Predictions - Global vs. Individual

[Sportsguy_USA]

Had a good chat with a bud over some beers the other night and couldn't agree which approach would work best.

We have our expert individual capper who spend night and day crunching the numbers, etc. etc. on figuring our their prediction for the game that week. One 'expert' guy.

Would this 'traditional' approach perform better, worse (or maybe the same) than if we could take a GLOBAL poll from players/individuals from around the world,and compile their their overall prediction.

Both we felt had pros/cons, but statistically one should outperform the other (in theory). Anyone have a brain-fart on which one would, in theory, work better over time?

[FourLengthsClear]

It is known that 'the public' hits at below 50% on ATS and Totals bets.
Any true "expert" would therefore beat the consensus.

[LUSabres]

It is known that 'the public' hits at below 50% on ATS and Totals bets.
Any true "expert" would therefore beat the consensus.

If he's an expert, he's better than the public.

[mjespoz]

There are valid arguments here on both sides. The concept of "wisdom of crowds" proposes that, under certain conditions, the average of the public can be more accurate than a single expert...

[Data]

Q: In general, who is sharper, the line maker or the market?
A: The market.

[Sportsguy_USA]

I should clarify - take the 'global' or 'crowd' opinion a step further rather than just a poll of their opinion. If this was done over time, consistently, then 'experts' would begin to arise from the crowd. This would help further refine the 'crowd' model.
Another interesting point we brought up as we both play the markets, if you look at the big successful companies in the market, their not making decisions based on a one person expert, but from an 'executive board' of experts (albeit 5-12 people) and in many cases, extend this to a larger market audience to test if their prediction/theories are sound decisions, or predictions. So if its working for the big boys as your Apple, Microsoft, etc. wouldn't the same theory apply to any types of predictions - specifically sports in this case.

Call it the 'crowd' approach or 'borge' (collective decisions - star trek) approach, but I think, in theory this would work better. Anyone know of any case studies, tests, etc, that have been performed on this ?

[CHUBNUT]

as usual with these topics situations are seen as static and an all together view of things, in the real world of gambling everyone is different and have different reasons for betting. Certainly one individual with superior knowledge of the sport with sound betting sense is best but if thats not you then the market is your best port of call. Too many of the so called models you read about on here are nonsense as they have no recent info compared to what the market is getting, its all down to how much weight you put on that recent info that creates perceived value. The bottom line is accepting the consecquences of your betting rationale, which is always more palatable in losses when its your own reasoning rather than someone elses like fading etc.

[Pokerjoe]

as usual with these topics situations are seen as static and an all together view of things, in the real world of gambling everyone is different and have different reasons for betting. Certainly one individual with superior knowledge of the sport with sound betting sense is best but if thats not you then the market is your best port of call. Too many of the so called models you read about on here are nonsense as they have no recent info compared to what the market is getting, its all down to how much weight you put on that recent info that creates perceived value. The bottom line is accepting the consecquences of your betting rationale, which is always more palatable in losses when its your own reasoning rather than someone elses like fading etc.

Good post, Chub.

[Sportsguy_USA]

I agree chub. A 'system' though that could compare the two...now that would be sweet.

thanks both.