My Knowledge Of A Bookie To Be Successful.
- Thread starter rainbow
- Start date
This is one of the most INFORMATIVE threads I have ever read about bookmaking...
wantitall4moi,
You are one stubborn guy who is NOT thinking as a bookmaker and it shows...
To suggest that a bookie should balance all of his games is ludicrous. Providing one has a sufficient bankroll, you will be PROFITABLE dishing out minus $1.10 even holding lopsided amounts wagered on ONE team or event over the long run because of VOLUME, period!
:+signs9-1 for me...
wantitall4moi,
You are one stubborn guy who is NOT thinking as a bookmaker and it shows...
To suggest that a bookie should balance all of his games is ludicrous. Providing one has a sufficient bankroll, you will be PROFITABLE dishing out minus $1.10 even holding lopsided amounts wagered on ONE team or event over the long run because of VOLUME, period!
:+signs9-1 for me...
rainbow
EOG Master
JC, I will ARGUE with you on this till I am BLUE in the face, I am sure STEVE at WSEX couldnt do it, but I know BOOKIES that know which way the line is going to go on the MAJORITY of the games and trust me I seen it out my own EYES, you trying to say if a BOOKIE knows which way the MAJORITY of the games are going they still cant do it, so untrue.
Very interesting thread. It should be pinned for atleast a week.
Shrink---
How many books run -110 thses days? How many books don't have enough money at any given time to pay what player balances are?
You QUALIFY it yourself, "as long as they have a sufficienty bankroll". What is sufficient? Being able to pay half the people that have balances? Being able to keep their neteller float at 10K? Being able to pay their advertisers so they get enough post ups from LOSERS to pay off what they owe?
I am also well aware of volume and how profittable it is. But if books are consistantly lopsided AND running GREATLY reduced holds, that volume is basically MEANINGLESS. Especially if they have a few bad weeks, and people start withdrawing money.
Why is it that these books create creative options to players they can't afford to pay. Like offering them a bonus or some free money that they have to roll over? BEcause they NEED people to LOSE to stay in business.
What part of NEEDING LOSERS isn't gambling?
I think I am the only one that is thinking as a true bookie is supposed to think. A bookie is supposed to balance action s best the can to eliminte being over exposed one way or the toher.They do that by moving lines or odds, to entice action the other way. Or they post outrageous numbers both ways to halt action altogether.
I know in the real world it isn't that simple, but it has to be adhered to alittle better than a lot of places do it.
I am sure most don't have a clue to what their threshholds are, and probably some don't even care.They are under the vast assumption that players lose, and they are correct. BUT enough players win in streak to ruin their theoretic expectaions.
Meaning I might lose 25K in a month. But I also might win 100K over the course of the year. So they are in the red to ME for 75K. Also on the flip side, some other player mightlose 50K to them in a year, but if he gets on a streak he might win 20-30K off them and actually might want that money. Just because the guy loses 75K doesn't mean he is going to lose it to THEM.
That is why a lot of books don't like scalpers, especially when they are on the losing side of the scalp. Because there is no gurantee that that player will play with them enough for them to recoup that loss.
If they had laid off that number somewhere (and it being a scalp they should be able to find one better) that book would have reduced their risk greatly. But they choose to hold the overexposure and GAMBLE, and they lost.
I think I have made some very solid points inhere. I have made some ravings for sure. But for the most part I understand what is going on,and I understand the volatility that exists. I just don't bury my head in the sand when I see it condoned. Especially by people that know better as well, and are backing these places knowing full well that what I say has credence.
Basically these places are run on paper, and ultimately they rely on losers to help keep them profittable. But as I mentioned before, the sheer number of books helps and hurts. ust because people lose more than they win, doesn't mean they are losing at all the same books. But having the number of books to split up the exposure also helps. So it isa give and take.But don't call what these places do bookmaking.
Even the guy on the street corner isn't going to sit on a 10-1 side no matter how much he likes the side he is shorton. If he loses and doesn't have enough to pay then what? That was something a lot of places offshore forgot about awhile ago, worrying about paying every balance they might have to pay out if everyone won the bets they placed.
How many books run -110 thses days? How many books don't have enough money at any given time to pay what player balances are?
You QUALIFY it yourself, "as long as they have a sufficienty bankroll". What is sufficient? Being able to pay half the people that have balances? Being able to keep their neteller float at 10K? Being able to pay their advertisers so they get enough post ups from LOSERS to pay off what they owe?
I am also well aware of volume and how profittable it is. But if books are consistantly lopsided AND running GREATLY reduced holds, that volume is basically MEANINGLESS. Especially if they have a few bad weeks, and people start withdrawing money.
Why is it that these books create creative options to players they can't afford to pay. Like offering them a bonus or some free money that they have to roll over? BEcause they NEED people to LOSE to stay in business.
What part of NEEDING LOSERS isn't gambling?
I think I am the only one that is thinking as a true bookie is supposed to think. A bookie is supposed to balance action s best the can to eliminte being over exposed one way or the toher.They do that by moving lines or odds, to entice action the other way. Or they post outrageous numbers both ways to halt action altogether.
I know in the real world it isn't that simple, but it has to be adhered to alittle better than a lot of places do it.
I am sure most don't have a clue to what their threshholds are, and probably some don't even care.They are under the vast assumption that players lose, and they are correct. BUT enough players win in streak to ruin their theoretic expectaions.
Meaning I might lose 25K in a month. But I also might win 100K over the course of the year. So they are in the red to ME for 75K. Also on the flip side, some other player mightlose 50K to them in a year, but if he gets on a streak he might win 20-30K off them and actually might want that money. Just because the guy loses 75K doesn't mean he is going to lose it to THEM.
That is why a lot of books don't like scalpers, especially when they are on the losing side of the scalp. Because there is no gurantee that that player will play with them enough for them to recoup that loss.
If they had laid off that number somewhere (and it being a scalp they should be able to find one better) that book would have reduced their risk greatly. But they choose to hold the overexposure and GAMBLE, and they lost.
I think I have made some very solid points inhere. I have made some ravings for sure. But for the most part I understand what is going on,and I understand the volatility that exists. I just don't bury my head in the sand when I see it condoned. Especially by people that know better as well, and are backing these places knowing full well that what I say has credence.
Basically these places are run on paper, and ultimately they rely on losers to help keep them profittable. But as I mentioned before, the sheer number of books helps and hurts. ust because people lose more than they win, doesn't mean they are losing at all the same books. But having the number of books to split up the exposure also helps. So it isa give and take.But don't call what these places do bookmaking.
Even the guy on the street corner isn't going to sit on a 10-1 side no matter how much he likes the side he is shorton. If he loses and doesn't have enough to pay then what? That was something a lot of places offshore forgot about awhile ago, worrying about paying every balance they might have to pay out if everyone won the bets they placed.
rainbow
EOG Master
DRUNK said:
COLD MONEY is when a BOOKIE SAYS I am 20 or whatever out to one side and you ask him is it anybody you respect, the BOOKIES I know will say its all COLD money, that doesnt mean its not COLLECTIBLE, atleast in my eyes it doesnt, a BOOKIE might say I am loaded on one side but the money is very SHAKEY, meaning the BOOKIE want get a good nights reast if he wins the WAGER because he will be sweating to get PAID.
Drunk, those were the days my friend.
I think this whole thread is kind of stupid.
And now I will demostrate why.
Why does the book maker wins over the players in the long run?
Answer: Bookie has the vig on his side and the player usually screws up.
Therefore, the only job is for the bookmaker is risk managmement.
There are two ways to manage risk.
In the case of one game such as the super bowl, balance the action on both sides, usually by moving the line.
The better way is to diversify your risk. Spread the risk out over 14 NFL games, and more.
There was a comment that Steve from WSEX still don't know how to be a bookie. That is not surprising because the Wall Street types are bringing a whole new (or the same) risk management model (learned somewhere else) applied to the off-shore gambling industry.
Traditional booking is obsolete.
Other than the salient points, the rest of the talk is all sophistry.
And now I will demostrate why.
Why does the book maker wins over the players in the long run?
Answer: Bookie has the vig on his side and the player usually screws up.
Therefore, the only job is for the bookmaker is risk managmement.
There are two ways to manage risk.
In the case of one game such as the super bowl, balance the action on both sides, usually by moving the line.
The better way is to diversify your risk. Spread the risk out over 14 NFL games, and more.
There was a comment that Steve from WSEX still don't know how to be a bookie. That is not surprising because the Wall Street types are bringing a whole new (or the same) risk management model (learned somewhere else) applied to the off-shore gambling industry.
Traditional booking is obsolete.
Other than the salient points, the rest of the talk is all sophistry.
Got a feeling that RB is a West Coaster.
Baker
EOG Member
Man I have tried to keep myself from responding to this thread but I just can't do it. Wantitall you must have signed up for only English classes in college, because it is very evident that you didn't major in mathematics or economics.
You are the same guy that spouts off that there is no such thing as a winning gambler (on the betting side) and now you say that bookmakers shouldn't hold action when they have a theoretical edge (negative expected value for bettors). How can it be both ways? You act like the only way to make safely money is through arbitrage and that simply isn't the case. The problem is that some books think like you and those are the ones that get themselves in trouble.
The kindergarten logic you use to try and support your points is ridiculous and lowers the IQ of this forum in general. Take some time off from writing another manifest and go research a couple things for starters. First take a look at the theory of Risk of Ruin.
After you fully understand Risk of Ruin and how that would apply to this thread go read about the Kelly Criterion as you have shown that you have no understanding of the theory or practice of utilizing that formula. What you will find is that when using Kelly, streaks (which you claim or the inherent flaw within the Kelly Criterion) are actually of no consequence when using the Kelly Criterion.
There are so many errors in what you are suggesting is the "right" way to book bets that it is hard to pick on just one. The main problem I see with your methodology is that you won't know how balanced your action is until very close to the time when the event starts. First off you assume that there is a magical sportsbook out there that can and/or will take all of your action (for the record if I owned a sportsbook I would definitely take all of your action), let us remember that for your money to be safe you would have to be confident that the sportsbook you were laying off action with would prescribe to the same methodology of balancing action. I don't think it is logical to assume that a sportsbook that will take your action and have balanced action themselves would exist.
Next you seem to assume that you as the bookmaker will be able to lay of action at the same line and price as you took it, and I don't think that this is the case either. If you take 110,000 in action and lay off 88,000 of it to have 11,000 balanced on each side you are in effect lowering your hold to .9091% (from what was a hold of 4.54%).
Lines and prices move. The chances that you will be able to lay off at the same line and price that you took the wagers at is definitely less than 100%. This is where the fatal flaw in your even action theory is. The effect of losing one polish middle would probably break you. Even a 1 point middle with a 1% chance of hitting creates a situation where you have negative expected value. Using the example above the equation would look like this:
1000-(.01*(88000+80000))=-600 a negative expected value of -.545%
The same could be said if you had to lay off at more vig than you charged. If you had to pay even one cent more in vig your expected value would all but disappear.
In this wantitall4moi length post all I am really doing is agreeing with Ironlock, now go read a book Voice of Treason.
You are the same guy that spouts off that there is no such thing as a winning gambler (on the betting side) and now you say that bookmakers shouldn't hold action when they have a theoretical edge (negative expected value for bettors). How can it be both ways? You act like the only way to make safely money is through arbitrage and that simply isn't the case. The problem is that some books think like you and those are the ones that get themselves in trouble.
The kindergarten logic you use to try and support your points is ridiculous and lowers the IQ of this forum in general. Take some time off from writing another manifest and go research a couple things for starters. First take a look at the theory of Risk of Ruin.
After you fully understand Risk of Ruin and how that would apply to this thread go read about the Kelly Criterion as you have shown that you have no understanding of the theory or practice of utilizing that formula. What you will find is that when using Kelly, streaks (which you claim or the inherent flaw within the Kelly Criterion) are actually of no consequence when using the Kelly Criterion.
There are so many errors in what you are suggesting is the "right" way to book bets that it is hard to pick on just one. The main problem I see with your methodology is that you won't know how balanced your action is until very close to the time when the event starts. First off you assume that there is a magical sportsbook out there that can and/or will take all of your action (for the record if I owned a sportsbook I would definitely take all of your action), let us remember that for your money to be safe you would have to be confident that the sportsbook you were laying off action with would prescribe to the same methodology of balancing action. I don't think it is logical to assume that a sportsbook that will take your action and have balanced action themselves would exist.
Next you seem to assume that you as the bookmaker will be able to lay of action at the same line and price as you took it, and I don't think that this is the case either. If you take 110,000 in action and lay off 88,000 of it to have 11,000 balanced on each side you are in effect lowering your hold to .9091% (from what was a hold of 4.54%).
Lines and prices move. The chances that you will be able to lay off at the same line and price that you took the wagers at is definitely less than 100%. This is where the fatal flaw in your even action theory is. The effect of losing one polish middle would probably break you. Even a 1 point middle with a 1% chance of hitting creates a situation where you have negative expected value. Using the example above the equation would look like this:
1000-(.01*(88000+80000))=-600 a negative expected value of -.545%
The same could be said if you had to lay off at more vig than you charged. If you had to pay even one cent more in vig your expected value would all but disappear.
In this wantitall4moi length post all I am really doing is agreeing with Ironlock, now go read a book Voice of Treason.
Ironlock
2
Wow. Dare I say, "SHARP POST".
Baker-I guess I've missed you're previous posts, but I swear you will not have one on this site slip by me in the future.
EOG steps up a notch in quality every time a poster like this steps up and shares a bit of wisdom.
Welcome to EOG Baker.
Iron.
BTW: I do think Want has alot to offer(he is obviously a generator of profit), just not in the area of bookmaking or mathematical theory. I hope you continue to post, in all areas. Discussions like this are the primary benifit of these forums if you ask me, and it wouldn't have been possible without want's contribution.
Baker-I guess I've missed you're previous posts, but I swear you will not have one on this site slip by me in the future.
EOG steps up a notch in quality every time a poster like this steps up and shares a bit of wisdom.
Welcome to EOG Baker.
Iron.
BTW: I do think Want has alot to offer(he is obviously a generator of profit), just not in the area of bookmaking or mathematical theory. I hope you continue to post, in all areas. Discussions like this are the primary benifit of these forums if you ask me, and it wouldn't have been possible without want's contribution.
DRUNK said:
You sure fooled me...
THE SHRINK
HI. Not sure that their is much of a differance between bookmakers and market makers on the floor of either the stock exchange or more so the option exchanges. they cannot take the 9-1,s and say it is part of the over all because one loser or in the example 2 losers ay denver had won that book would be out 180k would take half aseason to get back right? professional market makers who want to be there a long time will hedge to limit exsposure. you have to keep money to stay in the game. ask glacken hes a bond guy same stuff.
seems if a bm dosent move the number especially 3 or 7 he has to move the line a tad otherwise he is gambling not in business. Just some thoughts from a small bettor who works in the brokerage business.
this is really a good thread.
seems if a bm dosent move the number especially 3 or 7 he has to move the line a tad otherwise he is gambling not in business. Just some thoughts from a small bettor who works in the brokerage business.
this is really a good thread.
Baker I make the same challenge here as I did at the RX, and at Stanford Wong's place( which was mysteriously lost/deleted) If ANYONE thinks they can PROVE to me that betting Kelley will be more profitable than bettinga flat rate I make the same 5K challenge.
Start with an imaginary bankroll, and PROVE that Kelley is more successful than other methods. I told those there that swear by it that if they could show a profit after a full year of using it, I would give them 5K, if they went broke before the yearwas up, or didn't show a profit they could donate the 5K to a charity of my chosing.
That thread was mysteriously deleted. So it told me all I needed to know. People like to talk about a lot of theorwtical stuff but when it comes to actually proving it, most turn tale and run, especially when real money is involved.
I also explained how losing bettors has no relationship to losing books. Most guys do not lose all their money at the same books. I could theoretically breaka book if all I did was scalp winning sides with them. I might make one place rich by always losing with them, bt the place I was winning at would be busted soner or later. There are obviously books that ""win" more than they lose. Just because 98% of people that bet lose, doesn't mean all the books are geting their equal share of the pie.
Some obviously get more than others.
As for layingoff and polish midles, et al. If books are doing this behind the scenes or as a courtesy these are none existant. If Pin, CRIS, and OLY all have -3 for a game, not all are going to have the same hold. Pin might have -105, CRIS might have -110, and OLY might have -115. So they could EASILLY swap money around between them, especially if OLY was getting guys that wantedthe dog and the +105. Pin could easiily square the action for say a nickle. OLY would still be making a nickle and they could balance the action. If books got together and did this they would never have to move a spread unless a ton of steam hit a few books at once. The reason lines move for the mmost part is because one or two places get hit, they move, then EVERYONE moves. Some places could e totally lopsided on the dog, but be forced to move if everyone else does. They should of course hold a little longer since if they have the best line on the fave they will get their balance. but a lor of books thihnk like players here. They think steam means a winner. So they not only gamble by taking the lopsided action, they think they have a solid position with their lopsided money, and they want more so they follow the flow. So not only are they lopsided at the old number(worse), but they will probably be lopsided on the newer number as well.
Moving spread lines is always is a danger to books. Plain and simple, we can all agree on that. if they got together and made the balance more true they would lessen the exposure they have, not only by moving a line (opening up for middles and sides), but by balancing thier action as well. It would definately hinder the players, but would surely make it a little more stable.
So balancing action isn't nearly as difficult as people think. I wouold say that there are 4-5 books that will actually move BECAUSE of money, all the other books are simply followers and move on air. That is also something that is pretty scarey. These places are relying on the "reputation" of the books and the players at the books thatthey are following. Hoping that these moves are "smart money" moves that makes the postions they are holding/gambling on more solid than they actually are.
They are replacing mathematical probabilty. the one that says there is a 50/50 chance of something ocuring, and saying that because this book or that bettor thinks it is this, they now have a better than 50/50 chance of winning. (similar to Kelley) Whne it just isn't the case. A person's opinion is only as good as their ability. Over the long term it is BARELY above 50%, nomatter how good it is.Even if it is 54%, that is stillonly a 4% better chance, and since you canot determine WHEN that 4% chance will kick in, it becomes convoluted, especially when you are trying to get a short term result from some that took a very long term "test" to figure out.
Think of it this way...
You enter a contest to win a car. they have 3 placards yo have to chose from. On one of those 3 placards is written the word CAR. So you pick, they flip one over, that in not a winner. Theythen give you thhe option of changing your original pick or keeping it. What option gives you the best odds of winning, keeping your original pick or changing it?
Start with an imaginary bankroll, and PROVE that Kelley is more successful than other methods. I told those there that swear by it that if they could show a profit after a full year of using it, I would give them 5K, if they went broke before the yearwas up, or didn't show a profit they could donate the 5K to a charity of my chosing.
That thread was mysteriously deleted. So it told me all I needed to know. People like to talk about a lot of theorwtical stuff but when it comes to actually proving it, most turn tale and run, especially when real money is involved.
I also explained how losing bettors has no relationship to losing books. Most guys do not lose all their money at the same books. I could theoretically breaka book if all I did was scalp winning sides with them. I might make one place rich by always losing with them, bt the place I was winning at would be busted soner or later. There are obviously books that ""win" more than they lose. Just because 98% of people that bet lose, doesn't mean all the books are geting their equal share of the pie.
Some obviously get more than others.
As for layingoff and polish midles, et al. If books are doing this behind the scenes or as a courtesy these are none existant. If Pin, CRIS, and OLY all have -3 for a game, not all are going to have the same hold. Pin might have -105, CRIS might have -110, and OLY might have -115. So they could EASILLY swap money around between them, especially if OLY was getting guys that wantedthe dog and the +105. Pin could easiily square the action for say a nickle. OLY would still be making a nickle and they could balance the action. If books got together and did this they would never have to move a spread unless a ton of steam hit a few books at once. The reason lines move for the mmost part is because one or two places get hit, they move, then EVERYONE moves. Some places could e totally lopsided on the dog, but be forced to move if everyone else does. They should of course hold a little longer since if they have the best line on the fave they will get their balance. but a lor of books thihnk like players here. They think steam means a winner. So they not only gamble by taking the lopsided action, they think they have a solid position with their lopsided money, and they want more so they follow the flow. So not only are they lopsided at the old number(worse), but they will probably be lopsided on the newer number as well.
Moving spread lines is always is a danger to books. Plain and simple, we can all agree on that. if they got together and made the balance more true they would lessen the exposure they have, not only by moving a line (opening up for middles and sides), but by balancing thier action as well. It would definately hinder the players, but would surely make it a little more stable.
So balancing action isn't nearly as difficult as people think. I wouold say that there are 4-5 books that will actually move BECAUSE of money, all the other books are simply followers and move on air. That is also something that is pretty scarey. These places are relying on the "reputation" of the books and the players at the books thatthey are following. Hoping that these moves are "smart money" moves that makes the postions they are holding/gambling on more solid than they actually are.
They are replacing mathematical probabilty. the one that says there is a 50/50 chance of something ocuring, and saying that because this book or that bettor thinks it is this, they now have a better than 50/50 chance of winning. (similar to Kelley) Whne it just isn't the case. A person's opinion is only as good as their ability. Over the long term it is BARELY above 50%, nomatter how good it is.Even if it is 54%, that is stillonly a 4% better chance, and since you canot determine WHEN that 4% chance will kick in, it becomes convoluted, especially when you are trying to get a short term result from some that took a very long term "test" to figure out.
Think of it this way...
You enter a contest to win a car. they have 3 placards yo have to chose from. On one of those 3 placards is written the word CAR. So you pick, they flip one over, that in not a winner. Theythen give you thhe option of changing your original pick or keeping it. What option gives you the best odds of winning, keeping your original pick or changing it?
Baker
EOG Member
I need everyone to contact their local authorities to check and see if the local idiot is missing for I think we have solved the case. I am going to reply to wantitall4moi's last post piece by piece as I think almost every statement he made needs to be rebutted. I have to start with the last part of his post first as it is the most amusing. Note that wantitall4mois post is in bold.
Think of it this way...
You enter a contest to win a car. they have 3 placards yo have to chose from. On one of those 3 placards is written the word CAR. So you pick, they flip one over, that in not a winner. Theythen give you thhe option of changing your original pick or keeping it. What option gives you the best odds of winning, keeping your original pick or changing it?
For the love of everything holy please don't tell me you are somehow going to use this riddle in relation to sports betting. I really can't beleive you are trying to relate a riddle that you probably saw on the TV show NUMB3RS to books moving numbers. To answer your question you should always change your selection, doing so will provide for an expectation that you will end up picking the CAR placard 66.6% of the time. Do you even know why changing your selection would be the correct "play"?
Now you tell me what this has to do with sports betting.<!-- / message -->
Baker I make the same challenge here as I did at the RX, and at Stanford Wong's place( which was mysteriously lost/deleted) If ANYONE thinks they can PROVE to me that betting Kelley will be more profitable than bettinga flat rate I make the same 5K challenge.
Start with an imaginary bankroll, and PROVE that Kelley is more successful than other methods. I told those there that swear by it that if they could show a profit after a full year of using it, I would give them 5K, if they went broke before the yearwas up, or didn't show a profit they could donate the 5K to a charity of my chosing.
That thread was mysteriously deleted. So it told me all I needed to know. People like to talk about a lot of theorwtical stuff but when it comes to actually proving it, most turn tale and run, especially when real money is involved.
I don't doubt that I can't prove to you that Kelly is more profitable than flat betting, but that is only because you are too hardheaded to take a rational look at the evidence provided.
You seem to always ask people to show something in order to prove that it is true and/or possible (i.e. your statements that Kelly can't work because I haven't met any gambling multi-millionaires and your statement that middling isn't profitable because no one has shown you a long-term winning middling strategy or your statement above that a thread being deleted by some person on the internet acts as indisputable proof that the theory is invalid). Sooner or later you are going to learn that people aren't going to show you anything and you need to figure things out for yourself. If someone has something that works and is profitable they are not going to show you for a measly 5k.
Just because you can't comprehend something doesn't mean it isn't valid. JC made a mention of an widely accepted option pricing model that he used to price options when he was working as a market maker. I would presume that he was referring to Black-Scholes. I can almost guarantee you that you would find fault with this pricing model. The thing is that the guys who developed this model received the Nobel Prize in Economics. Who to beleive, the nice folks at The Nobel Foundation or wantitall4moi, who simply can't understand how it works.
Kelly is a little different as its developers weren't awarded a Nobel prize, but it is one of the most widely accepted load balancing formulas currently known. It just so happens that this formula is easily applied to bankroll management. What makes Kelly tough to use is that you must be able to accurately predict your expected winning % to take advantage of the formula. In actuality this is where most gamblers fail when utilizing Kelly, they simply cannot accurately predict the winning percentage of a particular play when placed within a subset of like plays.
As proof that Kelly will outperform straight betting I am confident that I could provide you mathematical proof but we would need to agree upon some assumptions first. We would need to agree to what the flat bet size would be (either as a product of the intial bankroll or a set dollar amount). We would need to agree on how accurately the Kelly bettor would be able to predict the winning percentage of each play and the variance of such prediction.
For example I can say that a person who flat bets 1% of initial bankroll should have an expected winning percentage of 52.9%. If they were good enough to accurately predict their winning percentage and it is >52.9% then if they used Kelly they would outperform their results obtained flat betting.
There is no error in Kelly, but as I said before most people who attempt to utilize Kelly don't realize that you must be able to accurately predict a plays expected winning % (again within a subset of like plays) before they should attempt to utilize the Kelly Criterion. The great thing about Kelly is that if a play has a negative expected value Kelly calls for a bet of $0.
I also explained how losing bettors has no relationship to losing books. Most guys do not lose all their money at the same books. I could theoretically breaka book if all I did was scalp winning sides with them. I might make one place rich by always losing with them, bt the place I was winning at would be busted soner or later. There are obviously books that ""win" more than they lose. Just because 98% of people that bet lose, doesn't mean all the books are geting their equal share of the pie.
Some obviously get more than others.
Explain to me how you could even theoretically only scalp winning sides with a particular book? If you could do this you are telling me that your expected winning percentage (theoretically??) would be 100%?
You did answer one of the questions to this thread yourself in this paragraph. It is true that all books aren't getting equal shares of the pie. In practise a books piece of the pie is equal to the volume in dollars of bets they book. It is also known as market share.
As for layingoff and polish midles, et al. If books are doing this behind the scenes or as a courtesy these are none existant. If Pin, CRIS, and OLY all have -3 for a game, not all are going to have the same hold. Pin might have -105, CRIS might have -110, and OLY might have -115. So they could EASILLY swap money around between them, especially if OLY was getting guys that wantedthe dog and the +105. Pin could easiily square the action for say a nickle. OLY would still be making a nickle and they could balance the action. If books got together and did this they would never have to move a spread unless a ton of steam hit a few books at once. The reason lines move for the mmost part is because one or two places get hit, they move, then EVERYONE moves. Some places could e totally lopsided on the dog, but be forced to move if everyone else does. They should of course hold a little longer since if they have the best line on the fave they will get their balance. but a lor of books thihnk like players here. They think steam means a winner. So they not only gamble by taking the lopsided action, they think they have a solid position with their lopsided money, and they want more so they follow the flow. So not only are they lopsided at the old number(worse), but they will probably be lopsided on the newer number as well.
Since you think it would be so easy to balance action and that books would be willing to give up action they may have attracted at a price premium to their competitors I can't beleive it is not the rage of the offshore business.
If Pinnacle is -105 on a favorite with 80000 in action on the fav and 20000 on the dog and Oly is +105 on a dog with 80000 in action on the dog and 20000 on the fav at -115 do you really think that Pinnacle would want to "swap money" with Olympic to take action with no hold whatsoever? If it is so easy why don't you explain how it could/should be done and what the holds would be.
Moving spread lines is always is a danger to books. Plain and simple, we can all agree on that. if they got together and made the balance more true they would lessen the exposure they have, not only by moving a line (opening up for middles and sides), but by balancing thier action as well. It would definately hinder the players, but would surely make it a little more stable.
Actually we don't agree. How would a book instantaneously balance action? Would you expect books to take a 1100 to win 1000 wager and instantly hedge 550 to win 500 on the other side? The result of this would be net zero. Obviuosly they only way to balance action is after all the action has been taken and by then it is too late as you will not get the same price you wrote the ticket at.
So balancing action isn't nearly as difficult as people think. I wouold say that there are 4-5 books that will actually move BECAUSE of money, all the other books are simply followers and move on air. That is also something that is pretty scarey. These places are relying on the "reputation" of the books and the players at the books thatthey are following. Hoping that these moves are "smart money" moves that makes the postions they are holding/gambling on more solid than they actually are.
The books that move without actually taking action are not doing so on the "reputation" of the books and players that they are following. These books are following what is known as Efficient Market Theory, another subject that you should go read up on.
They are replacing mathematical probabilty. the one that says there is a 50/50 chance of something ocuring, and saying that because this book or that bettor thinks it is this, they now have a better than 50/50 chance of winning. (similar to Kelley) Whne it just isn't the case. A person's opinion is only as good as their ability. Over the long term it is BARELY above 50%, nomatter how good it is.Even if it is 54%, that is stillonly a 4% better chance, and since you canot determine WHEN that 4% chance will kick in, it becomes convoluted, especially when you are trying to get a short term result from some that took a very long term "test" to figure out.
This is probably the last thing I am going to share with you. FYI probabilities don't "kick in", they are governed by the Law of Large numbers.
Think of it this way...
You enter a contest to win a car. they have 3 placards yo have to chose from. On one of those 3 placards is written the word CAR. So you pick, they flip one over, that in not a winner. Theythen give you thhe option of changing your original pick or keeping it. What option gives you the best odds of winning, keeping your original pick or changing it?
For the love of everything holy please don't tell me you are somehow going to use this riddle in relation to sports betting. I really can't beleive you are trying to relate a riddle that you probably saw on the TV show NUMB3RS to books moving numbers. To answer your question you should always change your selection, doing so will provide for an expectation that you will end up picking the CAR placard 66.6% of the time. Do you even know why changing your selection would be the correct "play"?
Now you tell me what this has to do with sports betting.<!-- / message -->
Baker I make the same challenge here as I did at the RX, and at Stanford Wong's place( which was mysteriously lost/deleted) If ANYONE thinks they can PROVE to me that betting Kelley will be more profitable than bettinga flat rate I make the same 5K challenge.
Start with an imaginary bankroll, and PROVE that Kelley is more successful than other methods. I told those there that swear by it that if they could show a profit after a full year of using it, I would give them 5K, if they went broke before the yearwas up, or didn't show a profit they could donate the 5K to a charity of my chosing.
That thread was mysteriously deleted. So it told me all I needed to know. People like to talk about a lot of theorwtical stuff but when it comes to actually proving it, most turn tale and run, especially when real money is involved.
I don't doubt that I can't prove to you that Kelly is more profitable than flat betting, but that is only because you are too hardheaded to take a rational look at the evidence provided.
You seem to always ask people to show something in order to prove that it is true and/or possible (i.e. your statements that Kelly can't work because I haven't met any gambling multi-millionaires and your statement that middling isn't profitable because no one has shown you a long-term winning middling strategy or your statement above that a thread being deleted by some person on the internet acts as indisputable proof that the theory is invalid). Sooner or later you are going to learn that people aren't going to show you anything and you need to figure things out for yourself. If someone has something that works and is profitable they are not going to show you for a measly 5k.
Just because you can't comprehend something doesn't mean it isn't valid. JC made a mention of an widely accepted option pricing model that he used to price options when he was working as a market maker. I would presume that he was referring to Black-Scholes. I can almost guarantee you that you would find fault with this pricing model. The thing is that the guys who developed this model received the Nobel Prize in Economics. Who to beleive, the nice folks at The Nobel Foundation or wantitall4moi, who simply can't understand how it works.
Kelly is a little different as its developers weren't awarded a Nobel prize, but it is one of the most widely accepted load balancing formulas currently known. It just so happens that this formula is easily applied to bankroll management. What makes Kelly tough to use is that you must be able to accurately predict your expected winning % to take advantage of the formula. In actuality this is where most gamblers fail when utilizing Kelly, they simply cannot accurately predict the winning percentage of a particular play when placed within a subset of like plays.
As proof that Kelly will outperform straight betting I am confident that I could provide you mathematical proof but we would need to agree upon some assumptions first. We would need to agree to what the flat bet size would be (either as a product of the intial bankroll or a set dollar amount). We would need to agree on how accurately the Kelly bettor would be able to predict the winning percentage of each play and the variance of such prediction.
For example I can say that a person who flat bets 1% of initial bankroll should have an expected winning percentage of 52.9%. If they were good enough to accurately predict their winning percentage and it is >52.9% then if they used Kelly they would outperform their results obtained flat betting.
There is no error in Kelly, but as I said before most people who attempt to utilize Kelly don't realize that you must be able to accurately predict a plays expected winning % (again within a subset of like plays) before they should attempt to utilize the Kelly Criterion. The great thing about Kelly is that if a play has a negative expected value Kelly calls for a bet of $0.
I also explained how losing bettors has no relationship to losing books. Most guys do not lose all their money at the same books. I could theoretically breaka book if all I did was scalp winning sides with them. I might make one place rich by always losing with them, bt the place I was winning at would be busted soner or later. There are obviously books that ""win" more than they lose. Just because 98% of people that bet lose, doesn't mean all the books are geting their equal share of the pie.
Some obviously get more than others.
Explain to me how you could even theoretically only scalp winning sides with a particular book? If you could do this you are telling me that your expected winning percentage (theoretically??) would be 100%?
You did answer one of the questions to this thread yourself in this paragraph. It is true that all books aren't getting equal shares of the pie. In practise a books piece of the pie is equal to the volume in dollars of bets they book. It is also known as market share.
As for layingoff and polish midles, et al. If books are doing this behind the scenes or as a courtesy these are none existant. If Pin, CRIS, and OLY all have -3 for a game, not all are going to have the same hold. Pin might have -105, CRIS might have -110, and OLY might have -115. So they could EASILLY swap money around between them, especially if OLY was getting guys that wantedthe dog and the +105. Pin could easiily square the action for say a nickle. OLY would still be making a nickle and they could balance the action. If books got together and did this they would never have to move a spread unless a ton of steam hit a few books at once. The reason lines move for the mmost part is because one or two places get hit, they move, then EVERYONE moves. Some places could e totally lopsided on the dog, but be forced to move if everyone else does. They should of course hold a little longer since if they have the best line on the fave they will get their balance. but a lor of books thihnk like players here. They think steam means a winner. So they not only gamble by taking the lopsided action, they think they have a solid position with their lopsided money, and they want more so they follow the flow. So not only are they lopsided at the old number(worse), but they will probably be lopsided on the newer number as well.
Since you think it would be so easy to balance action and that books would be willing to give up action they may have attracted at a price premium to their competitors I can't beleive it is not the rage of the offshore business.
If Pinnacle is -105 on a favorite with 80000 in action on the fav and 20000 on the dog and Oly is +105 on a dog with 80000 in action on the dog and 20000 on the fav at -115 do you really think that Pinnacle would want to "swap money" with Olympic to take action with no hold whatsoever? If it is so easy why don't you explain how it could/should be done and what the holds would be.
Moving spread lines is always is a danger to books. Plain and simple, we can all agree on that. if they got together and made the balance more true they would lessen the exposure they have, not only by moving a line (opening up for middles and sides), but by balancing thier action as well. It would definately hinder the players, but would surely make it a little more stable.
Actually we don't agree. How would a book instantaneously balance action? Would you expect books to take a 1100 to win 1000 wager and instantly hedge 550 to win 500 on the other side? The result of this would be net zero. Obviuosly they only way to balance action is after all the action has been taken and by then it is too late as you will not get the same price you wrote the ticket at.
So balancing action isn't nearly as difficult as people think. I wouold say that there are 4-5 books that will actually move BECAUSE of money, all the other books are simply followers and move on air. That is also something that is pretty scarey. These places are relying on the "reputation" of the books and the players at the books thatthey are following. Hoping that these moves are "smart money" moves that makes the postions they are holding/gambling on more solid than they actually are.
The books that move without actually taking action are not doing so on the "reputation" of the books and players that they are following. These books are following what is known as Efficient Market Theory, another subject that you should go read up on.
They are replacing mathematical probabilty. the one that says there is a 50/50 chance of something ocuring, and saying that because this book or that bettor thinks it is this, they now have a better than 50/50 chance of winning. (similar to Kelley) Whne it just isn't the case. A person's opinion is only as good as their ability. Over the long term it is BARELY above 50%, nomatter how good it is.Even if it is 54%, that is stillonly a 4% better chance, and since you canot determine WHEN that 4% chance will kick in, it becomes convoluted, especially when you are trying to get a short term result from some that took a very long term "test" to figure out.
This is probably the last thing I am going to share with you. FYI probabilities don't "kick in", they are governed by the Law of Large numbers.
The General
Another Day, Another Dollar
I am going to take a shot at answering my own question.
IF 10 top books were able to book action at even money as desrcibed, then this would harm Line Services because fewer players would need to worry about shopping for that xtra 3-5 cent savings.
Am I in the ballpark?
IF 10 top books were able to book action at even money as desrcibed, then this would harm Line Services because fewer players would need to worry about shopping for that xtra 3-5 cent savings.
Am I in the ballpark?
WOW I thought this died. But now that I see it...
Baker, you said it yourslef, most guys can't make Kelly work because they don't have a true prediction of their win expectations. You know why? Because it doesn't matter. Because it isn't the probability of the win, but simply the overlay or underlay of the odds themselves. Something that can be defined, not something that is guess work.
Kelly was first applied for BJ I believe where a person knew his odds, and his probabilities of winning in a certain situation (after counting enough cards to gain an advantage). Sportsbetting is totally different from cards or dice. Especially when you have a built in system that makes the players expectaion less than 50%, no matter what they think they can pick winners at.
THAT is the built in flaw. Your expectaion is ALWAYS less than 50%. Because that is how the pay odds work. So you should by definition of Kelly be betting $0 on every bet.
The only way people can predict their "values" or expectations is by comparing lines at differing books and playing the best line for whatever side their OPINION has made the position to take. So they are taking two man made ideas; a line, and their opinion and then try to put MATH into it.
So with Kelly there is a constant up and down of bet sizes going on. Obviously if you win your big bets and lose your smaller bets (the ideal Kelly result) you are going to make more money than someoe flat betting. That is common sense. The thing is that, as you say, most people do not make enough bets to get a good determination of their expection. Also line shopping can fluctuate that win expection 1-2% over the long haul, so again, there is a built in buffer. Especially when the whole theory is based on a person's win expectaions. Most guys are going to be 50%, simply because that is the true expectaion. But by getting the best numbers and best odds they can manipulate their returns. They are not really picking more winners, they are just getting better pay offs and better odds.
So a 54% player isn't really that much better thansomone who is picking 52%. He probably just got on the better side of the bad beats along the way to determining his expectaion. But if the guy that won 52% got BETTER holds, he would theoretically make more profit, utillizing Kelly or not.
Kelly simply puts too much pressure on the more highly opinionated games to succeed.
I read a paper about it once, and the guy said that because he had made 101 wagers and shown a profit that it was proven. That is so laughable it is barely worth mentioning. But when people hang their hats on these "scientific" studies it surely is worth mentioning. 101 bets is BARELY a month for 80% of gamblers. How is that any indicator? A streak here or a lucky win there, ANY betting method can succeed in that amount of bets. Try it for 2-3 thousand bets and see.
There are simply too many things that are in Kelly to make it work over the long haul. It is just a fancy verion of the star system. 1 star if you think the game is close, but you stilllike the side. 5 stars if you like the side quite a bit but they aren't a lock, and 10 stars for a team you don't THINK can lose.Oversimplified, but same difference.
Also as for Black Scholes... another economic theory that has been introduced to sportsbetting.
Stocks are not sports. Their "values" are not pulled out of thin air like a teams lines are. Most have had years and years of data collected on them, they go up and down, you have miles and miles of data to use to evaluate and predict moves. They hve a track record that can be tracked back to their inception.
With lines you have 5-6 days to evaluate and determine, big big difference. And as far as buying options and selling on teams, the whole panic factor and quick trun around makes it impossible to determine. Similar to what is happening in the oil market now. Try applying Black Scholes to what is going on there, with the war, a lame duck president, and hurricanes up the ass. The come back in a few months and look to see how it worked out. Or look back at what Black Scholes predicted for this quarter.
Sports is full of unseen and volatile potholes. Trying to determine and apply complex economic theories to them is not going to work. Because they are just too volatile. While options are also volatile, they are not nearly as unpredictable as a sporting event. That is why I point out all the factors influencing the oil market now. It is akin to what happens in every sporting event that ever happens.
And it is also still THEORY, even when applied to the arena it was written for.
Obviously this got way off topic, but the thread died a long time ago, so I see no harm in commenting.
Baker, you said it yourslef, most guys can't make Kelly work because they don't have a true prediction of their win expectations. You know why? Because it doesn't matter. Because it isn't the probability of the win, but simply the overlay or underlay of the odds themselves. Something that can be defined, not something that is guess work.
Kelly was first applied for BJ I believe where a person knew his odds, and his probabilities of winning in a certain situation (after counting enough cards to gain an advantage). Sportsbetting is totally different from cards or dice. Especially when you have a built in system that makes the players expectaion less than 50%, no matter what they think they can pick winners at.
THAT is the built in flaw. Your expectaion is ALWAYS less than 50%. Because that is how the pay odds work. So you should by definition of Kelly be betting $0 on every bet.
The only way people can predict their "values" or expectations is by comparing lines at differing books and playing the best line for whatever side their OPINION has made the position to take. So they are taking two man made ideas; a line, and their opinion and then try to put MATH into it.
So with Kelly there is a constant up and down of bet sizes going on. Obviously if you win your big bets and lose your smaller bets (the ideal Kelly result) you are going to make more money than someoe flat betting. That is common sense. The thing is that, as you say, most people do not make enough bets to get a good determination of their expection. Also line shopping can fluctuate that win expection 1-2% over the long haul, so again, there is a built in buffer. Especially when the whole theory is based on a person's win expectaions. Most guys are going to be 50%, simply because that is the true expectaion. But by getting the best numbers and best odds they can manipulate their returns. They are not really picking more winners, they are just getting better pay offs and better odds.
So a 54% player isn't really that much better thansomone who is picking 52%. He probably just got on the better side of the bad beats along the way to determining his expectaion. But if the guy that won 52% got BETTER holds, he would theoretically make more profit, utillizing Kelly or not.
Kelly simply puts too much pressure on the more highly opinionated games to succeed.
I read a paper about it once, and the guy said that because he had made 101 wagers and shown a profit that it was proven. That is so laughable it is barely worth mentioning. But when people hang their hats on these "scientific" studies it surely is worth mentioning. 101 bets is BARELY a month for 80% of gamblers. How is that any indicator? A streak here or a lucky win there, ANY betting method can succeed in that amount of bets. Try it for 2-3 thousand bets and see.
There are simply too many things that are in Kelly to make it work over the long haul. It is just a fancy verion of the star system. 1 star if you think the game is close, but you stilllike the side. 5 stars if you like the side quite a bit but they aren't a lock, and 10 stars for a team you don't THINK can lose.Oversimplified, but same difference.
Also as for Black Scholes... another economic theory that has been introduced to sportsbetting.
Stocks are not sports. Their "values" are not pulled out of thin air like a teams lines are. Most have had years and years of data collected on them, they go up and down, you have miles and miles of data to use to evaluate and predict moves. They hve a track record that can be tracked back to their inception.
With lines you have 5-6 days to evaluate and determine, big big difference. And as far as buying options and selling on teams, the whole panic factor and quick trun around makes it impossible to determine. Similar to what is happening in the oil market now. Try applying Black Scholes to what is going on there, with the war, a lame duck president, and hurricanes up the ass. The come back in a few months and look to see how it worked out. Or look back at what Black Scholes predicted for this quarter.
Sports is full of unseen and volatile potholes. Trying to determine and apply complex economic theories to them is not going to work. Because they are just too volatile. While options are also volatile, they are not nearly as unpredictable as a sporting event. That is why I point out all the factors influencing the oil market now. It is akin to what happens in every sporting event that ever happens.
And it is also still THEORY, even when applied to the arena it was written for.
Obviously this got way off topic, but the thread died a long time ago, so I see no harm in commenting.
OMNIVOROUS FROG
EOG Master
I may be declared clinically brain dead after attempting to read and comprehend this thread. There is so much verbage, yet the concepts get lost. Plus you have tangents going every direction possible. I am so confused I forgot what I was going to say. Maybe a nap or some Gingko Biloba will help.
Best Wishes...OF :+waving-5
Best Wishes...OF :+waving-5
OMNIVOROUS FROG said:
That says a lot coming from YOU:doh1
Baker
EOG Member
Wantitall4moi it is obvious to me that you are simply at a different level than I am in regards to the application of mathematics (not to mention statistics and econometrics) towards problems that sports betting presents. In regards to the Kelly Criterion the main difference between you and I is that I believe that it is both possible to have an expected chance of success of over 55% and be able to estimate your you chance of success with a small acceptable variance with a confidence interval of 95%. Will I prove this to you or show you? Not a chance in hell.
What concerns me about you is that you rather eloquently spread misinformation. Up until your last post you claimed that the inherent flaw in the Kelly Criterion was that streaks could create a situation where a person would be wagering a tremendous amount of money relative to their beginning bankroll. In actuality streaks are of no consequence when using Kelly. Meaning that you could go 51 wins followed by 49 losses or 49 losses followed by 51 wins and end with the same bankroll. My contention is if you took a specific set of results and calculated the P/L using flat betting and then took the same results and correctly applied the Kelly Criterion you would have made more money using the Kelly Criterion.
You have now changed your stance that the inherent flaw of Kelly is that people always have a 50 chance of success. How is this a flaw of Kelly? If a situation doesn't have a positive expectation Kelly says to bet $0. The old saying Garbage in Garbage out comes to mind in this regard. If someone plugs invalid values into the Kelly Criterion that is no more a fault of Kelly Criterion than if someone measured a side of a triangle wrong when utilizing the Pythagorean Theorem.
It seems to be your contention that bettors always have a 50/50 chance of winning, which leaves them with negative expectations. This is interesting as it proves that books should always hold all action.
The original point of contention in this thread was whether bookies should keep all action or lay off in order to get balanced action. If every wager has a 50% chance of winning and there is no correlation between the winning side and the side that has received unbalanced action then it can be said that whether a each certain side wins ATS is a random occurrence. If there is no correlation to the side that wins and the side that has unbalanced action than the balance of the action must be either a random or normal distribution with a mean equal to balanced action.
This being the case it can be said that for every time when Team B wins where 10% is on Team A and 90% is on Team B there will be an occurrence of Team A winning when the action is balanced 90/10 in favor of Team A. For this not to be true there would have to a correlation seen between the balance of action at a particular book and the winning side. If this correlation existed the proper methodology would be to take unbalanced action of the "light side" of the action, even to the point that the book would want to take more action (by placing bets) on the "light" side than it is holding in action on the heavy side. It is not very possible that this correlation would exist, even if it did exist it still wouldn't be correct to try to balance action.
The Law of Large Numbers (note LAW, even you can't argue with a LAW can you?) states that as the number of trials increases the ratio of successes to converges upon the expected ration of success.
What does all of this mean? This means that the more events that a book books where Team A has a random 50% chance of winning ATS and the action is unbalanced in a random or normally distributed manner the closer their profits will converge on their expected profits where (volume * hold% = expected profits).
Now I don't completely prescribe to the above theory in that I believe there are ways to create a situation where the expected probability of success is >55%. In actuality there are so few people that can do this it would only lower a books hold be at most a very modest amount. In actuality I think books that do a good job of profiling their clients should tail sharp bettors for even more of a profit, but hell that would be gambling.
What concerns me about you is that you rather eloquently spread misinformation. Up until your last post you claimed that the inherent flaw in the Kelly Criterion was that streaks could create a situation where a person would be wagering a tremendous amount of money relative to their beginning bankroll. In actuality streaks are of no consequence when using Kelly. Meaning that you could go 51 wins followed by 49 losses or 49 losses followed by 51 wins and end with the same bankroll. My contention is if you took a specific set of results and calculated the P/L using flat betting and then took the same results and correctly applied the Kelly Criterion you would have made more money using the Kelly Criterion.
You have now changed your stance that the inherent flaw of Kelly is that people always have a 50 chance of success. How is this a flaw of Kelly? If a situation doesn't have a positive expectation Kelly says to bet $0. The old saying Garbage in Garbage out comes to mind in this regard. If someone plugs invalid values into the Kelly Criterion that is no more a fault of Kelly Criterion than if someone measured a side of a triangle wrong when utilizing the Pythagorean Theorem.
It seems to be your contention that bettors always have a 50/50 chance of winning, which leaves them with negative expectations. This is interesting as it proves that books should always hold all action.
The original point of contention in this thread was whether bookies should keep all action or lay off in order to get balanced action. If every wager has a 50% chance of winning and there is no correlation between the winning side and the side that has received unbalanced action then it can be said that whether a each certain side wins ATS is a random occurrence. If there is no correlation to the side that wins and the side that has unbalanced action than the balance of the action must be either a random or normal distribution with a mean equal to balanced action.
This being the case it can be said that for every time when Team B wins where 10% is on Team A and 90% is on Team B there will be an occurrence of Team A winning when the action is balanced 90/10 in favor of Team A. For this not to be true there would have to a correlation seen between the balance of action at a particular book and the winning side. If this correlation existed the proper methodology would be to take unbalanced action of the "light side" of the action, even to the point that the book would want to take more action (by placing bets) on the "light" side than it is holding in action on the heavy side. It is not very possible that this correlation would exist, even if it did exist it still wouldn't be correct to try to balance action.
The Law of Large Numbers (note LAW, even you can't argue with a LAW can you?) states that as the number of trials increases the ratio of successes to converges upon the expected ration of success.
What does all of this mean? This means that the more events that a book books where Team A has a random 50% chance of winning ATS and the action is unbalanced in a random or normally distributed manner the closer their profits will converge on their expected profits where (volume * hold% = expected profits).
Now I don't completely prescribe to the above theory in that I believe there are ways to create a situation where the expected probability of success is >55%. In actuality there are so few people that can do this it would only lower a books hold be at most a very modest amount. In actuality I think books that do a good job of profiling their clients should tail sharp bettors for even more of a profit, but hell that would be gambling.
Ironlock
2
Baker-
We need a new topic. More of your brain needs to be spilled onto the forums. We both know wantiallformoi will respond in some manner that appears to be in disagreement with your above post. I, however, cannot fathom how a reasonable human being cannot appreciate the truthful, concise, well thought out, perfectly crafted, mathematiclly agreeable, nature of your post.
Thanks for the post. :+thumbs-2
Iron.
We need a new topic. More of your brain needs to be spilled onto the forums. We both know wantiallformoi will respond in some manner that appears to be in disagreement with your above post. I, however, cannot fathom how a reasonable human being cannot appreciate the truthful, concise, well thought out, perfectly crafted, mathematiclly agreeable, nature of your post.
Thanks for the post. :+thumbs-2
Iron.
There might be some semantics issues here. I don't think I ever said that all bets have a 50/50 chance of winning. I think I said that most bettors will, over a long stretch regress to 50/50. Maybe not in those terms, but similar.
Some sides due to the human influnce of a spread will have some sort of bias. I am not sure what that is, not sure anyone could define it. Because of the nature of sports betting and the flukes that happen.
But I still stand by the streakiness aspect, I just figured I had stated that one enough so it didn't need to be repeated again. It was alluded when I cited the paper I had read about someone saying Kelly was "proven" after 101 bets or something like that.
Now as for the whole laying off, and what not. I also never said that what books did was mathematically incorrect either. I simply said that they are running the risks, the same risks bettors run, when they have too many eggs in one basket. If they had an eternity to get back to even then it would not be an issue. My whole concern is that when they keep such lopsided positions ( I know these have been exaggerated, but still some have close I am sure) then if a book gets their brains beat in and there is a run on the bank where is that moneygoing to come from?
They take the lopsided action as if no one is ever going to ask for a pay out. Which in some case is probably not a bad theory. But some guys will win eventually here and there, and I am sure some expect to get paid.
What we both agree on is that long term everything is eventually going to even out, regress towards the mean, whatever fancy handle you want to cal it by. What we don't agree on is how gamblers can manipulate their positions to take advantage of biases they may think they see. You think these advantages are easilly discernable and easy to spot. While I say they are hit or miss, and that any one person's opinion is still only as good as his overall opinion.
Now what you say about getting thebalances with Team A and Team B, that would make sense in a vacuum. But sports beting isn't a vacuum. Again we see things differently there.
Same as with the Law of Large Numbers, again that "law" primairilly holds to things that are constant and unbiased. Betting lines are certainly not that. When human influnces get introduced all mathematical laws are thrown out the window. They can MAYBE be used to some degree, but the human factor makes it impossible to figure out just what kind of advantgae or disadvantage will be gained/lost when you apply the law or theory.
That is why I say Kelly is flawed. Probably semantics yes. But if noone can get an accurate probability of success then how can it work?
Name me one person who can say what their EXACT win/loss expection/ratio is. It can't be done. too many factors come into play. Something as simple as getting a half point can change their win rate 1-2%. They are not always going to get that extra half point either, so that in and of itself makes Kelly nearly impossible to apply to sprts. Not to mention emotions of a player on a win streak or a losing streak, the spur of the moment bets, the buy backs, the scalps, the half time middle shot taking. All thses things come into play, making it impossible to even figure out the players/bettors win probbaility.
Then there is the game itself. A turnover, a bad call, a missed tackle, a bad penalty. Those things wreak havoc on a player who is trying to handicap a game, so "bad beats" have to account for something, just as do lucky wins. I know that theoretically they will even out, but they won't EVEN out if a bad beat was a high level play (as applied with Kelly) and the lucky win was a low level play. While the wins and losses may even out, the actual money probably won't.
We both have our stances. I know what works for me. I make money doing this. So I have to think I am doing something right. But I am also smart enough to realize that there might be abetter way out there. I just do not think Kelly is it.
I think I have a decent handle on my expectaions on opinionated plays, I make about 1500-1600 of them a year or so. Was making more when I was betting NBA and doing this everyday. But even so that is a lot of bets, and after 15 or so years of doing it I think I have an idea of what to expect.
I think that 51-53% is attainable, and when those percentages include ML dogs in the varying sports, then that is more than a high enough percentage to profit. I have never beena believer in records and win rates. I keep score by my bankroll. If I could go 50-50 and have a 1-3% advantage on every game I would be very happy.
That was one thing you mentioned earlier that I didn't comment on. the whole scalping between 2 books. In theory a scalper could break a book, and yes it does mean that the players advantage is greater than 100%.
If You could get a daily game with odds at one place of +120, and the other team at -110 at another books. If one books lost their position enough they would eventualy go broke. Now as we have discussed sooner or later due to the law of large numbers they are supposed to split wins and losses. But as I contend, when the odds are man made and they are at a higher ratio than +100 (50/50) then that law becomes unuseable. If the book that has to pay +120 for every "loss", they need to have their side win 55% of the time just to make a slight profit. If their side only "wins" 50% of the time (as the law of large numbers would suggest) they will lose everything sooner or later. Especailly if they are unbalanced to boot.
Robbing Peter to pay Paul, THAT was my point about not balancing action, and having too much imbalance with some games. Sooner or later people have to get paid, and when the books are "losing" people are winning. So that only compounds the issue, since when people win enough they want money, and to win enough, it stands to reason that the book "lost" enough to get them there. So if they don't have it somewhere, then the problems start.
I think we bot agree on the math, I just think we apply it a lot differntly.
Some sides due to the human influnce of a spread will have some sort of bias. I am not sure what that is, not sure anyone could define it. Because of the nature of sports betting and the flukes that happen.
But I still stand by the streakiness aspect, I just figured I had stated that one enough so it didn't need to be repeated again. It was alluded when I cited the paper I had read about someone saying Kelly was "proven" after 101 bets or something like that.
Now as for the whole laying off, and what not. I also never said that what books did was mathematically incorrect either. I simply said that they are running the risks, the same risks bettors run, when they have too many eggs in one basket. If they had an eternity to get back to even then it would not be an issue. My whole concern is that when they keep such lopsided positions ( I know these have been exaggerated, but still some have close I am sure) then if a book gets their brains beat in and there is a run on the bank where is that moneygoing to come from?
They take the lopsided action as if no one is ever going to ask for a pay out. Which in some case is probably not a bad theory. But some guys will win eventually here and there, and I am sure some expect to get paid.
What we both agree on is that long term everything is eventually going to even out, regress towards the mean, whatever fancy handle you want to cal it by. What we don't agree on is how gamblers can manipulate their positions to take advantage of biases they may think they see. You think these advantages are easilly discernable and easy to spot. While I say they are hit or miss, and that any one person's opinion is still only as good as his overall opinion.
Now what you say about getting thebalances with Team A and Team B, that would make sense in a vacuum. But sports beting isn't a vacuum. Again we see things differently there.
Same as with the Law of Large Numbers, again that "law" primairilly holds to things that are constant and unbiased. Betting lines are certainly not that. When human influnces get introduced all mathematical laws are thrown out the window. They can MAYBE be used to some degree, but the human factor makes it impossible to figure out just what kind of advantgae or disadvantage will be gained/lost when you apply the law or theory.
That is why I say Kelly is flawed. Probably semantics yes. But if noone can get an accurate probability of success then how can it work?
Name me one person who can say what their EXACT win/loss expection/ratio is. It can't be done. too many factors come into play. Something as simple as getting a half point can change their win rate 1-2%. They are not always going to get that extra half point either, so that in and of itself makes Kelly nearly impossible to apply to sprts. Not to mention emotions of a player on a win streak or a losing streak, the spur of the moment bets, the buy backs, the scalps, the half time middle shot taking. All thses things come into play, making it impossible to even figure out the players/bettors win probbaility.
Then there is the game itself. A turnover, a bad call, a missed tackle, a bad penalty. Those things wreak havoc on a player who is trying to handicap a game, so "bad beats" have to account for something, just as do lucky wins. I know that theoretically they will even out, but they won't EVEN out if a bad beat was a high level play (as applied with Kelly) and the lucky win was a low level play. While the wins and losses may even out, the actual money probably won't.
We both have our stances. I know what works for me. I make money doing this. So I have to think I am doing something right. But I am also smart enough to realize that there might be abetter way out there. I just do not think Kelly is it.
I think I have a decent handle on my expectaions on opinionated plays, I make about 1500-1600 of them a year or so. Was making more when I was betting NBA and doing this everyday. But even so that is a lot of bets, and after 15 or so years of doing it I think I have an idea of what to expect.
I think that 51-53% is attainable, and when those percentages include ML dogs in the varying sports, then that is more than a high enough percentage to profit. I have never beena believer in records and win rates. I keep score by my bankroll. If I could go 50-50 and have a 1-3% advantage on every game I would be very happy.
That was one thing you mentioned earlier that I didn't comment on. the whole scalping between 2 books. In theory a scalper could break a book, and yes it does mean that the players advantage is greater than 100%.
If You could get a daily game with odds at one place of +120, and the other team at -110 at another books. If one books lost their position enough they would eventualy go broke. Now as we have discussed sooner or later due to the law of large numbers they are supposed to split wins and losses. But as I contend, when the odds are man made and they are at a higher ratio than +100 (50/50) then that law becomes unuseable. If the book that has to pay +120 for every "loss", they need to have their side win 55% of the time just to make a slight profit. If their side only "wins" 50% of the time (as the law of large numbers would suggest) they will lose everything sooner or later. Especailly if they are unbalanced to boot.
Robbing Peter to pay Paul, THAT was my point about not balancing action, and having too much imbalance with some games. Sooner or later people have to get paid, and when the books are "losing" people are winning. So that only compounds the issue, since when people win enough they want money, and to win enough, it stands to reason that the book "lost" enough to get them there. So if they don't have it somewhere, then the problems start.
I think we bot agree on the math, I just think we apply it a lot differntly.
tons of solid information here.........
Rainbow....knows his stuff and so does Judge....
wantitall4moi.... has some valid points but imo they are from a "players perspective" not a bookmakers....
the basic gambling philosphy of a "player" is different in many aspects verses a bookmakers......
such as this........many a sharp player can see "value" in laying -250 on a team..(or whatever chalk #).....whereas a bookie by nature wouldnt think of laying more than a buck fifty on whether or not the sun was going to rise tommorrow.......
successfull shops know who is betting sharp action..and act accordingly.....
Rainbow...btw i havent had a chance to talk to dk or herby about their trip down there last month....
Rainbow....knows his stuff and so does Judge....
wantitall4moi.... has some valid points but imo they are from a "players perspective" not a bookmakers....
the basic gambling philosphy of a "player" is different in many aspects verses a bookmakers......
such as this........many a sharp player can see "value" in laying -250 on a team..(or whatever chalk #).....whereas a bookie by nature wouldnt think of laying more than a buck fifty on whether or not the sun was going to rise tommorrow.......
successfull shops know who is betting sharp action..and act accordingly.....
Rainbow...btw i havent had a chance to talk to dk or herby about their trip down there last month....
The General
Another Day, Another Dollar
Welcome to EOG Dime.
I hope to see you around often.
BOL
I hope to see you around often.
BOL
very nice forum ya got here boys:+waving-5
Dime- Great to see you here. You make a good point on how you see things according to what side of the table that you are on.
Northern Star
EOG Enthusiast
Want:
It is apparent you don't understand how Kelly works. Just as an example with how Kelly works. Lets just show you a goofed up example of how it works versus straight betting the same amount. Assume you start your flat bet and keep it the same for every bet. Your bankroll is $100 and you wager $16 each bet. For simplicity lets say the kelly % you use is also 16% or a first wager of $16. You end up winning 60 plays and losing 40 plays. But just say you lose the first 10 plays and go 60-30 on the following 90 plays. The straight bettor is broke after 10 plays. The kelly bettor is still playing another day and wins at the end of his 100 plays.
With Kelly you never go broke. The two things Kelly does is to 1. Never go broke 2. Maximize your profits. Simple as that. If you predicted 60% win rate and played 15% of your bankroll or 17% of your bankroll, your would not have as big a bankroll as if you would have played 16% of your bankroll. The other thing youare totally off on is when you win or lose. I could lose 40 games in a row and then win 60 in a row in the above example and my bankroll would be exactly the same if it happened in the reverse order (win 60 in a row and lose 40 in a row). There is a formula which tells you exactly where your bankroll would be after some number of plays. You can donate the $5,000 to the Heiffer Foundation or if you prefer to help the people in New Orleans.
Northern Star
It is apparent you don't understand how Kelly works. Just as an example with how Kelly works. Lets just show you a goofed up example of how it works versus straight betting the same amount. Assume you start your flat bet and keep it the same for every bet. Your bankroll is $100 and you wager $16 each bet. For simplicity lets say the kelly % you use is also 16% or a first wager of $16. You end up winning 60 plays and losing 40 plays. But just say you lose the first 10 plays and go 60-30 on the following 90 plays. The straight bettor is broke after 10 plays. The kelly bettor is still playing another day and wins at the end of his 100 plays.
With Kelly you never go broke. The two things Kelly does is to 1. Never go broke 2. Maximize your profits. Simple as that. If you predicted 60% win rate and played 15% of your bankroll or 17% of your bankroll, your would not have as big a bankroll as if you would have played 16% of your bankroll. The other thing youare totally off on is when you win or lose. I could lose 40 games in a row and then win 60 in a row in the above example and my bankroll would be exactly the same if it happened in the reverse order (win 60 in a row and lose 40 in a row). There is a formula which tells you exactly where your bankroll would be after some number of plays. You can donate the $5,000 to the Heiffer Foundation or if you prefer to help the people in New Orleans.
Northern Star
The General
Another Day, Another Dollar
Welcome to EOG Northern Star<!-- / message -->
Why would I wager 16 if I started with 100? No one that i know of would be betting 16% of their bankroll on a single play. When you start that small you might as well bet it all and hope.
Not to mention a percentage bettor wouldn't be betting $16 (16%) after he lost the first bet. He would then have $84, and even at 16% he would only be betting $13 (13.44) on his next play. Thus if he lost he would still have $71 left after going 0-2.
Your post makes no sense, and niether do your examples.
Not to mention a percentage bettor wouldn't be betting $16 (16%) after he lost the first bet. He would then have $84, and even at 16% he would only be betting $13 (13.44) on his next play. Thus if he lost he would still have $71 left after going 0-2.
Your post makes no sense, and niether do your examples.
Northern Star
EOG Enthusiast
It was just an example. So lets go real real slow and maybe you will learn something.
You determine what the straight bettor will wager based off a starting bankroll of $100, also let me know what the gamblers record will be after a 1,000 games.
Then I will show you are wrong.
You determine what the straight bettor will wager based off a starting bankroll of $100, also let me know what the gamblers record will be after a 1,000 games.
Then I will show you are wrong.
The only wat to SHOW is to bet kelly out at the percentages it calls for at whatever a person thinks is win probability is.
There is no way to "proof" it other than DOING it.
I have no doubt that in a vacuum the formula would work, ONLY if the bettor's OPINION was as good as he THOUGHT it was.
When you flat bet you take all that guess work out of it.
What really gets kelly followers in trouble is betting MLB dogs. Especially anything over +150. How hard is it for someone to bet what they SHOULD bet on a +160 dog? or the most part people would probably think they had less than a 50% chance to win those games so they wouldn't bet them at all. Since that is the foundation of Kelly.When you have less than a 50% chance of winningyou bet 0%.
As a flat bettor, I am betting my same amount on that +160 dog, so if it wins, I am netting 8.35/5 (-107 standard) of any other bet (-107) I might make.
There is no way I am going to change, and people that follow Kelly aren't going to change either, no matter if they win or lose.
It isn't really the system that is flawed it is the way gamblers utilize it that is. But even in saying that Kelly was not made for sportsettors to use, so its not working really isn't surprising.
There is no way to "proof" it other than DOING it.
I have no doubt that in a vacuum the formula would work, ONLY if the bettor's OPINION was as good as he THOUGHT it was.
When you flat bet you take all that guess work out of it.
What really gets kelly followers in trouble is betting MLB dogs. Especially anything over +150. How hard is it for someone to bet what they SHOULD bet on a +160 dog? or the most part people would probably think they had less than a 50% chance to win those games so they wouldn't bet them at all. Since that is the foundation of Kelly.When you have less than a 50% chance of winningyou bet 0%.
As a flat bettor, I am betting my same amount on that +160 dog, so if it wins, I am netting 8.35/5 (-107 standard) of any other bet (-107) I might make.
There is no way I am going to change, and people that follow Kelly aren't going to change either, no matter if they win or lose.
It isn't really the system that is flawed it is the way gamblers utilize it that is. But even in saying that Kelly was not made for sportsettors to use, so its not working really isn't surprising.
Northern Star
EOG Enthusiast
The kelly system is very sound. One of the bigest mistakes gamblers make is poor money management. The fact that a gambler is making an informed decision to reduce his play when he is losing and increase his plays when he is winning is exactly what you should be doing.
Kelly is like having compound interest on your money. Let me give you one more example. Assume you start with a bankroll of $1,000. You will wager your single best play every day. After 1500 plays your record is 870-630 (58%). Your % of bankroll wagered is 6%.
Now if I followed your method of straight wagering $60 you would have a nice bankroll of $11,620.
Now take the same scenario and use the Kelly criteria and wager 6% and play those exact same 1500 plays and your bankroll would be
$1,372,000
The way I see it that is over a million reason why the kelly criteria is better than straight betting. There is no arguing that your method is better and maybe you should get a better understand of a great money management system.
Northern Star
Kelly is like having compound interest on your money. Let me give you one more example. Assume you start with a bankroll of $1,000. You will wager your single best play every day. After 1500 plays your record is 870-630 (58%). Your % of bankroll wagered is 6%.
Now if I followed your method of straight wagering $60 you would have a nice bankroll of $11,620.
Now take the same scenario and use the Kelly criteria and wager 6% and play those exact same 1500 plays and your bankroll would be
$1,372,000
The way I see it that is over a million reason why the kelly criteria is better than straight betting. There is no arguing that your method is better and maybe you should get a better understand of a great money management system.
Northern Star