Playing MarioKart 64 From a Browser

David Perry, of GaiKai has a demo up of GaiKai's technology. This allows users to play any video/computer game they have installed on their servers from any web browser on any computer. That's right, it reads your clicks, sends them to the server, which generates video and audio which is streamed back to the client.

David says:

(1) No installing anything. (I'm running regular Windows Vista, with the latest Firefox and Flash is installed.)

(2) This is a low-spec server, it's a very custom configuration, fully virtualized. Why? To keep the costs to an absolute minimum. We had 7 Call of Duty games running on our E3 demo server recently.

...

(6) We designed this for the real internet. The video compression codecs change in realtime based on the need of the application (or game), and based on the hardware & bandwidth you have. (For Photoshop we make sure it's pixel perfect.)

(7) Our bandwidth is mostly sub 1 megabit across all games. (Works with Wifi, works on netbooks with no 3D card etc.)

Make no mistake, this is an impressive technological acheivement. I'm just not sure what its good for.

David claims a ping time of 21 ms from LA to Fremont, CA (20 miles from here, 400 miles from where he sits, in LA). I think that's probably only achievable with commercial service DSL or T1 or something. Well, I checked ping times to Google (83ms) and Yahoo(15ms) just now.

You see, your mileage may vary. This gets to the main point of my skepticism. It's called control loop latency. The longer it takes for you to see the results of your changes in control show up, the harder it is to control. And remember, the games already have some control latency, it's inevitable. Computers are fast, but not instantaneous. At 30 fps, there is a guaranteed latency of about 33ms and video systems these days often introduce another 33ms of delay with buffering systems built into the monitor or projector.

Let me highlight the issue with a game that should be familiar to most of you. Just today I adjusted the sync on RockBand and it measured at about 70ms. I had it set for about 40ms, and the game felt wrong. I couldn't tell what was wrong, but I was missing a lot of basic repeated beats that should have been easy, but I'd keep breaking streak on them. It was just off. After I reset the calibration, BAM! Stuff that should have been easy, was, in fact, easy. That was from an adjustment of about 30ms.

So, this stuff matters. But the less synchronization a game needs, the less it will matter.

My second concern is the cost model, David said they had 7 copies of Call of Duty 3 running on one server for an E3 demo they gave. Meaning that for every 10 players (lets be generous) you need another server. And there's a fairly high maintenance cost, since all the software must be installed and updated, on perhaps all the machines.

If you go with servers that are dedicated to a particular game, it makes install and patches much easier, but then you have sort of a forecasting problem. After all, how is a customer going to feel when he or she wants to log in on, oh let's say, the 4th of July weekend from Mom and Dad's computer and there's no server available.

Long-term, I could see mmo type companies using GaiKai as their primary hosting, provided their game design avoids the need for high responsiveness in the control cycle. Just as a for instance, Illusionist DPS depends greatly on speed casting, at a rate of one spell about every 800ms. This could have big problems with Akai. A slower caster, like my Defiler, would have less of a problem. But timing your spells with your autoattack swings could become trickier. Still, that already has to deal with control latency.

In any case, that's a lot more computing power per player on the server side of the internet than is there today. How will that cost be supported? Ultimately, the players of games will pay for it, so the question becomes how much will players be willing to pay for no download/install/patch time and play-anywhere?

Yeah, I'm not thinking it's all that much, either. The answer has to be "considerably less than I'd pay for nice laptop that can play the game". The upside is that the platform requirement under this just became really low, and Blizzard has shown what a good thing that is.

Still, I think there has to be a killer app for this technology. Maybe it will kill off flash games? Invent a whole new genre of games? Hard to say.

What do you think?

Taking Fun Seriously

Serendipity abounds. I've been slow posting, and in the meantime, several items on the topic of "It's just a game, isn't it?" have come to my attention. First up, bunny ears.

It seems that to celebrate Easter, WoW introduced a new achievement which required players to find a female character from one of each of the races and catalog put bunny ears on her, presumably using some sort of gizmo. No permission of said female toon was required.

Some of the women objected. To them, a line was crossed and their avatars were sexualized without their permission. Many in the gaming community reacted just as is expected when challenged: It's just a game. It's just an avatar, it doesn't mean anything.

Credit where credit is due, Raph got here first:

But boy, avatars is a pretty special case. We have a lot of “specialized hardware” around this in our brains, and avatars tend to trigger a lot of it. For example, the fusiform face area or FFA is a part of the brain that seems to be involved in facial recognition, and also seems to fire off when identifying specific objects with fine distinctions (for example, it fires in birdwatchers when identifying birds, and in car aficionados when recognizing specific makes and models). The interesting thing is that the FFA activates even with iconified faces — with stuff that we just think of as a face.

It may be a game, but it isn't just a game. Your body and your mind reacts to it pretty much as if it's life. Some people manage to be pretty detached about the experience, but then that's true of life AFK, too.

Here's how I feel about it. People take their appearance seriously. WoW has lots of "zap someone's appearance" toys, but most of them only work on group members. So there's some form of implied permission there. The height of a players's avatar will affect their behavior, and so will the amount of "eye contact" another avatar is giving you.


So, I think WoW blew it here. It's not the moral equivalent of murder, but the moral equivalent of slapping a "hottie" sticker on the back of women at the park. I have little patience for those who argue that "people wear bunny costumes all the time, it's celebrating the Easter bunny." There are two problems with this, first, the people who wear the costumes choose to do so. Second, that's not the only meaning of bunny ears, which have been associated with sexual availability for at least 50 years.

Geez, just ask anyone who is a fan of manga and anime. The acheivement didn't require you put ears on male avatars, did it? I think there's be a lot less issue if it had. No, the female characters had to blow kisses at males. Doesn't that strike you as being a bit, umm, one-sided.

Anyway, not a mortal sin, but a mistake on Bizzard's part.

Where's the Pipeweed?

The New York Times reported today on new data concerning the fossils of hobbits found in Indonesia.

These bones were first found in 2004, but controversy remains over whether the bones represent a different species altogether, or just pygmy versions of homo sapiens.

A new study being released in Nature recounts many anatomical differences between the bones and homo sapiens, most notably the extremely large feet. Hair, of course, would not survive the 17,000 years. Still, some questions remain...

Dr. Jungers and his colleagues raised the possibility that the ancestor of the species was not Homo erectus, as had been the original assumption. H. erectus is known as the earliest hominid to leave Africa and make its way across Asia. At a symposium two weeks ago, several scientists edged toward the view that the hobbits emerged from another, more primitive hominid ancestor.

Well, nobody has ever claimed to know where hobbits come from, not even John Ronald Ruel himself. And any red hair would be long gone by now

Sunday Reading

Raph has a very interesting essay on game mechanics and the sort of things that make MMO's so highly addictive, by way of an Easter Egg game he did recently.

Now, there’s no reward in this game, there’s no winner or loser, and there’s no endgame. Yet even during testing, I had to tear myself away, and when put into Metaplace Central, average session length for the day went up 50%. But… in some sense, it’s a crummy game. Why this effect? Because the Easter Egg hunt is a confluence of a lot of highly manipulative tricks.

What are the tricks? And how do they operate in MMO's? Well, read the whole thing to find out why folks will raid four days a week or more, and why I go around with the title Exalted in front of my name.

New Classes for EQ3

I had a discussion last night with my guildies about what they wanted to see in EQ3. We were not in what one would call a serious mood. I'm not sure I've got everything exact, update me in comments.

Karaya

Vomitoriums (which, boringly, turn out to be part of structure that allows large numbers of people to leave at the same time), STD's, nuclear weapons. Her favored new class would be Attorneys, who would have an ability "Throw Book". A weaker, (NPC?) class would be "Public Defender".

Milia

Milia wants to see Farmer become a class in EQ3. I told her that's old ground, what with the Harvest Moon franchise and all. She then said she wanted to see in-game outbreaks of swine flu, which would be called "The Aporkalypse".

Phritz

The character class Phritz most wants to see for EQ3 is Gigolo, or Manwhore, as we sometimes call them. Another favorite was Telemarketer. I think I'd like to see NPC's be telemarketers, certainly. I want to be known as Slayer of Telemarketers.

Toldain

To much acclaim, I declared that the class I most wanted to see added to EQ3 was Game Show Host.

All this was discussed while we went collecting sparklies and named kills in Karnor's Castle in force. Don't jog Karaya's elbow, though. I did this once, and ended up with a fine view of the tile flooring. I think it's basalt, for the record.

Also all agreed that we wanted Nathan ("I'm gonna join the Qeynos Guard when I grow up. Have you ever SEEN a gnoll?") Ironforge to be a permanent fixture in any future sequels. All they need to do is make him aggroable.

Keen on Everquest 3

Keen, of Keen and Graev's Gaming Blog has offered his take on what EQ3 should be. He says it should stay pure, true to what makes Everquest Everquest, but still be radically evolutionary. This gives a better sense of what he means.

I’ll oppose the idea of stepping too far out of the boundaries of traditional MMO design. I think tying all the EQ games together, or adding mini-games, or streamlining it for console/multi-platforming, and other alterations are wrong. Going sci-fi, action oriented, or in another new direction aren’t right either. I guess I’m a MMORPG purist when it comes to certain things.

And here's some of the stuff he likes:

Obviously EQ3 should not be a PvP oriented game, but it wouldn’t hurt to bring back the pvp ruleset servers of old. Boats, vast cities, open world (not necessarily sandbox), group oriented, a rethought out death penalty, player cities, biiiiig monsters and raids, lots of gear (but not a treadmill), alternate forms of character progression instead of or on top of a level system, prestige classes and features of this ilk could be looked at.

I don't have a lot to quibble with, though I think regular readers know that I think any major increase in death penalty is doomed. And mini-games have always been a part of Everquest, remember gems?

But here's the thing:

Tipa isn’t alone in thinking that what hurt EQ2 so badly was that it had too much EQ. I believe quite the opposite though. I believe it was a combination of diverging too far from the original EQ, making changes that were not thought out or feasible at the time for the sake of change, as well as making a game that could not compete with its competitors at the time who were striking out to steer the industry in a new direction.

Every single change in EQ2 from EQ had a very solid justification, that they weren't popular and didn't compete well with WoW is hindsight. Let's look at some of the major changes from EQ to EQ2, and why they were made. Bear in mind that I'm guessing a little here, but only a little.

• Moded combat/non-combat.This got rid of mages and priests having to sit to med, and in consequence groups having to stick to one spot and camp. This change was huge and mostly welcome. It made combats faster paced, and it made for lots less down time, LOTS less. It's had the unintended consequence of making groups adventures less social, since with less downtime, there's less chat. Could you have predicted that. In any case, this change has been successful and imitated widely.



• No buffs out of group. There was a clear reason for this and it was a reaction to, if not outright hostility to Enchanters and C3 as well as the mass buffs of Aegolism and whatever else a cleric could throw out there on PoK. It got so that my enchanter couldn't get a group because everyone already had their C3. Buffs out of group encouraged soloing, and they wanted EQ2 to encourage more group play. By the way, lots of folks posting are saying they want EQ3 to focus more on group play. I think it's going to take some powerful designer kung fu to accomplish this. This rule still holds in EQ2.



• No interference in combat, no trains. This cuts way down on the amount of griefing that can be dealt out. Which was seen as a good thing, am I right. But again, it had unforseen consequences, and is now pretty much dead.



• Progression from archetype to class to subclass. This was an idea that I think people thought would be fun. You had these cool solo instances at levels 10 and 20 to be allowed to advance. I thought this was pretty cool, but it's been ditched as not providing enough replay value. Nobody is copying it either.



• Highly predictable combats. Fewer adds, easier to tell what would pull. In the early days, encounters would run on rails. I think this was done to expand the customer base, to make things easy for the brand new players. And to break the stranglehold of the "Holy Trinity" Tank,Healer,Enchanter. In the first few years of EQ, there was no good use of mezzing in a group. Folks would AE and break mez constantly, and typically, there was no need for it.



• Gear that leveled with you. This seemed like it would be fun, and avoid twinkage, but it ended up just confusing people, so it was dropped. This kind of gear was tried first in EQ, though.



• Highly interdependent crafting. It's supposed to be an MMO, right? So you should be interacting with other players. Why not with crafting too? This was extremely unpopular with the players, and for some reason, the markets never worked right. I think perhaps if the batch size had been made bigger, it might have worked better.



• Many (difficult) subcombines per product. This sort of drama is about making you feel you've really accomplished something when you complete the final product at the best level of quality. Which it did. Making just one spell for someone could take maybe half an hour. And you might blow the last combine. I remember the first time I made a rare combine for someone, I was pretty scared that I'd blow it. I didn't, and I felt great. But the risk and time meant the price was high. And you wouldn't just "oh sure, I'll make all your spells for you", it might take a week. So this has been dropped, too.



• Guild leveling. We had several guild leveling rules that have been ditched. Guild XP decayed over time, but the decay rate was way too big, leveling felt like a huge millstone around certain peoples neck. Only designated guild members could level the guild, though everyone could do writs for personal status. And the amount contributed depended on the number of mentors, or guild size. None of that stuff worked, though there was a solid theory behind it, a theory around making things more fair. Give the newer guilds a leg up, or give the smaller guilds a boost, etc. But like the anti-griefing, it didn't really work. And what's "fair" is actually pretty subjective.



• Online selling I don't know if this was a design or a technical constraint, but you would have to stay logged in if you wanted to sell things. It's pretty much exactly the same as Everquest, except you could be in your own room while selling, which avoided all that horrible Bazaar lag. But the community wanted to move on. So it's a case of something that wasn't changed that should have been. I prefer EQ2's marketplace to the timed auctions in WoW and LOTRO. By a lot.

That's all I can think of for now. This business of "keep the stuff I love but bring it up to date" is a lot trickier than you might think. There's a lot of folks out there who were mad that they couldn't do X (driveby buff, port people, play a beastmaster, mez meaningfully) from Everquest. That's why I think it's important that it not be called EQ3...too many expectations to meet. But we can keep a lot of what we liked about EQ: Lore, focus on group play, sparklies, housing, and of course, Fippy Darkpaw.

Musings on Everquest 3

Since Tipa presented her wish list for Everquest 3, I did some slumming on discussion boards, looking for what folks are saying about a possible EQ3.

Here is a completely unfair summary of the sorts of things folks are saying, presented in a completely one-side manner, with no attempt to be fair or anything. [My response to each of this is in brackets and italicized]

• ...Make it darker then anything people have seen. Im talking full out carnage. Time of War or somthing.
  
  [Somehow going to Lavastorm and slaughtering void beasts isn't carnage enough for some people, I guess. Or maybe they mean there should be no civilization left, no Qeynos, no Nathan Ironforge saying "have you ever seen a gnoll?" Wait, was I making fun?]



• ALSO sex and violence sells. All the die hard mmoers are adults now and we arn't kiddies anymore we want to see the violence and blood. Well at least us americans do.
  [ This die-hard mmoer has always been an adult (remember I'm 3000 years old). To me the hallmark of a child is one who wants his sex in a videogame, as opposed to the real thing. And if you want more gore in your life, either enlist, or work in a hospital. A slaughterhouse will do in a pinch. Oh, and keep working on growing up.]
  
  
• Consoles are the way to go. PC gaming has been a dying breed for some time now. Its too expensive to maintain a high end gaming PC to play the latest and greatest. I mean I have seen several threads of people complaining about not wanting to upgrade their PCs anymore to play a certain game. This will be eliminated by moving to consoles.
  [There's some credibility to this. The ante keeps getting upped, developers are locked in an arms race. Of course, THE most successful MMO ever runs on a PC and has a very minimal system requirement. Why is it that nobody seems to want to copy that? Maybe because they want more realistic flesh and carnage?]
  
  
• SOE has openly said that consoles are the way forward.
  [Well, probably somebody at SOE said that once. I would be deeply disappointed if it were true. I have four game-playing computers in my house, and only one TV with console attached. That said, a console is not a moving target. A console means not having to eat whatever flavor of DX microsoft is serving. Sometimes its good, sometimes not so much.]



• Vanguard was EQ3.
  [This one got shot down by other posters. No lore in common, makes it hard to to think of it as EQ3.]



• you want a new game from sony first they need a crew or at the very least a young person with lot of imagination cause great game come from great idea.
  [ Because obviously old people don't have much imagination. There's a germ of an idea here, they need fresh ideas. Duh. A sequel without a new idea is a trip to skip.]
  
  
• EQ3 won't happen, it was said by somebody from $OE that they knew calling EQ2 EQ2 was a mistake, there may be an EQ3 but it won't be called EQ3.
  [I think this is spot on. The game devs had some problems with expectations at the launch of EQ2, precisely because it was called Everquest 2, not Norrath Rebooted or something. People expected all the things that they knew and loved to be there. Putting Vox and Naggy back in the game weren't a problem, but busting up the Holy Trinity, and killing driveby buffs were a BIG problem with player expectations.]



• [Paraphrasing]Make the game harder, more like Everquest. Trains, stiff death penalty. No in-game tutorial. Slower paced, less button mashing. Drive away all of the people who don't like to play the way I like to play.
  [Yes, let's totally spend 10 million dollars to develop a game that 100 people want to play. And I thought I was an old fart.]



• I want to adventure in Middle Earth, not be forced through the chapters of the Lord of the Rings. [Uhh, weren't we talking about Everquest, not LOTRO?]



• The EQ series is perhaps, the only lasting game which has encouraged cooperation among players, it's a market that is virtually untouched (as many other games focus on competition through pvp or gear progression). If they can create an environment that encourages player interaction in small groups, allow people to progress in both small groups and small raids they will create a game which is fun and entertaining to a great many people. One thing they could improve on over EQ2 - is the ridiculous number of skills each class gets. They could get rid of half the skills and still have too many. Make a few important skills, and let people enjoy one another's company over focusing on spamming skills like an idiot.
  [Wow, this is the reason I still read message boards. I want to play that game. Wait, I AM playing that game, as Everquest 2. I've made other posts though about how game designers can enable leadership among the players. Of course, I'm an illusionist, and nobody but nobody casts more spells faster than we do, except maybe coercers with Jester's Cap.]



• Make better graphics, EQ2 had graphics that were soulless and ugly.
  [I'm a high elf, and I'm impressed with the level of snobbiness embedded in that statement. That's the kind of thing you hear at the opera in Neriak.]

I think the worst design decision (that's still in the game, anyway) in Everquest 2 was to make 24 classes. It's just too freaking many, each class is constantly stepping on the other's toes, or seeming to. And there's only 24 slots in a raid. In the abstract, 24 slots in a raid isn't bad, but with 24 classes in game there's no room to add a new one.

The second worst decision was to call it Everquest 2. So it had a huge burden of expectations.

At launched much was made of the fact that EQ2 was designed to go to level 200. I don't think that lasted. I'm still having fun, and that fun is mostly predicated on having interesting things to do with my friends or new people I meet.

But the most important thing that must be done for Everquest 3 is not to have better graphics, or more sex, or slower gameplay, or faster gameplay. No, the most important thing is for it to be fun.