A: SocialPredictor is similar to SocialPredator which is what I thought when I first read it.

B: "You must be logged in to use this feature!" = Fail.

I won't touch on the name so much, but there is a much better way to approach the login issue. I assume that logging in is only to keep the votes uncontaminated, but you said that you wanted to make this site more fun than statistically valid, and seeing the login requirement was not fun for me.

Later impressions:

This site has the potential to be an interesting/fun way to spend some time, and if I could A: see the predictions before voting (ie. just being able to browse and see what most people think about a specific subject), and B: not have to log in except maybe to participate in a discussion afterwards, then I would be very likely to use this site as a way to pass some time.

Also, a thumbs up for having a "correct users" section. I think the possibility of getting the predictions "right" and being able to be respected everywhere (on the site) as an accurate predictor is another fun angle that will attract and keep users.

Final impressions:

This is a startup, and by definition is designed to earn you money. Is there a plan for this in the future? Do you think ads are the way to go, or maybe sell the results to marketing agencies (for targeted subjects and users)?

I was thinking of adding the current results via graphs and it sounds like you think I need to communicate results in a better way as well - I'll get on that, thanks for pointing it out.

As for the money, that can come later. I'm more interested in seeing if it's a viably high-traffic site idea before I slap on the ads.

I completely understand why you'd want to limit your voting to actual registered users. Nevertheless, I'd try to find a way to let visitors pretend to participate. How about tracking votes by non-registered users in a separate tally? When I click to vote, I get a screen that says "Thanks for voting!" and has two sets of results: A "low-quality" set that includes all the non-registered user votes for the last 2 hours (or 12 hours, or 24 hours, or -- if there haven't been too many -- the last week) and a "high-quality" set that only includes registered users. Next to the registered user data is an explanation ("these are the results that really count") and a link ("join our site if YOU want to cast votes that really count!").

What you're describing is what I was thinking to almost a 'T'. Have two voting ranges, one for registered users and one for anonymous users. Sounds like I need to jump back into that and start building that out as a feature.

Thanks for the comments guys, these really make a difference.

  * I want to try stuff without any signup-process.

  * How will you make money?

  * Looks good

1. Pretty good presentation. Looks clean, professional.

2. Easy to understand what your site does and the value proposition. This is a huge step in front of most major sites.

----- Here is my criticism:

1. Your demo page is too tailored towards first time users. You need to look at models like reddit and digg to come up with a nice balance.

2. Like everyone else is saying, don't require sign-ups. And if you must, use OpenID and FB Connect to streamline the process.

3. Pretty pictures. Use some nice flash charts to give a better understanding on how your predictions are trending.

4. To be brutally honest. i don't see this site making critical mass. The problem with web 2.0 is that most people don't have time to join so many social networks, you need to consolidate this as either (a) an API for other companies to use and make your business a b2b model (there are a bunch for predication markets already) or (b) make it a social network application (but there should be a couple of these already too)

http://www.cnn.com/ELECTION/2008/veepstakes/index.html

That's a good example of how to leverage some cool algorithms and a funky API to integrate with major sites.

Perhaps I'm a dick wad, but I think honesty is better than sugar coating. If you want to talk further, e-mail me charlesju[at]gmail[dot]com.

You've pointed out the biggest issue I'll have going forward - hitting critical mass. Most of the other commentators have suggested removing the login requirements for voting. I think it's likely that will result in more interaction from casual visitors. However, I'm not sure how much it will do for my user count - we'll see.

Thanks.

The idea with SocialPredictor.com is that it's meant to be fun. There were two angles I knew I could take with a prediction site - fun or statistically valid predictions. I chose the former. Was I at least marginally successful?

I get that you were going for fun over market, though.