taking a picture during the fadeout until I noticed the same user having posted 0:06 as an Emerald Hill Zone Act 1 time. This time is as of now a three-way tied world record and I can tell you right now that without a shadow of a doubt, there is no way that this time is possible under normal conditions. In fact, there are only 5 world record times in Sonic 2 of the 19 that are listed, which are actually known to even be possible, and these are listed as proven! While you can completely eliminate any fake times with video proof, picture proof can work as long as you have a proof moderation team that knows what to look for. If you would like me to list which times aren't possible, let me know.

Anyway, thank you for reading and hopefully you'll take these suggestions and criticisms into account.  

  • Djon
  • avatar of Djon
  • Proof Admin
  • 3134 messages
  • Gender
  • Location
  • Offline  
2019-11-07 18:24:28  
Hi,

First, thanks for your participation in the forum about this main topic.

Sure you can help us with your expert knowledge on Sonic, if you can spot TAS do it please.

Emulation was first banned because of TAS possibility but soon VGR will upgrade to VGR V7 with a better system of rules by game and charts only be proved by video to disable this kind of case.
But you're right we are already in a changing world about video games and tomorrow will be other thing too. So in a future point in time, we maybe accept normal use of emulation like NES Mini for example only with video proof to show you play in a normal way. etc

  • GTi
  • avatar of GTi
  • Member
  • 252 messages
  • Gender
  • Location
  • Offline  
2019-11-07 19:39:20  
I am not against emulation. I also believe this will attract more gamers. And VGR should evolve and attract more regular (retro) gamers.
Like MAME arcade games which I also love to play.

But it should be a separate track. Or have emu and real hardware in the same table (just to compare) but the attribution of points/medals/etc should be separated. Playing with a real joystick makes a huge difference then playing with your xbox console pad.

I hope v7 of this site also makes clear rules per game. Like all the settings, no point leeching, no autofire, etc as currently I am in a schmup tournament and it's like I am the only one playing with no autofire. Autofire is allowed in this tournament but I prefer to play it like it was ment to be: button bashing  
Hence my scores  

Strong evidence is required, like with real hardware. Although I know with handhelds it's not that easy/possible or even fun to play. I don't demand videos for that.
It should be first fun to play. I do play at other places than behind my desk where I can record easily.
And of course create some trust. For me that is the most important part.
After some time/years you know each other a bit and you know to whom you can compete.
cheers

  • DanielHall
  • avatar of DanielHall
  • Member
  • 2 messages
  • Gender
  • Location
  • Offline  
2019-11-07 23:34:34  
Hi Djon! I have to be completely honest with you and admit that I wasn't being fair when I was calling them 'fake times'. I first saw these ridiculously low times that were full seconds ahead of the TAS with no video proof, got the wrong impression, when by the looks of things it's an abuse of the 'Time Over' mechanic where getting the last checkpoint and waiting until the time over resets the clock to 0:00, allowing you to respawn with a reset timer and an easy time. As far as I know, the rules don't necessarily restrict these scores so it would be a complete injustice to call these times illegitimate, it's just a bit strange when times outside of continuous play are recorded.  

For instance:

Emerald Hill 0:06 w/ Time Over Glitch: https://www.youtube.com/watch?v=WALSIiALW1I
Emerald Hill 0:18 (SRC WR): https://www.youtube.com/watch?v=MAQeCxT0Fgc

GTi :

But it should be a separate track. Or have emu and real hardware in the same table (just to compare) but the attribution of points/medals/etc should be separated. Playing with a real joystick makes a huge difference then playing with your xbox console pad.


Good point! Just got to be careful nowadays as the joysticks aren't necessarily replaceable lol.

  • Sellyme
  • avatar of Sellyme
  • Member
  • 38 messages
  • Gender
  • Location
  • Offline  
2019-11-08 11:30:38 - Last edit the 2019-11-08 11:31:07  

GTi :


But it should be a separate track. Or have emu and real hardware in the same table (just to compare) but the attribution of points/medals/etc should be separated.


This shouldn't be necessary for most emulated games. Back in 2003 it was an issue, but most emulators are extremely accurate to the original console now. Definitely something that would need to be done for some games/consoles, but for most there's no competitive difference at all.

The controller issue could be solved by simply clarifying the rules to allow any standard controller, even if it's not the native controller for that platform. My interpretation of the rules is that it bans modifying the actual controller itself, not the controller's connection to the console/device. If I plug my keyboard into my phone I'm hardly "modifying" the controller (which is what the rules specify). I'm pretty sure the intent was just to ban macros or turbo buttons, which can be done more explicitly and with much less ambiguity over what you're allowed to use on more open platforms.

  • Smoka
  • avatar of Smoka
  • Proof Admin
  • 176 messages
  • Gender
  • Location
  • Offline  
2019-11-08 12:43:00  
I'm also not against emulation personally, if that is allowed.
We need some info on the game leaderboard to see that the score is made on emulator. Or have a different leaderboard for emulation scores. Also need to have rules what emulator is allowed for each platform and maybe some settings? I saw some youtube video before about speedrunners using some software to make games run on the same speed even if everyone have different hardware. Don't remember what it was. Running accurate emulators most likely need a decent PC, so can't use any potato PC to play.

I think savestates was probably one of the biggest concerns about emulators and most scores are proven with pictures on VGR. So I think full video proof of the run would be needed for all emulator scores.

  • GTi
  • avatar of GTi
  • Member
  • 252 messages
  • Gender
  • Location
  • Offline  
2019-11-08 19:39:26  
[quote=Sellyme][quote=GTi]
Good point. And reconsidering my point. Same tracks are ok. Maybe just a small notification to show on which platform it was played.

  • Sellyme
  • avatar of Sellyme
  • Member
  • 38 messages
  • Gender
  • Location
  • Offline  
2019-11-09 04:33:29  

Smoka :

Running accurate emulators most likely need a decent PC, so can't use any potato PC to play.


Depends on the console you're emulating. PS2 games typically need decent hardware to avoid frame drops because the emulation still isn't entirely fantastic, but GBA games will run on a Pentium II with absolutely no issue.

I think savestates was probably one of the biggest concerns about emulators and most scores are proven with pictures on VGR. So I think full video proof of the run would be needed for all emulator scores.


You could just use an emulator like the RA suite that sets a flag showing whether or not savestates have been used.

  • Smoka
  • avatar of Smoka
  • Proof Admin
  • 176 messages
  • Gender
  • Location
  • Offline  
2019-11-09 08:06:32  

Sellyme :


Depends on the console you're emulating. PS2 games typically need decent hardware to avoid frame drops because the emulation still isn't entirely fantastic, but GBA games will run on a Pentium II with absolutely no issue.


Sure, and it depends how accurate emulator you want to run. There was NES emulator back in the late 90's that could run on 486. But compared to today's emulators and hardware, it most likely wasn't that accurate. I doubt that the emulator running on Pentium II is that accurate either, but maybe I'm wrong.

  • Sellyme
  • avatar of Sellyme
  • Member
  • 38 messages
  • Gender
  • Location
  • Offline  
2019-11-09 13:49:24  

Smoka :


Sure, and it depends how accurate emulator you want to run. There was NES emulator back in the late 90's that could run on 486. But compared to today's emulators and hardware, it most likely wasn't that accurate. I doubt that the emulator running on Pentium II is that accurate either, but maybe I'm wrong.


It's definitely not the one that was around in the 90s, but I have a cycle-accurate NES emulator on my phone that could run at 6000fps and still perform exactly the same as a real console (just 100x quicker). Emulation accuracy has come a long way since the 1990s, and for the older consoles that's almost entirely thanks to software breakthroughs rather than stronger computers.

Basically every major console up to and including fifth-generation ones has either cycle-accurate emulators available, or emulators that are functionally validated on the majority of the games in the console's catalogue with a small number that have documented issues (this is typically to do with rarely-used accessories). Additionally, Dolphin is extremely accurate for most Gamecube games, so you can stretch it into some sixth generation games as well.

Those emulators will run on basically any hardware that people are still using today. Dolphin is by far the most resource-intensive emulator for those generations, and the age of its system requirements are measured in decades - it requires a CPU that supports SSE2, which could theoretically be as old as a 2000 Pentium 4, and the most demanding sysreq (for GC emulation) is a GPU that supports OpenGL 3.0, which is any consumer PC hardware since mid-2008.

With that in mind, it's probably fair to say that for the first five generations + Gamecube, emulation can be considered sufficiently accurate on all hardware by default and only split off or disallowed for games with known flaws (most emus have a detailed list of these), and then the inverse can be true for other consoles where inaccuracies are more common and resource requirements much higher. That would be pretty easy to do with the v7 individual game rules pages and would drastically increase accessibility.

Page 1/1