Regulation

grumbler

la531983 said:

Its random. What more of an explanation do you need?

oldernonethewiser said:

neillaw said:

Hello,

My father and I both have 50k in premium bonds and without fail he wins more than I do every single month (for about 50ish months)

I'm sure it's all above board...  But I am thinking the odds of this are higher than winning the lottery!  Anyone heard anything similar to this happening before?  I did question this once to NS&I and they assured me everything was fine, but seems obvious there is some kind of glitch / bug (without thinking more sinister!)  Either that or it's truly a very lucky, or unlucky, chance

Thanks,
Neil

As above, it is random.

DullGreyGuy said:

neillaw said:

My father and I both have 50k in premium bonds and without fail he wins more than I do every single month (for about 50ish months)

I'm sure it's all above board...  But I am thinking the odds of this are higher than winning the lottery!  Anyone heard anything similar to this happening before?  I did question this once to NS&I and they assured me everything was fine, but seems obvious there is some kind of glitch / bug (without thinking more sinister!)  Either that or it's truly a very lucky, or unlucky, chance

People are very poor at understanding randomness, 

dunstonh said:

Anyone heard anything similar to this happening before?

NS&I Premium bonds are like a tombola.    Have you ever been in a large tombola where you get several people win multiple times, whilst most win nothing and the remainder win once?      

but seems obvious there is some kind of glitch / bug (without thinking more sinister!)  Either that or it's truly a very lucky, or unlucky, chance

It is a game of chance.   You are either lucky or unlucky.

Well-well...
If A and B have exactly the same amounts, the chances are the same.
Every month A has 50% (1/2) chances of winning bigger amount than B.
The probability of A winning more than B 50 times in a row is (1/2)^50= 1/(10^15)=0.000000000000001.
This ignores the chances of winning exactly the same amount, including zero.

Billion is 10^9
10^15 is million of billions

subjecttocontract

friolento said:

jimjames said:

If you want guaranteed earnings every month then move the money into a savings account currently paying 6%. That way you'll get £250 every month without fail. Not sure how much you get from your PBs but I suspect the average is less than that.

£250 before tax

A 6% return isn't really a 6% return if you pay tax on it:-
....its 4.8% if you pay 20% tax.
....it's 3.6% if you pay 40% tax

At that level of return premium bonds look very attractive.

DullGreyGuy

grumbler said:

la531983 said:

Its random. What more of an explanation do you need?

oldernonethewiser said:

neillaw said:

Hello,

My father and I both have 50k in premium bonds and without fail he wins more than I do every single month (for about 50ish months)

I'm sure it's all above board...  But I am thinking the odds of this are higher than winning the lottery!  Anyone heard anything similar to this happening before?  I did question this once to NS&I and they assured me everything was fine, but seems obvious there is some kind of glitch / bug (without thinking more sinister!)  Either that or it's truly a very lucky, or unlucky, chance

Thanks,
Neil

As above, it is random.

DullGreyGuy said:

neillaw said:

My father and I both have 50k in premium bonds and without fail he wins more than I do every single month (for about 50ish months)

I'm sure it's all above board...  But I am thinking the odds of this are higher than winning the lottery!  Anyone heard anything similar to this happening before?  I did question this once to NS&I and they assured me everything was fine, but seems obvious there is some kind of glitch / bug (without thinking more sinister!)  Either that or it's truly a very lucky, or unlucky, chance

People are very poor at understanding randomness, 

dunstonh said:

Anyone heard anything similar to this happening before?

NS&I Premium bonds are like a tombola.    Have you ever been in a large tombola where you get several people win multiple times, whilst most win nothing and the remainder win once?      

but seems obvious there is some kind of glitch / bug (without thinking more sinister!)  Either that or it's truly a very lucky, or unlucky, chance

It is a game of chance.   You are either lucky or unlucky.

Well-well...
If A and B have exactly the same amounts, the chances are the same.
Every month A has 50% (1/2) chances of winning bigger amount than B.
The probability of A winning more than B 50 times in a row is (1/2)^50= 1/(10^15)=0.000000000000001.
This ignores the chances of winning exactly the same amount, including zero.

Billion is 10^9
10^15 is million of billions

No, it's not 50% chance of one winning more than the other and 50% chance of it being the other way around. The most likely outcome in any month is that both win nothing; you cannot assume one or the other will win every month. You are also ignoring the chance that both win but win the same amount. 

eskbanker

DullGreyGuy said:

grumbler said:

Well-well...
If A and B have exactly the same amounts, the chances are the same.
Every month A has 50% (1/2) chances of winning bigger amount than B.
The probability of A winning more than B 50 times in a row is (1/2)^50= 1/(10^15)=0.000000000000001.
This ignores the chances of winning exactly the same amount, including zero.

Billion is 10^9
10^15 is million of billions

No, it's not 50% chance of one winning more than the other and 50% chance of it being the other way around. The most likely outcome in any month is that both win nothing; you cannot assume one or the other will win every month. You are also ignoring the chance that both win but win the same amount. 

Ignoring both winning the same was conceded in that post (see bolded wording), but two full holdings drawing blanks in the same month is far from the most likely outcome, with odds of about 1 in 100 last time I looked (1 in 10 each)!

Linton

eskbanker said:

DullGreyGuy said:

grumbler said:

Well-well...
If A and B have exactly the same amounts, the chances are the same.
Every month A has 50% (1/2) chances of winning bigger amount than B.
The probability of A winning more than B 50 times in a row is (1/2)^50= 1/(10^15)=0.000000000000001.
This ignores the chances of winning exactly the same amount, including zero.

Billion is 10^9
10^15 is million of billions

No, it's not 50% chance of one winning more than the other and 50% chance of it being the other way around. The most likely outcome in any month is that both win nothing; you cannot assume one or the other will win every month. You are also ignoring the chance that both win but win the same amount. 

Ignoring both winning the same was conceded in that post (see bolded wording), but two full holdings drawing blanks in the same month is far from the most likely outcome, with odds of about 1 in 100 last time I looked (1 in 10 each)!

Both scoring nothing in a month is not the most likely outcome....

The figures I get are for 1 person in 1 month
Wins, probability, probability squared
0: 0.092  0.0085
1: 0.220 0.048
2: 0.262 0.069
3: 0.208 0.043
4: 0.124 0.015
5:0.059 0.003
6: 0.0230.0005

These figures seem reasonable since on average with 1 in 21000 winning and you have 50000 bonds the average number of wins/month should be around 2.

So the chance of both people getting the same number of wins totals to about 0.19.

Having been invested for 50 months and given most prizes are £25 the chance of never having an even result would seem to be very low.

grumbler

DullGreyGuy said:

grumbler said:

la531983 said:

Its random. What more of an explanation do you need?

oldernonethewiser said:

neillaw said:

Hello,

My father and I both have 50k in premium bonds and without fail he wins more than I do every single month (for about 50ish months)

I'm sure it's all above board...  But I am thinking the odds of this are higher than winning the lottery!  Anyone heard anything similar to this happening before?  I did question this once to NS&I and they assured me everything was fine, but seems obvious there is some kind of glitch / bug (without thinking more sinister!)  Either that or it's truly a very lucky, or unlucky, chance

Thanks,
Neil

As above, it is random.

DullGreyGuy said:

neillaw said:

My father and I both have 50k in premium bonds and without fail he wins more than I do every single month (for about 50ish months)

I'm sure it's all above board...  But I am thinking the odds of this are higher than winning the lottery!  Anyone heard anything similar to this happening before?  I did question this once to NS&I and they assured me everything was fine, but seems obvious there is some kind of glitch / bug (without thinking more sinister!)  Either that or it's truly a very lucky, or unlucky, chance

People are very poor at understanding randomness, 

dunstonh said:

Anyone heard anything similar to this happening before?

NS&I Premium bonds are like a tombola.    Have you ever been in a large tombola where you get several people win multiple times, whilst most win nothing and the remainder win once?      

but seems obvious there is some kind of glitch / bug (without thinking more sinister!)  Either that or it's truly a very lucky, or unlucky, chance

It is a game of chance.   You are either lucky or unlucky.

Well-well...
If A and B have exactly the same amounts, the chances are the same.
Every month A has 50% (1/2) chances of winning bigger amount than B.
The probability of A winning more than B 50 times in a row is (1/2)^50= 1/(10^15)=0.000000000000001.
This ignores the chances of winning exactly the same amount, including zero.

Billion is 10^9
10^15 is million of billions

No, it's not 50% chance of one winning more than the other and 50% chance of it being the other way around. The most likely outcome in any month is that both win nothing; you cannot assume one or the other will win every month. You are also ignoring the chance that both win but win the same amount. 

The OP's statement was "without fail he wins more than I do every single month (for about 50ish months)".
You are right - chances of winning more are even lower than 50%. 

Linton said:

eskbanker said:

DullGreyGuy said:

grumbler said:

Well-well...
If A and B have exactly the same amounts, the chances are the same.
Every month A has 50% (1/2) chances of winning bigger amount than B.
The probability of A winning more than B 50 times in a row is (1/2)^50= 1/(10^15)=0.000000000000001.
This ignores the chances of winning exactly the same amount, including zero.

Billion is 10^9
10^15 is million of billions

No, it's not 50% chance of one winning more than the other and 50% chance of it being the other way around. The most likely outcome in any month is that both win nothing; you cannot assume one or the other will win every month. You are also ignoring the chance that both win but win the same amount. 

Ignoring both winning the same was conceded in that post (see bolded wording), but two full holdings drawing blanks in the same month is far from the most likely outcome, with odds of about 1 in 100 last time I looked (1 in 10 each)!

So the chance of both people getting the same number of wins totals to about 0.19.

If this figure is correct, then chances of winning more (or less) are 0.405.
0.405^50=2*10^(-20) = 2/10^20
10^20= one hundred million trillion
Right?
That said, I think the OP exaggerated, but everybody blamed the randomness instead of questioning the OP's statement.

Jonty6262

Is it like picking heads in a coin toss and tails coming up 50 times in a row? 

Middle_of_the_Road

Notepad_Phil said:

neillaw said:

...
 Either that or it's truly a very lucky, or unlucky, chance
...

Are you relying on your father telling you how much he's won or do you have access to his holders number to check directly? If the former then perhaps he's being a bit mischievous and kidding you on (it's something I could well imagine my late father doing if he had been lucky enough to have had that amount of money in premium bonds).

If he really has won more than you every month, which also means that he has never had a month with no winnings, then it is a very small percentage chance, but with the number of participants in the draw it's going to happen to someone and it's as likely to be you as someone else in the same situation.

This seems unusual to win every month, for over four years. I would be very surprised if this is the case.

As the OP has not included amounts won, or percentage returns, it could be that both bond holders are 'unlucky' if one wins 50 and the other 100, or both very lucky if it's 250 and 500.

mebu60

Until we are shown the itemised monthly winnings going back the 4+ years for both the OP and his father I feel it is highly unlikely the latter has beaten the former every month for that duration, particularly given the relatively modest prize payout in the earlier years of the span. As others have said, you'd very much expect the father to have had some blank months thereby nullifying the original claim.

That said, I've had an excellent 2023, another £200 this month (2 x £100, £50k holding). My brother, his wife and my ex-wife have all done well this year too. Checked on an elderly aunt last week expecting much the same but she's had an appalling year to date. Very kind of her to subsidise us! 

metrobus

I had 50k in it for about 8 months, did not have a single win .
When I put it in to the calculator it worked out that I would have had a far greater chance of winning a million than getting NOTHING.
It’s fixed for sure.