The PC gaming utility, Raptr, is normally used to optimize in-game settings, chat and socialize, and record game footage. It also keeps track of game-hours and aggregates them into a list once per month, which makes it one of the few sources for this type of data on the PC. We were late on it last month, which means that another was posted just a week later.
April marks the release of Grand Theft Auto V for the PC. It went live on the 14th and, despite only counting for half of the month, ended up at 4th place. Next month's survey will tell us whether the post-release drop-off was countered by counting Grand Theft Auto for a full month, which is double what they have now. It was just 0.17% of global play time behind CS:GO. Despite an error on the graph, it knocked DOTA 2 down to fifth, and Diablo III down to sixth. In fact, just about everything below Grand Theft Auto V dropped at least one rank.
Only three games actually gained places this month: ArcheAge, Warframe, and Spider Solitaire. Yes, that game is now the 19th most played, as tracked by Raptr. You could sort-of say that Hearthstone gained a rank by not losing one, but you would be wrong. Why would you say that?
League of Legends dropped less than a percent of total play time, settling in at about 21%. This is just about on target for the game, which proves that not even Rockstar can keep people from having a Riot.
#1Check It ..someday I’m sure
#1Check It ..someday I’m sure
Hah! We’ll see. (AlbertJ’s
Hah! We'll see. (AlbertJ's referencing a game that I am developing on Twitch.TV.)
Steam’s stats:
1. Dota2 –
Steam’s stats:
1. Dota2 – 391,313,250 hours (last 30 days)
2. CS:GO – 216,615,986 hours (last 30 days)
3. GTA V – 109,194,052 hours (last 30 days)
Raptr’s statistics are bull.
Honestly why do you still post this stuff?
It’s embarrassing enough for Raptr to release these statistics, as all it shows is that their sample size (their user base) is not large enough to reflect the real data.
I believe you somehow forgot
I believe you somehow forgot that steam doesn’t track any non-steam game, and thus wouldn’t register anyone playing a game like League of Legends.
You missed i think what he
You missed i think what he was getting to. According to steams stats, DOTA2 has almost 2x the # of hours CS:Go has yet raptr has CS #3 and dota as #5. Those 2 games are steam based games so it can call in to question a lot of other games on that list if they are really where they should be.
Their sample space is plenty
Their sample space is plenty large enough to get a great statistical precision. The question is whether there's "systematic error" that throws it off. Specifically, is there something different about a Raptr user than whatever group you associate it with?
Likewise, a Steam user is also different from a typical PC gamer because it doesn't account for other publishers, including Blizzard and Riot. For instance, I think Steam's figures do not track Grand Theft Auto V play time if the sale was made through Rockstar Warehouse or from a retail store (physical copy). If 99% of users purchased the game on Steam, that will probably be a minimal impact. If only 50% did, then your data is pretty much worthless, regardless of how many Steam users you include.
You really do not need a large sample size to represent a whole space. What you need is for the sample space to be as random and unbiased as possible. If that's the case, you will converge to "the real data" pretty quickly. The question, again, is whether a Raptr user is somehow different, on average, from whatever user you believe the study portrays. If you want to account for those who purchase physical copies or use other stores, then Raptr might be more accurate than your figures.
You may be correct about GTAV
You may be correct about GTAV not being fully tracked by steam, however DOTA2 and CS:GO can only be played through steam.
Considering DOTA2 and CS:GO are the most played games on steam (by a ton), you would hope that Raptr’s user base could at least reflect those two properly, but it cant.
This causes me to worry in 2 ways:
A) If Raptr can’t even get an accurate measure of steam’s largest two games, how can I be confident they have any accurate measure of non-steam games?
– e.g. I know DOTA2 is bigger than CS:GO by a long way, so why would I trust Raptr saying that WOW is bigger than DOTA2?
B) The statistics of smaller games recorded by Raptr (such as Spider Solitaire that you keep mentioning) probably do not reflect PC gamers as a whole in any accurate way.
– e.g. if DOTA2’s 391,313,250 hours = 4.96%, that means Spider Solitaire has 65,481,854 hours, which I’m pretty sure is bull
So from this I conclude: Raptr’s user base is not random enough or unbiased enough to rank these games in an order. So why bother post their statistics?
Raptr’s statistics is misleading when it comes to DOTA2, and CS:GO, and calls in to question the rest of the data. (Yes these statistics are regarding raptr’s user base, but people on the internet are going to use it assuming it reflects the entire pc user base.)
Presumably Steam’s stats also
Presumably Steam’s stats also fail to include people running cracked/bootleg versions of Steam games without a Steam account. I’d assume neither Raptor nor Steam tell the whole story.
Basically yes, but be
Basically yes, but be careful. The point is what do you think the whole story is? Are you assuming this represents PC gamers as a whole? It doesn't. It represents an average Raptr user just like Neilson Ratings represents a Neilson user, not a "TV watcher".
If a Neilson user is very similar to an average TV watcher then you can pull some interesting data out of it, even with just a few thousand users. If it doesn't, then you need to think about how it's different before you interpret the data. You would then test your hypotheses by creating new experiments that cut around your assumption and see whether it behaves in the way that you expect. If it doesn't, then either your hypothesis or your experiments need adjusting.
You'd be surprised how few professionals, who are supposed to be doing this for a living, get this even remotely right.
raptr DOES work on alot of
raptr DOES work on alot of bootleg/pirated games.
To get the full story with
To get the full story with complete stats we really should ask Microsoft or the NSA.