Casual controversy

[nolem]

i will still use jsave melon

 

Also, JumpBOT is pretty sweet if any of you guys haven't seen it before and it would be awesome to see where it can go with this.

[Torii]

Despite completely disagreeing with passing off the TAS as legit runs, I'm genuinely interested to see how this will benefit the making of TF2 jumping videos. How exactly does it work?

[nolem]

I have this picture https://i.imgur.com/blynWpZ.jpg but other than that I am clueless and I'm really interested in knowing how it works

[RNC1839]

Hi, I am an independent plugin developer collaborating with JoinedSenses' ++RJump East Coast jump servers that has been creating additions to the server's plugins repertoire.

My recent addition has been a JumpBOT rocket- and sticky- jump recording/playback plugin for tutorial purposes.  The description of the Casual's plugin surrounding this controversy appears to run in a similar nature as mine, and I would like to shed some light around positive uses of such plugins, especially for beginner jumpers so as to not let controversy drag bot plugins into becoming a taboo topic.

I currently have approximately 2.5k separate jump recordings, including alternates, spanning 116 maps.  I will post more about this in the plugins section and set the system up for demonstration on the servers shortly.

Yeah joined told me about this. It seems really cool and from what I can tell it's mainly going to be used as a tutorial kinda thing and be server side. I'm pretty sure that Casual's is client side so that's WAY more sketchy in terms of what can be lied about. I love the idea of this plugin. And it has the word BOT in the name so there is no thinking it's legit.

[Exile]

With great power comes great circlejerks

[Oatmeal]

Can someone please change the title of that ROTW 100 to include TAS somewhere? Also a link to this thread in the description wouldn't hurt.

Same goes for that void run. People shouldn't still be deceived into believing this is legit.

[RNC1839]

Can someone please change the title of that ROTW 100 to include TAS somewhere? Also a link to this thread in the description wouldn't hurt.

IMO the videos (including ultra and aperture) should be taken down entirely. But that's all on Afterglow (or whoever the hell runs the channel now).

[EXC0]

Just want to say I called it day 1, but I feel like this could make some really neat things, being able to see what is "possible" but not realistically for individual jumps. Seeing every jump get bounced, and making frame perfect edgeshots on every platform to get unimaginable times would be pretty cool too.

Another use could be to make really in depth tutorials for jump optimization, seeing what perfect technique is and comparing it to your own and others. Could easily explain beginner mistakes and they can see for themselves what is happening and why they arent doing as well as they should.

Good things can come from this as long as its applied properly

[Squishy]

Ok, so I'd like to point out a few things.

First off, I'm glad to see how accepting people are of the fact that these runs were TAS.  I think most people have handled it very well, especially Casual himself.  This is exactly how Raiin should have handled his runs:  admitted the truth and then done what he could to right his wrongs.

Second, this is really a revolution in jumping.  There are so many uses for Casual's tool that can be used for good.

1. Jump Map Showcases

No longer will you have to spend hours trying to get your cuts to look good.  With this tool, you can put together a flawless run of a map, improving showcases and videos of jump maps.  Since the tool will make it clear that it is a TAS run, people will be unable to pass off runs as speedruns done legitimately.

2. Theoretical perfect runs

Everything Casual did in all of his runs was possible in normal TF2.  He just didn't do them all in one run.  This opens up new possibilities for non-TAS perfect runs.  Someone could get a perfect or near-perfect run of a map and use it as a standard to aim for without using the tool since all the runs are possible without assistance.  As far as I know, nobody has flawless jump_soar, let alone gotten a perfect run of it.  With Casual's tool, someone could get a TAS run of the map and then use it a benchmark for their normal runs and something to aim for.

3. TAS speedrunning

Maps could be created with TAS in mind, leaving room open for ridiculous strats that would otherwise be near-impossible.  And since everyone would have the tool, everyone would be on even ground and it would come down to creativity, not consistency and execution.  What always amazed me the most with Casual's runs was the creativity of the strats, not the execution of the strats.  People would never have tried some of the things he did in his runs because they were too risky to do during a good run.  However, since everyone could use the tool in a TAS run, new possibilities would be open and people would be limited only by their creativity and the game itself, not necessarily how consistently then can grind out a jump.

There are even more possibilities than the ones listed here, such as co-op TAS running in segments and other things that would have previously been impossible, but I'll save those for later.  For now, I'd just like to thank Casual for revolutionizing jumping.

Good points Obsidiian, especially 2. and 3. Also, I have to admit, I really believed in these runs.

Edit: Removing my thank you for the honesty. Looking through more posts/info I realize it was pretty messed up not to explain or talk about this earlier, and in another way than this.
Also, that alle knew it all along. Why hide it if you knew, that is just stupid and it should have been obvious that it would some day lead to a big drama that we are having now.

Edit 2: Looks like cody didn't really know about it for much longer than anyone else, glad to hear that. I hope i am wrong about alle knowing it all along aswell.

Why couldn't the people who knew, including Casual just told the community about the plugin once it was made. Just say: "hey, I made this, look what is possible with it. is anyone interested in this, and how can we develop this further?"

To me it's just so stupid to hide it for so long. There were no need to fake runs to prove a point in my opinion. Should have told about it when you made it.

[Casual]

I took some time to address some of the questions raised in this thread. If there's any more which I haven't answered or if you have any more questions feel free to ask.

he curves his aim perfectly on that last jump too, which was weird.

's just me moving my mouse trying to make a nice strafe around that corner

This run is fake imo. Unless Casual, himself, can explain all of my notes.

alias +rjump "+attack 107; +jump 65; +duck 83"
alias -rjump "-attack 107; -jump 65; -duck 83"
You cannot rely on the key number to differentiate scripted from actual keypresses.

The recrouch is because I use mouse2 to rjump, then switch to shift (meaning release mouse2 -> press shift) for further walljumps

left/right movement switch is airstrafe, not that hard to have them near each other

@@moves to the left without any console commands

It's a bot, it does not skip keypresses.

@@4 wallshots no a or d

+moveright is active the whole time, probably not ideal for walljumps but hey, I made it

@@5050 really fast 1 tick turn to the right

One tick commands aren't impossible to do occasionally, and in this case I was tapping the keys to try to align the rockets as best as I could, it's a tas after all.

@@relatively sloppy mouse movement

I agree.

The mouse movements are fucking weird

But why the alt account?

Because they want to check your account for being alts of banned people, for this you need to put your friendslist to public. If you are ever forced to do this maybe you should reevaluate your policies... So I made an alt because private account 4 lyfe.

It looks like every speedshot he hit was intended.

Yes, even more, even though I do other stuff in between, it all the speedshots might as well have been sequential. What I'm saying is that the 'chance' to hit X speedshots in row is the same as doing X intended speedshots while doing other stuff in between.

We also must understand that Casual has cheated runs before.

This run is done using the same plugin, nothing was cheated or faked. However the plugin had a bug that would make the character lag one tick behind, creating the occasional wrong hitmarker.

Maybe he's going to do B2 maps during the competition, and is just showing everybody his "skills" so it's less suspicious.

Nice theory but not true.

Looking at his aliases, it must be his own site.

It is my website, note the lack of anything ever being sold. (VAC too strong)

Either way, it still looks awesome.

think of all the contests that could be cheated

I'd like to point out I was never given winnings for aperture even though I 'won'.

So I got invited to join a IP casual put up....in STV

god dammit, default port on stv that's only meant to record stuff wrecked that. Ah well, I've put up a password on the stv now.

HE DID A FLAWLESS MAP WITH A SHOTGUN.

I mean if CasualHacks.net didn't give it away in the first place..lol

owning a website with a fishy name doesn't imply anything.

I'm not sure he could perfect a run in 1 week.

I spent a lot of time on these speedruns (except ultra, spent less on that). For skylight I spent roughly 20 hours running the map twice (once for 'practice' and testing out strats, for which I got a 1:58.X time and a 2nd fixing all mistakes and improving) spread over 2-3 weeks. It certainly wasn't easy.

I still feel like we aren't getting the truth.

There's nothing more to it, sorry.

How exactly does it work?

Records: all key presses (after they've been translated to commands), viewangles, current position, current velocity, whether you're on ground for every tick. During playback, just overwrite with recorded stuff.
Except, source engine isn't 100% deterministic. I think the main problem lies in the fact that you are recording keypresses at a certain time (since map was loaded) and small variations cause the exact same input at a later/earlier time to give slightly different results. This causes annoying desycs after ~20seconds.
To fix this I put a hook on where the server processes the player's input and enforce (teleport) the earlier recorded positions & velocities. Since these corrections are tiny they are not noticable. I don't know if this disqualifies the whole process for actual TAS. I have some other ideas I'll explore to get things properly played back.

I have this picture https://i.imgur.com/blynWpZ.jpg

The green line shows the 'track', the numbers show the time in seconds every time you hit or leave the ground. This allows me to compare different runs and see if I'm making an improvement or losing time.

Same goes for that void run.

Void is on a completely different channel with different rules, the point of that video was showing the funny reactions of people to the run. I'd say the void run deserves its own proper video though...

IMO the videos (including ultra and aperture) should be taken down entirely.

Doing so will lose the rating/comments/history. Idk if that's the right solution but I'm not exactly unbiased. In any case I've got them downloaded for safekeeping >.>

Also, I have to admit, I really believed in these runs

This is one I haven't addressed yet. The hardest part for me with this deal (where I'd submit the demos as single segment speedruns) was the people that supported me. I'm terribly sorry to disappoint but hope something great can come from this.

[Dr. Heinz]

@casual
on the ultra run, you fucked some shit up at the end, if I remember correctly, there appears a -jump just after you finish the run, even though there was no +jump before. Also, suddenly there is a nullcancel script. And your keys are switching around a bunch. Like, foward goes from W to E and crouch from ctrl to shift, stuff like that. That all happend when the run ended.

I never really carefully looked at the aperture run or this skylight run (never had the demo of void), but I am sure there are minor mistakes there as well

[Oatmeal]

if you have any more questions feel free to ask.

There are a few that come to mind. 

Have you given your TAS plugin to anyone else before today?
Would you have released more TAS runs and passed them off as legit runs, if people wouldn't have noticed it and complained about it here?
You said that you are going to release the plugin (with ways to make sure that people don't abuse it for passing them off as non-assisted runs). On the other hand do you have a history with (only the intention of?) selling tf2 hacks on your website casualhacks.net. How do we know that you aren't going to sell your vanilla plugin to people who want to become popular with fake runs?

[flo]

Trust a hacker, what could go wrong.

[Combustion]

Its scary when you are inadequate. But when you become powerful beyond measure to get that close to passing off run as legit. Its even more scary. When people find a way to cheat the system, they do it. I still don't understand why u would record these runs and try to pass them off as legit. Now people can see how believable this can be. Probably a lot more casuals will pop up in the future. Who is to say people won't create a plugin better than the current one you use and start passing runs off as legit. Its unfair to ppl like guyyst who speedrun a lot and get great times then suddenly a person with TAS comes and destroys it. So many people are disappointed about this especially those that thought you were actually that good. 

[Combustion]

3. TAS speedrunning

Maps could be created with TAS in mind, leaving room open for ridiculous strats that would otherwise be near-impossible.  And since everyone would have the tool, everyone would be on even ground and it would come down to creativity, not consistency and execution.  What always amazed me the most with Casual's runs was the creativity of the strats, not the execution of the strats.  People would never have tried some of the things he did in his runs because they were too risky to do during a good run.  However, since everyone could use the tool in a TAS run, new possibilities would be open and people would be limited only by their creativity and the game itself, not necessarily how consistently then can grind out a jump.

There are even more possibilities than the ones listed here, such as co-op TAS running in segments and other things that would have previously been impossible, but I'll save those for later.  For now, I'd just like to thank Casual for revolutionizing jumping.

Good points Obsidiian, especially 2. and 3. Also, I have to admit, I really believed in these runs, but as others have said: Thanks for being honest in the end Casual.

Firstly. The whole point of speedrunning a map is to see how fast you can complete it. How consistent you are to complete the map. Not how lucky you are. If you throw consistency out of the equation, all the hard work to gain speedshot consistency, sync consistency is all thrown away. And also, TAS running will also mean how lucky you are, nothing else. Creativity will hit a ceiling before you know it and it just comes down to luck after that. Also, i didn't see any creative strats in casuals runs just hard as fuck strats. Creativity is the reason why JOTW exists is it not? I don't think it will revolutionize jumping. If TAS its not contained properly, the exact opposite would happen.