Some new stats have surfaced regarding player usage data and playing habits. The stats are corralled to look only at people who log into Xbox Live to play online and only people who play their games online. But according to the stats it looks like backward compatibility isn't played much by those who like to play their games online. Which would seem to support recent claims made by Sony about why they have never focus on it.
Ars Technica collected the data from September, 2016 to February, 2017 from an API site that monitored millions of Xbox Live accounts before Microsoft privatized the data. The article is prefaced with plenty of caveats about the margins of error.
With all the caveats out of the way, they revealed that of the 930,000 Xbox One Xbox Live accounts they monitored, only 1.6% of their time was spent on Xbox 360 games. 54% of Xbox One owners spent their time on Xbox One games, while the rest played non-gaming apps and activities. This would presumably make it seem as if what Sony's chief financial executive, Jim Ryan, said was true about backward compatibility not being very popular.
He claimed that the old PSX and PS2 games just didn't look very good and that based on their own data it wasn't a popular feature at all. While on the outset the Ars data might seem to back up his point, there's a lot of other factors to take into consideration before thinking that no one likes backward compatible games.
For instance, some of the data was over-represented from the sample sizes, such as Halo 3 and Kinect Adventures, while other data seemed underrepresented, like the lack of play-time and ownership of LEGO games on Xbox consoles. I tend to doubt that Kinect Adventures is more popular among Xbox owners than LEGO titles... right? As they mention, according to the data provided by the API, the motion-based title would mean that more than 21% of all Xbox owners have it in their library, more-so than Halo 4, Assassin's Creed, Battlefield 3, Skyrim and Left 4 Dead 2. If we were judging by those raw numbers it would mean Kinect should have been way more popular than it is.
Of course, this is sample data based strictly on people who log online, people who have these titles attached to their Xbox Live profiles, profiles that aren't set to private, and consoles that were regularly pinging updates to Microsoft's servers for the data to be collected by the API.
This also doesn't account for the fact that a while back there was an error some Xbox One owners had encountered where they couldn't connect to their Xbox Live accounts to save the data of their Xbox 360 games. It would go to download the profile and then stop due to a networking issue that some gamers had run into. So anyone who played their backwards compatible games (even though they couldn't save the data) also wouldn't have counted toward the aggregate figures.
Gamers also playing across multiple profiles where not every profile may be signed online would also fail to show up. Kids playing offline games also would not register across the million or so user tags collected in the stats. Alternatively, kids playing online games through their parents' account would also bloat the data, so in a household where a family shares an XBL profile, and one kid plays Call of Duty disproportionate to the other members of the household, Call of Duty would show up as the preferred game on that profile due to the high usage.
Essentially, what the data tells us is that not a lot of people actively playing online put in a lot of play-time with backward compatible games. This would make sense given that the data Ars collected from XBL profiles showed a huge variance of play-time toward mostly online-only games or games with regularly updated multiplayer, such as Call of Duty, FIFA, Battlefield and GTA Online. However, all of those games are dwarfed by the amount of time that Xbox One owners spend using their console to watch Netflix and TV.