Forget poll widgets - Prediction markets right in your website | Hacker News Reader

Forget poll widgets - Prediction markets right in your website

blog.askmarkets.com

14 pointsgtzi11 comments

The reason why polls are so popular is because they're easy to understand. You have one vote, you select the option you think is the correct answer to the question and the result show the spread of people's opinions.

Adding a layer of complexity to this by "investing" makes it harder for the user to understand. It took me a good two minutes and a read of the blog post to understand how eaxctly to vote (or invest). If this was a widget on another website, I wouldn't know what it was or to do with it. There's no obvious way to select an option and it's not particularly clear that you have to click on the row then select the amount to invest.

There needs to be a better explanation of what exactly the widget does within the widget itself, otherwise users will get confused and not use it.

Riklomas, thank you for your time and comment, I think you're right. We believe that an answer to a poll is the binary truncation of one's belief, and what we're trying to grasp is the full answer. Still, it's harder to extract it indeed, and the main reason is that we are used in submitting our views in polls, not markets. However, given that simplicity and directness is our main target with producing the widget, do you think that a different terminology (like betting or playing instead of investing, or what?) will make things more straightforward for the busy user?

It's not exactly the same idea as yours, but I think UserVoice do something quite similar, you get a set number of votes and you can pick how many to set the feedback you want:

http://feedback.uservoice.com/pages/1-general?lang=en&re...

Still a vote-like, 0/1 mechanism (with a pre-defined number of votes for each user) - it's hard to find the market mechanism in "everyday usage" anywhere around, that's our more innovative part :)

Just an UI improvement: How about using sliders to control the invested amounts?

Use some Javascript to ensure that all sliders add up to 100% (or so). If you need an inspiration for the finer points, look at how the game `Master of Orion' (the original, not the sequel) does resource allocation.

I still believe that entering a number is a more straightforward way to introduce/implement the concept, than having 'n' sliders etc. Also, the step of extra thinking before "you put the -exact amount- money where your mouth is" results into the added value of the service, a more insightful representation of users' belief.

I agree with the above comment -- I wouldn't use it to conduct a traditional poll on a web site.

But the 'Complexity' might be more like 'Entertainment' if you turn into a Facebook app and/or a Hunch-like destination site.

I think it is obviously a GAME which doesn't mean it can't have utility, too.

Also,

1) Nice graphic design. Nice UI design.

2) Way too much copy.

3) A lot of the copy has a great playfulness to it but what's with bludgeoning polls? It feels like you're protesting too much and I bet PollDaddy isn't your competition, as much as social prediction market sites, etc.

4) Cool site!

Hi bkhn, thanx for your comment.

No, poll sites are not our "enemies" or so, we just think that it's the shortest path to introduce the market/bet/game concept to our users. Your ideas on this are more than welcome.

Also, what do you mean by "copy"? :)

copy = text. Too much of it! :)

The input (1k of play money) and the output (a change in % on a table) are both fairly ethereal. Aside from claiming the benefits of market efficiency, I don't see a reason to call a system where you get 1,000 different votes a market.

I agree with gtzi that the terminology is getting in the way of my use of the widget. The idea that this kind of system gives a more detailed view of user preference than polling is great, and I don't think you need to make it do more than that. You can also probably cut the votes by a factor of 10 (or 100); don't imagine there's a way I can be mixed up about 4 options that requires this granularity to express.

Interesting concept.

I would refine it as a gambling widget (that's what it is) and offer a percentage to the host.

Grow it and sell the service to bookies, they'll love it.

Reply on news.ycombinator.com