Mnementh said: 68soul said: TWRoO said: Oh I see... so do we only get a profit if our predictions are higher than the average of all peoples predictions? Yes, i think that's the trick... The bad thing is, if by an incredible chance (or misfortune), the average prediction is 100% (nearly the case here with the PS3 and a global 96% accuracy), you'll loose your money anyway, even if your prediction was higher than 90%... Is it me, or is it a little bit unfair? Loose a bet when you're 90% or 95% correct? And in another case, win with a 30% accuracy, if the global one is 0%? It's a really, really strange system... not really rewarding, imho... If you have 90% accuracy and the average accuracy is 95%, than you underperformed. If you have 50% accuracy and the average is 30%, you're a great predictor.

Allright Mnementh, maybe... but i think that system is too dependant on luck: the average prediction is... average...

Like i already stated, the PS3 has 96% average accuracy in that game, but it doesn't mean most of the predictors were totally right... i was 95% correct on the PS3, and i loose... just a little bit, ok, but i loose... I think that's not fair...

And what about Wii play? If the average is 0%, how's the formula then? Cos 100 divided by 0 and 0 divided by 0 is always the same result...

Anyway, i find strange to win more money on my Wii Play 37% prediction than on my PS3 95% prediction... and i find strange that most of the best overall predictors of that round actually loose money...

I mean, if the goal is not to win vgchart$, but not to loose too much, what's the point? It's not really rewarding, and it doesn't give the best predictors what they desearve, or at least that's the way i see it...

 

ioi said: rendo, it's not over my head I just don't like your attitude (again)...

WAAAAH.  I'm sorry I call you out when you screw up and I won't kiss your ass like most people.  You seem to get pretty irritable pretty easily, so maybe you should take a breather once in awhile. 

As for it being over your head, that opinion still stands.

 

Then perhaps you should be a little more diplomatic when you respond to people instead of throwing in some smilies to be smug. I treat people who they treat me, you want to get up all in my grill dawg, I'll show you how to step. Werd.

@ioi

Well, you will still lose when you get 90% right? So that's not a big of acomplishment. From first 10 predictiotors everyone lost on PS3 bet.

@Mnementh
Not really, I doubt many people won on PS3 bet. I checked the first 10 predictors and none of them got 96% accuracy (thus everyone lost money on that bet). Just becouse average predicition is close to real number doesn't mean that people actually bet very close to that number. When one guy bets 60k for a game, second 100k, and game sells 80k than average prediction accuracy is 100% but both guys lose money. The fact is that this rules drain money from system.

ioi said: rendo, it's not over my head I just don't like your attitude (again)... Like i already stated, the PS3 has 96% average accuracy in that game, but it doesn't mean most of the predictors were totally right... Please explain how this one works

Many people could be way too low, or way too high... and some people could be very close... it's just an average, but if that average is too close to 100%, everyone loose, even the people who were very very close, just because of fate...

Imagine if thousands and thousands of vgchart$ are bet: it's not "some win what the others loose"... everyone loose then, even if some way more than others...

Perhaps there should be a way of stretching and squashing the money given depending on accuracy (I haen't seen the exact formula used so perhaps this already happens) For instance in PS3s case the average was 95% or whatever... and if you get 95% you get squat... this is very fair because it was obviously easy to predict. If you got half way between 0% and the average 95%, you get half your money back, wheras if you get halfway between 100% and the average (95%) you get 1.5X your money back.

kamil said: @Mnementh If you have 90% accuracy and the average accuracy is 95%, than you underperformed. If you have 50% accuracy and the average is 30%, you're a great predictor. Not really, I doubt many people won on PS3 bet. I checked the first 10 predictors and none of them got 96% accuracy (thus everyone lost money on that bet). Just becouse average predicition is close to real number doesn't mean that people actually bet very close to that number. When one guy bets 60k for a game, second 100k, and game sells 80k than average prediction accuracy is 100% but both guys lose money. The fact is that this rules drain money from system.

You really have a point here, that all actually can lose on a bet. On the other hand, the fixed breaking even point (at 50% or 60% whatever) brings easy money in the system. Because most of the people are able to reach a 50% accuracy. More a problem is, it's easy to foresee, which prediction could be near and which could be off. Predictions for console are mostly very near the real stuff (remember the old prediction league, console-predictions brought the points). Same with titles with great legs: They sell every week nearly the same amount. Hard to predict on the other hand are new releases. With a fixed accuracy as break-even point it means I would put all my money on console-predictions and minimum bet on new releases.

@ioi: Ok, so everything is good. I actually asked a question how you calculate accuracy eariler but you must have missed it.
I think this system is better than the one with predefined 50% accuracy level.

ioi said: no, kamil it doesn't work like that. If you are over or if you are under then it doesn't cancel out. The average accuracy is the average of all the individual accuracies, not the accuracy of the total of all estimates.

Ah, thanks for the clarification.

ioi said: Entroper - thus discouraging people from playing it safe with easy predictions.

IMO there are two things going on here. The first thing is that the rules were changed AFTER we had already made predictions on the first round. So people bet big on safe predictions and then had the rug pulled out from under them with a different set of rules. That's highly annoying, but I can understand that if this is the system you want to use, it would be equally annoying for you to try to implement two different systems.

But the other thing is that it's now close to impossible to determine a good allocation. With a fixed system, you bet big where you're confident you'll be accurate. It still doesn't guarantee success, because anything can happen. But with the averages system, you really get penalized where the average prediction is close.

I understand that the averages system is more likely to encourage risky betting, but it makes the game a lot harder, and not usefully so IMO. It may be easier to win money with the fixed system, but you are still competing with everyone else on who can win the most money. I think the averages system is going to discourage large betting, because people will be too likely to lose their money. I haven't participated in any of the rounds since you announced the change for this reason.

IMO, it is not a bad thing if the system results in slightly more winnings than losses on average. After all, you're not a casino, you can print infinite money. :) In fact, I think an upward trend encourages greater participation in the betting, and like I said, the essence of competition is still there because you don't just want to win money, you want to win the most money.