[the_logos]

There's been some good discussion lately about reaching out beyond text MUDs. I don't know how many of you actively try to spread the message beyond existing text MUD users but I thought I'd share the specific results of an experiment we just ran.

We want to start paying an ad agency per registration they deliver to us. To be clear though, by registration we don't mean someone who just creates a character. In our games, registration is when you're asked for your rl details, which doesn't happen until you've finished the newbie intro for the game. So, a player has to stick it out through the newbie intro before we'll pay this company for that player.

In order to set a price per registration what we did was run a small test where they'd deliver 11224 click-throughs to us (there's a reason we used 11224 but it's not interesting or relevant for this discussion). I believe it took about 1.4 million impressions to deliver those clicks. (For reference, the actual creatives we used can be found here.)

Each of those clicks went to the Iron Realms portal site and from there to opening our Nexus client, then to creating a character, then into the game and finally registering. The numbers worked out like this:

• 11224 click-throughs to the portal site.
• 2192 of those clicked on Play Now and opened Nexus.
• 1297 of those finished the character creation process.
• 51 of those played through the game long enough to register.

Those 11224 clicks cost us 5 cents/click, which means we paid $561 for 51 registered players, or about $10/player. However, the source site mattered greatly. For instance, on one set of sites we got closer to about $5/player while on others it was more like $30/player. The small sample size of registered players probably guarantees large statistical errors here though.

We don't have any idea yet if those registered players are actually worth $10 each of course and won't truly know for over another year (as we have to watch how people spend first).

The first big problem I see here is our portal. A ~20% rate to just click on one of the 'Play Now' buttons is terrible and we're going to completely redo our portal to try and improve this. Shame, as I rather like our portal but the numbers are pretty unequivocal.

60% of players completed the character creation process which is fairly good I'd say, so I'm not too worried about this aspect of things. I'm guessing it's high because the creation process is graphically-driven in Nexus, with character portraits and that sort of thing to select from.

Then things get ugly again. 51 out of 1297 players is about 3.9%. That means 96.1% of the players who finished character creation dropped out soon after being put straight into a mainly text environment (our Nexus client puts at least somewhat attractive graphics around the text output area but it's still mainly a text output area, of course).

It's very hard to know what the issues are in the last part. The fact that players are suddenly in a text environment when they were previously on a graphically-oriented webpage or in graphically-driven character creation is almost certainly a big part of it. Who knows what else plays into it though. There are so many factors that could affect this, from the style of prompt a MUD uses to the colors (or lack thereof) it chooses to the length of room descriptions, down to whether it uses UK or American spellings (colour vs. color), etc etc etc. In theory we could change one of these at a time and run tests to optimize but that's not really economical sadly.

What we're considering is possibly having illustrations done that give the general 'feel' of the general area you're in when you start playing, and display those in Nexus. We'd then display illustrations less and less as the person goes on in the game until finally (probably once you're done with the newbie introduction) they're not getting any at all. The idea here is to kind of slowly shepherd someone from a mainly graphical environment (character creation) to a fully text environment and minimize how 'jarring' the experience of going from one to another is. I have no idea how effective this will be (or even, at this moment, of whether we'll go ahead with it or not) but it's what's come up in discussion after seeing these numbers.

Anyway, just thought I'd share.

[KaVir]

Originally Posted by the_logos It's very hard to know what the issues are in the last part. The fact that players are suddenly in a text environment when they were previously on a graphically-oriented webpage or in graphically-driven character creation is almost certainly a big part of it.

In the past few weeks I've had a number of newbies log on and start asking how they can "view" things - once they discover that the mud is text-based, they start screaming about how it sucks (and quit shortly afterwards). I'm guessing that someone mentioned my mud on a forum for graphical muds without bothering to mention that the mud is text-based.

It's made me wonder if I should put together some sort of blurb about the advantages of text-based gaming - the old "book vs movie" comparison can be used to make text-based muds sound a lot more appealing, IMO, although as always it'll depend on the sort of audience you're targetting (I could see it working quite well for your Feist mud, for example, if you were trying to appeal to fans of his books).

I think if players log on to a text-based game with the expectation of playing something graphical, they're likely to have a more negative response than if they're eased into it more gently with phrases like "interactive novel".

[the_logos]

Originally Posted by KaVir It's made me wonder if I should put together some sort of blurb about the advantages of text-based gaming - the old "book vs movie" comparison can be used to make text-based muds sound a lot more appealing, IMO, although as always it'll depend on the sort of audience you're targetting (I could see it working quite well for your Feist mud, for example, if you were trying to appeal to fans of his books). I think if players log on to a text-based game with the expectation of playing something graphical, they're likely to have a more negative response than if they're eased into it more gently with phrases like "interactive novel".

We were just discussing this today too. Is it better to keep someone as long as possible with shiny graphical goodness in the hopes that they will come to like or tolerate the text interface and suffer losing the ones who feel misled or is it better to drive home up front (and lose a bunch of people by doing so...but you very probably would have lost those people anyway) that it's a text game so that you're not creating false expectations?

I really don't know frankly. We may create two similar versions of the portal - one that pushes that the games are text and another that doesn't - and run split-tests on them to see which works better.

One of the big problems with telling them it's text is that it is very hard to effectively to communicate that text MUDs offer things graphical ones do not, especially because you only have a couple of sentences to do it before most people lose interest. You can say all the obvious things (greater depth, roleplaying, immense worlds, systems the big guys don't implement, etc) but it's hard to be convincing. It's so much easier to communicate 'coolness' graphically.

Random thought that someone had today was to stick a little flash movie on the front page that shows text flying across the screen "Matrix-style" with a label below it letting the viewer know that he's watching the intense speed of combat (this only works for MUDs whose combat is very fast/spammy), and then freeze the text (with some nice Flash effects of course) on a line indicating you just killed another player. "You reach down and rip still-beating heart out of your victim." (or whatever). As I said, random idea for trying to show that text MUDs can be 'cool' as opposed to telling people, which is rarely as convincing.

--natt

[scandum]

When I was a kid I used to pull random books out of the bookshelves in the library, check the summary, and if the summary seemed interesting read the first page, which had to be both interesting and top notch in order to take the book home.

While creating an interesting summary should be easy, as kavir pointed out, I assume the actual first few screens of text the players have to read might be disappointing. Keep in mind the reading type of people are spoiled. Hiring a skilled author, Ian Livingstone for example, to write the introduction area, might yield better results.

[the_logos]

Originally Posted by scandum When I was a kid I used to pull random books out of the bookshelves in the library, check the summary, and if the summary seemed interesting read the first page, which had to be both interesting and top notch in order to take the book home.

I'd say that's a pretty insightful take on how people get interested in things in general.

Well, that and repeated exposure I suppose, which is the only way to explain the appeal of phenomenons like the sport of cricket, Barbra Streisand, Lindsay Lohan, and MySpace.

--matt

[prof1515]

My staff voted last year that when we're ready to open we'll try recruiting from ONLY non-MUD sources because of the general low-quality of players we've seen coming in from those venues as far as knowledge of our setting and willingness to learn a setting. We're going to try targetting those interested in the theme rather than those interested in MUDs in an effort to attract players with the maturity and knowledge to RP characters within the setting of our game correctly rather than those who can RP but not beyond standard medieval-fantasy cliches.

Won't be able to say how that works out until we open in 2009/2010 though.

Jason

[Brody]

I've been trying to draw from the population of WoW players on the Earthen Ring RP server for text-based games, but, as one might expect, that has met so far with limited success. Mostly, we've gotten interest from people who knew what MUDs were before and played for a sense of nostalgia (and, secondarily, an affinity for the game themes).

I'm hoping to find more success through exposure of the OtherVerse Wiki and MU*Wiki at Wikia.com. My feeling is that by showcasing these sites in a mix that includes comic book fandom sites, Star Wars and Star Trek canon sites, and EQ/WoW fan sites, we'll have better luck snaring 1) people who are already rather Net savvy and 2) people who are fans of text-based environments and 3) people from interests outside the existing talent pool of MUDers.

[Spoke]

I have been playing MUDs since '96. When I started I was starting college in Colombia (Spanish speaking country for those who did not know) and my English proficiency was comparable to the Spanish proficiency of the average American high-schooler. This said, I plunged into MUDs because there was someone in the computer lab who was playing too, and so I was walked through the rough beginning. Now, I took the fact that the game was in English and that I needed to do well in the TOEFL to graduate as an excuse to keep playing and soon enough there were about 6-8 of us playing at some point or another. After all this time (graduating, going to graduate school, etc) I think of the 20ish people I have introduced to MUDs I think only 3 remain playing, that I know of. This is even after they spent considerable time in the games mind you, some of them had played for a few years with me to just banish after some cool graphical game hit the shelves (AoE, AoE II, WC3, SC, WoW, etc.) I myself have been largely absent from MUDs for about a year and a half.

Why do I say all this? I do believe that it is very hard to retain a large number of players in a consistent basis, in part because of what the_Logos points out about Graphics vs Text, but also, I believe, because there are numerous graphic alternatives that do not require your undivided attention for long spans of time (I am excluding games like WoW, Ultima on Line etc) or the graphic multi-player games make it easier for you to share with friends (in the sense that since graphics appeal to a greater pool of people, there is a bigger chance you can play with RL friends than there is with text games.)

The other thing is that any new MUD I have tried takes a while to get used to, even after playing them for 12 years, hell, even coming back to a game you played for several years (to an already established character) takes some getting used to. I do think the challenge is more with how people are used to receive/obtain information now (or rather, how many are used to be spoon-fed everything with the smallest effort on their part), you not only need to entice people to remain in the game, but if you want someone to stay long, you need to make sure this person is going to be able to help him/her-self in the long run.

What I have done as a player (during the times I cared about it) to try to keep people in is to go an extra step when answering questions by doing it in detail, but following up immediately with a link (finger pointed ...) to a help file so that they know the answer was at their disposal anyway. This is much like the approach I take with a toddler, not just say: "Do it yourself for Christ's sake!" but rather, "Here, let me do it WITH you the first time, I will show you how you can do it next time." Another big turn-off is the initial spam many games have, you are inundated with information that is very much meaningless to you at the time, even if you are experienced MUDder but new to the game, I think letting people choose the amount of information they will receive (maybe in an indirect way) might do the trick, like, if you have an INFO channel, maybe do not threat all INFO the same way but allow for different levels of detail, that way you can turn off much of the spam while still keeping the important stuff. But then again, there is the other extreme, and this I found the two or three times I tried to start a player in IRE games (not a flame but more like feedback), I did not find human interaction during the newbie process, it is very nicely done, well written and the storyline is appealing, but maybe because I just was checking them out and not actually eager to play or because of the feeling of disconnect from the actual world, I ended up breaking the link before actually trying the game.

Sum-up:
1.- I think it is to be expected that retaining players on text games is a difficult feat, mainly because graphical games are also available to those people and its easier to play with groups of people you know on graphics than on text.
2.- I think one of the big advantages but at the same time disadvantages of MUDs is that they require you to be immersed deeper and for longer time in the game. I believe most MUDs require more knowledge of details than the graphical counterparts (should I whirl whirl slash or whirl whirl thrust against this dragon? -GW2- vs should I click on sword or spear?)
3.- The initial spam may be overwhelming, a judicious review of how much of what a new player sees is actually important to that player during his first couple of hours in the game can make a difference between me staying in or out of a game.
3a.- Public channels are one of the first thing that may turn me off, if they are not reasonably moderated. Public channels that consist on illiterate morons insulting each other's mother most of the time mean Alt-F4 (or Ctr-C) for me.
4.- Both beginners and old timers can benefit from a helping hand in the first few minutes of the game, as an old timer I would ask for key help files so my start can be faster, as a beginner I would usually ask questions about how-to, and would relish when the person would give me links and helped me help myself.

I hope this is worth something for those of you studying ways to keep people in

[Hephos]

Good clients is the way to go! Easy to use, nice for the eye, and features that just don't drop you into a disguised telnet window.

Graphical maps, equipment windows, status bars, etc etc. Those things attract players more than a dull black window with spamming text and may keep them playing just a little bit longer to get the chance to become hooked.

Take a look at the BatClient we've built for batmud. Beautiful!
The official BatMUD site

IMO, adding illustrations into the client sounds like a great idea. Even if it is just in the starting areas or similar, just to keep people playing more of those few first critical minutes.

[shadowfyr]

Think I will throw in my two cents here and say that Hephos is right. The client makes a difference. Damn near everyone, except the purists who like "simple", who have ever tried Mushclient won't use anything else. Why? Because you can practically code a fracking mud using its script system, once you get around some of the quirks in how it handles certain situations. Mind you, that is an exaggeration, but one recent discussion has been to make a sort of single player "mud" as a plugin, which could walk the player through the steps of setting up the client to connect to a mud, configuring triggers to handle things that happen from a mud, etc. All possible because of recent additions to the debugging capabilities, specifically the "world.simulate" function, which sends data through the client as though it was being recieved in a packet from a server. The few things that bug the hell out of me is a) no GUI support, for things like specialized windows, etc., b) MXP support is limited, so more than one font, or inline images is not possible, and c) while some people have worked in the path search algorythms for one, no one has had the patience to actually code a mapper for it.

I would also like to add on option I think could help, even if it was just a support dll. A Scene Description Language based render engine. Yeah, I know, I know, its not a 3D game, but a text game. But... If you are going to use images at all, it doesn't make sense to me to rely on hundreds of megabytes (or gigs) of static images, or even what is almost certainly going to be gigs of image based texture data and mesh data, when the one thing that makes muds superior to graphical ones, supposedly, is that you can change things on the fly and quickly design new areas. Sure, it still takes a bit of time, but its a lot easier to alter the "color" or "size" of an object, or add a few extra bits to them, using primitives (or prims as SL calls them), than to spend hours adjusting the "shape" of a curve in something made for DirectX or the like. And still images... You definitely have to pay someone for, since that is as time consuming, and poorer quality a lot of the time.

Point being. Why use something that requires paying an expert to do it, and/or which you can't change when you *need* to? I really don't think either "standard" option for adding graphics makes much sense in a game designed to be changed "on the fly".

[the_logos]

Originally Posted by Hephos Good clients is the way to go! Easy to use, nice for the eye, and features that just don't drop you into a disguised telnet window. Graphical maps, equipment windows, status bars, etc etc. Those things attract players more than a dull black window with spamming text and may keep them playing just a little bit longer to get the chance to become hooked. Take a look at the BatClient we've built for batmud. Beautiful! The official BatMUD site IMO, adding illustrations into the client sounds like a great idea. Even if it is just in the starting areas or similar, just to keep people playing more of those few first critical minutes.

That is indeed a pretty client, though Batmud needs to work on their download speed to take advantage of it in the case of hooking people new to text MUDs. It took me over 5 minutes on a fast broadband connection to download their client and it's only a 5 meg client.

One question though: I notice it requires Java 1.5. What kind of penetration does Java 1.5 have, any idea? I found this breakdown of penetration but it's from 16 months ago. It's not pretty for Java. Java vs. Flash: Which technology dominates on the client side? According to the study that site ran, Java had slightly over 50% penetration, but half of that was Java 1.1.4. I can't comment one way or another on the likely accuracy of that study or how much the situation has changed in the last 16 months but that makes even Java 1.1.4 (which is what Nexus uses due to its higher penetration) less-than-ideal and makes Java 1.5 quite a bit less than ideal.

After looking into this some I'm starting to wonder if one reason our conversion rate between people clicking through to the IRE portal and people opening Nexus by hitting one of the 'Play Now' buttons is so low is because people are being asked to download Java and just leaving. Any time people have to download and install anything you're going to lose a large percentage of them.

We're going to run another test and measure how many people click Play Now vs. how many people actually manage to load up Nexus. Presumably most of the people lost between those two actions are lost because they don't have Java and choose not to download it.

I have to say, what would be awesome would be a Flash mud client. Flash is nearly ubiquitous (97%+ penetration for Flash 8 I think). Unfortunately I'm told that the cross-section of people who are both excellent Flash developers and who are good at software development is not very high. It's one thing to display pretty text with Flash and an entirely other thing to deal with telnet implementation, and possibly a server-side proxy due to the same kinds of permission problems that plague Java when run in a browser.

--matt

[the_logos]

Originally Posted by shadowfyr Think I will throw in my two cents here and say that Hephos is right. The client makes a difference. Damn near everyone, except the purists who like "simple", who have ever tried Mushclient won't use anything else. Why? Because you can practically code a fracking mud using its script system, once you get around some of the quirks in how it handles certain situations.

None of that really matters for the sake of interesting new players in text MUDs though. Asking prospective new text MUD players to download a third-party application is a no-go from a marketing angle. You need a really big 'carrot' to overcome the resistance people have to installing software and part of the marketing problem with text MUDs is that we lack a way to really effectively communicate what the 'carrot' is (depth, roleplay, etc).

Point being. Why use something that requires paying an expert to do it, and/or which you can't change when you *need* to?

I'm not sure how to respond to that. We pay experts to do all sorts of things, from art to engineering/coding to world building to facilitating roleplay to website creation to dealing with taxes, legal issues, and so on. I realize hobbyists are not in a position to do the same though and that it seems to be a lot harder to get volunteer help on the art side than it does on the building or coding side of things.

I really don't think either "standard" option for adding graphics makes much sense in a game designed to be changed "on the fly".

Our newbie intros are only changed every few years so that's not much of an issue (recall that I was proposing we might layer in illustrations only for the newbie intro).

it doesn't make sense to me to rely on hundreds of megabytes (or gigs) of static images

Hundreds of megs? Gigs? God no. For the purposes of attracting and keeping more true text MUD newbies there can be no extended download, which hundreds of megs or gigs would certainly require.

--matt

[Rathik]

Originally Posted by the_logos One question though: I notice it requires Java 1.5. What kind of penetration does Java 1.5 have, any idea? I found this breakdown of penetration but it's from 16 months ago. It's not pretty for Java. Java vs. Flash: Which technology dominates on the client side? According to the study that site ran, Java had slightly over 50% penetration, but half of that was Java 1.1.4. I can't comment one way or another on the likely accuracy of that study or how much the situation has changed in the last 16 months but that makes even Java 1.1.4 (which is what Nexus uses due to its higher penetration) less-than-ideal and makes Java 1.5 quite a bit less than ideal.

You can never be sure about statistics, but I have a feeling that site is completely off regarding Java. My own sites have been above 90% for both Java and Flash penetration for quite some time now. Adobe has the most reputable stats I could find for Java and Flash in March 2007 and April 2006 (bottom of page) with 98.7% Flash and 87.5% Java penetration for Mar '07, although there is no mention of the Java versions.

Personally I do think Flash mud clients would be more appealing than a Java client. I have seen so many more interesting effects done in Flash over a Java applet, and with Flash, I don't get that few-second hang as I do whenever I start a Java applet.

[scandum]

Originally Posted by the_logos I have to say, what would be awesome would be a Flash mud client. Flash is nearly ubiquitous (97%+ penetration for Flash 8 I think). Unfortunately I'm told that the cross-section of people who are both excellent Flash developers and who are good at software development is not very high. It's one thing to display pretty text with Flash and an entirely other thing to deal with telnet implementation, and possibly a server-side proxy due to the same kinds of permission problems that plague Java when run in a browser.

I got windows telnet working better than the average mud client for my own mud, but Microsoft decided to no longer enable telnet per default in Vista.

I'm not sure if luring players with pretty graphics is going to work very well. There are several free mmorpgs out there and someone might actually think, wow, this is the suckiest mmorpg I've ever seen!

Some nice ascii art could be impressive though.