Raph's Website » Voice vs text

Voice vs text

May 8th, 2007 (Visited 5814 times) Tags:

The debate always rages, and it always seems silly to me. I often say that technologies accrete, which means that usually, new technology doesn’t replace old technology. In this case, the kinds of live text chat that we consider to be the “old way” are in fact the new way, having arisen during the 70s, long after the rise of party line telephony. An adaptation to limited bandwidth, text chat has stuck around because it has different affordances than voice chat does, and even now that bandwidth has increased, it still offers and forever will offer some things that voice cannot.

Text over voice:

Text is far better at history and at asynchronicity. You can step away, and then check backscroll. You can search logs. Hell, you can have logs, practically speaking.
Text is always linear, and issues with simultaneity are easily resolved; users cannot “stomp” each other as much as with voice.
Despite the horror we all have of “chat box spam,” text handles high volumes of content much better than voice, including multiple conversations at once. Voice breaks down in large groups without mandated restrictions on who can speak.
Text can be filtered better, including not just outright blocking or word replacement, but also interface niceties like color-coding, splitting of multiple channels, etc.
Text is more accurate for entering input, though this may change as voice recognition gets better.

Voice over text:

Voice has a far higher emotional bandwidth than text does. Text will never convey emotion as well as tone of voice does. This right here is the defining virtue of voice for social applications — its power cannot be overstated. This is why many claim voice to be more ‘immersive’ (a badly abused word…).
Voice has a far higher rate of output than text does — it’s faster and easier to say something than to type it. This is why it’s prized in high-end gaming situtations.
Voice permits expression not possible in text: singing, whispering, shouting, etc. While conventions exist to represent some of these in text, you’ll never have the same experience watching someone type that they are singing versus actually hearing them do it.

In glancing over these lists, it’s clear that just as there are some sorts of information better presented in a letter than a phone call, there’s some sorts of situations where text trumps voice and vice versa. Text’s sweet spot is large volumes of data, whereas voice’s sweet spot is smaller groups with high emotional content.

You can follow any responses to this entry through the RSS 2.0 feed. Responses are currently closed, but you can trackback from your own site.

Dearest Copernicus: Joost Schuur's blog wrote on May 9th, 2007 at 7:56 am:

MMO development studio Monumental Games has struck a deal to include the Vivox Precision Studio in Monumental’s MMO middleware software development solution. Just in time, Raph Koster discusses the pros and cons of voice vs. text chat. Oh, that Spring dashboard update on Xbox Live that adds Messenger integration and a hojillion other things? Comes out tomorrow, May 9th at around 2 PM Eastern. Halo 2 for Vista, on the other hand, which was supposed to be the debut for Games for

Sierra Kilo wrote on May 8th, 2007 at 6:39 pm:

you just received and other features. I think I’ve mentioned this before, but I have an odd fascination with instant messaging and text-based communication. I’m not much of a talker, but I feel pretty comfortable writing or typing. Raph has an interesting article on text-based vs. voice communication. Text works much better in a chaotic communication situation (such as a chat room or online game) than voice. You can visually filter out text better than you can audibly filter out voices, so it

constihill wrote on May 8th, 2007 at 6:31 pm:

Raph’s Website » Voice vs text

Zen of Design wrote on May 25th, 2007 at 11:09 pm:

The ironic thing is that we designers had a long list of complaints about voice chat, which mostly turned out to be invalid. And the people who loved Voice Chat didn’t quite nail why it works so well. Raph has discussed the latter. I’m interested more in the former, because it’s actually instructive from a designer’s perspective – how’d we go so off-track? Myth: “No one wants to find out the elven hottie in your group is actually a 42-year old truck driver.

Random Witterings wrote on May 18th, 2007 at 5:53 pm:

Voice vs text article. They are both interesting reads about using voice chat though I would state that Raph’s list, though good, seems to fail to pick up on voice being able to happen concurrently with actually being able to continue to control your character

Casually Hardcore devlog wrote on November 30th, -0001 at 12:00 am:

Voice vs. Text argument NPD gamer study My favourite web game – Kingdom of Loathing! Casual games from investor POVs

Omen of Clarity wrote on November 19th, 2007 at 8:34 pm:

typing, and it’s more social. This utility has traditionally required an external client running alongside the game to use, the most popular client being Ventrilo, a PC-oriented program. Raph Koster goes over the pros and cons of voice chat of in hisblog, commenting on these reasons. While ease and speed are positive aspects of voice chat (barring issues with clarity), the social aspect of voice chat is more ambiguous. The utility can exacerbate existing problems with communication and create new ones.

Voice Chat - Page 4 - StarTrek-Online.NET wrote on May 8th, 2007 at 8:55 pm:

[...] Koster has a blog entry on "Voice vs text" at his website. It’s an interesting take on what makes each approach useful for different [...]

Zen of Design»Blog Archive » Voice Chat Ready To Come Of Age? wrote on May 17th, 2007 at 5:28 pm:

[...] to be invalid. And the people who loved Voice Chat didn’t quite nail why it works so well. Raph has discussed the latter. I’m interested more in the former, because it’s actually instructive from a [...]

Talaen said on May 8th, 2007 at 8:46 am:

From a gaming perspective, I believe that voice gets used as a crutch to make up for a lack of player skill and coordination far too often. Don’t get me wrong, when I’m playing Gears of War over Xbox Live you can bet I’m on voice chat with my team. Likewise if myself and a couple of guildmates are flying a tough mission in SWG (although there aren’t many of these). The main reason that we use voice in these situations is that the controls don’t allow us freedom to type. A gamepad isn’t a keyboard, and neither is a flightstick.

But a lot of people take voice too far. You don’t need voice chat to group in EQ or WoW, and not even to raid. I used to run 6-group pickup raids on Lady Vox in EverQuest with only the shout channel, individual groupchat, and a (physical) notepad to keep track of group assignments, and we would win without player death most of the time. It took practice to be sure, but raid encounters (in any game) really haven’t gotten all that much harder in the past 6 years. There’s absolutely no requirement for voice in almost all traditional western MMOs. The only ones where voice is ever really needed (like in SWG) are more because the game’s interface/controls intefere with your ability to type.

On immersion, voice chat can just as easily break immersion as it can enhance it, the same as ooc text chat can do that. It’s just a lot easier to filter ooc text chat out. Add to the list of cons for voice that voice is more intrusive – every try to watch TV or read a book while someone nearby is talking loudly on the phone? Hard to do unless you’re a zen master. And finally, a lot of players (of all types of game) often have to get up and walk away from the keyboard or console for a short time. This is why games have pause features. Text handles afk very well, voice (because of it’s synchronous, real-time nature) does not.

My stance on voice is that it’s a nice-to-have, until PC games start coming with a headset or at least an integrated voice client (like with XBox Live) then I don’t plan on really using it heavily. Basically, my peer group in multiplayer games needs to all be using voice for me to use it, and as long as text chat is the default and there are people who can play the game without having voice enabled, that’s not likely to happen simply because I don’t want to be the one to tell the guy who doesn’t have a microphone that he can’t play in my group because he hasn’t spent the money on it.

Tim said on May 8th, 2007 at 9:06 am:

One more to add. Text works for people with hearing loss, which is my own personal issue with voice. Without cranking up the volume I have a hard time getting full comprehension out of voice (or TV or Radio). Hearing aids (at $1900 a pop) help. Text (or captions) helps even better.

If you had a good speech to text system, it really would give the best of both worlds.

Corax said on May 8th, 2007 at 10:10 am:

More things to consider: Text is not neccessarily always linear, as well as voice doesn’t have to be simultaneous; it’s only a matter of implementation.
Also, voice chat isn’t always more immersive – how many of the sweet little elven girls have low, middle-aged male voices?
And there’s no real reason why there couldn’t be voice backlogs as well.

TextToSpeech and SpeechToText might indeed be the way to go.

Naladini said on May 8th, 2007 at 11:22 am:

Text is also quite silent, useful around sleeping children, regardless of where your intended audience may be.

Raph said on May 8th, 2007 at 11:55 am:

Hmm… I do think text is always linear. Even when you attempt to present simultaneity, you have to read it linearly.

As far as voice being non-simultaneous, or having logs — yes, they are technically feasible, but they aren’t practical.

Ilsoap said on May 8th, 2007 at 1:58 pm:

Voice may have a higher rate of output, but text has a higher rate of INPUT. I usually read what someone has said faster than they would have said it.

And even in the output category, text chat has obviously adapted with its own forms of shorthand.

Those points aside, text also offers the advantage of being less immersive. I think a lot of players go into these worlds or environments comfortable in the fact that they’re still in the real world, and talking with text is one of those big things that separates the virtual world from the real one. If gamers wanted a truly immersive world, they’d turn the monitor off and go outside (even up close, the textures look amazing!).

Wolfe said on May 8th, 2007 at 11:28 pm:

The rise of voice chat can probably to some degree be attributed to the phenomenon Chris Crawford calls Asymmetric Interactive Relationships. – http://www.erasmatazz.com/library/JCGD_Volume_8/Asymmetric_Relationships.html

When things get complicated in a game and the player needs to spend a large volume of the available brain power on playing mostly visual clues of various sorts that need to be processed by the player. For a typical mmorpg its stuff like where everyone is standing, which direction the mob is facing, how many hitpoints the players have, who is losing hitpoints the fastest, who is building the most threat right now and so on. The list keeps going on and its far longer than the 7 recommended things.

The games use audio for primitive interactive response and ambience, a serious gamer dont need either of these so communication based on brain bandwidth from the ears of the players allow designers to build more challenging game content in multiplayer games. The trend set by WoW goes deeper and deeper into raising the bar for players and trying to coordinate a raid without voice chat is a severe disadvantage. Of a random selection of “skilled players” maybe half has learned the skills of reading text chat while also performing well enough to contribute to a raid.

WoW has set the path which all gamey mmorpg’s will have to follow, or they will instantly be trivalized by WoW players. If you reduce the difficulty of challenges to where a group larger than 10 players can successfully defeat the upper end of challenges without using voice chat it wont be a challenge to anyone used to communicating through voice chat.

One way out of the problem would be to make important game feedback sent from the game through the sound channel, but not have any visual component, and it needs to be different for most participants in a raid. I dont think that would be a functional sollution but it could be an experiment worth making anyway.

Richard Bartle said on May 9th, 2007 at 12:09 am:

Both text and voice say something about the user. When I use text on a US server, people know I’m not American (I use words like “armour”, and I don’t split infinitives). When I use voice, people know I’m English, which part of England I’m from, and that I’m male.

It’s what real voice does to virtual identity that puts me off using it for virtual worlds.

Richard

Noche Kandora said on May 9th, 2007 at 12:17 am:

I don’t think people can be nearly as sincere with voice as they can be with text. There’s a certain innocence and genuineness that comes out in people when they are communicating via text in a virtual world. I just don’t think people have the courage to express themselves that way with voice.

Talaen said on May 9th, 2007 at 9:15 am:

Disagree. WoW raids aren’t any more challenging in my experience than any other game’s raids, and the presence or absence of voice chat wasn’t really a determining factor in success of the raid. What is important is teaching players about positioning, aggro control, timing, and so on, which WoW does a good job of with instance progression up to the top end. By contrast, most other games don’t start forcing players to pay attention to this stuff until the latter half of the game, which results in the perceived disadvantage.

If you start making things like positioning, resistances, aggro control, dynamic ability timing, and so on important in “boss” fights early on, players will be able to utilize those skills in high-level raids with or without voice chat.

If voice were really required, a skilled raiding guild wouldn’t be able to do raids without voice, and many guilds do raids with just text chat in a lot of games all the time. The reason voice gets used by many is that people are naturally lazy, and are more apt to pay attention to someone screaming in their ear than they are to someone typing in all caps in a little bitty chat window.

Tess said on May 9th, 2007 at 9:29 pm:

Another point in favor of text:

It doesn’t drive everyone around you completely insane.

You can text while your roommate is sleeping. You can text in a room full of other people playing games. You can text on the subway. You can text at work. You can text in a restaurant. You can text in a library. You can text by your baby’s crib. You can text while your spouse is watching TV.

Voice chat can be a major problem if there is anyone else within earshot of you who is not engaged in the same conversation you are. It is noisy, distracting, annoying, and may even include content which is inappropriate for the ears of little ones who may be toddling nearby. If everyone in a LAN cafe was voice chatting all at once, everyone would go deaf from the sheer cacophony. It’s simply not practical in many situations.

Wolfe said on May 9th, 2007 at 11:55 pm:

Learning how to chat while playing is not something todays WoW player find important. The tests I have made over the last few years of WoWing gives me a hunch that only about half of a normal raid group actually read chat messages while an encounter is ongoing, and the numbers are getting worse by the month as more and more people get groomed into a voice chat culture.

Depending on players to already understand how to treat an encounter is likely to be successful if you have really hardcore raiders, but those are not common enough. Most ordinary players also fail to learn the lessons that WoW tries to teach about how to play while levelling, they need to be told about resistances and even after having been told they promtply ignore the instructions of collecting resists and vendor it as junk to make room for other things in the bank.

I honestly doubt that any group of 10 “first timers” would be able to defeat an encounter like Netherspite or the Prince in Karazhan without using voice chat, or spend multiples more time to practice on it which is discouraged by repair costs and trash respawn. (Unless they seriously out gear the encounter or spend hours developing and executing a detailed plan, but there is too much randomness to these encounters to do the waterfall type of sollution before its on farm status.)

Kerri Knight said on May 16th, 2007 at 10:32 pm:

Funny that WoW gets brought up here, as within the last few days we’ve learned WoW is likely to implement a voice chat interface into the game.

I’m taking bets on how long after implementation it will take until ‘must have voice chat’ becomes a standard suffix to the typical LFG. How wonderful that we’ll have yet another issue for the community to divide into the ‘haves’ and ‘have nots’ similar to the degree that the “you don’t raid because you’re a scrub”/”I’m not a basement dwelling loser” camps have.

Though I could comment on the issue of different gameplay styles having disparate levels of reward structures, I’ll wait for a more appropriately themed discussion to do so :9.

Voice vs text

18 Responses to “Voice vs text”

Meta

Recent Comments

Categories

Tags

Recent Trackbacks

Archives

Twitter @raphkoster

Academics

Comics

Game devs

MMO devs

News/Forums

Other

Players