collective intelligence in sports gambling

[snufflyjoe]

ok I searched and could not find any articles online about using collective intelligence in sports gambling.

i am no expert at all in this but am very interested in discussing if this princiiple can or is able to be applied to sports betting.

I watched this interesting program not so long ago where theydemonstrated this principle in work. A guy went around with a huge jar of jelly beansand went around to about 600 office workers and asked them to guess how many jelly beans were in the jar he got wildly different estimates. He then took all the responces figured out the average of responces he got and the average was the exact amount of beans that were in the jar.

could this same principle be applied to sports betting? and how?

[mathdotcom]

Justin7 apparently is a huge believer ...

[Justin7]

Justin7 apparently is a huge believer ...

In fading the public? Yes. Collective intelligence in sports betting is a disaster, unless each member of the group wins independently. The, collective opinions show a significant gain.

[mathdotcom]

i was referring to your point in another thread that to get an accurate model you necessarily require a group and could never do it alone

[Justin7]

i was referring to your point in another thread that to get an accurate model you necessarily require a group and could never do it alone

Sure, but I don't want the group's opinions. I want their resources: databases, current data scraping, historical lines, current line scrapes, current injury scraping, betting, bookkeeping.

[MonkeyF0cker]

Sure, but I don't want the group's opinions. I want their resources: databases, current data scraping, historical lines, current line scrapes, current injury scraping, betting, bookkeeping.

Those are your reasons for not being able to do it without a group? I still see no need for a group. What am I missing?

[Justin7]

Those are your reasons for not being able to do it without a group? I still see no need for a group. What am I missing?

1. Timing. What if you want to spend a Sunday with your kids, but limits don't go up until 4am on Saturday morning? Or in the middle of your kid's birthday party? Let someone else deal with that.
2. Credit. I note that you worry about committing capital to futures. You're a pro; don't you have virtually unlimited credit? If not unlimited, it virtually is when in a group.
3. Allocation of energy: each day, I can spend so many hours concentrating on difficult problems. Can you really put in four good hours of research/modeling, and do another ten hours of shopping/betting/other bullshit? The more attritional factors other people deal with, the better you can focus on long-term projects.
4. Camaraderie. Don't undervalue that, unless you are a sociopath. And even sociopaths can benefit from this in different ways.
5. Reality checks. Other group members will call you out if they think an approach is full of shit. I'm sure it will shock you that teammates have called me out on new ideas once in awhile. Feedback on new approaches is near instantaneous. Perhaps this is of less value to you, especially if your analysis is conservative and more "traditional". But that approach also misses numerous opportunities.
6. Lowering of volatility. Sharing plays with other people focusing on single (or several) markets that win (in exchange for giving up a piece of your winning plays) mean you have much more volume, and lower relative variance. It is also eye-opening, seeing what markets they defeat.
7. Problem/information exchange. Different people or groups have different problems to solve. Fresh minds and approaches can often change impossible problems into more manageable questions.

With point 7, the modern syndicate isn't just a rigid team in a tight formation. There are tentacles, like strands of a spider web that connect to other groups to various degrees. If you are working alone, you are working much harder than you have to in order to make the same earn.

[MonkeyF0cker]

1. Timing. What if you want to spend a Sunday with your kids, but limits don't go up until 4am on Saturday morning? Or in the middle of your kid's birthday party? Let someone else deal with that.

No kids. I can take a day off if I want to. I'm not splitting my money with other people to begin with.

2. Credit. I note that you worry about committing capital to futures. You're a pro; don't you have virtually unlimited credit? If not unlimited, it virtually is when in a group.

Credit is part of your bankroll. I wouldn't bet season-long futures with cash or credit. I can roll that money over many, many times in the course of a season. It's worth far more to me not tied up for the entire season. See Opportunity Cost.

3. Allocation of energy: each day, I can spend so many hours concentrating on difficult problems. Can you really put in four good hours of research/modeling, and do another ten hours of shopping/betting/other bullshit? The more attritional factors other people deal with, the better you can focus on long-term projects.

Why would I need to put in so many hours? Computers do more timely and reliable work than any group of humans.

4. Camaraderie. Don't undervalue that, unless you are a sociopath. And even sociopaths can benefit from this in different ways.

Friends don't need to be made at work.

5. Reality checks. Other group members will call you out if they think an approach is full of shit. I'm sure it will shock you that teammates have called me out on new ideas once in awhile. Feedback on new approaches is near instantaneous. Perhaps this is of less value to you, especially if your analysis is conservative and more "traditional". But that approach also misses numerous opportunities.

I prefer to test my methods computationally rather than through hunches and guesswork.

6. Lowering of volatility. Sharing plays with other people focusing on single (or several) markets that win (in exchange for giving up a piece of your winning plays) mean you have much more volume, and lower relative variance. It is also eye-opening, seeing what markets they defeat.

You're making assumptions about others' volume and wagering patterns and projecting your preferences on their individual utility.

7. Problem/information exchange. Different people or groups have different problems to solve. Fresh minds and approaches can often change impossible problems into more manageable questions.

This is the heart of the matter. If you're incapable of figuring things out on your own, then you certainly do need a group.

With point 7, the modern syndicate isn't just a rigid team in a tight formation. There are tentacles, like strands of a spider web that connect to other groups to various degrees. If you are working alone, you are working much harder than you have to in order to make the same earn.

Another assumption. More people doesn't necessitate more efficiency. And what if people enjoy working on solving the problems to begin with?

You've made a list of PREFERENCES. These are NOT REASONS why people as solo entities CANNOT profitably model sports as well or better than "large groups."

[mathdotcom]

[snufflyjoe]

In fading the public? Yes. Collective intelligence in sports betting is a disaster, unless each member of the group wins independently. The, collective opinions show a significant gain.

I kind of thought fading public opinion may actually work. I would like some data to back this up. Be interesting to compare poll info (like on sports websites that have a poll of team A vs Team B) I will look into this to see how often the polls get it right and if you would make more long term going with or against the poll results

[allin1]

I think justin7 is not that into programming, scraping and other stuff. Why would he bother with those when he can leave that to somebody else? And why not benefit of someone elses info? It's just like in the stock market, it's good to have friends with additional info and resources IF YOU'RE THAT TYPE OF GUY. If you are not, then I can understand those computer geniuses that would rather be lonely wolfs than hunt in packs.

I remember reading Jack Schwagger's books with interviews with the best financial market traders of the time and it was funny to see how different some of them were, working in the way it suited them individually. One was isolated in kind of an office/library doing some old school research with pen and paper. Others had connections all around the globe calling eachother to exchange info, having employees around to do the the phone calls, research etc. Bottom line is the traditional "lonely wolf" couldn't have worked in the other type of enviroment and vice versa for the social type of trader with partners and employees but they were all making millions in profits. If you are making money and you are comfortable then it's OK

[TomG]

Sure, but I don't want the group's opinions. I want their resources: databases, current data scraping, historical lines, current line scrapes, current injury scraping, betting, bookkeeping.

So what do you feel are the biggest strengths you bring to a betting group? Specifically, why should a betting group give up a share of their action and potential trade secrets in exchange for your work?

[Justin7]

So what do you feel are the biggest strengths you bring to a betting group? Specifically, why should a betting group give up a share of their action and potential trade secrets in exchange for your work?

Analysis and modeling... But I not looking for new contacts now -- I have more projects than I can do, and have to turn down work.

[mathdotcom]

Analysis and modeling... But I not looking for new contacts now -- I have more projects than I can do, and have to turn down work.

Then what are they bringing? Beards?

[allin1]

From what I understood from Justin7's book is that the top of the food chain in sharps "hunt in packs" aka syndicates. It is actualy the only way to place huge bets at differeny books at the same time right?

[allin1]

Then what are they bringing? Beards?

He already said it in a previous post

Sure, but I don't want the group's opinions. I want their resources: databases, current data scraping, historical lines, current line scrapes, current injury scraping, betting, bookkeeping.

[TomG]

Then what are they bringing? Beards?

They bring the data. J7 brings the analysis. The result is profit.

[Darkside Magick]

http://www.intelligentbettingtips.com/how_it_works

[HUY]

ok I searched and could not find any articles online about using collective intelligence in sports gambling.

i am no expert at all in this but am very interested in discussing if this princiiple can or is able to be applied to sports betting.

I watched this interesting program not so long ago where theydemonstrated this principle in work. A guy went around with a huge jar of jelly beansand went around to about 600 office workers and asked them to guess how many jelly beans were in the jar he got wildly different estimates. He then took all the responces figured out the average of responces he got and the average was the exact amount of beans that were in the jar.

could this same principle be applied to sports betting? and how?

That's the way sports betting works. People estimate the probabilities of events happening; the average of their predictions should be the true probability (there are actually some theorists which define the probability of an event as the equilibrium of a betting market on that event), otherwise there would be a bias in the estimate which would be soon spotted by the bettors and eliminated.

[mathdotcom]

Can't believe J7's comparative advantage is modeling... who are these clowns collecting the data? They must be really clueless about offshore and/or statistics.

[Wrecktangle]

4. Camaraderie. Don't undervalue that, unless you are a sociopath. And even sociopaths can benefit from this in different ways.

LMFAO. Excellent point for the troll cult at this site.

[MonkeyF0cker]

Can't believe J7's comparative advantage is modeling... who are these clowns collecting the data? They must be really clueless about offshore and/or statistics.

It all makes sense, Mathy. He uses other people's data that he could hire a third world programmer to retrieve for about $10/hour since he's apparently incapable of doing it himself. Then, he creates profitable models with that data and splits everything those models make in the future with them.

Pretty simple.

Not just anyone can profitably model sports. J7 can though. That's why he has a group.

[big0mar]

It all makes sense, Mathy. He uses other people's data that he could hire a third world programmer to retrieve for about $10/hour since he's apparently incapable of doing it himself. Then, he creates profitable models with that data and splits everything those models make in the future with them.

Pretty simple.

Not just anyone can profitably model sports. J7 can though. That's why he has a group.

Do you seriously trust virtual assistants with your data?

[MonkeyF0cker]

Mathy, some day when you're a big baller like Justin and wouldn't sell 40 minutes of your time for $30, you'll understand.

Until then, keep quiet.

[MonkeyF0cker]

Do you seriously trust virtual assistants with your data?

I'm not exactly sure where you're coming from, but if you're asking whether I would trust other people to accurately mine the data, I don't have to. I do it myself.

Either way, Justin is trusting someone else with the integrity of the data. What's the difference?

[big0mar]

I'm not exactly sure where you're coming from, but if you're asking whether I would trust other people to accurately mine the data, I don't have to. I do it myself.

Either way, Justin is trusting someone else with the integrity of the data. What's the difference?

Yes, I think the difference is in deciding who you trust. Certainly thats a decision that will vary from person to person. Personally, I wouldn't trust a virtual assistant with a database for these purposes. I use them regularly, but not for sports-wagering purposes.

In regards to whether doing something yourself is profitable or not, its simply a matter of opportunity cost.

[MonkeyF0cker]

Yes, I think the difference is in deciding who you trust. Certainly thats a decision that will vary from person to person. Personally, I wouldn't trust a virtual assistant with a database for these purposes. I use them regularly, but not for sports-wagering purposes.

In regards to whether doing something yourself is profitable or not, its simply a matter of opportunity cost.

For any programmer worth his/her weight in horse poo, web scraping (even asynchronously) is a simple task.

Again, why is there a difference in trust? The information is publicly available. It's not like you're hacking into a CIA mainframe and obtaining classified documents.

If you're talking about the effort made to verify the contents of the data, then how do you know you can trust it unless you verify the data yourself anyway?

[big0mar]

For any programmer worth his/her weight in horse poo, web scraping (even asynchronously) is a simple task.

Sure. But many simple tasks eventually pile up. If you're scraping for 15+ different markets, it becomes time consuming.

Not to mention that there are very good programmers that can do very complex things which can be very beneficial for sports wagering.

Again, why is there a difference in trust? The information is publicly available. It's not like you're hacking into a CIA mainframe and obtaining classified documents.

Because accuracy is of the utmost importance. I wouldn't trust someone overseas that has no financial stake to manage a 150k entry database with 100% accuracy.

[big0mar]

Really my point here is that, yes you can do all this stuff yourself. I know plenty of people that clear $100k/yr working by themselves. But at some point if you want to scale your operation enough, you must bring in help. And if you're deciding you want to scale your operation, you need to be able to trust your help as much as you trust yourself. The more you scale, the more expertise you require. Sure its easy to manage your money using whatever system you prefer when you're doing $100k volume, but when you are doing $1M+, you need to have a solid grasp on the finance considerations. I've seen many people throw away points because they haven't properly evaluated their cost of capital.

[MonkeyF0cker]

Sure. But many simple tasks eventually pile up. If you're scraping for 15+ different markets, it becomes time consuming.

Or about 5-10 minutes per night after pressing a button.

Not to mention that there are very good programmers that can do very complex things which can be very beneficial for sports wagering.

Is this downside? Not sure what you're implying.

Because accuracy is of the utmost importance. I wouldn't trust someone overseas that has no financial stake to manage a 150k entry database with 100% accuracy.

No financial stake? They're doing it for free now?

[MonkeyF0cker]

Really my point here is that, yes you can do all this stuff yourself. I know plenty of people that clear $100k/yr working by themselves. But at some point if you want to scale your operation enough, you must bring in help. And if you're deciding you want to scale your operation, you need to be able to trust your help as much as you trust yourself.

Scaling is only necessary as a function of outs.

[big0mar]

Or about 5-10 minutes per night after pressing a button.

Yes. How are you defining how much your time is worth?

Is this downside? Not sure what you're implying.

I'm saying when I travel, I need access to many things through my iPhone. I don't have the time to develop the proper applications for my iPhone. There are programmers that do however.

No financial stake? They're doing it for free now?

They are being paid regardless. You aren't going to a virtual assistant and asking for a refund 6 mos. after they've made an error you're just catching.

[big0mar]

Scaling is only necessary as a function of outs.

I do not agree.

[MonkeyF0cker]

Yes. How are you defining how much your time is worth?

Not following. You said it's time consuming. I'm saying that I can automate the vast majority of it. My time is apparently worth more than yours.

I'm saying when I travel, I need access to many things through my iPhone. I don't have the time to develop the proper applications for my iPhone. There are programmers that do however.

Can't afford a laptop? I'm confused. There are wireless Internet providers.

They are being paid regardless. You aren't going to a virtual assistant and asking for a refund 6 mos. after they've made an error you're just catching.

And what do you do in the opposite situation after you've invested 6 months with your "partner's" data? Why are you assuming that the integrity of the data is any better? What's to say it's not worse with a partner?

[big0mar]

Just as an example of how brain power can help scale your operation, I knew a gentleman that would follow the PGA around. He would chart practice rounds and range balls for particular players. Over time as the data accumulated, it became very valuable. Now certainly anyone is capable of walking around a golf course and charting data, but it would not be cost-beneficial.