Law of Large Numbers

[Seoulmate]

Traditionally bricks and mortar casino's used the law of large numbers to profit over the long run. It is more difficult for players to use this mathematical principle in physical casinos, unless they live in the casino (and some do!).

However, online, players can autobet 24/7, at high speed, which allows them to also employ strategies that are based on the Law of Large numbers.

Has anyone used a strategy that employs the use of the Law of Large numbers?

[cryptonesiac]

I have tried (just a little bit), but it's not as easy as it looks

you need a lot of patience and time, because that law means you need a shitload of numbers. And I still think the house edge eats up a fair amount of the predictions you can make

EDIT: for example, I tried to count the aboves and belows at 50% chance and then bet against the leading number (so, 1000 above 50.00 and 950 below 50.00 >> bet on below). The law of large numbers is however relative, not absolute. So you can end up with more losses, even though the ratio above and below is getting closer to 1.

[Seoulmate]

I do find that games like dice and limbo have a tendency to balance out, long streaks below 10x for example will often end with 10x showing more frequently than would be expected on average. A low median does resolve to a higher median over time and vice versa. Similar to trading online the median does appear to trend upwards or down over the longer term. In the shorter time frames 100 rolls for example it can be quite volatile.

Some sites provide a flash bet facility which allow 10k rolls in about 5 seconds. I used this once to take 1st place in a 24hr volume wager race. I started 10mins before the end of the race but used flash bet to take 1st place and prize.

 

[sheenazbay]

honestly I do not really understand what you mean, I do not understand about strategies that use the big number law. I often play dice and limbo. I just play and play according to my calculations for each payout that I use

[wngo123]

Could you explain how we could use law of large numbers to our advantage here on stake? How many rolls it would require and do you think that this is possible here? I'm not really sure how to do it and I have doubts this could work. I would need more details.

[dolphinpuke]

reverse martingale is good for this

[mrsino]

Honestly, I don't really understand what you mean, I don't know about strategies that use the bulk laws. I just play and play according to my calculations for each payment I use

[Andrew_]

Scalping is a nice strategy. I'm trading of this strategy

[iamsheikhadil]

This law sounds interesting..but as far as I know, doesn't it require like a huge bank roll to win and balance the losses ?? But obviously, this is the finest strategy if I'm correct about assumption.

[Seoulmate]

16 hours ago, sheenazbay said:

honestly I do not really understand what you mean, I do not understand about strategies that use the big number law. I often play dice and limbo. I just play and play according to my calculations for each payout that I use

Well for example in Limbo if you bet on odds that are 50/50 you must target 1.98x. Theoretically you should win half and lose half as the odds are 50/50. So the median over a large number of rolls must resolve to 1.98x, but the median rises and falls, similar to trends in a stock market. So if you measure the median as you play you can bet larger when the median is low (knowing it will rise over time) and make smaller bets when the median is high, knowing that it will come down eventually returning to 1.98x. 

In a nutshell you win more bets when the median is rising and you lose more bets when the median is descending. I think a thousand games should be enough to identify an upward or downward moving median.

[sheenazbay]

1 hour ago, Seoulmate said:

Well for example in Limbo if you bet on odds that are 50/50 you must target 1.98x. Theoretically you should win half and lose half as the odds are 50/50. So the median over a large number of rolls must resolve to 1.98x, but the median rises and falls, similar to trends in a stock market. So if you measure the median as you play you can bet larger when the median is low (knowing it will rise over time) and make smaller bets when the median is high, knowing that it will come down eventually returning to 1.98x. 

In a nutshell you win more bets when the median is rising and you lose more bets when the median is descending. I think a thousand games should be enough to identify and upward or downward moving median.

I understand now, but I have never noticed more between low and high, so I only play with a percentage on loss. because even though a lot is at low, bets will return if hit, even though 5 is compared to 1 (five lows and 1 high)

[nuuuitsjdragon]

Actually this method requires a lot of bet tracking and a whole lot of data to sieve through before any strategies can be employed.

Anyway, it gets increasingly difficult as it trends towards a larger number of bets and closer to the actual house edge.

In any case, I don’t think it’s possible unless we are discussing very high multiplier (5-6 digit) hits as that helps to balance out your eventual losses.

[Seoulmate]

2 hours ago, nuuuitsjdragon said:

Actually this method requires a lot of bet tracking and a whole lot of data to sieve through before any strategies can be employed.

I've used this strategy and it is one of the most effective, The 4 ways I've used it are;

1: Luck Score: use the luck score as a substitute for median if it's available on the site you are playing

2: Script: use a script that employs the median. The most successful script Ive ever used was one that used a median strategy to place bets

3; Bet History and google sheets: use api to pull your bet history data into a spreadsheet.

4: Onsite bot: Some sites run bots that provide median stats on request.

Quote

"Anyway, it gets increasingly difficult as it trends towards a larger number of bets and closer to the actual house edge."

No this isnt an issue as you only need sample the median from the last 100, 300, 500,1000 bets, or whatever range you are working with.

Quote

 

In any case, I don’t think it’s possible unless we are discussing very high multiplier (5-6 digit) hits as that helps to balance out your eventual losses.

 

It's possible. You may be confusing mean with median here. High numbers don't affect the median, as its more robust than the mean, which is why we use it.

16 hours ago, wngo123 said:

Could you explain how we could use law of large numbers to our advantage here on stake? How many rolls it would require and do you think that this is possible here? I'm not really sure how to do it and I have doubts this could work. I would need more details.

Believe me it works, all the casinos use it to work out the house edge and calculate profits on any game.

The law of large numbers is a very interesting mathematical principle as it gives the ability to predict quite accurately outcomes over a large set of trials/games. Think of Tossing a fair coin 10 times the odds are 50/50 but we don't necessarily get  5 head (we may get 8 tails and 2 heads), but toss it 1000 times and we end up with a much closer ratio of heads to tails.