Statistical significance of betting system

[bettingman6]

Honestly I don't think people should just mindlessly bet based on what their systems say, and at the end of the day I think almost any successful bettor is going to have to bet somewhat based on their arbitrary whims. You just have to hope that your whims are better than most people's whims. Something tells me that Billy Walters didn't bet exclusively based on what his computer models said.

Therefore I'd suggest that people spend at least a year betting $20-100 on games before you bet extremely serious money. After 1-2 years, you should have an idea if you can make profits if you bet a lot of money. Backdating your computer system to past years might tell you how well you would have done if you had mindlessly followed your computer system, but I don't recommend doing that.

[danshan11]

no system works unless you have an edge. you need to see an edge everything else is a waste of time and definitely money

[gojetsgomoxies]

I still believe in the stink test hypothesis. Lost 5 in a row, got shutout last game, 2-8 last 10, 0-5 last 5 ats. Stuff like that where no one would bet on them

dan, in a previous thread, when i mentioned liking streaks i meant stuff like that. "deep value" it would be called in stock market. although sports betting and stock market are very different in many ways.

[danshan11]

the thing is I try and find some bad smell bad criteria that would cause people to bet the other side and throw the line off.

Say power rankings say Yankees -150 this game, but they lost 10 straight, did not score last game, are at home ( because touts talked all these idiots into thinking road is a sharp play), and the starter lost his last 3 games. I would need to verify this correctly, but I mean in general you would get the idea that the crowd would frown on betting on the Yankees in this situation and would bet on the other side so it might be possible to get some line bias here and hence value in the line. I am speaking hypothetically here, this is not an exact scenario just trying to share my idea of my thinking on this!

in soccer, they did a premier league study that if a team was a heavy favorite, that the draw had value because no one thought the draw had a chance and the pricing was not accurate. I am sure this edge has probably dried but at a time it made sense.

People have tendencies of who they think will win
low price dogs on the road is the new popular bet, so if you can get that scenario and add a few stink factors to the other team, its a good chance this line could hide some value and the line could underperform

[peacebyinches]

So, as I mentioned, this got me thinking again, and I'd be interested in knowing if there is anyone that has looked at the emphasis bettors place on certain types of information, given a certain game line. Lets say team A is a big market team, at home, is on a winning streak, but has not done well against the team they are playing. The big market, home field, and overall recent performance all factor as a positive in terms of how much this team would be a favorite. The recent performance specifically against the team they are playing would hypothetically bring the line closer. If each of these factors were the only 4 factors used and were all weighted evenly, there could easily be some 'bias metric' that is 75% weighted towards team A. Would the majority of bettors be biased to put their money on team A, therefore biasing the line to be inaccurately large on for this team, or would this be completely nullified by emphasis on the 4th criteria, therefore washing any value out of the line, again, only in this specific situation. In an ideal world we could literally measure the weightings that bettors in certain markets are giving each of these factors and potentially pinpoint situations where a line would be biased AND what the driving factor for the bias is.

[Waterstpub87]

Honestly I don't think people should just mindlessly bet based on what their systems say, and at the end of the day I think almost any successful bettor is going to have to bet somewhat based on their arbitrary whims. You just have to hope that your whims are better than most people's whims. Something tells me that Billy Walters didn't bet exclusively based on what his computer models said.

Therefore I'd suggest that people spend at least a year betting $20-100 on games before you bet extremely serious money. After 1-2 years, you should have an idea if you can make profits if you bet a lot of money. Backdating your computer system to past years might tell you how well you would have done if you had mindlessly followed your computer system, but I don't recommend doing that.

You're whims are wrong. It is what every behavioral finance paper says. Literally. Read fooled by randomness by Taleb.

The only way to win is to build models. And test, and improve and test and improve. Hoping your whims are good is a recipe for disaster. Now, if your model has an nba spread off by 9 points, you are probably wrong. Anyone who is serious about this is going to wave that play off and move on. But if you don't follow your model at all, then don't have one.

[Waterstpub87]

the thing is I try and find some bad smell bad criteria that would cause people to bet the other side and throw the line off.

Say power rankings say Yankees -150 this game, but they lost 10 straight, did not score last game, are at home ( because touts talked all these idiots into thinking road is a sharp play), and the starter lost his last 3 games. I would need to verify this correctly, but I mean in general you would get the idea that the crowd would frown on betting on the Yankees in this situation and would bet on the other side so it might be possible to get some line bias here and hence value in the line. I am speaking hypothetically here, this is not an exact scenario just trying to share my idea of my thinking on this!

in soccer, they did a premier league study that if a team was a heavy favorite, that the draw had value because no one thought the draw had a chance and the pricing was not accurate. I am sure this edge has probably dried but at a time it made sense.

People have tendencies of who they think will win
low price dogs on the road is the new popular bet, so if you can get that scenario and add a few stink factors to the other team, its a good chance this line could hide some value and the line could underperform

It depends on your religion. The open should be relatively where they think the close will be. If you believe that beating the close = winning, which is what I've gathered by your posts before, you have to take a position on how the line is set. If square money moves the line, the Yankees are probably not the theoretical right side, as the square money will come in on the other side. If sharp money moves the line only, ignoring the public money, and the yankees fundamentally are correct then this is a good play.

Yankees are a bad example though, because they tend to be a public team anyway.

But it is also a chicken vs egg problem. If they yankees should be -200 because of fundamentals, and then open at -150 because of random, unimportant factors like the starter losing the last 3 games or not scoring the game before, then it is a sharp play. But if the fair price is yankees -150, then the other factors are unimportant. Without knowing a proper price on the yankees, you are unable to determine whether they are undervalued or not.

[u21c3f6]

It depends on your religion. The open should be relatively where they think the close will be. If you believe that beating the close = winning, which is what I've gathered by your posts before, you have to take a position on how the line is set. If square money moves the line, the Yankees are probably not the theoretical right side, as the square money will come in on the other side. If sharp money moves the line only, ignoring the public money, and the yankees fundamentally are correct then this is a good play.
Yankees are a bad example though, because they tend to be a public team anyway.But it is also a chicken vs egg problem. If they yankees should be -200 because of fundamentals, and then open at -150 because of random, unimportant factors like the starter losing the last 3 games or not scoring the game before, then it is a sharp play. But if the fair price is yankees -150, then the other factors are unimportant. Without knowing a proper price on the yankees, you are unable to determine whether they are undervalued or not.

W, I like your explanation but with one exception. It is not necessary to know the proper price. I have never created a line.

Think of it this way, at one time you could play home dogs and make a profit. No knowledge of proper line needed and no need to beat the close as home dogs were being underestimated from the gitgo. Things are certainly not that simple any more but the concept is the same. Find criteria for events where the line tends to be under or over estimated and then your job is to figure out at what time during the wagering would be the best time to wager to get the best price (and that does not mean you have to beat the close as oftentimes the best price IS the close). Joe.

[danshan11]

I don't understand how you can bet a game and not know what the line will be (i.e. model)

I model Yankees -150
if they are -130 I bet
if they are -180 I don't bet

I don't understand how you would know to bet if you don't have a line

when I say line I mean probability
if I model the Yankees have a 60% chance to win
and the current line probability is below that including margin, that would be a bet
how can you know when to bet what if you don't have a probability
if Yankees were -260 would you bet them still
if they were -120 would you bet them still
I don't get how you can say I am taking the Yankees if you don't know the probability. I assume it is common knowledge they won't win all their games, so you can't just assume they will win, right? Am I missing something or is my thought process wrong on this?

[danshan11]

It depends on your religion. The open should be relatively where they think the close will be. If you believe that beating the close = winning, which is what I've gathered by your posts before, you have to take a position on how the line is set. If square money moves the line, the Yankees are probably not the theoretical right side, as the square money will come in on the other side. If sharp money moves the line only, ignoring the public money, and the yankees fundamentally are correct then this is a good play.

Yankees are a bad example though, because they tend to be a public team anyway.

But it is also a chicken vs egg problem. If they yankees should be -200 because of fundamentals, and then open at -150 because of random, unimportant factors like the starter losing the last 3 games or not scoring the game before, then it is a sharp play. But if the fair price is yankees -150, then the other factors are unimportant. Without knowing a proper price on the yankees, you are unable to determine whether they are undervalued or not.

my point was public has an influence on high publicity and tv games

say the Lakers are playing on the road and as dogs
lets assume public money is dumb money for this example
Model power rankings whatever say examples ONLY
Lakers +3
now the largest team following in the NBA (public money) take off .25
major market team take off -.25
Report on Espn this AM Lebron is the KING has had a triple double last 10 take off a .25
Lakers just won 5 straight take off a .25
Oppenent lost last game by 30 take off a .25

so you got to move the line down to 1.5
so now the public has brought it to 1.5
a pro needs the line to be off by at least say 1 pt to bet it
so now the pros jump on it and move it back to 2 but they stop at 2 because the margin tells them 2.5 to 3 is fair so 1.5 is a good bet and 2 is where you stop

this game closes at 2 and the probability is actually 3 over a zillion games

this is what I would call a stink effect
where the line has been moved by the public to 1.5 and pushed back by pros to 2 and the fair actual line is probably 3

if dumb money pushes a line down, the pros will push back but they wont push all the way back because they have to think of the margin so they leave space

this is all random in the air numbers dont pick at them I just used some numbers as an example

this also is the hole in the CLV theory, you can find these or other type games where the line might not truly be efficient. this is the same reason the Super Bowl is rarely efficient, dumb money push without care and pros only push back so far because of margin. I can argue that this is where you can see a gap in efficiency with a good reason
with all that said, I still think you have to beat the line overall!

[danshan11]

also a side note
Sharp money does not mean winning it means the line was wrong, they still have no clue who will actually win that game

Yankees are -150 pros are like it should be -200 so they bet it and bet it and bet it and it closes at -200 and you come in and say all the pro money pushed this so the pros think the yankees will win, sure they do but they think it will at -150 -160 -170 not -200, -200 is too late.

sharp money does not pick winners, sharp money shows inefficiency in the line
good bet at -150, lose the juice bet at -200
sharp money does not pick winners, sharp money picks lines that are not correct and has nothing to do with winning or losing on that game

I think a ton of people get this confused

[danshan11]

W, I like your explanation but with one exception. It is not necessary to know the proper price. I have never created a line.

Think of it this way, at one time you could play home dogs and make a profit. No knowledge of proper line needed and no need to beat the close as home dogs were being underestimated from the gitgo. Things are certainly not that simple any more but the concept is the same. Find criteria for events where the line tends to be under or over estimated and then your job is to figure out at what time during the wagering would be the best time to wager to get the best price (and that does not mean you have to beat the close as oftentimes the best price IS the close). Joe.

dont you believe if there is any over or under estimation that a sharp player will come in and clean that mess up quickly?

[u21c3f6]

dont you believe if there is any over or under estimation that a sharp player will come in and clean that mess up quickly?

No, otherwise I couldn't show a profit. Based on your previous posts, you believe that the line always gets back to perfect for all events and that is simply not true and the reason I can still be profitable even if I don't beat the closer. Joe.

[danshan11]

No, otherwise I couldn't show a profit. Based on your previous posts, you believe that the line always gets back to perfect for all events and that is simply not true and the reason I can still be profitable even if I don't beat the closer. Joe.

I posted two posts up that I don't think the line is efficient in every single game.
I believe in anchor bias and line bias
anchor bias
a line that was set wrong to begin with and it is anchored to that opening price and does not come down to perfect
line bias
dumb money having more weight than pro money because of a few factors but mainly discussed above

I do not believe the line is perfect in all instances. I think overall it is efficient.
in most cases, people that can't beat the line say the line is wrong of course there are exceptions

One of my highest regards for line value has to do with the fuel gauge. Say you don't care about the line then how do you know when to stop betting or keep betting after losing sessions.
I bet 100 games win like crazy
I bet another 100 games lose like crazy
when do I know to adjust or stop without using CLV as the indicator of skill?

[danshan11]

No, otherwise I couldn't show a profit. Based on your previous posts, you believe that the line always gets back to perfect for all events and that is simply not true and the reason I can still be profitable even if I don't beat the closer. Joe.

you often refer back to profit as a measuring stick. do you think someone could be profitable and still not have a clue in a large sample size?

[u21c3f6]

I posted two posts up that I don't think the line is efficient in every single game.
I believe in anchor bias and line bias
anchor bias
a line that was set wrong to begin with and it is anchored to that opening price and does not come down to perfect
line bias
dumb money having more weight than pro money because of a few factors but mainly discussed above

I do not believe the line is perfect in all instances. I think overall it is efficient.
in most cases, people that can't beat the line say the line is wrong of course there are exceptions

One of my highest regards for line value has to do with the fuel gauge. Say you don't care about the line then how do you know when to stop betting or keep betting after losing sessions.
I bet 100 games win like crazy
I bet another 100 games lose like crazy
when do I know to adjust or stop without using CLV as the indicator of skill?

First, the posts you refer to you is certainly a step back or two from your pre-previous posts where you were adamant that the line was always perfect and therefore if you didn't beat the close the only way you could be profitable was by luck or survivorship bias.

Which brings me to my second point, CLV is not necessarily an indicator of skill. What good is a skill of finding wagers that on average beat the close but do not produce a profit? Your answer has always been: variance. My answer would be that you may be identifying those wagers that are actually inflated and do not perform at the rate expected by the close of those wagers.

Also of interest is that you linked a paper that shows a method for trying to find profitable wagers without creating a line. Joe.

[danshan11]

first my opinion always changes, nothing is in stone especially at my low IQ and my low betting IQ. I learn new things everyday.

if CLV does not measure skill what does? I got a friend has over 1500 records and is a winner has a 5% ROI and he could not pick a winner in a million years and he thinks a model is the chicks on TV, is he skilled? I will tell you how he picks games, he looks at last time the teams played each other and how both teams done in their last 3 or 4 games and trusts his gut, LOL


that paper I linked was an example of a study of the method these guys were discussing in this thread, I did not write or endorse that paper in any way or have I even evaluated it to determine whether I believe in it or not. I know it was written by someone with higher sports and real IQ than me.

[danshan11]

which is more likely long term huge sample size winner?

1. beat the line consistently
2. pick his spots negative line value
3. pick his spots and has even line value

[danshan11]

you often refer back to profit as a measuring stick. do you think someone could be profitable and still not have a clue in a large sample size?

can you answer this so I can better understand how you think?

[u21c3f6]

which is more likely long term huge sample size winner?

1. beat the line consistently
2. pick his spots negative line value
3. pick his spots and has even line value

This is where I believe you get tripped up. Yes, if we randomly select wagers and set them to beat, equal or have negative line value, then the beat wagers will outperform the equal wagers which will in turn beat the negative value wagers.

The problem is that you don't get to randomly select wagers and automatically get to beat the close. Once you select wagers using some criteria, you now have a non-random selection of wagers that may or may not perform like the aggregate even if your selection of wagers in aggregate beat the close. In other words, someone who on average beats the close with their non-random wagers may not be long-term profitable while someone who does not beat the closer on their non-random wagers may be long-term profitable.

Joe.

[danshan11]

I consider myself a loser if I dont beat the closing line by over 60% above margin and 75% overall.

I do not care wins, losses, or draws

what do you use as a measuring stick

bet Yankees at -150
A they close at -180 and lose
B they close at -180 and win
C they close at -120 and win
D they close at -120 and lose
for me A and B are winners LONG TERM
C and D are losers LONG TERM

if you go by win loss record my friend is the best handicapper in the world and his dumbass spends 12 hours a day at a tech company like an idiot for 80k a year, when he could make that a month with his savings as a bankroll.

[danshan11]

This is where I believe you get tripped up. Yes, if we randomly select wagers and set them to beat, equal or have negative line value, then the beat wagers will outperform the equal wagers which will in turn beat the negative value wagers.

The problem is that you don't get to randomly select wagers and automatically get to beat the close. Once you select wagers using some criteria, you now have a non-random selection of wagers that may or may not perform like the aggregate even if your selection of wagers in aggregate beat the close. In other words, someone who on average beats the close with their non-random wagers may not be long-term profitable while someone who does not beat the closer on their non-random wagers may be long-term profitable.

Joe.

also can you answer how you can choose a team if you dont have a number on the game

if Yankees are -150 now
would you bet at -490 or +112 ???? I mean who cares the line you know you can pick winners?

[danshan11]

I think our back and forth on the CLV is gold is a waste of our time. My message was that you can conceivably find value because of some public bias and the fact that the sharps will only push back so far, which could possibly leave a hole that might have value in it.

[u21c3f6]

also can you answer how you can choose a team if you dont have a number on the game

if Yankees are -150 now
would you bet at -490 or +112 ???? I mean who cares the line you know you can pick winners?

First, let's clear up a point. You are the one that constantly brings up the pick winners argument. You will not find a post where I claim to pick winners, in fact I have posted many times that I specifically do not try to pick the winner. I also do not have best bets because I am not trying to pick the winner. One selection is the same as another selection to me.

Read the article that you posted for an example of how one can make a selection without creating a line. If that answers your question, great, if not, ask what didn't make sense to you.

Joe.

[danshan11]

you are a system bettor? how do you make a bet selection

[danshan11]

First, let's clear up a point. You are the one that constantly brings up the pick winners argument. You will not find a post where I claim to pick winners, in fact I have posted many times that I specifically do not try to pick the winner. I also do not have best bets because I am not trying to pick the winner. One selection is the same as another selection to me.

Read the article that you posted for an example of how one can make a selection without creating a line. If that answers your question, great, if not, ask what didn't make sense to you.

Joe.

I wasnt asking you how the article can choose without a line, I was asking how you do it?

I dont want your secret sauce, I am asking in general how can someone select a bet without having a number on it?

a hunch
nice uniforms
gut feeling
like dark colors

[u21c3f6]

I wasnt asking you how the article can choose without a line, I was asking how you do it?
I dont want your secret sauce, I am asking in general how can someone select a bet without having a number on it?
a hunch
nice uniforms
gut feeling
like dark colors

I am a just the facts kind of guy, no hunches considered in making selections.
There are uniforms I like and there are uniforms that should be burned IMO like the old-time Pitt Steelers bumble bee/ prison uniforms and not considered in making selections.
I don't have gut feelings because I don't know enough about the teams to have gut feelings! What?!
I prefer lighter colors but again has nothing to do with making selections.

As far as the secret sauce, from my point of view you really are kind of asking for more than I am willing to share because if you just wanted to know how it could be done, the article you posted is a perfect example. However, I will try to expand on my answer in another post.

Joe.

[danshan11]

sure yeah I dont have any secrets, I try to beat the line by modeling and nothing else. I just think its real simple question
how do you know when to play the Yankees at -150 -200 -400 +200, how do you know where the bet spot is without a number (fair line you created)

I am not asking for the things you use but the how do you know without a fair number to start with

[danshan11]

like for example cbb today
at what number would you pick Texas Tech -2 -5 +10 +3 what number?

[Gaze73]

I wasnt asking you how the article can choose without a line, I was asking how you do it?

I dont want your secret sauce, I am asking in general how can someone select a bet without having a number on it?

a hunch
nice uniforms
gut feeling
like dark colors

System bets don't need to know the true odds, they just need to exploit a bias. If e.g. my system X is profitable in the +100 to +299 range, I don't care at all what odds the dog in that range has, because I know it's profitable due to certain factors that cause the bias.

[u21c3f6]

sure yeah I dont have any secrets, I try to beat the line by modeling and nothing else. I just think its real simple question
how do you know when to play the Yankees at -150 -200 -400 +200, how do you know where the bet spot is without a number (fair line you created)

I am not asking for the things you use but the how do you know without a fair number to start with

The question is simple but to give an answer that would be understood without spelling it out is far more complicated. I have written about my focus in the past as well as mentioned it in a post above but your Yankees and Texas Tech examples are stated from your focus which is not how I do things.

Joe.

[danshan11]

The question is simple but to give an answer that would be understood without spelling it out is far more complicated. I have written about my focus in the past as well as mentioned it in a post above but your Yankees and Texas Tech examples are stated from your focus which is not how I do things.

Joe.

LOL

[danshan11]

System bets don't need to know the true odds, they just need to exploit a bias. If e.g. my system X is profitable in the +100 to +299 range, I don't care at all what odds the dog in that range has, because I know it's profitable due to certain factors that cause the bias.

so basically you think the probability is 100%, line is irrelevant, nice that sounds very cool and profitable!

[danshan11]

The question is simple but to give an answer that would be understood without spelling it out is far more complicated. I have written about my focus in the past as well as mentioned it in a post above but your Yankees and Texas Tech examples are stated from your focus which is not how I do things.

Joe.

so Joe a "pro" capper or a long term winner do they need to create a fair line to be able to win long term?

[u21c3f6]

so Joe a "pro" capper or a long term winner do they need to create a fair line to be able to win long term?

No.