Monte Carlo simulation to predict expected profit

Pages: [1]
Author Topic: Monte Carlo simulation to predict expected profit  (Read 1509 times)
AlexNotman1
Becoming a Pro
***

Karma: 6
Posts: 267



View Profile

AlexNotman1

Monte Carlo simulation to predict expected profit
« on: March 04, 2019, 10:10:47 AM »

I noticed quite few questions about being breakeven valuebetting only after decent number of bets so I started looking for a way to predict expected profit after particular number of bets. That must be possible and I don't even doubt it. I started digging about Monte Carlo simulatios as I read it may be useful for that kind of calculations. Anyone with idea how to make such simulations? I mean something which looks like this one (17.53 for example):

https://www.youtube.com/watch?v=tsdFrOLnTXw

I'm crap about Excel, even basic stuff not even talking about more complicated formulas, would take me ages to find out how to build such simulation.

The idea is, as abovementioned, to build formula where we could input some data like: our average +EV, number of bets placed (let's assume we talk about flat stake of 1 unit each bet) to calculate what are chances to achieve particular profit after - let's say - 1000 bets.

I'm sure that tool would be extreme useful for community as these questions about expected profit comes up frequently lately but I just dont know how to do it. Anyone smarter if we talk about Excel/programming?
Logged
dealer wins
Totally Pro
*****

Karma: 12
Posts: 908



View Profile

dealer wins

Re: Monte Carlo simulation to predict expected profit
« Reply #1 on: March 04, 2019, 10:41:53 AM »

Variance takes longer than a 1000 bets to even out, and you could experience wild deviations either positive or negative.  After 50000 bets it would be more closer to expected, and after a million bets it should be bang on the expected value.
Logged

Never trust a goose!!!
AlexNotman1
Becoming a Pro
***

Karma: 6
Posts: 267



View Profile

AlexNotman1

Re: Monte Carlo simulation to predict expected profit
« Reply #2 on: March 04, 2019, 11:03:32 AM »

1000 bets was just an example. I know how variance works. All I need are numbers what can I expect. I asked similar questions before and this is what I got: "1000 bets is too small sample etc etc". Allright, everyone with some experience know that and this is not the issue. I'm asking about exact percentages having fixed, average +EV after particular number of bets. I know that if there is - let's say - 95% to earn 5 units after 1000 bets with 5% +EV (or even 99%) it's not sure thing but cmon guys, if you can help, post some advices not "1000 bet's its not enough". It looks like no one can answer this question but there are always a lot of of people commenting. I searched internet all over and all I got was Monte Carlo, weird that no one can make such tool or formula or I can't search properly.
« Last Edit: March 04, 2019, 11:05:42 AM by AlexNotman1 » Logged
pythonic
Gaining experience
*

Karma: 6
Posts: 61


View Profile

pythonic

Re: Monte Carlo simulation to predict expected profit
« Reply #3 on: March 05, 2019, 08:17:31 AM »

Monte Carlo simulations are actually pretty straightforward to implement but depend on what exactly you want to simulate.
You write a loop with numbers from a random number generator in the programming language of your choice and count the results.
You just need to understand the concept well and know how to apply it on your problem.
Logged
Wolfie
Pro
****

Karma: 26
Posts: 399



View Profile

Wolfie

Re: Monte Carlo simulation to predict expected profit
« Reply #4 on: March 05, 2019, 12:48:41 PM »

Graphic in photo below is for 4% EV , 100000 simulations.
Logged
AlexNotman1
Becoming a Pro
***

Karma: 6
Posts: 267



View Profile

AlexNotman1

Re: Monte Carlo simulation to predict expected profit
« Reply #5 on: March 05, 2019, 12:57:40 PM »

I guess this is pretty close to what I was looking for, thanks Wolfie. But what Frequency means? Shouldn't be achieved profit on the left side? Is if variance in percentages? Btw. what about odds? From what I know higher odds = higher variance. So shouldn't be odds implemented there?
« Last Edit: March 05, 2019, 01:05:47 PM by AlexNotman1 » Logged
Wolfie
Pro
****

Karma: 26
Posts: 399



View Profile

Wolfie

Re: Monte Carlo simulation to predict expected profit
« Reply #6 on: March 05, 2019, 01:05:48 PM »

I guess this is pretty close to what I was looking for, thanks Wolfie. But what Frequency means? Shouldn't be achieved profit on the left side?

For example in 15210 bets graphic:
Frequency 3% means 3% of simulations resulted in exactly 4% yield.
Logged
Wolfie
Pro
****

Karma: 26
Posts: 399



View Profile

Wolfie

Re: Monte Carlo simulation to predict expected profit
« Reply #7 on: March 05, 2019, 01:21:07 PM »

The graphic is for the purpose of skepticism on our way of finding value. For example if your Expected Value is +4% and after 1521 bets you have a yield lower than -2% than you should worry. If you have a yield lower than -6% than your way of finding value is definitely wrong. This is for random odds.
Logged
Pages: [1]
Print