[jvrossb]

Adding email registration can be a mistake in situations when users who would have signed up through Facebook now sign up through email instead. Depending on what % of your growth is driven by virality vs organic discovery and how viral your fb users vs email users are (could be a 10X difference or more) you can be better off losing say 50% of players but having the remaining 50% sign in through Facebook than having 80% sign in through email and 20% through Facebook. It really depends on your numbers.

[furyofantares]

I have a gut-level reaction to this reasoning, which is that I believe that when you provide less value to your users you will likely get fewer users no matter how well you've justified the opposite case to yourself.

But I also have some specific reasoning that leads me to believe it's not a poor choice to provide another way to sign in. Signing in with facebook is extremely easy compared to normal signup paths. What this means is that most people who trust you with facebook will still sign in via facebook. And I don't think there's much of an argument that people who don't trust you with facebook will be signing in with facebook if it's the only option.

You can always allow them to link with facebook later, after they know what the game is, and after you've explained to them why linking will be valuable to them.

[jvrossb]

The correct answer doesn't exist. There's a methodology you can follow to find an answer for every product, and it will vary case-by-case and sometimes over the lifetime over the product. Your gut reaction will be right many times but wrong sometimes as well. It's entirely possible to run the numbers and figure out what the right decision is right now for this game.

[steveklabnik]

> that I believe that when you provide less value to your users

You're assuming that more options always increases value, but this is not the case. See Apple and, let's say, the Samsung Jitterbug for counter-examples.

[furyofantares]

Yeah, forcing users to make a decision or navigate a cluttered UI are both ways in which options decrease value. So adding options can suck if people aren't going to exercise them (which is what I see with a lot of options that get added to software -- only a very tiny percentage of users ever tweak them.)

However the claim I was debating was that a second sign-in method would be bad because people would exercise it. The argument was basically that it's bad because it's increasing value for users in a way that decreases value for you. And I don't think adding a second sign-in method to a page that currently only has one button really causes a difficult or confusing decision or creates a cluttered UI.

I suppose I still do see your point, though, since I personally wouldn't go much beyond that -- "sign in with facebook or twitter or google or browserid or github or create an account or or or" sounds like a miserable experience.