Lottery numbers (OT, of course!)

[GB]

5, 6, 7, 8, 9, 10 drawn in South African lottery.

So, obviously, it cannot be random. ;)

[NY]

Why is it that some people persist in thinking that 5, 6, 7, 8, 9, 10 are any more or less likely than 1, 6, 11, 23, 37, 41 - or any other combination of numbers? The fact that they are consecutive is noteworthy - but it's not a fix.

[Andy Bennet]

On average those 5 balls and powerball will occur once in 42,375,200 draws, same as any other valid combination.

[williamwright]

He was joking, yer daft pillock!

Bill

[GB]

Ok, but there'd have been the same furore over any consecutive sequence. So, there are are 53 similar consecutive draws possible. (The UK has 59 different balls, although that may vary in other countries.)

So, really, the chance of a sequence like this is more like 1 in 700,000.

At a guess, there are maybe 500-1000 major lottery draws per week around the world. (The UK has two draws a week, and there are about 170 countries in the world, many of which have lotteries. Plus, there are other lotteries, not just the national ones.)

So, there have possibly been 500k - 1m draws in the last 20 years. And, a 1 in 700k event has occurred. Not so astounding really! :)

[Brian Gaff (Sofa]

That is right, randomness is not the same as non consecutive. There should be an equal possibility of that happening but the odds increase with each drawer, but the device or electronic device, does not know what the last number was. Randomness is never truly random as you need infinity for that!

My old Sinclair Spectrum used a seed created from a count of the number of TV frames displayed between two reads, but since it was held in two bytes that meant it zeroed out at 65535, or FFFF in hex. Brian

[Dave Liquorice]

No they don't! The *odds* of any particular set of numbers being produced in a random draw is the same for each draw. It's perfectly possible for the same sequence to be drawn on consecquative draws with no "fixing" required. The *probability* that a particular set of numbers being produced in consequative random draws is very, very, low but not zero.

The odds are fixed by the maths of the number of selctions to be made from the size of available selections. ie the 6 from 49 or 1 from 6 for a dice or 1 from 2 for a coin toss.

Lotteries tend not to use electronic means of generating "random" numbers(*) but very carefully constructed rotating drums full of very carefully matched, size, shape, weight, evenness of density etc numbered balls.

ERNIE (Electronic Random Number Indicator Equipment) for Premium Bond numbers is a notable exception. The latest ERNIE 5 uses quantum effects rather than thermal noise to produce the randomness.

[Jethro_uk]

Why do you think ? I know what I think ...

[Reentrant]

That actually happened in the Bulgarian national lottery in 2009; two consecutive draws a week apart produced the same numbers (6 from 49).

The following year Israel had the same winning numbers three weeks apart.

[Andrew]

But there is still a requirement to convert whatever random serial stream of ONES and ZEROS that it produces, into an alphanumeric string representing a bond number. Then all the invalid bond numbers must be ignored (encashed) and further checks needed to make sure the holder isn't dead. Is this done in sofware ?. Often wondered.

Depending on the exact point in time that this serial bit stream is converted into bond numbers has a massive effect on the outcome.

[Jethro_uk]

The was an apocryphal story on Qi of a lady who picked the winning numbers in the NY and NJ lotteries in the same week.

Sadly her NY winning numbers were on the NJ lottery, and the NJ winning ones on the NY lottery ....

[Andy Burns]

Nothing stops your estate winning the premium bonds after you're dead.

[Tim Lamb]

:-) My sister, dead 4 years, won £25.00:-)

You would imagine somebody responsible for getting an estate probated would check National Savings!

[newshound]

But were they drawn in that order? That makes a lot of difference to the probability.

[newshound]

And there are other fairly "odd" sequences:

1, 11, 21, 31, 41, 51

3, 6, 9, 12, 15 etc.

[Jethro_uk]

I have a memory that a US lottery had quite a lot of winners one week, as the numbers formed a cross on the slip ...

[Chris J Dixon]

Wasn't it the iPhone that had its "random" mode adjusted so that it would not play tracks from the same album close together, as this wasn't what its users expected.

Chris

[whisky-dave]

The numbers that are most likely to not come up seem to be those I have chosen :-(