Kelly Re-Visited

[danshan11]

all staking methods should lead back to risking the least amount of money on plays you want to play to win the same amount
you want to bet less to win the same amount, that should be everyones goal.

[Derailer]

you cannot have a good bet if the market goes against it
if you bet the Bears -7 and they go to -5 your bet was bad on that particular game win or lose
think of it like this
you risk 100 to win 100 at -7 and now you risk 60 to win 100 at -7 with the new line change and you think you still have a good bet when you put 40 more dollars at risk for the same bet. This is a bad bet no matter how you sum it up, no matter what your model shows.

Not if the Bears win by 17 with no turnover luck.

[HeeeHAWWWW]

and how to determine positive expected value? the only way I know to determine positive expected value is to bet a better line than the closing line

Since you can't know the closing line beforehand, you're saying any form of successful betting is impossible.

[danshan11]

Since you can't know the closing line beforehand, you're saying any form of successful betting is impossible.

NO, I am saying the ONLY forum of successful betting is getting a better line than the closing line.

[Derailer]

NO, I am saying the ONLY forum of successful betting is getting a better line than the closing line.

That is true on average. You project that to specifics.

[danshan11]

example Normal Lineups no injuries or anything weird
Barria LAA
Verlander Astros
Astros at home
the current line is Astros -525
two questions
who will win?
who do you bet?

[Derailer]

example Normal Lineups no injuries or anything weird
Barria LAA
Verlander Astros
Astros at home
the current line is Astros -525
two questions
who will win?
who do you bet?

Again, you have to argue your point based on a professional level of gambling. There is no point going on and on about not being able to beat a closing line because nobody bets closing lines.

[danshan11]

I bet the Angels because I have the Astros at -270ish on that game so the value is on the Angels, who do I think will win, probably the Astros but that does not play into my decision in anyway!

I mean look at my spreadsheet and look at my losses and CLV, do you think I am good capper with this spreadsheet, would you tail me after looking at the sheet and 99 out of 100 would say HELL NO and I would say hell yes!

[danshan11]

Derailer dude you are saying one thing like you agree but than saying things like if the Bears win, that is not how it works it dont freaking matter if the Bears win ever! I am stating my opinion these are not facts dude, I wish and hope people come on here and tell me I am completely wrong and show me the right way to think if so.

[Derailer]

All sports are subject to small sample sizes so there are no numbers that are exceptionally accurate. They are accurate enough. But there is an opportunity to beat the market on information it is not correctly recognizing and this is what separates the winning gamblers from everyone else. They have better data or better methods for evaluating a game. The market move is not always right.

[danshan11]

I agree that individual games can be not effecient but overall the market is effecient very effecient and if you bet against it you will lose by exactly the margin like everyone else does that does not have an edge

[Derailer]

OK, that is what everybody believes so you can just stop arguing. But it is not the case 100% of the time and winning gamblers will beat line moves. That is why they win.

[Bsims]

I agree that individual games can be not effecient but overall the market is effecient very effecient and if you bet against it you will lose by exactly the margin like everyone else does that does not have an edge

Bingo. I bet on individual games not the overall market.

[danshan11]

I hope you dont believe that Bsims

[Derailer]

I hope you dont believe that Bsims

You just said individual games were not particularly efficient. Why do you continue to believe that people that have models do not believe in CLV? It does not occur to anyone that you would think they would not believe in the value of CLV so they think you are arguing specific points when you continue to try to tell someone CLV has value.

[HeeeHAWWWW]

NO, I am saying the ONLY forum of successful betting is getting a better line than the closing line.

.... which you cannot know in advance, unless you're psychic.

Here's a little example of how a proper modelling effort works: My last 1108 bets I've been recording my estimated EG. The sumproduct of those is 2.17. Meanwhile, my bankroll has gone up by 2.15x. Not a coincidence.

[danshan11]

.... which you cannot know in advance, unless you're psychic.

Here's a little example of how a proper modelling effort works: My last 1108 bets I've been recording my estimated EG. The sumproduct of those is 2.17. Meanwhile, my bankroll has gone up by 2.15x. Not a coincidence.

if you dont have a good idea (not specific) where the lines are going to go, you are in the wrong business. Line moves need to be in your direction 7 or 8 out of 10 times or you are in bad shape. What is EG, sorry for my ignorance? 1100 bets is a good chunk and should weigh heavy but not certainly. I think you should see how your line value looks on those same bets I think if EG means you made money I bet unless you were just lucky that you had good line value. I mean simply do you think the books are good cappers too, and if they are why do they take bets on both sides? why would you "fred" not take action on a game if Pinnacle will? They take action on everything because they have margin they dont know or care who is going to win, they know margin wins, they take bets at 1-2% off the implied and do very well without knowing in advance who will win. You ever seen anyone more successful at sports betting than a book.

[HeeeHAWWWW]

if you dont have a good idea (not specific) where the lines are going to go, you are in the wrong business.

So you should be able to predict a closing line ..... but you've also said the need to predict a probability to feed into Kelly makes using it impossible. But you've also said they're the same thing! A closing line is an implied probability.


EG = expected growth, it's what Kelly aims to maximise. It's a little like EV, but for your bankroll.


ps those 1100 are just those I have with EG recorded. Full spreadsheet has near 8k bets.

[danshan11]

impossible to accurately predict a closing line this is why Kelly is not good for sports betting because before the game actually closes we do not know the closing line.
you cant predict the closing line accurately but what you can do is identify where 8 out of 10 times the line will move in your direction. it kinda all is big talk but in reality you are just buying cheap.
in my opinion all bets are good on the board, it is in theory suppose to be a toss up on the board right?
so I dont just want to flip coins at -110 so how can I play and still win
well only 1 way
find coin flips where the line will move enough my direction to cover the margin and that is it.
Any other method is flawed because the effeciency of the market has proven that capping a game is I say impossible (but I am sure there are exceptions) so if we assume it is we can only beat the number the game is random.

think of a coin flip game if they pay -110 you wll lose
sports are in theory the same a coin flip and if you pay -110 you will lose
BRady is good on 3rd down its in the line
Lebron plays strong at home in the line
Raiders are great in red zone defense in the line
Johnson wont play its in the line
29999 out of the last 30000 games they won at home its in the line
there is nothing you can cap that is not in the line

you can actually see it, watch a baseball line move if the wind changes direction or watch a nfl game line change cause they change a star running back. 99.99% is in the closing line so capping the game is useless but capping the number is a great thing. People dont choose bad teams they choose bad numbers and that is where the hole is and the market potential. The Red sox will probably beat the tigers 99% of the world can tell you that but what they cant tell you accurately is how many times out of 1000 and that is the magic bullet.

you keep capping games and you will pay margin over time UNLESS you are capping something that matters and the market does not know about it yet.
you learn to cap numbers and you will have a slight chance as long as the Nerfs keep thinking that the line does not count some obvious thing that actually matters

[danshan11]

So you should be able to predict a closing line ..... but you've also said the need to predict a probability to feed into Kelly makes using it impossible. But you've also said they're the same thing! A closing line is an implied probability.


EG = expected growth, it's what Kelly aims to maximise. It's a little like EV, but for your bankroll.


ps those 1100 are just those I have with EG recorded. Full spreadsheet has near 8k bets.

what is your CLV on those 8000 bets and that will say a ton. remember the game changes too so even if you were good with some method 7000 bets ago it might be dead now or evolving and your past may not be indicative of your future

[danshan11]

you have to keep a living breathing line value status, You need to make sure if you have any edge it is not slipping away and the line will tell you that on every single bet. you might have a bad line here or there but overall your clv will show you where you are. I have wins losses and good bad and ugly but what is fairly consistent on all my sports and picks is the line value. You also can see I am better at basketball and the WL and CLV show it. its obvious to me that clv is for sure the Horsepower rating of what you are doing on the board

[Derailer]

Any other method is flawed because the effeciency of the market has proven that capping a game is I say impossible (but I am sure there are exceptions) so if we assume it is we can only beat the number the game is random.

What is the process by which you think a number is beaten?

[danshan11]

bet Yankees -140 and it closes at lower than -145 that is beating the line because that is enough to cover the margin and in this scenario you are now flipping coins at +EV

[Derailer]

bet Yankees -140 and it closes at lower than -145 that is beating the line because that is enough to cover the margin and in this scenario you are now flipping coins at +EV

How does the line move from -140 to -145 and what causes it to reverse and close -130?

[HeeeHAWWWW]

impossible to accurately predict a closing line this is why Kelly is not good for sports betting because before the game actually closes we do not know the closing line.

Nah - I've posted this before, but Kelly is robust to errors.

[danshan11]

How does the line move from -140 to -145 and what causes it to reverse and close -130?

well in most cases some sort of news or update but that is why you need to see 7 or 8 out of 10 move your way, if not you are doing something wrong. the real way to know if your capping is any good is to create your own implieds and compare them to the actual closing lines, the closer you are to actual closing the better you are doing, dont focus on anything else, this is the key IF there actually is a key.

[danshan11]

Nah - I've posted this before, but Kelly is robust to errors.

kelly is not the error, the error is not accurately knowing your edge. Kelly will always be flawed if you dont know your edge and you dont BEFORE it closes or at the time you bet.

[Derailer]

well in most cases some sort of news or update but that is why you need to see 7 or 8 out of 10 move your way, if not you are doing something wrong. the real way to know if your capping is any good is to create your own implieds and compare them to the actual closing lines, the closer you are to actual closing the better you are doing, dont focus on anything else, this is the key IF there actually is a key.

Don't you basically believe that you have better numbers than the market and that the market is going to move in your direction? Presumably, you believe that the market is dominated by sharp programmers like yourself and this would mean that you all have basically the same numbers. So aren't you going to need for someone else to bet the games in your direction?

And aren't you also going to need to have someone else create betting opportunities by causing the market to be wrong for at least as long as it takes the programmers to correct it?

And if this is the case, some of this activity is going to be sharp otherwise the numbers would always go in your favor.

[Derailer]

The reality is if you look at the results of 5 different winning bettors in any sport, you will find that they bet entirely different games and for different reasons. There would be little similarity in their bets. This is because there is lots of opportunity across the board but it is a challenge to recognize that opportunity. Some recognize things that others do not for different reasons because the things you think have no value actually have a lot of value.

If the market was truly efficient, winning gamblers would basically all have the same bets.

[danshan11]

what makes you think the market is not effecient, what info brings you to that conclusion, I can show you 1000s upon 1000s of games and ATS they are nearly 50% does that not prove the market is overall effecient

[danshan11]

the market number we see I think is a few things
openers come out as to draw attention.
Rec bettors move the lines on local games or high profile games
Nerfs move the lines by betting useless stats and noise
Sharps come in and clean up where the above made mistakes and books kinda allow it to help reduce overall game risk

[Derailer]

what makes you think the market is not effecient, what info brings you to that conclusion, I can show you 1000s upon 1000s of games and ATS they are nearly 50% does that not prove the market is overall effecient

We all know the market is efficient on average. You can beat the market because it is not sufficiently efficient on all games or even close to efficient which is why 5 different winning gamblers can bet entirely different games and come out with winning returns. If the market was so inefficient that the lines were beaten on average, there would be no gambling market.

[danshan11]

who wins at sports betting? I say its the books and how do they win? by beating the line

[Derailer]

who wins at sports betting? I say its the books and how do they win? by beating the line

No, they win with the vig.

[danshan11]

No, they win with the vig.

exactly the VIG and the vig is exactly that beating the line