Alex

@ResearchVoyager: Thanks for the comment! One reason I chose to tackle gaming was that there's a ton of structured data that hasn't been leveraged well. A lot of the stats and probability models that power Gamerloop could be extended to other industries. There's a ton of stuff coming down the pipe, but the way I think about things is in terms of data derivatives (1st order = achievements, 2nd order = difficulty / completion rates, 3rd order = sequence/need inference, 4th order = matching based on need, 5th order = coming soon...).

@IRB: Easy/Normal/Hard does use relative completion rates as a signal; there was a lot of experimenting to figure out what the right criteria should be.

In a nutshell: Easy = casual gamers, Normal = mid-core gamers, Hard = generally only hard-core gamers. These values can change as Gamerloop gathers more info about a game or the algorithms get tweaked (the end goal is to provide useful recommendations to people).

Just wanted to say thanks for having me on the show.

Happy to answer any questions people might have on the technologies behind Gamerloop!