When I started my research, I had only one measure of success available – the average number of players from Steam Charts. I later obtained data about owners from Steam Spy hoping to simply infer owner numbers from player numbers. However, there is not enough correlation between these two types of metric. I don’t believe it’s due to an error. There are, for example, games sold in bundles or just very cheap games. These get high sale figures but not necessarily profit or player base.
A bit of an extreme example could be Unknown Battle which had around 130,000 owners two months after its release. However, this is a $1 game and the store page suggests that possibly less than 90 % of users bought the game on Steam. Most activations likely come from giveaways etc.
Another example is Gun Rocket. Again, a $1 game but people actually bought it which may be part due to it being in two bundles. It looks like the presence of trading cards is an important factor. It can, in fact, be statistically proved that games with trading cards have generally higher player numbers.
Is it then a good idea to measure how successful a game is by its owners? 3rd party activations and bundles can make this metric very unreliable. It would be better to know the profit rather than just number of sold copies but this would be close to impossible to obtain, especially for older games. Hence I came to the conclusion that sticking to average number of players will be better. Not only is it more reliable but can also help in areas such as multi-player support.