I had a little free time on my hands time so I felt like making some considerations about life cycles in MMOs.
(off topic: the text formatting functions on these boards are not really optimal - euphemism :))
I already had the opportunity to mention or quote Dr Nick Yee several times on these forums.
Nick is a social researcher with strong statistical background.
In the early 2000 (being an EQ player himself) he started a project called Daedalus with target to analyse demographics and motivations of MMORPG players.
I have also participated on his project since Day 1. As Nick was player and researcher at the same time, I consider that his studies belong to the best compared to other academics who talk about MMOs without having ever played one.
During the first years he was using EQ to get his (statistically significant) samples and after 2005 he changed for WoW.
Contrary to the popular belief, this introduced little bias because the WoW population was not fundamentally different from the EQ population - admittedly his
samples didn't contain asiatic players that make up more than half of the world wide WoW player basis.
For example actually the average age of WoW players is slightly higher than the one of EQ players what puts a big hole in the myth that WoW players are immature kids :)
Today I would like to talk to you about the life cycle analysis.
It was done by Nick on a significant sample of 1860 EQ players.
He defined 5 player's life cycle stages.
Starting (a newbie player discovering the game)
Ramping up (the player knows the basics, has a clear goal and is actively progressing towards it)
Mastery (the player is in the end game and eventually raiding)
Burn Out (the player wonders where all the fun went)
Casual (the player spends relatively little time in game and logs in, well, casually. He is more motivated by the friends (social interaction) than by the game. Females more than males.)
First a very interesting remark about the "casual" category.
It appears that only a small number of players start as "casuals".
This population which finishes by largely overwhelming the total game's population is rather representing an asymptotic behaviour pattern to which players converge from all other categories and especially from the burn out category.
That's why Nick is calling it also "those who recover".
So how do these stages play out in time for the sample of players ?
Month 1 : Not suprisingly 90% are starting or ramping up. Only 8% are casual – those are what could be called the "hardcore casuals".
Personnaly I probably belong to this category.
Months 2 to 3 : The players considering themselves newbies crash down and are now only 12%.
The casuals increase to 20%. This period is dominated by 56% of ramping up.
Months 4 to 6 : No player considers himself as starting after 4 months (this doesn't mean that there are no newbies in game, don't forget that we are working on a fixed sample of 1860 players). The ramping up category is still majority with 40% but the casuals are strongly increasing to 31%.
Months 7 to 10 : The casuals take the lead with 37% and won't loose it untill the end.
Months 11 to 18 : The mastery has its peak with 35%, the casuals are 43%.
Beyond 19 months : Asymptotic regime. Casuals dominate with 54% followed by mastery at 29% and burn out with 12%.
What conclusions can be drawn from these observations ?
I do not want to bore you with a PhD a thesis so will put them as they come in no particular order :)
Even for a "difficult" game like EQ the players consider that 1 month is enough to know if they will continue or not.
Obviously for less difficult games this time is even less and may be measured in a few weeks.
After about 1 year of playing, the players sharply separate in a majority of casuals and a minority of mastery.
Other categories don't exist anymore besides about 10% of burn out.
Even if the sample are players, it is likely that the demographic structure of the game itself mirrors these time evolutions too.
The major consequence I see is that all expansions after about 1 year should privilege the casual players while providing something to the
mastery (raiding) population as second priority too. Many MMO I know get this priority wrong – they spend too much development time
with the mastery and not enough with casuals
The ramping up population which could be considered most stable because it is probably the one that
has most fun in the game doesn't automatically transform in mastery.
More than a half transforms in casuals and less than a half in mastery. After about 1 year no player considers himself as ramping
up so that the definitive demography of a game may be considered achieved in 1 year. This stresses (if it was necessary) the
paramount importance of the state in which the game is delivered during the first year. That means that the time invested in the
initial development has much more value than the same time invested in development after 1 year. In other words there is more value in
terms of customer retention to release a better game 1 month later rather than release it on schedule and invest the 1 month
development in an expansion 1 year later.
I guess I fit the curve in most MMO's I have played - strong ramp up period, end up casual or burnt out on it.
For me personally, EQ was the exception (friends dragged me away from VG for other pursuits and I only had an on-again, off-again involvement with that game... probably a year or so total mixed over the lifespan of the game.)
In EQ though - I can honestly say that I was always on that upward trajectory - always seeking to ramp up, get better... and that was over about 5.5 years or so.
Contrary to the popular belief, this introduced little bias because the WoW population was not fundamentally different from the EQ population - admittedly his
samples didn't contain asiatic players that make up more than half of the world wide WoW player basis.
For example actually the average age of WoW players is slightly higher than the one of EQ players what puts a big hole in the myth that WoW players are immature kids :)
I just wanted to comment on this statement....the age of the two player bases may have been similar but the maturity level of the games was worlds apart...In my short (one week about) play time in WoW it was the most vile idiotic and totally Chuck Norris filled stupidity I have ever seen. Dont recall the server I played on but it was an RP server as those are the only ones I play on if they are offered...The culture of the two were just so different its like comparing apples to oranges. Not saying WoW didnt have mature friendly people but the instant gratification crowd was just so overwhelming in that game. The Trolls were allowed to flourish and spew racist and homophobic remarks constantly in general chat...I mean honestly 'Barrens chat" didnt get that reputation from being all hugs and rainbows. I am not a WoW hater, even though it may seem like it, but the community of that game really did ruin it for me.
I actually started my MMO career as a casual, as in non raider, but I was hard core time wise. I only got into raiding because the group I played with starting getting into it and I was the group cleric so I just followed them. I spend lots and lots of time on my MMO of choice just have no desire to be "the uberest"
Sevens said:
Contrary to the popular belief, this introduced little bias because the WoW population was not fundamentally different from the EQ population - admittedly his
samples didn't contain asiatic players that make up more than half of the world wide WoW player basis.For example actually the average age of WoW players is slightly higher than the one of EQ players what puts a big hole in the myth that WoW players are immature kids :)
I just wanted to comment on this statement....the age of the two player bases may have been similar but the maturity level of the games was worlds apart...In my short (one week about) play time in WoW it was the most vile idiotic and totally Chuck Norris filled stupidity I have ever seen. Dont recall the server I played on but it was an RP server as those are the only ones I play on if they are offered...The culture of the two were just so different its like comparing apples to oranges. Not saying WoW didnt have mature friendly people but the instant gratification crowd was just so overwhelming in that game. The Trolls were allowed to flourish and spew racist and homophobic remarks constantly in general chat...I mean honestly 'Barrens chat" didnt get that reputation from being all hugs and rainbows. I am not a WoW hater, even though it may seem like it, but the community of that game really did ruin it for me.
I actually started my MMO career as a casual, as in non raider, but I was hard core time wise. I only got into raiding because the group I played with starting getting into it and I was the group cleric so I just followed them. I spend lots and lots of time on my MMO of choice just have no desire to be "the uberest"
I firmly believe this has everything to do with games fostering self reliance over community reliance. Even early on in WoW when there was at least some challenge to the game - I remember clearing to a quest mob or whatever, and as you are killing the mob right before the boss - someone running by you and attacking the boss. Say something, and you'd get a "It will respawn again in 3 minutes, just shut up and wait for it" - because, you know, THEY couldn't sit there and wait the 3 minutes.
But because they know they will never need you (or anyone else for that matter) - why do you matter? And so you can be "behind the keyboard rude" in WoW and the rest of today's MMO's.
Some people seem to think it's because of WoW's playerbase size that this happens - but I disagree - given that it happens in all the rest of them to some extent - regardless of player population size. No - it's definitely self reliance.
Counter that with a community reliance based game - where now you know you might need that person - or - that you know a bad reputation in general might hurt you... now you behave better, or you leave the game.
I think it's just that simple.
Sevens said:
Contrary to the popular belief, this introduced little bias because the WoW population was not fundamentally different from the EQ population - admittedly his
samples didn't contain asiatic players that make up more than half of the world wide WoW player basis.For example actually the average age of WoW players is slightly higher than the one of EQ players what puts a big hole in the myth that WoW players are immature kids :)
I just wanted to comment on this statement....the age of the two player bases may have been similar but the maturity level of the games was worlds apart...In my short (one week about) play time in WoW it was the most vile idiotic and totally Chuck Norris filled stupidity I have ever seen. Dont recall the server I played on but it was an RP server as those are the only ones I play on if they are offered...The culture of the two were just so different its like comparing apples to oranges. Not saying WoW didnt have mature friendly people but the instant gratification crowd was just so overwhelming in that game. The Trolls were allowed to flourish and spew racist and homophobic remarks constantly in general chat...I mean honestly 'Barrens chat" didnt get that reputation from being all hugs and rainbows. I am not a WoW hater, even though it may seem like it, but the community of that game really did ruin it for me.
Yeah i still play WoW when not raiding in EQ, but i hate the community for the most part. I don't remember it being too bad in Burning Crusade, but when i came back and started playing during Lich King i was disgusted. Then again, EQ's general chat isn't that much better over the last couple of years. I avoid general and server chat channels for that reason.
I think this kind of attitude is also caused as much by game design as it is by a maturity gap. There is no accountability for your actions in WoW, i am not talking about EULA enforcement but rather there are no real in-game consequences for acting like a jack wagon, i.e. your reputation on your server has has almost no impact on your game play, unlike EQ where a bad reputation could ruin a player's game play experience causing them to quit or server transfer or name change in hopes of a fresh start. Remember name change drama in EQ? Bad players would pay or beg to change their name (stalking, harassment, etc) not realizing that all the name changes were posted on the SOE boards.
Cross server group and raid finders removed any real fear of being black balled from guilds and groups or being put on no rez/help lists that many people maintained in old EQ. Furthermore the proliferation of smaller guilds in WoW, which came about because you could raid anything with 10/25 players down from the 40 man raids in vanilla wow also meant that guilds didn't need to police their players like the 100+ person sized guilds that were the norm in EQ in order in order to maintain the guild's reputation. EQ could be a harsh mistress, people learned to play nice or they got shunned. Another interesting thing about EQ i haven't seen in any other game was that it spawned huge server specific forums that were not moderated by SOE where people could air out all the dirty laundry, so it was never missed or easily swept under the carpet.
Sevens said:
Contrary to the popular belief, this introduced little bias because the WoW population was not fundamentally different from the EQ population - admittedly his
samples didn't contain asiatic players that make up more than half of the world wide WoW player basis.For example actually the average age of WoW players is slightly higher than the one of EQ players what puts a big hole in the myth that WoW players are immature kids :)
I just wanted to comment on this statement....the age of the two player bases may have been similar but the maturity level of the games was worlds apart...In my short (one week about) play time in WoW it was the most vile idiotic and totally Chuck Norris filled stupidity I have ever seen. Dont recall the server I played on but it was an RP server as those are the only ones I play on if they are offered...The culture of the two were just so different its like comparing apples to oranges. Not saying WoW didnt have mature friendly people but the instant gratification crowd was just so overwhelming in that game. The Trolls were allowed to flourish and spew racist and homophobic remarks constantly in general chat...I mean honestly 'Barrens chat" didnt get that reputation from being all hugs and rainbows. I am not a WoW hater, even though it may seem like it, but the community of that game really did ruin it for me.
I actually started my MMO career as a casual, as in non raider, but I was hard core time wise. I only got into raiding because the group I played with starting getting into it and I was the group cleric so I just followed them. I spend lots and lots of time on my MMO of choice just have no desire to be "the uberest"
It still amazes me that folks STILL tell the old tired Chuck Norris jokes after all these years. I didn't like the Wow crowd, I saw conversations where people were being nasty to others because they could, they were anonymous. I saw flame wars on their forums that were wayyyy over the top. Nerf this class, they beat me in pvp! (lol)
Discussions of date rape drugs and how to go about it, over General chat.
People who didn't like to use their own heads, and would spam chat for EVERY step in the particular quest they were on. Kill stealing en masse, just generally bad behaviour over all that turned me off completely from that game.
Person standing beside me yelling over chat..hey where's the mailbox. "At the Inn" (we were standing in front of the Inn....go figure. (duh?)
I wasn't impressed with the maturity level of the people I encountered there, not a bit.
I prefer the EQ communities, there your server rep meant something, you didn't get your group killed on a whim and then shrug it off "oh well". You thought before you acted.
The one part of his research that is glaringly missing, based solely upon what you outlined, is the segment that goes from burned out back to Mastery and back multiple times. I have been playing EQ since 2000 and followed his first 4 life cycles in a similar fashion, but that's where the similarities ended. I went from burned out to quitting to come back to Mastery immediately for a couple of years to quitting again to Mastery again where i have been for the last 4 years. I actually became more hard core after my first burn out phase than i was before and then even more hard core this last time around. I think his research is good, but i don't think that it is complete thought no fault of his own, i just don't think he thought people would continually leave and come back in the decade since he changed over to WoW, i know i didn't. I think if he stuck with EQ, he would have had to had another couple of life cycles for all the players who inevitably come back to the game after their life problems, school, real-life obligations, stints playing other games or burn out fades.
Sevens said:
Contrary to the popular belief, this introduced little bias because the WoW population was not fundamentally different from the EQ population - admittedly his
samples didn't contain asiatic players that make up more than half of the world wide WoW player basis.For example actually the average age of WoW players is slightly higher than the one of EQ players what puts a big hole in the myth that WoW players are immature kids :)
I just wanted to comment on this statement....the age of the two player bases may have been similar but the maturity level of the games was worlds apart...In my short (one week about) play time in WoW it was the most vile idiotic and totally Chuck Norris filled stupidity I have ever seen. Dont recall the server I played on but it was an RP server as those are the only ones I play on if they are offered...The culture of the two were just so different its like comparing apples to oranges. Not saying WoW didnt have mature friendly people but the instant gratification crowd was just so overwhelming in that game. The Trolls were allowed to flourish and spew racist and homophobic remarks constantly in general chat...I mean honestly 'Barrens chat" didnt get that reputation from being all hugs and rainbows. I am not a WoW hater, even though it may seem like it, but the community of that game really did ruin it for me.
I actually started my MMO career as a casual, as in non raider, but I was hard core time wise. I only got into raiding because the group I played with starting getting into it and I was the group cleric so I just followed them. I spend lots and lots of time on my MMO of choice just have no desire to be "the uberest"
One can make "Statistics" say anything.
That is a wrong factoid. By 2005, EQ was already 6 years old and most of the oldschool peeps where leaving, while the PoP uber macro pumped-up testosteroned filled kids were grabbing hold of Smedley and adding bunny ears to EQ.
Or the fact, that nearly EVERY EQ PLAYER, also owned a WoW account. The whole world tried it..
while the study seems interesting i think the sample figures are waay to small to say that it in any way can give an accurate representation of WoW players. esp for things like average player age. 1860 players is not even 1% of the assumed 5.5 million players who had subs in 2005
http://www.statista.com/statistics/276601/number-of-world-of-warcraft-subscribers-by-quarter/
the sampling is roughly the same as me asking 20 people in my town of 50k a question and then stating that their answers are representative of the populace.
also i would argue that since the survey would be something they signed up for, your average 10 - 15 year old simply would not have any interest.
I know the data is skewed, because at the time I had done my own collection of data, for months. (I helped a ton on MMOCharts back when).
We even had age polls on WoW's forums, that showed that average age of a WoW player was in early twenties and growing younger. Blizzard themselves was helpless in obtaining the actuals, because someone 18 or older had to buy the game and continue the subscription, even if a grandchild was playing.
Gawd said:
while the study seems interesting i think the sample figures are waay to small to say that it in any way can give an accurate representation of WoW players. esp for things like average player age. 1860 players is not even 1% of the assumed 5.5 million players who had subs in 2005
That's why I said that Nick had a STRONG statistical background. Btw the 1860 sample was used for the player's life cycles. Statistics about age used an even larger sample.
the sampling is roughly the same as me asking 20 people in my town of 50k a question and then stating that their answers are representative of the populace.
No this would not be at all the same even not roughly. The size of a representative sample is not simply proportional to the size of the population (why should it be ?) . So the size of a representative sample for 50 k population is not 100 times smaller than the size of a sample of 5 million population. The dependence goes actually like a square root.
The representativness is not so much in the sample size but in the way of selecting the sample. In this instance when the life cycle is a sequence of 5 states randomly distributed in the overall population, 1860 is actually overkill.
A bit of technique is now necessary to appreciate this kind of studies.
People who are not familiar with probability maths often think only in binary terms : "This statistic is right." or "This statistic is wrong".
This is not how it works.
When deriving a result from a given sample, this result will have a margin of error (an interval where you want to be almost sure that the variable for the whole population lies).
That means that if your sample A says Average age = 25 years with a margin of error of 5 % then you want to be sure that the Average age of the whole population will be between 23.75 years and 26.25 years.
If you use then another sample B of another population with a margin of error 5% and find Average age = 27 years then the average of the whole population will lie between 25.65 and 28.35
Now if you suppose that A is EQ for some period (f.ex 2002) and B is non asiatic WoW for some period (f.ex 2007) you can conclude with a high confidence that non Asiatic Wow players after 2 years of game were older than EQ players after 2 years of game because there is only a very small overlap between the 2 intervals.
As one cannot be 100% that the whole population lies inside the error margin, one uses the confidence level which says what is the probability that your population will lie inside the error margin interval computed for the sample. This level is set generally at 95% of 99% (you want to be pretty sure that you know what you are talking about)
So as an example if you want to know the answer of 1 million people on a yes/no question with an error margin of 5% and with confidence level of 95%, you will need a sample of ... only 400 people !
The sizes of representative samples are in most cases far, far below of what people imagine.
Hieromonk said:
I know the data is skewed, because at the time I had done my own collection of data, for months. (I helped a ton on MMOCharts back when).
We even had age polls on WoW's forums, that showed that average age of a WoW player was in early twenties and growing younger. Blizzard themselves was helpless in obtaining the actuals, because someone 18 or older had to buy the game and continue the subscription, even if a grandchild was playing.
Well it was not really skewed. To get a 95% confidence level for a population of a few thousands of people, around 200 sample is enough.
Back then I helped Nick with the age question by simply asking randomly in chat around 200 people on 2 EQ servers I was playing on.
This didn't take much time at all.
The number I got was not only consistent between the 2 servers but also with Nick's figure (within an error margin of 5%).
I didn't do this check for WoW but there is no reason to think that the result would have been different.
A forum would work too provided you got at least 200 answers. However the error margin would be bigger because the sample is not 100% random.
Hieromonk said:
I know the data is skewed, because at the time I had done my own collection of data, for months. (I helped a ton on MMOCharts back when).
We even had age polls on WoW's forums, that showed that average age of a WoW player was in early twenties and growing younger. Blizzard themselves was helpless in obtaining the actuals, because someone 18 or older had to buy the game and continue the subscription, even if a grandchild was playing.
yeah and this data is further skewed by a tendency of the younger generation of gamers to lie about their age in order to run with the older gamers. nearly all of the raid guilds in every MMO i have played require you to be 18+ which develops the habit of younger gamers to lie about their age.
Gawd said:
Hieromonk said:
I know the data is skewed, because at the time I had done my own collection of data, for months. (I helped a ton on MMOCharts back when).
We even had age polls on WoW's forums, that showed that average age of a WoW player was in early twenties and growing younger. Blizzard themselves was helpless in obtaining the actuals, because someone 18 or older had to buy the game and continue the subscription, even if a grandchild was playing.
yeah and this data is further skewed by a tendency of the younger generation of gamers to lie about their age in order to run with the older gamers. nearly all of the raid guilds in every MMO i have played require you to be 18+ which develops the habit of younger gamers to lie about their age.
I was in a guild in Wow where there were a couple of 8 year olds, with fully capped characters. Its not something that can really be regulated if their parents sign them up.
That's life.
I have a couple of questions about his study. Did he give a measure of hours per week that each life cycle was played? For instance, many would call me a hard core gamer because I spent close to 40 hours a week or MORE playing EQ for almost 6 years but I classified myself as a casual gamer because I had no interest the contest of seeing which guild was best by doing such and such a raid/boss. I spent most of my time doing quests and enjoying the company of in game and real world friends.
It is possible to have life cycles of players without knowing how long those said players actually spent playing the game. Of course it would help tremendously if there was a break down of the time it took to maintain said life cycle in hours per week or such...
Deadshade said:
Gawd said:
while the study seems interesting i think the sample figures are waay to small to say that it in any way can give an accurate representation of WoW players. esp for things like average player age. 1860 players is not even 1% of the assumed 5.5 million players who had subs in 2005
That's why I said that Nick had a STRONG statistical background. Btw the 1860 sample was used for the player's life cycles. Statistics about age used an even larger sample.
the sampling is roughly the same as me asking 20 people in my town of 50k a question and then stating that their answers are representative of the populace.
No this would not be at all the same even not roughly. The size of a representative sample is not simply proportional to the size of the population (why should it be ?) . So the size of a representative sample for 50 k population is not 100 times smaller than the size of a sample of 5 million population. The dependence goes actually like a square root.
The representativness is not so much in the sample size but in the way of selecting the sample. In this instance when the life cycle is a sequence of 5 states randomly distributed in the overall population, 1860 is actually overkill.
A bit of technique is now necessary to appreciate this kind of studies.
People who are not familiar with probability maths often think only in binary terms : "This statistic is right." or "This statistic is wrong".
This is not how it works.
When deriving a result from a given sample, this result will have a margin of error (an interval where you want to be almost sure that the variable for the whole population lies).
That means that if your sample A says Average age = 25 years with a margin of error of 5 % then you want to be sure that the Average age of the whole population will be between 23.75 years and 26.25 years.
If you use then another sample B of another population with a margin of error 5% and find Average age = 27 years then the average of the whole population will lie between 25.65 and 28.35
Now if you suppose that A is EQ for some period (f.ex 2002) and B is non asiatic WoW for some period (f.ex 2007) you can conclude with a high confidence that non Asiatic Wow players after 2 years of game were older than EQ players after 2 years of game because there is only a very small overlap between the 2 intervals.
As one cannot be 100% that the whole population lies inside the error margin, one uses the confidence level which says what is the probability that your population will lie inside the error margin interval computed for the sample. This level is set generally at 95% of 99% (you want to be pretty sure that you know what you are talking about)
So as an example if you want to know the answer of 1 million people on a yes/no question with an error margin of 5% and with confidence level of 95%, you will need a sample of ... only 400 people !
The sizes of representative samples are in most cases far, far below of what people imagine.
If you want to know what a million people would answer in a yes or no question survey, you simply ask one million people.
What you see in MMOs is really no different from what you see in any of life's activities that require a commitment over a long period of time. On any normal distribution, you'll get the left side being the ADHD folks who get bored of everything in 5 minutes or less, and at the other end you have folks who have that infinite patience to play a game because well, they rolled a character so that's the next ten years because well, I rolled a character after all.
MMO players generally tend towards that long term commitment end of the scale, same as the folks who play 4X strategy games, RPGs like Skyrim/Dark Souls that actual completion requires 500-1,000 hours, etc. The number of people still playing AO, EVE, EQ1, and other non-WoW timesinks shows that even if older and less shiny, the right side of that commitment bell curve will stay hooked if they have a reason to be.
Expecting "first 4 years of WoW" numbers is wildly unrealistic, because the nature of the game P:RotF is trying to be will simply not have the same ADHD mass appeal that Blizzard's loot pinata mindset has. So P:RotF is going for the segment of the study that starts off pretty committed, stays committed, and doesn't burn out easily but does get bored. How to keep that player interested, not feeling like they are on an endless treadmill, but are doing some manner of meaningful advancement or progression the majority of their logged in time...that's the challenge.
Lots of PC gaming niches. P:RotF will speak to one particular niche. The entire MMO playerbase is not the market for this game. Never was.