Outside The Box

[katstale]

Since this area seems to get the less goofball traffic from the "bet the lunch money bunch" I thought I might start a thread which might help some of the more serious minded on here think abt some issues related to gambling in ways which might help there bottomline. As Ganch mentioned in another thread--I will not be giving away proprietary information--but I believe I can point you in some directions maybe you never thought abt.

If the thread gets invaded by punks and/or charlatans I will just discontinue posting in it.

Let me start the first post by saying that someone was asking abt developing a model to pick winners. Let's define that?! The boys in Vegas have developed a model also. It is the model that sets the line--so you are saying you want to develop a model that is better, more sophisticated somehow, than theirs? Good luck at that. More than likely a fruitless and costly enterprise.

Instead, lets take their model and assign it the status of the "holy grail"! AND adjust our thinking from there. More to follow if any wish!!

[curious]

Since this area seems to get the less goofball traffic from the "bet the lunch money bunch" I thought I might start a thread which might help some of the more serious minded on here think abt some issues related to gambling in ways which might help there bottomline. As Ganch mentioned in another thread--I will not be giving away proprietary information--but I believe I can point you in some directions maybe you never thought abt.

If the thread gets invaded by punks and/or charlatans I will just discontinue posting in it.

Let me start the first post by saying that someone was asking abt developing a model to pick winners. Let's define that?! The boys in Vegas have developed a model also. It is the model that sets the line--so you are saying you want to develop a model that is better, more sophisticated somehow, than theirs? Good luck at that. More than likely a fruitless and costly enterprise.

Instead, lets take their model and assign it the status of the "holy grail"! AND adjust our thinking from there. More to follow if any wish!!

Sounds like an interesting hypothesis.

[TLD]

I’m not 100% sure what you’re getting at, but let me take a stab at it.

Most sportsbettors, certainly most newbies, assume that the idea is to handicap a sporting event skillfully enough to recognize which side of the line offers value. In effect, you’re trying to beat your book by being better than it at projecting the outcome of a given sporting event.

On the other hand, many more experienced bettors—in fact, I would venture to guess the majority of sharps and people who think they’re sharp—have come to believe that this is, as you say, a “fruitless and costly enterprise” because you will never be better at making such projections than those who set the line (whether they be thought of as the sportsbooks, linesmaking services such as Las Vegas Sporting Consultants that the sportsbooks work with, the sharp sportsbetting syndicates whose wagers have the most influence on line movement, or simply all market forces collectively).

So then the idea would not be to surpass in handicapping ability these other forces, but instead to find a profitable way to make money off their superior ability.

I think “steam” betting is a prime example of this. A steam bettor does not attempt to find lines that differ from his own judgment of what the correct projections of the outcomes should be. Instead, he attempts to piggyback onto the bets of those whose judgment of such matters he recognizes as being superior to his own.

That is, if the best evidence available to him of when the sharpest and most knowledgeable betting syndicates have a certain opinion on a game is that the lines at virtually all the books move in the same direction in a very short span of time, then he jumps to bet on the same side he infers from the movement they are betting on, preferably at a “straggler” book that hasn’t moved yet and so is momentarily off the (new) market line.

Here he’s not handicapping the game at all, but is simply interpreting market movement to tell him what those far sharper than himself have concluded about this game, and he’s trusting their opinion.

So I take your question to be, what can we come up with like that, that instead of trying to beat the opinions of the experts on both sides of the counter (the “Holy Grail”), tries to win by ascertaining and following those opinions?

(Not that I have an answer. I’m just trying to make sure I understand the question.)

[katstale]

You are going in the right direction. Welcome to this thread. I will go to the next step tomorrow, but I never "assume" anybody is smarter than the wise guys. In fact, for right or wrong, I am assuming the wiseguys are smart and everyone else is a wannabe. It's this thinking which will form your mathematical model.

Again, good start.

[TLD]

It’s not my approach, but I respect it and will read with interest whatever you and others choose to share in this area.

I’m one of the deluded masses seeking to make my own projections and bet them when they differ significantly from the best line that happens to be available when I’m ready to bet. (And my results have far surpassed chance for many years running, so I can’t help thinking that perhaps the impossible is possible after all.)

But like I say, I respect those who instead are seeking to play little angles based on market movement and such. And for that matter, it’s not like the approaches are incompatible. One could, for example, use what information one gleans from watching the market as just another piece of evidence in one’s handicapping, without it being the sole or controlling factor. Thus it would augment, rather than replace, one’s own handicapping.

[katstale]

Glad you are keeping an open mind. I don't want to get too far off track, but you are correct that if you have "local" knowledge you can beat the line at times for those specific games. If a guy was gonna limit his action to only taking shots like that--I would say he could be successful. I am trying to reduce the research and make it as easy and forthright as possible.

For example, anybody with local knowledge abt the K-State basketball team this year knew the two freshmen on their team are NBA quality right now. It has taken Vegas a while to catch up with this information. A guy could have put down large coin on some of those early lines and cleaned up.

So you are correct, in that a guy could use both approaches as long as he doesn't begin to think he is Lefty Rosenthal.

OK, tomorrow!! lol

[20Four7]

Using horse racing as an example, the markets become more efficient as post time approaches. IN the Para mutuel system favorites win the most, 2nd favorites win the 2nd most etc. This implies that the closer the event is to happening the more efficient the market becomes. I have found sports lines follow the same pattern. The "sharpest" line is to be found just before game time.

[HedgeHog]

Would a contrarian model be considered "outside the box" thinking? The purpose of the Lines is to get equal action on both sides, allowing the Books to make their vig with little or no risk. As such, lines are often too heavily influenced by recent results, yielding value to betting on out of favor teams. Going against the herd works for me.

[katstale]

24/7 and HH--both of you are correct. (From this point on I will use "baseline" to refer to the opening line.) Late action will tend to go "herd mentality". I like to think of that in this way. Have you seen those National Geographic shows on TV where they show the Wildebeast migration? OK, once you have that visual--now take that to where they have to get across that river where all those Democrats errrrr crocodiles are waiting for them. What happens?? Well, the first few brave ones make it--the crocs generally get a few of the ones who are trying to cross later.

See the baseline in these terms and you will be getting a framework for what you need to get a few % points in your favor.

Before Pinnacle, a guy had to get to around 54% to overcome the vig. After Pinnacle, 54% winners could make a guy with a large roll a healthy living. What if we could just get to 56%???

I Imagine every level headed person in the world would be happy with that win rate, BUT--not the gambler!!! lolol

So in thinking outside the box we have to decide do we want to be a businessman (except instead of suit and tie you get to dress like Paulie on the Soprano's) or do we want that thrill of gambling? You think that is an easily answered question--but I assure you it is not.

[donjuan]

The purpose of the Lines is to get equal action on both sides, allowing the Books to make their vig with little or no risk.

This is incorrect.

[donjuan]

Before Pinnacle, a guy had to get to around 54% to overcome the vig.

52.38% is breakeven on -110 lines. 54% at -110 is an edge of almost 3.1%.

[HedgeHog]

This is incorrect.

It would help tremendously if you explain why this incorrect, if indeed it is. You disagree that lines are set to draw equal action? Books want to draw one-sided action? Please enlighten us instead of just saying wrong.

[kissthecup]

It seems clear that the only way to assign value to any "model" would be with respect to it's effect on one's bankroll.

Win percentage means nothing without also considering the average odds received or the number of wagers.

As for the model employed by odds makers, it is probably quite sophisticated, and is probably based on information not so easily available to everyone. However, it has a very different purpose than the bettor's model. It's sole concern is maximizing profit for the book. That is accomplished by minimizing profit for the entire betting community in general. This is key - an individual bettor's model considers only his or her performance against the book, and thus is not in direct opposition to the book. In other words, my model does not have to be "better" than the book's for me to be profitable. Surely if Pinnacle's only concern were to minimize MY profit, they could study my betting habits and do so. But instead, wisely for them, they study the betting habits of the entire population of gamblers. So the base issue here is this: Can a model be developed that will take advantage of the fact that lines are set with respect to a large general population? If you're going to bet on sports with an expectation of winning you'd better believe the answer is yes.

[donjuan]

It would help tremendously if you explain why this incorrect, if indeed it is. You disagree that lines are set to draw equal action? Books want to draw one-sided action? Please enlighten us instead of just saying wrong.

Books want to play optimally, too. So they use a combination of maxing value and managing risk. Simply trying to balance action on both sides is both unrealistic and illogical.

[katstale]

I give credit for the thinking and input and rather than take each point and discuss it, let me say this in general.

Go back to my wildebeast visual. You are not trying to beat the "book"--you are trying to beat the other wildebeast safely across that river.

In theory, if a book can balance his books and clear his vig--that is his goal. I have known the "real life" back room guys and I remember them calling the "bigger guys" laying off wagers at the last moment to balance the books.

Remember, your goal is not to beat the book--you just want to beat the guys betting their lunch money.

By the way, thanks for the math. We can all agree with Pinny like books out there a guy can do well between 54-56%.

[donjuan]

By the way, thanks for the math. We can all agree with Pinny like books out there a guy can do well between 54-56%.

Pinny is more efficient than your -110 books of the past so you will have trouble hitting 56% against them ATS.

In theory, if a book can balance his books and clear his vig--that is his goal.

Please see above.

[katstale]

Guys, I am not here to argue some of your long held beliefs--it serves no purpose. I will say this, if you are held hostage by stuff you read/heard/believe you will not be a winner. If you are already lifetime (minimum 3,000 wagers) hitting over 54%--don't waste your time reading this.

Ok, so we need a baseline to work from. There are alot of them out there. In the not so distant past, it was simple--the Stardust basically set the line for almost every one. Fortunately, the guys who set that line are still employed and still putting it out there. Some of you will know how to get that. Others will need to do the research. Hint: they syndicate it out to many of the major newspapers in the country. So you can read it in the morning with your coffee.
Will deal with this issue more later.

Final thought for the day. We discussed that with reduced juice books you could make a good living as a business man with 54%-56% winners. Ponder this for a day.....

If you were to look at the board for the last two days, NBA and NCAAb, and forget abt spreads--all you were asked to do was pick the winning team--what do you guess your winning percentage would be?

What would you guess the average knowledgable jock would do just picking who is gonna win?

later

[donjuan]

Guys, I am not here to argue some of your long held beliefs--it serves no purpose. I will say this, if you are held hostage by stuff you read/heard/believe you will not be a winner.

These are important concepts. If you understand what a perfectly efficient line is, you can't beat it without inside information. Since a book like Pinnacle is more efficient than a -110 book circa 1990, it makes a big difference.

[20Four7]

For the most part Pinny is a lot sharper than the "average" book. But I also believe pinny puts out opinionate lines on certain occassions. You just have to look to see pinny holding a +124 -134 line while every other book has moved the line a point and a half or so going with -110 both sides. They are trying to encourage action one way. They did this for a long time the other day on the raptors 1st half line. They did NOT move the line like every other book but held it giving the following lines. Pinny is a gambling operation first a book maker 2nd in my opinion.

[HedgeHog]

These are important concepts. If you understand what a perfectly efficient line is, you can't beat it without inside information. Since a book like Pinnacle is more efficient than a -110 book circa 1990, it makes a big difference.

I can't buy into this in general. Do you think the line makers are efficient in every line they put out? No way. If they were, lines would never change, after all they're efficient. A false assumption if ever there was one. Perhaps "efficient" needs to be defined.

[teazeman]

Books want to play optimally, too. So they use a combination of maxing value and managing risk. Simply trying to balance action on both sides is both unrealistic and illogical.

books make decisions on every game, their performance is tied greatly to the ability to make the correct ones.

[HedgeHog]

books make decisions on every game, their performance is tied greatly to the ability to make the correct ones.

They'd be just as happy to just make their vig...w/o risk. Books aren't in the business to gamble, as ironic as that may sound.

[katstale]

This is why I am going slowwwwwww. We are getting out in the weeds talking abt things that are "absolutely not important" in relation to our ability to make money.

See above, I am not trying to beat Pinny's line--or anybody elses. If you must look at it in those terms I am actually using Pinny's line in a comparison to the baseline that I have established.

Now, get out of the weeds and answer my last question. What your win % be if you were just asked to pick winners (non ATS)?

[HedgeHog]

This is why I am going slowwwwwww. We are getting out in the weeds talking abt things that are "absolutely not important" in relation to our ability to make money.

See above, I am not trying to beat Pinny's line--or anybody elses. If you must look at it in those terms I am actually using Pinny's line in a comparison to the baseline that I have established.

Now, get out of the weeds and answer my last question. What your win % be if you were just asked to pick winners (non ATS)?

Your just pick a winner question is ambiguous as some ML favs are huge. Who wouldn't pick the fav to boost his %. Picking S/U winners isn't relevant w/o the spread. They go hand in hand.

[Data]

Now, get out of the weeds and answer my last question. What your win % be if you were just asked to pick winners (non ATS)?

69.5% in NBA

[katstale]

Just making a point HH, it will be clear later why this % is important. How do you think you would do? Take the board today--what is there, 40 or 42 games ncaa and NBA? I wonder how many you would get right just picking the winner?

[chemist]

The boys in Vegas have developed a model also. It is the model that sets the line--so you are saying you want to develop a model that is better, more sophisticated somehow, than theirs? Good luck at that. More than likely a fruitless and costly enterprise.

Instead, lets take their model and assign it the status of the "holy grail"! AND adjust our thinking from there. More to follow if any wish!!

Sportsbooks post weak lines every day. Mostly in exotic propositions and obscure sports such as darts and snooker. Even golf and tennis have weaker lines than NFL and the EPL. Sometimes they deliberately post bad lines as a promotion, sometimes the weight of stupid money is too great for them to oppose.

[HedgeHog]

Just making a point HH, it will be clear later why this % is important. How do you think you would do? Take the board today--what is there, 40 or 42 games ncaa and NBA? I wonder how many you would get right just picking the winner?

Today, I'll take every college BB fav and expect to win at least 70% SU. But again, some are 1 point favs and some are over 20, a big variance.

[katstale]

This is for sure true--can't argue with that, just not germane to this discussion. I am going for the "general" outcomes related to the big sports and would certainly never claim that people don't put out some bad lines. Nail them when they do.

[TLD]

What your win % be if you were just asked to pick winners (non ATS)?

Your original question was about NBA and NCAAB, but I don’t handicap those, so the best I can do is answer in terms of NFL. In 2007 NFL, the team I projected to win won straight up 65%-66%. But 2007—though it was a winning season for me—was my worst season in about a decade. So I’d guess it would normally be slightly above that.

So high 60s?

[katstale]

Today, I'll take every college BB fav and expect to win at least 70% SU. But again, some are 1 point favs and some are over 20, a big variance.

Report back how that comes out--and of course it is not enough of sample to be definitive (significant)--but I bet that you and Mr. Data are correct. I was in a bowl game office pool (please don't disrespect me by thinking I work in an office for the man) this year (friend of mine works there and they let me in) and I think that 70% mark was pretty close.

[HedgeHog]

Just making a point HH, it will be clear later why this % is important. How do you think you would do? Take the board today--what is there, 40 or 42 games ncaa and NBA? I wonder how many you would get right just picking the winner?

[chemist]

They'd be just as happy to just make their vig...w/o risk. Books aren't in the business to gamble, as ironic as that may sound.

I don't think you're right. See, for example, Does Sportsbook.com Set Pointspreads to Maximize Profits? Tests of the Levitt Model of Sportsbook Behavior by Paul, Rodney J. and Weinbach, Andrew P., The Journal of Prediction Markets, Volume 1, Number 3, December 2007 , pp. 209-218(10):

The Levitt (2004) model of sportsbook behavior is tested using actual percentages of dollars bet on NFL games from the internet sportsbook, Sportsbook.com. Simple regression results suggest that Sportsbook.com sets pointspreads (prices) to maximize profits, as the Levitt model assumes, not to balance the betting dollars, as the traditional model of sportsbook behavior assumes. Sportsbook.com is found to accept significantly more wagering dollars on road favorites, larger favorites, and on the over for the highest totals in the over/under betting market. Bettor liquidity constraints and sportsbook betting limits may help explain this result.

[katstale]

I don't think you're right. See, for example, Does Sportsbook.com Set Pointspreads to Maximize Profits? Tests of the Levitt Model of Sportsbook Behavior by Paul, Rodney J. and Weinbach, Andrew P., The Journal of Prediction Markets, Volume 1, Number 3, December 2007 , pp. 209-218(10):

I don't know abt all of that--I know they just increase their profit margin by stiffing players!!!

Let's avoid the text book stuff. I promise we won't need it. I guarantee you "Big Al" down at the local bar bookin bets for the week didn't read any text books. Not puttin that stuff down, it just gets in the way. Fogs the vision of a simple path I am leading you down. Hang in there with me.

[donjuan]

I can't buy into this in general. Do you think the line makers are efficient in every line they put out? No way. If they were, lines would never change, after all they're efficient. A false assumption if ever there was one. Perhaps "efficient" needs to be defined.

I never once said that lines were perfectly efficient, just that Pinnacle's lines have to be more efficient than a -110 book in the past because they have high limits and low juice. Of course opening lines are inefficient, but I think we also need to define whether we are trying to beat openers, closing lines or somewhere in the middle of the week. To assume that you can have the same win rate against Pinnacle's lines ATS that you can against a book like SIA is not based in reality. Also, the closer you get to game time, the more efficient the line will become.

This is why I am going slowwwwwww. We are getting out in the weeds talking abt things that are "absolutely not important" in relation to our ability to make money.

So your ability to pick winners SU regardless of the spread/ML is somehow more relevant than discussing the fact that you can't hit the same percentage as you could against some super-square book 15 years ago?