The PC gaming utility, Raptr, keeps track of per-game play time across each of their of their tracked titles. Because it is not locked to Valve, Blizzard, Riot Games, Mojang, and so forth, it compares games that are from different publishers and distribution platforms as long as the software is running. Around once each month, the company shares their findings and gives brief explanations for notable results. Again, these are not sales or download figures. This ranking is decided by the number of hours played.
First, League of Legends continued its reign as most played PC game; in fact, it widened its lead to over one-fifth of all recorded game time (20.55%). This increase was mostly attributed to the game's 4.15 update. Second place, with a significantly less 7.62%, is World of Warcraft. Raptr believes it passed DOTA 2 for two reasons: WoW gained players from their Mists of Pandaria 50%-off promotion and DOTA 2 deflated a little bit after the swell from The International tournament.
Counter-Strike: GO held steady in fourth place and Smith Smite (Update 09/17/2014: Corrected typo), a free-to-play MOBA from Hi-Rez Studios, jumped five places to fifth place.
I don’t know anyone that uses
I don’t know anyone that uses Raptr, even AMD GPU owners.
Raptr claims to have 23 million users, while Steam claims to have 75 million users… Best case scenario, Raptr is still missing metrics for 50 million Steam users.
If the sampling is random
If the sampling is random enough, millions of users can be extrapolated. Heck, Ars Technica was close with their "Steam Gauge" sales figures at only ~100,000 users.
I guess the big question is whether "Raptr users" are characteristic of "every PC gamer". If not, who does it best represent and what's the systematic error introduced?
I believe the sampling of
I believe the sampling of users by Raptr over-represents “dumb” users who have no clue about how to use a computer. This tips the balance in favor of games like current-day WoW that would only be played by less educated people.
really? I hope you stub your
really? I hope you stub your toe.
actually i totally agree with
actually i totally agree with him there is no use for raptor
A few points, as the PR flak
A few points, as the PR flak for Raptr:
1) Raptr now has an installed base of >30 million, the 23 million figure is a bit outdated
2) Steam is great, we love Steam…but Steam only tracks Steam games, whereas Raptr tracks many thousands of games not found on Steam, as well as tracking virtually all Steam games
3) To Scott’s point, you don’t need to survey/track every member of a given audience (in this case, the PC gamer audience, which in fact is well over 1 billion people worldwide) in order to make projections about that overall audience. 30 million out of 1 billion is about 3% of the total audience. By comparison, CNN polls just 500 people in its surveys and calls that a “representative sample” despite the fact that 500 people only make up .000154% of the population.
And another thing to mention
And another thing to mention is the way statistics work, the bigger the sampling pool the more accurate the results usually are. Going from 23 million users to 24 million is pretty drastic for confirming accuracy.
The key is good randomization unless you really poll all accounts.
They are playing the wrong
They are playing the wrong games.
Nah. People can play what
Nah. People can play what they want. Most of these games are not for me, but I do get some that are.
http://en.wikipedia.org/wiki/
http://en.wikipedia.org/wiki/Margin_of_error
1,067 people polled give a margin of error of just 3%.
this is standard polling practice. any large sample may as well be “everyone” was polled.
That is only true if the 1067
That is only true if the 1067 people are chosen randomly. For example, if you randomly select 1067 raptr users, the result would be within 3% of the values presented here.
I just doubt that raptr is an unbiased sample of all gamers. If your selection method is biased, that 1067 people are just as good or bad as 100 or 30m. And if I look at the difference between 1st and 2nd place I think there’s a strong overlap between raptr users and LoL players. Maybe a LoL player could tell me why raptr is so popular amongst LoL players.
Well the main thing is “What
Well the main thing is "What is this study trying to find". Its goal is to rank total playtime. This will bias popular games with long and frequent playsessions.
Is this a bad bias? Of course not, it is the metric that they are looking for. If they were looking to replicate sales figures, it would be a bad study because it is weighted by session length. Likewise, if they were trying to find out how much time a game has been played, looking at sales figures would be a terrible idea in that case. Again, it all depends on what you're looking for and, in this case, what they're looking for is suited to MOBAs and MMOs getting a high rank.
Raptr is a pretty good
Raptr is a pretty good sampling of “all PC gamers” given that we have >30 million users in more than 100 countries worldwide. As one example, about 10% of our installed base is in China, so overall it’s a fairly global audience.
It sure makes sense to me that LoL would capture approximately 3x the “share” that WoW commands, given that per Blizzard’s and Riot’s own data, WoW has about 7 million MAUs and LoL has 10x that many. Of course, all of WoW’s users are paying customers, whereas just a fraction of LoL’s users spend money each month…but we don’t claim to be tracking ‘monetary success’ or etc, just popularity based on time spent.
Never heard of this Smith, a
Never heard of this Smith, a new free-to-play MOBA from Hi-Rez studios….is it like Smite 2.0 or something?
Hah, whoops! Fixed.
Hah, whoops! Fixed.
WOW is #2? Really?
WOW is #2? Really?
Smite is awesome, i highly
Smite is awesome, i highly recommend it to everyone. It’s free to play too so why not try it?
Absolutely gorgeous graphics too.