Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Quick Links
XFire - As MMO population estimation tool
Yamota
Member UncommonPosts: 6,593
The purpose of this thread is to discuss XFire numbers and from those try to extrapolate MMO playerbase numbers from. The mods of this site said they would support such a thread and sticky it so here is hoping they will.
So lets start by qouting what is said on the site: http://www.mmodata.net/ about X-Fire:
"I just got the Q4 quarterly report from EA, and they report 1,3m subs, so I will add this. My latest estimate of 1,41m subs was pretty generous, however I always try to be conservative in my estimates, whether they are downward or upward.
What it does show me is that the Xfire numbers are a good indication to base estimates off, purely based on Xfire I would have estimated 1,2m."
So with that in mind, let us start with some theory-crafting. This is the method I use to get estimated player base numbers. I take a game with a known playerbase and from those games calculate what I would call the X-Fire factor. As an example I will use Eve estimated playerbase of 360k subs (only counting the subs in the West as I believe X-Fire is not used much in the East).
Then I divide this sub. number with the X-Fire played per day which for Eve is currently at 690.
360k/690 = 522 (rounded)
So 522 is the X-Fire factor which I will use to calculate the playerbase of other games. It would be better if I had another game to calculate the same for but currently I dont know any known sub. numbers of any other game. But this is just an example.
So lets use this number to estimate another MMOs playerbase, such as SW:TOR which has a players per day number of 1.471. Multiply this with the X-Fire factor to get an estimated playerbase of SW:TOR:
1.471*522 = 768k (rounded)
Same method for GW 2 gives a playerbase of 2 million.
Reasonable? I think so. Keep in mind that the released sub. numbers for SW:TOR was way back in Q1 2012. And GW 2 has sales of 2 million a month or so after release.
Comments
As much as you tried to use Xfire as a reliable method to estimate subscriptions, it's actually a poor idea.. Xfire is a website that attracks a certain type of player.. So naturally the games that are played by that segment of players will be skewed.. Using Xfire as a general population sample is like Detroit to estimate the USA..
I think Xfire is best for showing population trends and generally a game on xfire with more hours played has more players than a game with less.
Look at WoW, GW2, SW:TOR being the top 3 MMOs...seems pretty accurate.
TSW just had a bump in hours due to B2P, Rift had a bump during storm Legion.
Xfire is very good for very general estimations.
8 people have walked past my window in the last hour. From that I can deduce that the town I live in has 55k population. The facts prove it!
Sadly this is more likely to be accurate than the usual x-fire numbers and MMO population bullshit.
Trying to take a tiny sample of a skewed pool and applying to a broad spectrum is inherantly flawed.
^^^ This ^^^
It in no way should be used to determine actual player numbers because as has been said, not everyone uses it.
Based on hours played from most populated to least it goes...
WoW, GW2, SW:TOR, Aion, EvE, TSW, Rift, TERA then like 50 dead MMOs after that.
That seems pretty accurate to me... it gives a general idea, a pretty good one for the most part IMO. It always has shown trands, when a game is gaining players or losing players. Every event GW2s numbers shoot up, F2P SWTORs shot up, B2P TSWs shot up. And they all remain in a realistic position based on their playersbase relative to the other MMOs.
I have read this argument on many of these xfire threads but I have seen anyone post anything that would back this up.
What type of player does XFIRE attract and how does this affect MMOs?
Which MMOs have skewed results?
I would not expect anyone to be able to produce any evidence, but at least a logical argument of why one MMOs population figures would be skewed positively and one negatively.
The thing you have to remember is that in order to use a sample to predict anything you have to understand that sample and be sure it is representative of the information you are trying to extrapolate.
I do not know a single person in RL, or after discussion in the MMOs I play have ever talked to anyone who used x-fire. Now that doesn't mean no-one uses it of obviously but it does mean that the people who use it are likely to be from a very specific and niche demographic.
This makes any use of this as a tool inherantly inaccurate unless you can understand the sample itself. When numbers change in x-fire it is only showing the trends of a specific demogrpahic, not a broad spectrum of trends.
Every survey is related to specific demographic.
I understand what you are trying to say but you are wrong and your point moot.
I have never seen anyone produce any evidence that would back up this statement.
I have seen plenty of evidence that would suggest this is not the case when applied to MMOs
Yet, if you do not understand that demographic anything you try to do with that data in inherantly flawed.
This has been discussed to death in many threads. The only people who try to cling to x-fire as a population estimating tool are the ones who want to use that data to make a point.
Anyone else who has had an education in statistics knows it simply cannot be used that way or as proof of anything without knowing the sample. I'll leave it at that as you do not seem to understand anything about statistics.
Ok seems the discussion been about XFire's validity or not which was not the point of the thread as that discussion has been done many times and never lead to any conclusions.
So to get back on track. around 800k for SW:TOR, 2 Million for GW 2. Sounds pretty reasonable to me.
SW:TOR had a big bump during its F2P conversion but has since then come back down again. And it seems GW 2 has pretty much stabilized, for now, at around 2 million.
Secret World got a big bump when it went B2P. Around 50k subs to now about 150k active playerbase.
My gaming blog
Again, point of this thread is not to discuss the validity of it. So why are you bringing it up if you think it is so pointless to discuss it?
Sounds like topic-hijacking to me.
My gaming blog
except we know this is complete crap.
EQ obvioulsy has a LOT more than 8874 people
EQ2 has a lot more than 17226
EQ2 is certainly over 100k and EQ1 is likely not thar far from 100k.
Rift likely has at least 3x the 62,000 your estimation would suggest.
Note that EQ1 has an extremely older player base, EQ2 far older than average. and Rift somewhat older than average. Your method woould be off on both EQ and EQ2 by a factor of more than 10, and Rift by a factor of probably around 4.
These three games are proof enough that one figure wont estimate all games populations.
This thread title is "Xfire - As MMO population estimation tool" so you can expect its inability to perform that function to be a the topic of dicussion.
If i went on to a heavy metal music forum and started a thread "The Beatles as a Heavy Metal band" how do you think those replies would go?
That's the thing, you are only parroting something that sounds smart regardless of application.
You do not always use and need complete sample in statistics...all is a matter of probability and accuracy.
If XFire can't be determined to be a valid source of information or a valid, repeatable process can't be determined to go from the XFire source to the goal of Game Population numbers, what exactly is the point of the discussion? It seems to me the validity of XFire as a population estimation tool is a core part of the discussion.
** edit **
That's why the mods want a stickied thread...so they can get all of that discussion in one place.
I can not remember winning or losing a single debate on the internet.
No, you do not need to use a complete sample, as long as you understand the sample you have whic is impossible with x-fire. That is quite simply the point I have been trying to make and it seems, deep down at least, you understand this yourself.
There is a reason WoW is not used along with these types of estimation. GW2 would have an active playing population of over 5 million if the OPs method was used so it is disregarded thus showing the inherant flaw.
GW is healthy but obviously not getting all those players so it illustrates exactly why this type of thread is white knighted by the few and understandably taken as unreliable by the rest.
Wrong.
"factor 522" represents % of EVE players using xfire and no more.
You basically say that because certain % of EVE players use xfire, same % is used by other game which is something you have no base for - you only proved xfire <-> EVE online relation.
Xfire numbers can indicate trends or very roughly population size within same sample.
Irony and funny is that your argument why xfire numbers are "impossible" is that the sample is incomplete...
I am not saying OP method is correct but I say that you are not correct either. Xfire numbers can and do provide indication of trends and population estimates within same sample(MMO) because there is empirical evidence of such correlations.
Actually I do think it can be used to look at trends within the sample. The part where we disagree is I do not feel the x-fire sample is representative of the people who play MMOs across the board.
I agree - we have discussed this AD NAUSEUM!! XFire IS NOT a good tool for populations. we don't know the % of players for each game using it - some maybe 0% some may be 100%. It is not useful at all. Only if you know the EXACT % of players of a particular game using XFire, then it may be useful. Otherwise, it is just a measure of XFire users.
I agree. It was clearly an infliuence of GW2 fans that were pushing to have these topics closed. However, I wonder if it's still too soon. Still a lot of rabid defenders around.
Xfire numbers are one way (not particularly good, mind you) to estimate population trends (as in, population incresing or decreasing), but not as a general population guide.
First off, a disclaimer, I don't work professionally with statistics, but I have taken several courses about it in college, so take with a grain of salt. For any data to be statistically significant it must comply with several criteria, a big one of them is that the sample size must be completely random, as in, any member of the total universe being studied must have equal chances of becoming part of the sample, and any serious discussion about statistical data must deal about what measures were taken (if any) to ensure this.
Xfire users are not nearly a random sample, since there is a specific target market for it. The fact that the numbers for total population for EA (he doesn't mention a specific game, does he?) fit the expected from Xfire only means that their average player fits the Xfire user profile. (just to mention 2 games), maybe players of WoW fit the profile, while maybe players of EVE don't, in which case the numbers will be skewed towards WoW. Likewise, while the numbers might be used to estimate players increase or decrease, it might also mean just increase or decrease within the players that fit the Xfire user profile.
What can men do against such reckless hate?