So you have to bet but a very small percentage of your bankroll. Do I calculate 0.1% so that it does not have a significant effect? This does not fit Kelly's criterion at all, which does not take into account such tiny bankroll percentages.
surebet.com and valuebetting
- Enkoff
- Gaining experience
Post
Re: surebet.com and valuebetting
- Arbfan
- Gaining experience
Post
Re: surebet.com and valuebetting
I made 175 bets - Settings are 50 bookmakers to compare and odds from 1.6 - 3.2 - and overvalue + 5 % , 2 hours to match , 10 different soft bookmakers and flat bet - up 20 units . But that is very small sample too. Just posting other side of coin . It would by good to share experience here .
- Attachments
-
- Screenshot_71.jpg (6.53 KiB) Viewed 633 times
- niel
- Gaining experience
Post
Re: surebet.com and valuebetting
To OP and guy posting above me, what software do you use to track your bets? I'm curious.
Personally i have been using the surebets service and i'm very happy about it. For me it works, so there could be a few things happening here like bad variance.
On average how much +EV do you aim for? After simulating lots and lots of bets, for me i try to use these settings:
min 30% probability (otherwise you get too much variance), min 2% +EV (otherwise you get too much variance), scale trades with a watered down kelly criterion, minimum odds for calculation select at least 20-30. This means how much bookmakers the tool compares to, the bigger the number the more accurate it will be.
If you are having trouble with valuebetting could also do some arbitrage instead. You won't have drawdowns since it's guaranteed profit.
Personally i have been using the surebets service and i'm very happy about it. For me it works, so there could be a few things happening here like bad variance.
On average how much +EV do you aim for? After simulating lots and lots of bets, for me i try to use these settings:
min 30% probability (otherwise you get too much variance), min 2% +EV (otherwise you get too much variance), scale trades with a watered down kelly criterion, minimum odds for calculation select at least 20-30. This means how much bookmakers the tool compares to, the bigger the number the more accurate it will be.
If you are having trouble with valuebetting could also do some arbitrage instead. You won't have drawdowns since it's guaranteed profit.
- Arbfan
- Gaining experience
Post
Re: surebet.com and valuebetting
As I wrote in first post Overvalue min 5% , Minimum Odds for calculation 50 , Hours to Event start 2 h , Odds 1.6/3.0 I dont use min/max propability . 200 bets so far flat bet.
- Attachments
-
- Screenshot_77.jpg (8.01 KiB) Viewed 495 times
- A2format
- Gaining experience
Post
The questioner's problem is a bit different: a small sample of data, and is it possible to judge the ROI?
Most people would say, no, you need a large sample of data. And they are wrong.
A small sample of data does not give you a way to know the exact ROI, but it does give you hypotheses about the probability of the ROI that might be. You can see from the author's figure, a strong downward slope: this means that the probability of a positive ROI is likely low, but the probability of a ROI <=0 is increasing. With a set of data, these hypotheses will be more and more refined, and according to the law of large numbers, later will give a precise answer about the ROI. (yes, on 1-2 bets pointless to judge, but on 100-200 can already make the first cautious conclusions, with still a large error).
Also, on the available small sample, you can conduct a Monte Carlo simulation and a little more refine the hypothesis about the value of ROI, on the available small sample.
That said, what I said is well illustrated by Surebet's drawing above your post.
Re: surebet.com and valuebetting
This is not a valid use of the law of large numbers. The law of large numbers allows you to determine, for example, the mathematical expectation as accurately as possible.CharlieSheen99 wrote: ↑Thu Feb 08, 2024 11:06 amhttps://en.wikipedia.org/wiki/Law_of_large_numbers
Anyone that wants to make successful valuebetting should understand well this.
The questioner's problem is a bit different: a small sample of data, and is it possible to judge the ROI?
Most people would say, no, you need a large sample of data. And they are wrong.
A small sample of data does not give you a way to know the exact ROI, but it does give you hypotheses about the probability of the ROI that might be. You can see from the author's figure, a strong downward slope: this means that the probability of a positive ROI is likely low, but the probability of a ROI <=0 is increasing. With a set of data, these hypotheses will be more and more refined, and according to the law of large numbers, later will give a precise answer about the ROI. (yes, on 1-2 bets pointless to judge, but on 100-200 can already make the first cautious conclusions, with still a large error).
Also, on the available small sample, you can conduct a Monte Carlo simulation and a little more refine the hypothesis about the value of ROI, on the available small sample.
That said, what I said is well illustrated by Surebet's drawing above your post.
- A2format
- Gaining experience
Post
Soft bookmakers are more likely to have value bets, but higher probability that they will quickly limit the player or close the account. That's why there is no point with soft bookmakers. The player either loses or they don't let you win.
And at sharp bookmakers, if there is no value bet, there is no point in betting. Although, in fact, sharp bookmakers have bets on value, but much less often.
From the point of view of mathematics, there is no difference: if the probability is 0.55 and the odds are 2, then the ROI will be the same, both at soft and at sharp bookmakers. The only difference is that the soft ones will most likely not let you win.
Re: surebet.com and valuebetting
In my opinion, value betting is independent of the bookmaker. There are odds and actual odds.arbusers wrote: ↑Mon Feb 26, 2024 8:17 pmOf course 150 bets are not enough to determine, but at the same time they are not insignificant to disregard. I see a negative trend unfolding which is not encouraging. But I m starting to see a reason.
Trend.png
In your last attached screenshot I am noticing a majority of sharp bookmakers/exchanges. This can only make your life more difficult. I m wondering why don't you add soft bookmakers in the project? At the same time the graph provided by Surebet.com above should include a number of soft bookmakers.
Soft bookmakers are more likely to have value bets, but higher probability that they will quickly limit the player or close the account. That's why there is no point with soft bookmakers. The player either loses or they don't let you win.
And at sharp bookmakers, if there is no value bet, there is no point in betting. Although, in fact, sharp bookmakers have bets on value, but much less often.
From the point of view of mathematics, there is no difference: if the probability is 0.55 and the odds are 2, then the ROI will be the same, both at soft and at sharp bookmakers. The only difference is that the soft ones will most likely not let you win.
- Arbfan
- Gaining experience
Post
Re: surebet.com and valuebetting
Little update - +400 bets - good variance last days .
- Attachments
-
- Screenshot_102.jpg (9.49 KiB) Viewed 287 times
-
- Screenshot_101.jpg (14.48 KiB) Viewed 287 times
-
