Hi, a virtual stock exchange for predicting stuff is a cool idea. But I also like the predictions league in the format it has now (especially as I start to gain better ranks). So I hope the old predictions league will stay also if a new format is introduced.

Just curious but what would be the use of this wealth? To be cooler than those with less than you? Or do you have something planned?

I'm not suggesting that you need prizes but I do think people could lose interest if there isn't a reason to earn "wealth".

Definitely has potential, but I think it needs some ironing out before I would say it is a good plan.

I would like to participate more, but hate having to predict both hardware and software at once for a region. In 'real' trading situations, you can choose what to risk money on. Could you make it so entering values in less fields isn't a relative loss?

I really like the idea of stock trading. It would make everyone feel involved.

Well, I think the big question would be, how to make such a thing, without looking like a cheap copy cat of SE.

I hope you know that this would take a lot of effort. There should be more automation, even in the league know. Why are the results not in with the numbers, e.g.. You know that I would like to see "community points", "Wealth" or however you wanna call it on this site.

I think you should keep the current format but expand it more... You could add:

1) The ability to choose 10 of 25 games you predict on but give a multiplier on each one based on how likely it is to be right like in sports betting... For example, a game that has been out for 3 months and is consistently the same sales could be a 1.1 multiplier, but a new game this week could be a 3.0 multiplier...
2) The ability to research more via the site would be nice... Things like a query to get the opening 6 weeks sales of games in some sort of table or graph for a given platform, or classify games as AAA, AA, A, and B and then be able to get the opening weeks sales for a given platform. Or tie games of sequels together and query what each version did over a time frame.
3) There should be consistency on the days that the predictions league input will be available so people know when to look rather then constantly visiting just to see if this week's is ready or not.
4) The altime scores should be there but the last year or half year should also be there. I find the last 10 weeks to be too short term, and the alltime to be too long term, there should be a middle term of 1 year or 6 months seems good to me.
5) Hardware bundled software should be exluded from predictions... WiiPlay will always be some portion of Wii sales, from 40-65% (sometimes out a bit but it tends to be consistent) so WiiPlay shouldn't be there because if it is right chances are your Wii hardware is right, and if it is wrong the same is true.

I like the current format as it is simple and based on how well you can predict period. You don't make decisions based on anything other than raw predictions. When you go to a wealth system, people manage their wealth more and it requires more time and effort to excel at it. Hedging, selling short, etc. would all be involved in a new one and to tell you the truth simExchange has it pretty down so you would be competing with them. The current format could be expanded and added to so that you are more different then them.

I would prefer something automated....It'd be easier, because many of the older #s change very often, and I can't always go back and hand-edit each week.


IMO, if you could allow a "snapshot" of the top-20 for both Japanese and American software, and the next-gen hardware each week, automated, and allow users to wager funds on what will happen.

IE, assign wealth values not only on what you predict, but when you predict. IMO, the Prediction League has gotten rather stale since we have Japanese Sunday estimates from the Source. If we could see a weighted ranking, that rewards people for predicting early, would be neat.

IE:

A person gains some weath (but not much) from post counts, or various forum contributions, as well as registering.

Then they can goto the league, and find a game they want to predict. Let's say Super Mario Galaxy this week. They want to guess BEFORE first day numbers are out. They have the ability to wager with......A 5.0x multiplier on Thursday, and as low as a 2x multiplier on Tuesday (last day of predictions).

When they wager their money, a 100% spot-on guess yields the full payout of 5x on a Thursday guess, or a 2x on a Tuesday guess. A 50% breaks even, and anything lower looses that percentage times 2 (therefore, a 10% accuracy yelds only 20% of your investment).

Thus far, we get nothing for predicting games under current league rules, and such, so accumulating wealth is just as relevant.

A weighted bet can be good too - a double or nothing bet. It could double the payout, but double the cost of a sub-50% vote.

Not only this, if it was automated, a accurate guess now, once numbers were corrected, could spell trouble for those that are poor at predicting.

As for what games to predict: I would ask that you allow some sort of user-submitted content for various games to predict the next week for opening-week releases. This could allow for extra wealth to be gained, and allow for many more fun, niche games to be able to be betted on.

Or, you could go the "weighted" way of defining releases based on strength and hype: an accurate guess on Zack & Wiki would garner a much stronger return than betting on Halo 3's 5th week. This could be inversely calculated based on a pre-defined "rank" of popularity, reviews, and such.

Those are just a few random ideas.

A really cool idea would be to treat it like a stock market where you buy stocks in games for a certain amount of units.

You earn by how much the game sells in a week/month.

GTA4 for example will sell much, therefore it should be a really expensive stock.
Unknown games should be really cheap.

It would become even better if everyone was forced to buy a certain number of games. That way you have to decide whether to buy 1 really great selling game and two really unknown or buy three medium selling games.

I would say go with allowing them to predict whatever they want, whenever they want, more like SimExchange. That might make it less hassle for people to mull through alot of things.

Or, go further: allow people to wager "on a line" - get a higher bonus if they predicted a box of games (ie, 10-20 games on a week for a given region). That way they can go with the easy accurate one or two-off guesses, or run the marathon, and possibly win safer, or win alot better.

I was never that interested in predicting software, but i remember there being a couple of prediction tables that were for monthly or quarterly hardware.

Assuming ioi you haven't deleted them i found my own way of predicting quarterly hardware for this generation well in advance.... but it would be nice to have something else set up where we can predict what the hardware market will look like well into the future.
So I might predict now what the LTD console sales will look like by March, june, september 08 etc, and it would record when i predicted it and what regions it's for. (no need to go more complex than quarterly, seeing as there would be the weekly prediction thing too)
Just for play though, not for these points or whatever...maybe we could have it totally seperate to the normal prediction league which is weekly with the software and "wealth" etc, because i like predicting but i have no interest in doing it every week with the software.

About the stock exchange thing: I thought a little bit about it: If it is implemented the way, that you buy shares of games, and get back profit for sold copies, then it means that accuracy in your predictions is not the most important. You always bet on the AAA-games (Halo, SMG) and you have nothing to lose. If it is implemented more like a bet on horseraces it could work. That means you an amount on a game AND an outcome in weekly sales. If you predicted the sales good... and the others did it too, than you will earn only a little bit more as what you have bet. If you predict good... and the others are totally off, then you will win a good amount. That way you dont need a multiplicator for games that was recently released against games that are out for a long time: If the game is more difficult to predict, more people will be offf with their prediction. I like the idea, that you will get more, if you predict early. I'm a little bit concerned about the possibility to lose all or most of the wealth. That would mean that it is very hard to get back into the game. A suggestion how it could work: Every week get's a player 100 coins (or whatever amount) if (and only if) he enters the prediction league. You can bet on previously chosen games (by game-mod) and all hardware. You can bet until your coins (100 for this weeks and your winnings from earlier games) are emptied. You enter your bet and your prediction. For every game and piece of hardware the bets will be summed up and used as winnings. You get higher winnings for better accuracy, earlier bet and higher amount of your bet.