Can you trust videos these days?

[Don]

Recently I find that when I go to Youtube or whatever site that shows some 'awesome playing' I start getting really suspicious whether they're tool assisted or not. After all it gets awfully suspicious how no one ever seems to get hit in a Megaman game or any similar feats of supernatural playing ability.

I'm guessing this all started with the Super Mario Brothers 3 in 8 minute video. It seems like people are now obsessed with artifical perfection, even though no one will really believe you're that awesome. And really if you slow the game down to 1/30th the rate and save every 2 seconds to redo mistakes, it's actually quite easy to pull off superhuman playing. Frankly, I find such videos very boring, especially when it's obvious that the person playing could not have done this normally. The most telltale sign is when you see people unnecessarily dangerous moves for no reason.

Perhaps gaming video has to be certified as real in the future.

[Zeus]

I'd like to think that you can tell when it's assisted or not. There's that SMB1 5 minute video. I watched it and I think it can happen...if you're damned good and FUCKING LUCKY. But it didn't look like it was doctored to me. When I was watching the SMB3 one, that guy was pulling off jumps that looked impossible.

Same with a lot of these speed runs that are everywhere. Watching the SM64 one in 16 minutes looks like it's fine. Same with the Metroid ones or the Ninja Gaiden 3 one in something like 16 or 20 minutes.

At the end of the day, if you've played the game enough, you can probably tell what's been doctored and what hasn't IMO

[Don]

Well the problem is you do have to understand the game pretty well to appreciate whether someone is just good or if they had help. And even fairly *modest* looking replays probably used some features to help (like load/save). Even if you didn't slow down time, just being able to undo arbitrary mistakes without starting over is huge.

[Kupek]

Most of the speed runs I've watched explicitly say "tool assisted." There's not trying to hide it.

[Don]

Most of the speed runs I've watched explicitly say "tool assisted." There's not trying to hide it.

Weird, most of these videos I found, the guy just claim he was that good.

I don't have a problem using tools as long as they tell you it is used.

This isn't limited to just speed runs either.

[Julius Seeker]

I am a fan of speed run competitions. If I were going to have a competition against someone online, there would be some specific thing I would have them do at the beginning of the game; like name Link a certain name, and then allow for only a single sitting to finish the game, then send screen shots to prove it.... from an emulator with no fast forward function.

[Kupek]

Maybe because most of the ones I've watched have been from here: http://tasvideos.org/

[Don]

I wonder if you can have stuff like authentic playing. Not sure who or how will certify this, though.

[Zeus]

With speed runs, there is an amount of faith you have in the person's honesty as to whether or not the speed run was assisted or not. I've come across a few sites which will also break it down as just "finished" or "100% complete" as well. But the whole save/load idea will be based on either honestly or just someone in the community who saw it saying "there's no way you can do this, here's why".

The only "authentic" thing I've seen as of right now is the records at Twin Galaxies. They will confirm the actual score and tell you how they've confirmed it. For instance, if you take a look at the Billy Mitchell record for Donkey Kong (yes, he's beaten it again since the King of Kong film came out), it says that it was witnessed by a referee (http://www.twingalaxies.com/index.aspx? ... 3852&vi=22). They've been recognized by the Guiness Book of World Records now (that's where the records for the Video Games edition came from) so it's as official as they get.

But other than that, it's more of the Wikipedia-like checks and balance system. For this type of stuff, I think that's OK.

[Don]

Well people have been trying to become professional gamers for a while, so having a recognized verification body would indeed by a good idea.

[Flip]

Well people have been trying to become professional gamers for a while, so having a recognized verification body would indeed by a good idea.

http://www.twingalaxies.com/

Have you not caught that King of Kong documentary? They seem to be mostly classic game based, though.

[SineSwiper]

We've had this discussion before. I like TAS videos. Human perfected games are boring because they actually have mistakes, but being able to produce a video like that as an artform takes a lot of time to do BECAUSE there are no mistakes. But, really, I don't care HOW it's produced, as long as it looks cooler.

Check out the link Kupek sent and look for some cool demos there. Trust me, it's entertaining to watch.

[Don]

I think saying perfect games produced in assisted situation is better than flaws games played in normal situation is like saying using steroids to beat world record is better than doing it the legitmate way because normal humans aren't as fast/strong.

I've been watching a bunch of Touhou (crazy side scrolling shooter) videos and there's a guy named Gil in Japan who seems to be widely accepted as the world's best player in this game (and probably the shooter genre in general) and I assume all his records are verified since he's the world record holder in almost everything related to this game. He actually dies in many of his world record setting videos, and there are some manuevers in this game where even the best player in the world cannot pull successfully over 75% of the time.

On the other hand you'll find an endless amount of replays of people who escape from a situation that will kill the best player in the world 25% of the time and they'll do it continously 20 times. Yet none of these guys will come close to his scoring record (2+ billion vs having trouble breaking 1 billion) so it's not like these guys are actually better. It's also pretty obvious these guys don't know where the blind spots on the insane patterns are, which is why they have to resort to crazy dodging skills (that's almost due to being able to slow down the game).

To me a good replay shouldn't make you wonder 'how the heck does he do that?' It doesn't mean you could do the same thing, but you should be able to tell how the guy did it.

A good example comes to mind would be the guy who parried a Super Arts in Street Fighter 3 (Ken vs Chun Li). Obviously that was real since it was captured live, and you don't have to wonder what magic was involved because even a novice can parry some hits. Now obviously no one can reilably parry all 15 hits of a Super Arts (if it was, his opponent would not have tried to cheese him with that), which is why seeing it pull off is amazing. On the other hand if you do something like say Soul Calibur where in training mode and CPU have 100% parry rate, you can get some indefinite parry chains going too. It might look prettier but no one really cares that a computer can parry 100% of the time, and I'd treat a tool that allows you to parry 100% of the time the same way. It's because people can and do fail that makes attempts close to perfection worth watching.

[SineSwiper]

I think saying perfect games produced in assisted situation is better than flaws games played in normal situation is like saying using steroids to beat world record is better than doing it the legitmate way because normal humans aren't as fast/strong.

Dumb argument. This isn't the Olympics. You don't get a gold metal. You don't even win one internets. The special Olympics kid doesn't teach you about arguing on the internet, either.

In the Olympics, seeing somebody run faster or jump higher doesn't really look cool at all. It's just doing what the other guys are doing with more of it. The TAS videos illustrate cool tricks that are either not possible or extremely hard to do consistently.

Again, I consider it art. A good example is the Gradius demo, which has the player take every speed up he can get. Any normal person would crash the ship into a wall at the slightest touch. However, with this TAS video, he's doing neat tricks with the speed boost to entertain you.

You should probably read the Why and How page. It explains a lot of this already.

[Don]

Dumb argument. This isn't the Olympics. You don't get a gold metal. You don't even win one internets. The special Olympics kid doesn't teach you about arguing on the internet, either.

In the Olympics, seeing somebody run faster or jump higher doesn't really look cool at all. It's just doing what the other guys are doing with more of it. The TAS videos illustrate cool tricks that are either not possible or extremely hard to do consistently.

Again, I consider it art. A good example is the Gradius demo, which has the player take every speed up he can get. Any normal person would crash the ship into a wall at the slightest touch. However, with this TAS video, he's doing neat tricks with the speed boost to entertain you.

You should probably read the Why and How page. It explains a lot of this already.

I'm not sure how you can arbitarily decide running faster or jump longer isn't cooler. Everyone was watching Michael Phelps because he swims faster than anybody else out there. By your logic there is nothing interesting about the Japanese guy who parried a Super Arts with no health left and come back and win it is very easy to pull off with even just a mere load feature, and would be trivial with any kind of slow time down feature. So by your defintion a good Street Fighter 3 movie would involve 2 guys parrying each other's hits forever, even though you can just set the training mode and have computer auto parry everything, and they will never mess up unlike a human player, even with tool assists.

I downloaded the movie and you might as well be watching the computer demonstration play where they can make arbitarily difficult manuevers and never screw up. Since you brought up a fast speed thing, this is the fast speed attack by Gil in TH8 http://www.youtube.com/watch?v=ce1Yxf8GDPE with Youmu, the fastest character in the game. You'll notice on the first pattern he was never at the center of the boss's hitbox because he needed that much room to clear the patterns, which results him having to time out the pattern instead of actually beating it, despite having the character with the highest firepower in the game as well.

The replay actually failed the last pattern at the 45s mark because he can't clear the star edges while going at max speed. Ironically this isn't a very hard pattern to clear, but you absolutely have to slow down when going between the circles or you can't clear it, which he did not.

And here's a no vertical (no movement in the vertical direction) by the same guy on one of the hardest patterns in the game http://www.youtube.com/watch?v=9bTnLf1j ... re=related

It's also interesting to note that he actually got spanked very bad on the next pattern which requires a good deal of vertical movement to clear.

It's not just the fact that this guy seems to be perfect when he wants to be, but rather the fact that he does screw up on his replay that gives thes replay a meaning.

[SineSwiper]

Well, we've had this argument last time, so we can just agree to disagree. I was going to post another example that you may have seen:

Dual Megamans

EDIT: Here's another good one with Castlevania:HoD. The speed that Maxim flies through the levels is amazing to watch.

[Zeus]

Well people have been trying to become professional gamers for a while, so having a recognized verification body would indeed by a good idea.

There have been professional gamers for a while now. If you're getting paid 6 figures to go to gaming events and just play, you're a pro

[Don]

Well those stuff are done live so obviously there's no question on authenticity, but even the stuff you play at home should have some validity. For example Gradius 5 (or whatever the one on PS2 was) has a place where you can enter your score verified by a code, and assuming people can't just use inviniciblity you can reasonably say someone who has a score of 3 billion is probably a lot better than someone with a score of 1 million. But if you can't be sure that people will cheat, then the whole system becomes meaningless.