[ggruschow]

Maybe you could run in the background and just pay attention to what the users actually use? It seems like the pattern of use is going to be more indicative of what the rest of us should take a look at than the people-who-bother-to-rate-bias will allow. For example, a few basic things I'd like to know: if someone downloads it, tries it, and uninstalls or never touches it again after a couple days, if it force closes a lot [on a class of devices], and if people are using it consistently over time, like multiple times a week for months.

That said, I don't really need recommendations as to what to try, I just need something that doesn't present me with a zillion clearly crappy options (and this applies to Apple's too.. Stupid-compiled-pamphlet-of-information I'm looking at you). After eliminating those, there's not so much left in the Android Market that I can't go through it in a reasonable amount of time.

[gryftir]

The app market finding problem differs depending on how you intend to monetize it. Running an app means more to an app that makes it money via advertising then one which is sold. Making money from analytics is different from targeted advertising and from having a paid app.

But if you want an interesting take on it, how about figuring out which people who are downloading the popular apps first, and tracking their behavior as a cohort. Use the trendsetters to predict the trends.

[cludwin]

Great advice I think you are on to something there...

One way to look at this problem is to create a "Best Of" app list and keep it really clean.

I'll look into that ;)