Does Online Multiplayer Always Make Sense?

[Rastien]

I play alot of super street fighter 4 i wouldn't claim i'm that good at it but i can pull of various combos and the like.

The AI in ssf4 works solely off counters as far as i can tell, the difficulty just increases the rate at which they counter and don't take your shit on the chin.

The real difference between playing against a player and an AI is the fact you really can get away with some very cheap shit. Playing against a person if they are playing very defensively keeping back not attacking you can 'Taunt' them make them angry so they get reckless and come flying at you leaving themselves wide open.

You could sit spamming taunt and dicking around the screen with Dan Hibiki 'Wahoo jumps' and the AI won't give a shit, it will play exactly the same. You do this to player and after the 6th time you dodged a shoryuken by gliding over it yelling 'WAHOOOOO' i can assure you they will be mad as shit and will come at you like bear that you just kicked in the nuts.

Another example is against an AI if your both crouching looking at each other, you dash forward for a cheeky quick throw the AI will nearly ALWAYS counter it. A player how ever may not expect this as it's a very nutsy thing to do and if you do it at an unexpected moment you will catch them off gaurd. You will never catch the AI off guard if you hit the AI it's because it let you or doesn't have an avaible counter.

AIs in fighting games are horribly hard to make properly, because the result is always one of the following;

1. The AI has one or more inherent flaws that can be easily abused, thus all you need is to abuse that precise weakness
2. The AI uses every move perfectly and is unbeatable. The exception is if there's a parry option, in which case you can usually figure out the pattern of the AI and beat it every time by parrying all its attacks, turning it into scenario 1.
3. The AI reads your inputs. This is done way too often, and it's annoying as fuck, as the AI is unbeatable.

You need that human factor to add guessing games and reads to the equation.

This the AI works off input and has no time for 'the guessing' game

 

[Voltano]

You know, your point might have made more sense if people didn't have microphones and a friends list.

If you agree to play with a friend who lives at the other end of the country, you both put the same game in and you play together, and you talk the same way you would if they were sitting next to you. When playing with silent, anonymous randoms then sure, they could easily be replaced by bots, but not every interaction occurs that way.

I agree on that, but that argument also implies a bad cup of coffee can be good if you add the right cream to it.

What I suspect Yahtzee is talking about is using these Online games as a way to socialize with strangers and build rapport through the game - which seems like where they are fumbling so badly at. Games like "Resident Evil 6" or "Anarchy Reigns" gives no way of communicating with online, anonymous people, and the games require the player to focus more on the challenges presented towards them rather than learning if there is anyone that has a similar interest in tween romance films. Your attention is focused more on scavenging ammo or inputting the right "password" for a combo rather than playfully insulting your friends while playing a local multiplayer game.

Sure the bad coffee might taste better with some cream, but it's still bad coffee. If bad online multiplayer game design makes it terrible to socialize with familiar and/or potentially new friends, then no amount of friends will make this online multiplayer component any better.

 

[MrToy]

Actually I liked his cartesian reference: even now there exist games in which AI opponents are more preferrable than the real ones. Because they are cleverer.
Wonder when a situation Yahtzee described will actually occur.

 

[yunabomb]

I can see the point Yhatzee was trying to make, with strategy games and shooters, it's perfectly understandable and okay to have online modes, but with fighting games, it kinda makes less sense seen by the perspective of an old schooler SNES era gamer, we've moved from an era of partying with our buddies mashing buttons and making silly tournaments to beating strangers online.

Just for fighting games, tournaments have become much bigger now while online play has become more popular as well.

 

[Waffle_Man]

Because I can't get 19 bots to crab walk down long hall.

You can!
Just enter the right commands in the console.

Getting 19 bots to crouch walk and go in a certain direction isn't the same thing. Mass crab walks necessarily include confused comments from the other team and quite possibly an auto-shottie rush response.

 

[crimson sickle2]

In certain fighting games, it is possible to set up an online room and everyone with a microphone can communicate with each other. Fighting games grew online to try to recreate the old arcade environments, a bunch of really smelly strangers crowding around a machine, taking turns to try to win and pass on. Some fighters still don't do this well, last I checked Street Fighter doesn't, but that experience without the smell is the overall goal.

As far as AI goes, current AI is made to counter the player's movements whenever possible (I think there might be counter to the amount of times it does in a row though), high level AI counters even before sprites move, they counter as soon as the input is made. However, most AI can't recreate large combos, in Street Fighter, AI can be pretty similar to human, but in any combo-centric fighter, fighting a human and fighting an AI is very different.

 

[Frankster]

Disagree with yahtzee on this one for the most part:

a) Fighting games benefit greatly from online and is one of the genres i find weird if it just has physical coop.
b) Don't find online fights and such to be as regularly one sided as yahtzees suggesting.
c) Do agree about there being less of a social element in online play, at least with the people you don't know.
So in those aforementioned fighting games, 95% of the time when the other player tried to communicate with me, it was to insult me rather nastily
d) Local multiplayer is cool but not always possible. 90% of my friends live in completely different countries to me so online coop is a big decider in the games i purchase.
e) As a guy with moody internet, can't disagree about lag being a negative effect.

And then i disagree big time. As mentioned ive got a lot of friends whom are in other countries then me.
One of the ways i regularly socialize with them is via online gaming.

So yeh dont knock online gaming, some of us like it and use it to socialize with their friends quite well

 

[RobfromtheGulag]

Brawlers - no experience
Fighters - Back when I played them the high difficulty setting computers would simply react instantly to everything you did, generally resulting in one or two known circuitous tactics being spammed if you hoped to win. In Street Fighter or Tekken the computer would simply block everything you did and break every throw attempt. In Smash Brothers Melee, if you tried to engage in a low aerial battle (the ideal method of combat for pro players) the computer would drop you like a sack of potatoes. If however you spammed smash attacks (which anyone playing more than an hour could dodge) the computer would walk into them and die.
-Case in point, Fighting game AI from ~2005 was not terribly advanced. I think if they worked at it they could make it better, but the developers are probably envisioning this as a party game to begin with and put <1% of the budget into AI programming.

I don't play Brawlers or Fighters, and I don't play consoles, so I have no idea the state of things, but assuming there is actually no social interaction I see this as a pertinent article. My biggest concern (back in the day) for online fighters was lag; in a game where you're only using moves that give you a 0.05 second advantage or better, latency can be a big issue. But besides people playing distant friends, the chatroulette version of fighters is somewhat pointless. I think some feature microphones so you can garble trash-talk in 32kbps audio streams.

 

[Okysho]

I actually agree with Yahtzee on this, specifically in terms of fighting games and inherently swapping out players for Ai character when a connection fails.

This is something Smash Brothers Brawl tried and I admit. I couldn't tell if my connection had died and was replaced with a lvl 9 character or if I'd actually beaten the player (A random player might I add) that I'd spent 15 minutes waiting for in the lobby.

Then again, I also play fighting games the way Yahtzee does. It's either a 4 player mayhem mashup with everyone in the room having a blast or something you play 4 or 5 rounds with in between something heavier.

Then again, I also hate playing online...

 

[microwavefriendly]

while i do firmly believe that not every game needs multiplayer, i feel you are slighting the concept of online gaming and multiplayer.

this link provides a counterargument that points out innovative ways in which multiplayer and online connectivityhave have been used to enhance games:
http://nitomko.hubpages.com/hub/Bridging-the-Gap-between-Single-Player-and-Multiplayer

 

[Maszynow]

Too bad Yahtzee pointed out that he is not a bot. I would love to read those paranoid comments all over the net. Bloggers crying about their dreams, of magcally meeting him, shattered. 4chan planning on hacking the Escapist network, just to confirm the identity of the guy, in all those photos online.

Sorry to interrupt this discussion gentleman, but I really felt the need to point out this missed opportunity

 

[Aaron Sylvester]

Yahtzee I thought we had a deal here - you keep your promise to stay away from multiplayer games (i.e. not talk about them) and we'll keep watching & reading your awesome stuff! Do what you do best, i.e. make game developers bang their heads on a wall figuring out how the fuck to impress you.

You already found out how it was a stupid idea to try and review an MMORPG in the span of 1-2 weeks, no-thanks to the WoW:Cataclysm review in which you told us about your leveling experience to 60 and utterly failed to understand why people raid together.

Fully understanding the multiplayer scene is beyond you, so can you just forget about them and get back to singleplayer town?

 

[baba44713]

I recently played a game of Awesomenauts where I started yelling at ally Clank for constantly running out of my healing range.. only to realize at the end of the match that the player has disconnected and it was a bot all along.

On the other hand, when people today have thousands of "friends" on their social sites, one can argue that every online multiplayer game regardless of genre is a game between friends.

 

[Thujal]

that was actually one of the points that he was trying to make that if the AI is designed in a good enough way that you can actually make it more difficult then a human...

I appreciate that. My one big issue is the effort vs. effect. If we take for example the entire roster of Capcom fighting games ever produced, which are basically entirely based around multiplayer, and it's balance, and have very little single player content, there is very, very little that can be taken away from the current games. What you (and Yahtzee) are then asking for is extra development, testing and QA time to get this in place, as nothing can be removed from the core product as it stands. Who is going to pay for this extra development? The players by raising the price of the core offering? And, realistically, how well is this expected to work when the job is finished, and can it in any way even approach the multiplayer already built in? And how big is the audience who actually cares about this feature?

Stuff like AI seems (to me, as an outsider) to be exactly the same tech that was in place for Street Fighter II (released in 1991) as is used in Super Street Fighter IV: AE 2012... I know this won't actually be the case, but it seems as intelligent to play against, and there has basically been zero advancement in this field at all... And I feel this is alright, because these aren't single player games in that regard! I love bosses like Gill in Street Fighter III because they require you to think outside of the traditional rules of fighting an opponent in SFIII including meter management, and he is in no ways cheap. I loved the four bosses of SFII because they all had personality and doing that on the SNES as a child and they all had their own personalities and unique attacks and weren't playable characters (until SFII:T)... I felt like in SFIV, Seth was cheap and lazy, and single player in fighting games in that regard was the worst it has even been in the series!

Mortal Kombat also got this right, with it's story mode and challenge tower! Tekken 6 with it's world tour type mode (reminded me of SF:A3). I've also heard a lot about the lab in Tekken Tag Tournament 2 which also trains players the fundamentals of fighting games while also being entertaining!

I think this is the direction developers need to spend their time to get more people interested in fighting games outside of it's multiplayer, not AI opponents that... I cannot see, from my point of view, from the perception I have (and I did a module on AI in the final year of my computer science course 10 years ago, so I'm hardly cold to the subject), that the effort put into the AI of fighting games would in any, shape, form or way, ever be worth the outcome, is a massive risk, and could only serve to get whatever fighting game series this is potentially put into, cancelled, due to a lack of return on investment.

Maybe I'm wrong. Maybe this is really easy to do and is a no brainer and every player who goes against this AI will love doing so and never need to go online to play multiplayer ever again.

I really don't believe this is the case. But would love to be proved wrong!

 

[StriderShinryu]

Well, at least he went into this article admitting he was ignorant about the subject matter...

Yeah, it's pretty obvious he doesn't get the appeal of fighting games. That was obvious and expected. He also seems ignorant of how AI works, though, which is a little more surprising given his role as someone who not only plays but also makes games. He throws around a lot of Ifs in relation to how AI can/could work but doesn't get to the point where it doesn't work like that at all, especially in the context of breaklng things down to a 1 on 1, actually thinking like a human level. The fact is, AI doesn't "think" like a human can and that, really, leads right back into the fighting game experience. They aren't about seeing your character beat up the other guy's character, they're about using your wits and skills to beat the other player.

Fighting games are all about a mental test combined with a physical challenge. It's like taking rock-paper-scissors, mutliplying it by one hundred and then then extrapolating that second to second decision making over minutes or even hours. Sure you, at some point in the future, may be able to code an AI that could play a reasonably accurate game of RPS but it wouldn't ever be able to capture the human element of tendency, adaptation and, perhaps most importantly, fallibility. That last aspect is really the key. Even the best fighting game players in the world make mistakes, they just make less of them due to skill and experience, and it's your goal as an opponent to force your opponent to make more mistakes than you do. It's your goal to take those minor victories and press the advantage. It's such a core concept to the experience that there's actually a term, Yomi, within the fighting game community that encapsulates this constant series of mind games. It's that circle of "I know what I want to do, but you know what I want to do too, so what will I really do" combined with the physical requirements of being able to consistently carry out the option you choose that makes the genre. It's a very human concept that not only trumps what any AI can be but also has the scope of organized competition, both offline and on, growing near constantly for the last few years.

 

[II2]

Not a fighting game sort. Don't know the community.

That said, wasn't so much of it about the arcade box and later console experience? I'd agree local multiplayer is much better for those types of games, but if you've got friends far away, or are the type for who Fighters/Brawlers are important games, it makes sense that the online component exists in a manner that would be asked for, by the people it serves?

Just spit in the wind though, I don't have any inside angle.

 

[uzo]

lol Holy shit there's some vitriole in this thread.

The beautiful thing is the same people hissing and spitting venom at Yahtzee for making fun of their particular beloved genre (in this case, fighting games) are probably the same people who, like the rest of us, laugh when he makes fun of someone elses' preferred genre (for example, J-RPGs).

I like KoF, SF, GG, DOA, etc. .... but holy crap people "hur durr I'm not going to watch anything Yahtzee ever does again as long as I live in my life for all eternity waa waa" is kinda .. well .. sooky. And if you read my prior paragraph (presuming you have if you made it this far), a touch hypocritical.

 

[Eclectic Dreck]

Fighting games have strategy. To imply there is none seems to me to be trolling. And I cannot believe I just signed up to this website (and got my password back in a plaintext email no less, a big security no-no), just to try and address this!

I vehemently disagree with this assertion. Any strategic element a fighting game happens is found in the larger metagame - at the character selection screen. Once in a game, there is no such thing as a strategic decision as that implies a capacity to foresee outcomes more than several moves into the future. This is a hopeless act.

But, that isn't to say the game is utterly devoid of clever play. Any fighting game worth it's salt exists simultaneously as two separate games. In the first, you have the game Yahtzee describes: a game wherein all parties desperately mash buttons and the outcome is left to little more than chance. Depending upon the game in question, one encounters varying difficulty mastering the mechanical inputs of the game. Some games pride themselves on technical mastery (For example, Street Fighter), while others set the bar of mastery incredibly low (Super Smash Brothers for example). If both players can reach such mastery, the second game is revealed.

Since a properly executed combination in these games can be devastating, and since most such games offer easy counter mechanisms, the game becomes more of a cold war where no party is free to fully commit to an action unless they force the opponent into a position of vulnerability. While the difficulty of doing this and the windows of opportunity one might create vary from game to game, the game becomes less about mechanical mastery than it is a game of bluff and counter bluff.

While one fundamentally endeavors to outwit the opponent, we can find that there are other arenas of human competition that offer a parallel. In Fencing, you are taught first to plan a single move in the future - into the second intention. One might thrust but, knowing that the enemy would parry the thrust, you thus know that you may disengage or simply counter-parry and riposte. Each action of blade and footwork is likely to draw a particular reaction and fencing masters over the years recognized this as little more than a physical dialog. But, because of the increasing unreliability of such prediction, (a thrust may be met instead with a bind, envelopment, displacement or counter attack for example) each degree of separation with intention and initial action becomes hazardous. Advanced fencers plan two moves in the future. World class fencers manage three. The great legends of the sport may, from time to time, achieve four. It is a game of bluff and feint and wit certainly but this is not strategy. If one were to attempt to assign such a base word it would be, at best, tactics. But other equivalent endeavors chose a more appropriate term - fencing calls this the language of the bout, boxing says it's the sweet science. Surely the genre of fighting games can manage something similar.

Of course, it should be noted, a great many games have similar levels of play. These are the games that tend to attract the "e-sport" audience. League of Legends and other MOBAs have far more in common with Street Fighter than Dawn of War. Starcraft is little more than a game of chess with an elaborate move set and synchronous play. This is why certain genres rarely achieve much of an audience for professional play. Call of Duty and other shooters rarely ascend beyond perfection of mechanical action; one's only requirement to dispatch an enemy is to know they exist and properly align the cross hairs and issue a fire command in a timely fashion. Similarly, it is these games that players tend to feel they have the least control over their destiny; glorious triumph or contemptible defeat rely on variables a player simply cannot control. Players in these games achieve domination not through guile but by ruthless calculation of odds. They throw a grenade through a door because people are often behind that door. They point their weapon down a particular corridor because people often travel through the space. They occupy a space because it confers a non-trivial advantage over common encounters. It is this that allows such games to decay into confused melees where each player fights a team alone while surrounded by comrades. That the larger game objectives offer little incentive to take risks on behalf of another is why one can so easily find endless complaint about a whole host of taboo behavior from attacking the enemy the moment they spawn to simply camping in a corner and gunning down passersby.

So, in a greater sense, the fighting genre is something of an anomaly in that it does offer something beyond mechanical master, a reward for knowledge beyond a slight edge in the calculus. But what it offers is not strategy. It is something else entirely. It's just that no one has bothered to name it yet.

 

[Eclectic Dreck]

The problem with Yahtzees comparison between an AI and a person is that they actually don't act the same. Unless the game is very specific in what actions can be taken, such as in a turn based game, a computer can not be accurately programmed to simulate human learning. It [em]can[/em] be programmed to be capable of learning, however.

To be fair, creating an AI that could pass a Turing test in the context of most video games is within the realm of possibility. The problem is that most AI techniques require tremendous amounts of system resources. Thus why you instead get AI that follows from waypoint to waypoint and acts in a predictable, if often efficient, way. Fuzzy logic and A* are computationally cheap and generally considered "good enough" to get the job done.