Nothing strange about consecutive Lotto numbers

[Swa]

No, the maths does not support the above. Those that try and analyse past draws and use that information in the belief that it will somehow influence future draws are the ones that fall into that trap. Hence why all the nonsense about so-called hot and cold numbers, seeking patterns and sequences and numbers that have not come up before, and belief in the dozens of "calculators" available on the Internet are just fooling themselves.

Each "draw" (assuming, of course, there is no tampering with the draw system, and that the RNG is truly random) is completely independent of what came before and the past does NOT predict the future.
The maths highlights that clearly if you take the trouble to work through it all.

Any number in the pool is equally likely to be drawn, (following the draw rules) and it is well within reality that a "draw" can repeat itself. The mathematics can be used to determine how many draws would be required for a specific set of numbers to be drawn and repeated.

The maths of Combinatorics does NOT predict actual numbers drawn or their order. Instead, it just shows up the probabilities of odds and evens and that some "patterns and sequences are more likely to occur.

The gamblers fallacy doesn't just apply between games but also within games. It is also stated as if odd/even numbers keep coming up the opposite is likely to increase in frequency.

In this case it would be saying that in an individual draw the odds are more likely that a combination of odd and even numbers will occur so such a combination is more likely to win. In a sense this is true. As odd or even numbers are removed from the pool of available numbers the bias shifts towards the opposite happening. But it's also the case there are more combinations of odd and even numbers than solely odd or even numbers. A combination of odd and even numbers has the same chance of winning as one that has only odd or even numbers.

 

[Geoff.D]

The gamblers fallacy doesn't just apply between games but also within games. It is also stated as if odd/even numbers keep coming up the opposite is likely to increase in frequency.

In this case it would be saying that in an individual draw the odds are more likely that a combination of odd and even numbers will occur so such a combination is more likely to win. In a sense this is true. As odd or even numbers are removed from the pool of available numbers the bias shifts towards the opposite happening. But it's also the case there are more combinations of odd and even numbers than solely odd or even numbers. A combination of odd and even numbers has the same chance of winning as one that has only odd or even numbers.

Within games is entirely predictable and can be calculated (the probabilities and odds.

We know how many numbers there are in a pool. We know how many odds and evens there are.
We know how many balls are going to be drawn before the game starts. We know if there is a power ball involved and whether it will be drawn from the main pool or a separate one, and how many power balls will be drawn. There is, therefore, nothing about the game that can't be statistically analysed EXCEPT the actual number values drawn.

The maths allows us to determine how each draw affects subsequent draws within the game. It IS after all the very design of a lottery game. It is therefore NOT what the gambler's fallacy is about.

Hence why one can use statistical mathematics to determine the probabilities. The probabilities however do not tell you what numbers are going to be drawn just some of their properties.

It is the very basis of the nPr and nCr formulas, series and distributions. If you reject the validity of these maths disciplines than ALL statistical analysis collapses.

 

[Swa]

Within games is entirely predictable and can be calculated (the probabilities and odds.

We know how many numbers there are in a pool. We know how many odds and evens there are.
We know how many balls are going to be drawn before the game starts. We know if there is a power ball involved and whether it will be drawn from the main pool or a separate one, and how many power balls will be drawn. There is, therefore, nothing about the game that can't be statistically analysed EXCEPT the actual number values drawn.

The maths allows us to determine how each draw affects subsequent draws within the game. It IS after all the very design of a lottery game. It is therefore NOT what the gambler's fallacy is about.

Hence why one can use statistical mathematics to determine the probabilities. The probabilities however do not tell you what numbers are going to be drawn just some of their properties.

It is the very basis of the nPr and nCr formulas, series and distributions. If you reject the validity of these maths disciplines than ALL statistical analysis collapses.

Ok, but it's still not the case that it's better to choose a combination of odds and evens rather than just odds or evens as the specific sequence still has the same chance of winning. Assuming of course the RNG has nothing to do with it.

 

[RonSwanson]

TRNG or HRNG versus PRNG is a huge debate.
My own view is the currently there are NO PRNGs that are NOT deterministic and therefore are open to manipulations and fraud. I don't care what the programmers do to try and disguise the deterministic nature of the algorithms. Humans are NOT random animals, we all think in very well defined ways and therefore nothing we programme can ever be truly random. Or best bet is a purely natural method to generate randomness.

HRNGs and TRNGs are the only ones that stand a chance of being free of manipulation. The move in SA to make use of PRNGs was a move backwards.
And it is also at the basis of why the "Quick picks" available should not be used at all by anyone wanting to play the games. They are too easily manipulated to favour the "house" and are not the player's friend at all.

Largely agreed, but do you have a source/s to validate this claim? Where was the use of PRNGs published?

 

[Blackhand]

HRNGs and TRNGs are the only ones that stand a chance of being free of manipulation. The move in SA to make use of PRNGs was a move backwards.

They are all susceptible to manipulation, depending at what level a skilled attacker is able to breach.
Assuming the worst case, which would be an attacker being able to inject code, the RNG method becomes irrelevant, they can push out any numbers they want and still get a clean audit (overwrite/manipulate the audit at insertion).
Granted, unless there is a glaring security issue in the lotto system, this is a very difficult attack.

As for PRNGs as a whole, well they are used for almost everything in security at this point. I know of no major tech company that is not using "cryptographically sound/secure" PRNG for key/secret generation.

 

[cguy]

They are all susceptible to manipulation, depending at what level a skilled attacker is able to breach.
Assuming the worst case, which would be an attacker being able to inject code, the RNG method becomes irrelevant, they can push out any numbers they want and still get a clean audit (overwrite/manipulate the audit at insertion).
Granted, unless there is a glaring security issue in the lotto system, this is a very difficult attack.

As for PRNGs as a whole, well they are used for almost everything in security at this point. I know of no major tech company that is not using "cryptographically sound/secure" PRNG for key/secret generation.

I would also hope that for the purposes of a lottery the entire RNG process is logged, so that the logged input seeds should be able to reproduce the lottery number. It’s still possible to inject fake seeds of course, but this now becomes a much harder attack (attacker needs to know the algorithm, reverse engineer plausible seeds and inject them). I am sure some of the seeds would be based off a nanosecond time stamp, which could also be verified to be within a tight range - faking this would require a very fast inversion, which would be very tricky.

 

[RonSwanson]

They are all susceptible to manipulation, depending at what level a skilled attacker is able to breach.
Assuming the worst case, which would be an attacker being able to inject code, the RNG method becomes irrelevant, they can push out any numbers they want and still get a clean audit (overwrite/manipulate the audit at insertion).
Granted, unless there is a glaring security issue in the lotto system, this is a very difficult attack.

As for PRNGs as a whole, well they are used for almost everything in security at this point. I know of no major tech company that is not using "cryptographically sound/secure" PRNG for key/secret generation.

PRNGs typically use input from the system itself and manipulates it by truncating everything that isn't random and then mixes it with a seed (entropy) that it itself creates. It's cheap and nasty. Many exist, but most are unsuitable for use as a CSPRNG, and even when they are, there are known issues such as NSA backdoors. TRNGs use other sources of entropy, such as atmospheric noise, radioactivity or "avalance noise" from a diode. The use case is slightly different when used for gambling, and some of the typical CSPRNG requirements do not apply (such as blocking when there is insufficient entropy). TRNGs are more expensive, yes, but the use case and budget is completely different when used for gambling vs crypto.

 

[RonSwanson]

Meet Alex, the Russian Casino Hacker Who Makes Millions Targeting Slot Machines

This Russian mathematician-turned-hacker has cracked slot machines worldwide. His secret: seeing through pseudo-random numbers.

www.wired.com

Just one example of PRNG use failure in the gambling industry.

 

[Geoff.D]

Largely agreed, but do you have a source/s to validate this claim? Where was the use of PRNGs published?

No it was not published. The management company changed and they changed all sorts of things. Including doing away with the draw machines. That is all we know.