Actually what you say is not true. Owise tracks your predictive success in each category. If you have built up a good track record in Entertainment that won't give you a high weight in Politics. (In fact, we've found people who are very good in Entertainment are often simply awful in predicting political events, but that's another issue...)
We do give you a small fraction of your "overall" weight if you try your hand in a new category you've never predicted before, because we've found having some track record does correlate a bit across categories, but we're refining how that part of it works as we gain more data.
We are also refining exactly how fine grained we go with this category idea. Right now we track weights by top level category (Politics vs. Weather vs. Entertainment) but in the near future we may go down the heirarchy more deeply if you have enough events within a category (Entertainment/TV vs. Entertianment/Movies for example).
Owise is very much a new, experimental system, very much in the data gathering and refinement stage. I'm not claiming it's perfected yet, but we've thought through quite a few of the issues.
Hi Jerf: maybe post them in the general discussion board over on the Owise site. I know of some potential ways to attack the system, it's just an order of magnitude harder than attacking things like hsx or tradesports or iem, since a new member has zero weight at our site, you can't simply buy your way in using money, and for other reasons. We have several audits in place to detect manipulation attempts, which we don't discuss publically for obvious reasons, but we're always happy to hear ideas on how to make the system more bullet proof.
Shutting of feedback would in theory be desirable, but it's not practical because that information could filter from one person to another anyway. Also, people are allowed to change their predictions at any time (based on new information).
Market based systems do not exist in this kind of vacuum, of course, since participants see the current price.
Also, because the Owise system weights members predictions using past performance, we believe it is less necessary to worry about feedback effects (although we don't have enough data yet to prove that). People who are correct get higher weight, regardless of what the blended prediction was. Those who are unduly influenced by an incorrect blended prediction won't profit from it in the long run, their weighs will go down. (Again, this is different from market based prediction systems where people can and do profit even by making clearly erroneous "predictions" through their transactions. Consider an arbitrager who makes money regardless of what the true underlying event probability is--he may not even have an opinion on the underlying probability!)
Our current evidence (site only up for 3 months) is that people with higher weights make dramatically better predictions than people with lower weights. Note that stock markets or simulations of stock markets also weight opinions by how much money a person has "bet". Presumably people who are successful at such things have more money to bet than people who lose tons of money in the market.
Owise.com is actually a lot different in underlying mechanism, making predictions by blending opinions like a nerual network does.
Our philosophy is: why bother with all the trappings of the "market". It just confuses people, and leads to all kinds of gamesmanship that has nothing to do with what you're trying to predict. We simply ask people what they think the outcome will be, and we use a mathmatically correct way of ranking their accuracy (called a "proper scoring rule" for those of you who like math).
People who are right get higher weights in our system, can win prizes like Amazon gift certificates, and gain "titles" such as "Senior Political Analyst, Level 3". Bloggers can use our data graphs free, just don't nuke our watermarks. The system has already been more accurate than experts at some predictions, although we are just starting out and need a lot more people.
Slashdot Mirror
Actually what you say is not true. Owise tracks your predictive success in each category. If you have built up a good track record in Entertainment that won't give you a high weight in Politics. (In fact, we've found people who are very good in Entertainment are often simply awful in predicting political events, but that's another issue...)
We do give you a small fraction of your "overall" weight if you try your hand in a new category you've never predicted before, because we've found having some track record does correlate a bit across categories, but we're refining how that part of it works as we gain more data.
We are also refining exactly how fine grained we go with this category idea. Right now we track weights by top level category (Politics vs. Weather vs. Entertainment) but in the near future we may go down the heirarchy more deeply if you have enough events within a category (Entertainment/TV vs. Entertianment/Movies for example).
Owise is very much a new, experimental system, very much in the data gathering and refinement stage. I'm not claiming it's perfected yet, but we've thought through quite a few of the issues.
Hi Jerf: maybe post them in the general discussion board over on the Owise site. I know of some potential ways to attack the system, it's just an order of magnitude harder than attacking things like hsx or tradesports or iem, since a new member has zero weight at our site, you can't simply buy your way in using money, and for other reasons. We have several audits in place to detect manipulation attempts, which we don't discuss publically for obvious reasons, but we're always happy to hear ideas on how to make the system more bullet proof.
Market based systems do not exist in this kind of vacuum, of course, since participants see the current price.
Also, because the Owise system weights members predictions using past performance, we believe it is less necessary to worry about feedback effects (although we don't have enough data yet to prove that). People who are correct get higher weight, regardless of what the blended prediction was. Those who are unduly influenced by an incorrect blended prediction won't profit from it in the long run, their weighs will go down. (Again, this is different from market based prediction systems where people can and do profit even by making clearly erroneous "predictions" through their transactions. Consider an arbitrager who makes money regardless of what the true underlying event probability is--he may not even have an opinion on the underlying probability!)
Our current evidence (site only up for 3 months) is that people with higher weights make dramatically better predictions than people with lower weights. Note that stock markets or simulations of stock markets also weight opinions by how much money a person has "bet". Presumably people who are successful at such things have more money to bet than people who lose tons of money in the market.
Our philosophy is: why bother with all the trappings of the "market". It just confuses people, and leads to all kinds of gamesmanship that has nothing to do with what you're trying to predict. We simply ask people what they think the outcome will be, and we use a mathmatically correct way of ranking their accuracy (called a "proper scoring rule" for those of you who like math).
People who are right get higher weights in our system, can win prizes like Amazon gift certificates, and gain "titles" such as "Senior Political Analyst, Level 3". Bloggers can use our data graphs free, just don't nuke our watermarks. The system has already been more accurate than experts at some predictions, although we are just starting out and need a lot more people.