Out of curiosity Warp, do you regularly go out anywhere on your own volition (not counting things like groceries and other necessities)?
I go every week to an MtG playgroup, where I play for 6-8 hours.
The problem is that it's a nerd hobby, and such hobbies are notoriously lacking of social interaction outside of the hobby itself. I'm not exactly sure why, but they tend to not be the sort of hobby where people become close friends and start interacting with each other outside of the hobby, like eg. inviting the others to unrelated events, or to simply hang out, and so on. Especially not at this age.
Flogging it out gives me 41,584 possible hands with an Ace, Jack and 10.
If that number is correct, and if "44 choose 5" is 1086008, then the probability of getting A, J, 10 is 41584/1086008 = 5198/135751, or about 1 in 26.
If that's correct, then the probability of 3 players getting KQ suited, and the community cards having AJ10 is (11/840601125)*(5198/135751) = 57178/114112443319875, ie. approximately one in 1996 million (ie. about one in 2 billion).
Is this correct?
Creating an additional tier does precisely nothing to address this question, so it's not going to help.
I think it would help.
Generally speaking there has always been a principle of keeping the number of different kind of TASes of a given game to a very moderate amount. (Once you start hitting like 5 or 6 different kinds of TASes, it starts being too much, and the discussion always turns to whether to accept a new one for it or not.)
Also another driving principle is that of obsoletion: A "better" TAS of a given game, using the same category, obsoletes the older one, and thus there always exists only one "official" TAS of that game in that category.
A demo tier would pretty much remove those limitations. You could have as many ACE (or whatever) runs of a particular game without having to worry about those limitations. If there are 20 amazing ACE runs using a particular game, it would be completely alright. It would also allow other types of TASes of that game besides ACEs, and there would be no need to worry about there being "too many" categories. As long as it's something notable and exceptional, it would be ok.
It would need to be a rather distinct tier, kept conceptually separate from actual TASes, so as to not get confused with them.
I just realized that right about this year, give or take, I have been now living alone for longer than with somebody else, during my entire life. I moved out of my family's house when I went to university, and have lived alone since then. Now for longer than my age at that point.
I had a couple of girlfriends at some point. The last one over ten years ago (I don't remember the exact time anymore, but maybe about 15 years ago.) Nobody since then. Nothing. That's like over 20 years living alone, and about 15 years without any sort of relationship whatsoever.
It's hard to explain why that's so. I'm not hideous-looking. I'd say I'm relatively decent-looking (you might have seen my recent photo in the new year's resolution thread). So I don't think it's that.
I sometimes get temperamental online, because the anonymity of online communication removes inhibitions, but I'm nothing like that in real life. I'm really collected in real life. Maybe even too collected. Maybe I just don't communicate all that well, especially to semi-acquaintances. Anyway, whatever the reason, it's just not meant to be.
It takes a really long time, but at some point one just stops caring. There becomes a point when one just accepts it. It's not even depressing anymore. It just is what it is. Life sucks, but it could suck a lot more. I'll just enjoy what I do have, and will just cope with that part of the human brain that craves companionship and contact with other human beings.
Nothing is accepted to the demo tier as it doesn't exist. My impression was that this thread discusses the suggestions, which I have plenty of.
As you might remember, I have been one of the biggest proponents of some kind of demo tier, separate from the actual speedrun tiers. But the discussions about that kind of tiered system has never produced any actual results, even though there have been many good suggestions. Will it ever happen?
My question is still this: Why is it bad to host such runs that our community members made exclusively for our community?
I'm asking what the limits and criteria of acceptance are for that. Do we just publish everything that anybody submits, or are there minimum requirements? Why are some submissions accepted and others rejected? How many "runs" of the same game is it reasonable to host?
The implicit, but unstated, rule motivating Warp's posts seems to be that a playaround can't use ACE, only a demo can.
In my books, once you start executing your own code, the game has ended. And since the game ended before reaching the end, the game wasn't completed.
The game isn't even relevant anymore at that point. It was simply used to inject code into the console. Which game was used for this becomes a moot point (other than by the technical details of how to inject code into the console using that particular game, but this has nothing to do with speedrunning and completing the game in question.)
Why do some ACEs even bother to jump back to the game's code at some point? It's completely moot. If they are pretending like they have actually completed the game, they did so in a completely suboptimal manner (unless the ACE does nothing else besides jumping to the end, which would be a somewhat different discussion.) In fact, it almost feels like they are trying to get published based on a technicality. (And, once again, not that the authors are deliberately trying to circumvent some rules. It's just effectively what's happening, even if it's in good faith.)
The requiremements do exist, they just weren't yet clarified, written down and agreed upon, as it happened to sports games in vault. This run's judgment note makes it clear that we won't accept everything that's ACE. Just wait.
I'm contributing by giving your answers that you ignore.
I was not criticizing that post. I was criticizing the post I quoted.
As for "answers", my original question remains unanswered: How many are too many? A hundred? A thousand? How many?
Is tasvideos.org going to become just a dumping ground for people's demos? Is tasvideos.org going to be to console demos what, for example, shadertoy is to shaders? Whatever people submit, we just accept? Or is there a limit?
It seems to me that completely different standards are applied to ACE demos than to actual speedruns. As you yourself acknowledged, all requirements are thrown out the window.
1) In a four-player game of Texas Hold'em, what are the odds that three of them will get King-Queen suited as their hand cards?
Let's see if I can figure this out, at least partially.
There are 52 cards in the deck, so the probability of the first player getting a king or a queen is 8/52. After that there is only one correspondent card with the same suit, so the probability of getting it is 1/51.
After this there are three kings and queens left, so the second player getting one is 6/50, and then the other 1/49. And then likewise for the third player 4/48 and 1/47.
So if we just had three players, the probability of all of them getting KQ suited would be (8/52)*(1/51)*(6/50)*(1/49)*(4/48)*(1/47) = 192/14658134400 = 1/76344450 (ie. approximately 1 in 76 million.)
Now, the fourth player must not get both the remaining king and queen, as per the spirit of the question (because else the question would have been about four players getting KQ suited.) This is a bit trickier.
Taking this into account changes the probability by a really minuscule amount, but let's do it for completeness sake.
Basically the fourth player can choose two of the remaining 46 cards, except for one particular pair. There are "46 choose 2" possible ways to get two cards, ie. 46!/(2!*(46-2)!) = 1035. This means that the probability of the fourth player getting a pair of cards that's not the remaining KQ is 1034/1035.
Thus the total probability is 1034/(1035*76344450) = 1034/79016505750 = 11/840601125.
Now the community cards need to be 10, J, A, and two other cards. For the spirit of the question to remain valid, the 10, J and A must either be unsuited, or if they are suited it must be the fourth one (ie. not one held by any of the three players). And the two other cards must not make a straight flush with the 10, J and one of the player's pocket cards. In other words, the community cards must not have 9, 10 and J suited, with the same suit as the pocket cards of one of the players.
Uh... this becomes a bit complicated.
(I suppose that the community cards could be allowed to make something better than a straight, as long as it causes a split pot. In other words, full house, four of a kind or a straight flush that's smaller than king high. Also flush, as long as it doesn't make one of the players win. This would simplify the calculations. Maybe.)
If we allow ourselves to simplify things and don't care if one of the players wins, rather than it being a split pot, then it becomes easier. After all, the question is "what are the odds that all get KQ suited, and all get an Ace-high straight?" For simplicity, let's just allow flushes and straight flushes that are a win for one of the players.
However, even after this simplification I'm stuck. Basically I need to calculate all the possible ways of choosing 5 cards from a deck of 44 in such a manner that three of them are A, J and 10, and divide that number by the total number of ways to choose 5 cards from a deck of 44. That second number is 44 choose 5, but I'm stuck with that first number. (Complicating things is that one or two of the aces, jacks or tens may be in the hand of the fourth player...)
I would have accepted that run in a heartbeat and worried about consequences and implications later.
And I worry it would lead to a slippery slope. Where do you draw the line? Do we suddenly start accepting all ACE demos with complete disregard to whether they even resemble a speedrun at all?
How many of these ACE demos would you accept? How many would be too many? Ten? A hundred? A thousand? Where's the limit? Or is there a limit, or any sort of criterion that defines acceptability? Should tasvideos.org just become a dumping ground for all console demos as long as they are injected into the console using ACE techniques?
Consider, for instance, speedrun.com. Every game may have one or more categories, but the requirements for those categories are really strict. They are not arbitrary and acceptance is not really up to people's subjective opinions on whether it's "worthy" or not. If the speedrun adheres to the requirements, and can be verified as being legit, its results are published in the top list for that category.
I have hard time believing that if you tried to submit, let's say, a glitchfest demonstration, they would accept it. It's not a speedrun. It's something else. It doesn't really belong there. It belongs to YouTube, Twitch, or somewhere else.
With this I'm not saying that ACE runs shouldn't be published here. I'm just questioning the criteria by which they are accepted. There seem to be pretty much none.
what makes the demo itself legit, is exactly the fact that it's not an homebrew ROM, but an arbitrary code executed via an official game, that was originally not supposed to do anything of this but
Basically the only difference is the method of insertion into the console. And yes, of course it's a million times more impressive how it's being inserted here, but that has relatively little bearing on the question of whether this is a speedrun at all, or whether it plays the game to completion.
at the same time the demo's content is deeply connected to the original game from the artistic point of view.
But it's not really a speedrun of the game. It doesn't even play the game except for a few seconds. It might perhaps be using assets from the game, ie. data from the game's ROM (although I have no idea to what extent), but it's not playing the game. That's what concerns me. That's why I'm so ambivalent about considering this any kind of TAS of the game.
In principle, tasvideos seldom accepts custom games or hacks, unless they are particularly notorious or famous. Moreover, I don't think tasvideos would ever accept a custom ROM that's not a game at all, but a demo.
It feels like ACE is somehow becoming a way to circumventing that principle. My intention is not to deride or insult the creators of these ACE demos (since I have enormous respect for their technical skills, and the amount of work they put into them), but it almost feels like the particular game that's being used to inject arbitrary code into the console is used more as an excuse to have a demo published on the site than anything else. The game itself isn't relevant. I'm not saying the authors are deliberately and actually thinking like that; I'm just saying that it effectively amounts to pretty much that. It's of course not the intent of the authors to circumvent any rules or princples, or anything of the sort, but in practice it's equivalent to it.
CardsOfTheHeart wrote:
Do we have a clear definition of "playaround" somewhere? I think the lack of consensus on ACE comes from people applying their own definition. I, for one, believe that a "demo" "plays around" with a game's coding, thus it could fall within the playaround guidelines.
I think that's stretching the definition of "play". I think "playaround" means actually playing the game in question. It does not refer to the game's executable code, the console hardware itself.
Anyway, what word you use and how you want to define it doesn't really matter. It's just a word. No principles and criteria should be formed by taking some word and trying to assign new meanings to it. The words and terms themselves are not the point. It's the intent behind them.
what makes it not a playaround, or not comparable to a playaround?
It doesn't actually play the game, but for a few seconds only. The rest is purely a demo.
Let me ask in turn: What makes this different from a custom ROM with a demo programmed in it, other than the method by which it's injected into the console?
Except that it's hardly even a playaround. It's a demo. It may have been injected into the console via a game, but it's still pretty much just a demo.
So, I ask once again, where do you draw the line? How many of these are too many? What would happen if ten more people were to submit similar demos using this particular game? Accept all of them indiscriminately? Accept only some of them? Based on what criteria?
The very fact that such runs are accepted at all, in one form of another, is because they are just movies made with usual TAS tools. This is what we call superplay in a general sense: hand-crafted emulator movie.
So is a 1-second movie that does nothing more than eg. start the game and immediately quit to the main menu, and ends. It's a "hand-crafted emulator movie". It plays the game. It's still unacceptable.
Why is it unacceptable? Because we have some minimum requirements for it to be even considered acceptable.
Minimum requirements that go out the window with ACE runs. Arbitrarily.
Why are you calling it devolving? You think the work dwangoAC and the tas block team showcased at the marathons is degrading the community?
Maybe "devolve" is not the perfect word, but it seems to me that the rules of what constitutes an acceptable submission are being loosened and loosened as time passes. At this point it seems that almost anything is acceptable, as long as it's "cool". There is no "it must break all existing records" nor "it must complete the game" requirements anymore. Heck, it doesn't even need to actually play the game anymore. The game itself has become almost irrelevant, in many ACE cases.
The site might "devolve" (for the lack of a better word) into just a demo compo website, with a twist.
That's why I asked where the limit is. If there is no limit, then that might just as well happen. Everybody will just start coding their old-console demos and inject them into the consoles.
Torje commented about this question in his stream that the TAS would have to use the floor clip inside Deku tree, which is exceedingly difficult to do, even when TASing, and it would essentially be a nightmare to get correctly. I'm not acquainted enough with the technical details to understand why it's so difficult.
I voted Yes after watching the run, and I would not change my vote
I would vote yes because the voting question is asking if I found the run entertaining, and I indeed did. Very entertaining, and amazing.
However, I cannot get rid of the feeling that answering "yes" also implies "I'm 100% ok with this being published". No matter how it's claimed that the poll is not asking that, it nevertheless gives that vibe.
If somebody submitted a video of an SNL sketch, I would probably have to vote "yes" because I would find it entertaining, but it would be completely irrelevant. A video of an SNL sketch doesn't belong here, and thus voting "yes" would feel just outright wrong.
With this I'm not saying that I would vote "no" if the question were "should this be published?" I'm undecided. Ambivalent. And I wanted to discuss the reason.
If you don't like the ACE category, Warp, don't follow it. There are other categories.
Why not: If you want to publish your console demo, do it somewhere more appropriate?
Speedruns with completely arbitrary goals are shunned, and generally not accepted. Why is an exception made for ACE demos?
This is not a speedrun. It's a demo. It's like speedrun.com suddenly started accepting completely arbitrary demo effects made by people using shadertoy. They may be fancy to look at, but rather off-topic at speedrun.com. It would make no sense.
But apparently anything goes here. You don't even need to complete the game. Or play it. Or anything. There's pretty much nothing left that could be called a "speedrun".
Will tasvideos.org just devolve into a publication site for people's console demos?
I'm actually beginning to think that ACE is now pretty much complete as a concept, forming a sort of dead end in the space of what you can do with games.
I would like to somewhat join this raining on everybody's parade, and I'm sorry to present a dissenting opinion, but anyway.
This demonstration is, without any doubt, extremely impressive from a technical point of view. It's the kind of demonstration that would get laughs, cheers and a standing ovation at GDQ.
But that is the point: It's a demonstration. This is demo coding. It's not a speedrun. Nor a superplay. Nor any other s-starting word you might want to apply as a backronym to "TAS".
There's nothing wrong with demo coding. Heck, I was a huge fan of demo coding in the 90's and early 2000's, and participated a bit in the demo scene myself. Kudos to the m4d l33t skillz of the author! And that was honest praise. I'm not trying to be funny or condescending. I am impressed, and this deserves praise.
But, once again, this is demo coding, not speedrunning. It's essentially modding a demo into the console. Sure, it happens via the control inputs alone, which is impressive from a technical point of view, but when we get down to it, it's just that: Impressive demo coding. Not really a speedrun.
We already know that you can take over many of the existing consoles and inject custom code into them. That has been amply demonstrated.
What I see happening here is demo programming being published in the guise of it being a "tool-assisted speedrun". Where do we draw the line? How many such submissions need to be submitted before we start putting limits? Ten? A hundred? A thousand? How many?
Once again, the demo is impressive. I'm just asking if it's "off topic" here, and how much is too much (or, rather, how many are too many).
I honestly apologize for being so lazy (and not checking the rest of this really long thread), but I was curious to know why the current OoT TAS is so old, and there isn't one using the newest known glitches. Could somebody write the short version of it?
I'll answer this later if no one else does, but I have some questions. For (1), does it matter what the fourth hand is? Can it be anything, including the last King-Queen suited?
Good question. Since in this situation the fourth player also getting a KQ would have been rather extraordinary, the question would have been about all four players getting it, rather than just three of them. I suppose that in the spirit of calculating the odds of this particular situation, the fourth player should not have KQ in hand. (Can have King or Queen, though.)
For (2), do we ignore the fourth hand completely, as in, we can assume that it has been folded and plays no further part in the deal?
Yes. Also in the spirit of the calculating the odds for the situation.
Also, as a question of interest, in the deal that you mentioned (three players with King-Queen suited, 10, J, A and two others on the table), do you remember if all three players have nut hands? That is, from their point of view, they all had a hand which from their point of view could never lose no matter what anybody else had? If so, that would have been very interesting. The usual result being that all three players go all-in and a huge amount of excitement is generated, only to end in a split pot anyway. :D
I happen to have a screenshot of the exact situation:
