An Post Prizebonds

FlyOver

Hi all,

Saw this on the An Post Prizebonds FAQ:

Q. What are the chances to win a prize if you have for example €1,000 invested?

A. A customer with €1,000 invested has a 3.8 to 1 chance of winning a cash prize in a 12 month period.

Simply wondering is this a correct statement?

seanybiker

Dunno to be honest. Me uncle used to buy prize bonds alot over the years. He had loads of them and never got much back from them.

Thomas_S_Hunterson

I'd interpret it to mean that there is a 79.167% probability of winning (3.8/(1+3.8)).

I don't know if it's correct or not. Instinctively it sounds a bit high.

FlyOver

I thought the same myself. Just under an 80% chance of winning a weekly draw within a year with €1000 investment sounds good, maybe too good.

With the factors involved, is there a method in which this can be proved or disproved?

henbane

Rough calculation from the prize page:

208844 draws per year
€834333333 is the total fund, 1 bond is €6.25 => 133493333 bonds total
€1000 is 160 bonds

Draws per year * (160/total bonds) = 0.2503

Pherekydes

Sean_K wrote: »

I'd interpret it to mean that there is a 79.167% probability of winning (3.8/(1+3.8)).

I don't know if it's correct or not. Instinctively it sounds a bit high.

Wrong way around.

Odds of x to y is equal to the probability y/(y+x).

3.8:1 = 21% approx.

Michael Collins

Surely it should be 1/(3.8+1) = 0.2083. Which is fairly close to henbane's calculation (and erring on the underestimation side).

A 100-1 bet is way less likely to come good than a 2-1 bet.

Edit: Pherekydes beat me to it - and got the same answer.

Thomas_S_Hunterson

Pherekydes wrote: »

Wrong way around.

Odds of x to y is equal to the probability y/(y+x).

3.8:1 = 21% approx.

I think you're thinking of bookie's prices.

Although this does depend on what the Prize bond company mean. (not sure they know themselves)

/edit: Consider a horse wich has a 10% chance of winning a race.

The odds of it winning are 1 to 9

The bookie will offer you 9 to 1.

red_fox

Sean_K wrote: »

I'd interpret it to mean that there is a 79.167% probability of winning (3.8/(1+3.8)).

I don't know if it's correct or not. Instinctively it sounds a bit high.

It's simply 1/3.8 ~ 26.3%, which is close enough to the 25% above to sound true. Odds are listed so the 1 one the right is you, think of odds 1:1, in terms of probability it's a sure thing (in racing they're not true probabilities, for example 0.5:1 is not defined as a probability)

Thomas_S_Hunterson

http://en.wikipedia.org/wiki/Odds

Pherekydes

Sean_K wrote: »

I think you're thinking of bookie's prices.

Although this does depend on what the Prize bond company mean. (not sure they know themselves)

3.8 to 1 means the same thing no matter who says it.

If the Prize bond company don't mean 'a 21% chance of winning' then they shouldn't be be stating 3.8 to 1. Because, by definition, they are equivalent.

Thomas_S_Hunterson

Pherekydes wrote: »

3.8 to 1 means the same thing no matter who says it.

If the Prize bond company don't mean 'a 21% chance of winning' then they shouldn't be be stating 3.8 to 1. Because, by definition, they are equivalent.

It can be 3.8 to 1 in favour or 3.8 to 1 against.

The way the prize bond company puts it, I would interpret it as 3.8-1 in favour.

The price you get in a bookie is the odds against an event.

I think the prizebond company may be confused.

Pherekydes

red_fox wrote: »

It's simply 1/3.8 ~ 26.3%, which is close enough to the 25% above to sound true. Odds are listed so the 1 one the right is you, think of odds 1:1, in terms of probability it's a sure thing (in racing they're not true probabilities, for example 0.5:1 is not defined as a probability)

That's wrong, I'm afraid. The calculation is 1/4.8 ~ 21%

1:1 is evens. i.e. A horse has an equal chance of winning and losing, i.e. 50:50

0.5:1 is the very same as 1:2 which is 2/3.

Thomas_S_Hunterson

red_fox wrote: »

It's simply 1/3.8 ~ 26.3%, which is close enough to the 25% above to sound true. Odds are listed so the 1 one the right is you, think of odds 1:1, in terms of probability it's a sure thing (in racing they're not true probabilities, for example 0.5:1 is not defined as a probability)

You're ocnfusing decimal odds here now.

/edit: beaten to it.

red_fox

Sean_K wrote: »

It can be 3.8 to 1 in favour or 3.8 to 1 against.

The way the prize bond company puts it, I would interpret it as 3.8-1 in favour.

The price you get in a bookie is the odds against an event.

I think the prizebond company may be confused.

As confused as I was, I read odd as against. As said earlier the minimum number of prizes as listed on the webpage gives a 25% chance of winning, the odds against interpretation gives 26% so it seems right to me, if phrased poorly.

Pherekydes

Sean_K wrote: »

It can be 3.8 to 1 in favour or 3.8 to 1 against.

Absolutely not!

4 to 1 means 1 chance of the occurrence happening and 4 chances of it not happening.

That is the very definition of odds.

If someone says 4 to 1 when they mean 1 to 4 then they are wrong, plain and simple.

Pherekydes

If they are trying to publish odds to represent the probability of winning which is given by henbane's calculation as approximately 25% then they should change the probability to the nearest simple fraction, i.e. 1/4 and then calculate the odds from that which is clearly 3 to 1. It seems they've made an arse of a simple calculation.

Thomas_S_Hunterson

Pherekydes wrote: »

Absolutely not!

4 to 1 means 1 chance of the occurence happening and 4 chances of it not happening.

That is the very definition of odds.

If someone says 4 to 1 when they mean 1 to 4 then they are wrong, plain and simple.

Read the wiki article.

Pherekydes

Sean_K wrote: »

Read the wiki article.

I did. I got my info from it.

But cross checking with mathworld shows that the wiki page is wrong in one calculation. The calculation to choose 1 day from seven is clearly 1/7 which equals odds of 6 to 1. I am right; the wiki page is wrong.

gillo_100

I saw a bit of a discussion about prize bonds here
http://www.askaboutmoney.com/showthread.php?t=96976
About 10 posts down is a very good post explianing a lot.

Thomas_S_Hunterson

Pherekydes wrote: »

I did. I got my info from it.

But cross checking with mathworld shows that the wiki page is wrong in one calculation. The calculation to choose 1 day from seven is clearly 1/7 which equals odds of 6 to 1. I am right; the wiki page is wrong.

You're still looking at the odds against choosing the right day.

Pherekydes

Sean_K wrote: »

You're still looking at the odds against choosing the right day.

That's the definition of odds. Google 'odds against'.

Thomas_S_Hunterson

Pherekydes wrote: »

That's the definition of odds. Google 'odds against'.

No odds are ratios of probabilities. "Odds against" and "odds for" are subsets of "odds".

FlyOver wrote: »

Q. What are the chances to win a prize if you have for example €1,000 invested?

A. A customer with €1,000 invested has a 3.8 to 1 chance of winning a cash prize in a 12 month period.

This, to me, is pretty clearly, and perhaps mistakenly, expressing the odds for winning.

FlyOver

To get back to the original question, is that statement from An Post Prizebonds correct or incorrect?

Pherekydes

Sean_K wrote: »

This, to me, is pretty clearly, and perhaps mistakenly, expressing the odds for winning.

To me, this is pretty clearly expressing the odds against winning.

Odd for winning would be normally expressed as 1 to 3.8, or 3.8 to 1 on. In the absence of the word 'on' at the end it is assumed to mean odds against.

Pherekydes

FlyOver wrote: »

To get back to the original question, is that statement from An Post Prizebonds correct or incorrect?

Yes!

:pac:

Michael Collins

The chances of anything coming from Mars are a million to one...but still they come!

LeixlipRed

Michael Collins wrote: »

The chances of anything coming from Mars are a million to one...but still they come!

Do they just visit you or what?

Michael Collins

LeixlipRed wrote: »

Do they just visit you or what?

Yeh but only after a few drinks

TommyGunne

The An Post statement is clearly meant to be ~20%. That is the intention, and that is how odds are used in common speech. If someone says something is 4 to 1 to happen, that means they think it happens 20% of the time, and if they think it happens 80% of the time they say either 1 to 4 or 4 to 1 on.

Odds are expressed as ratio of losses to total events whether you are talking to a bookie, a postmaster or whoever (Clearly not to someone with a background in stats/maths w/o a gambling background though :D)

From wikipedia: "the first figure represents the number of ways of failing to achieve the outcome and the second figure is the number of ways of achieving a favorable outcome: thus these are "odds against""

Wikipedia is perfectly clear on this. Dunno why you cite the wiki article when it says that everyday use of odds is the above.



To more answer the OPs question, yeah that seems about right. AFAIK, that should correspond to ~just less than the current interest rate to An Post. Dunno the distribution of prize sizes, but AKAIK, claims that aren't collected within a certain time period go back to An Post, except the large prizes, where An Post must contact the owner. Most people don't collect claims (ie nephews who received €50 in bonds and never checks.) That would suggest however that if you regularly check your prize bonds, they would be significantly better than the interest rate. I haven't done any sums, and I don't know the distribution of prizes, but it'd be interesting to check how it compares against the interest rate for a person that collects all claims. Instintively, if the stipulation on collecting in time is correct, then I would reckon from the 3.8 to 1, and knowing the distribution will be heavily right-skewed, then I would guess that it would be quite a bit better than the going interest rate.

Really depends on people that don't claim their small prizes lose out though.

rjt

TommyGunne wrote: »

To more answer the OPs question, yeah that seems about right. AFAIK, that should correspond to ~just less than the current interest rate to An Post. Dunno the distribution of prize sizes, but AKAIK, claims that aren't collected within a certain time period go back to An Post, except the large prizes, where An Post must contact the owner. Most people don't collect claims (ie nephews who received €50 in bonds and never checks.) That would suggest however that if you regularly check your prize bonds, they would be significantly better than the interest rate. I haven't done any sums, and I don't know the distribution of prizes, but it'd be interesting to check how it compares against the interest rate for a person that collects all claims. Instintively, if the stipulation on collecting in time is correct, then I would reckon from the 3.8 to 1, and knowing the distribution will be heavily right-skewed, then I would guess that it would be quite a bit better than the going interest rate.

Based on the figures above, and assuming that you collect any prize you're given, the expected return on a euro (per year) is pretty much exactly 3 cent. So it's less than 3% return for the average person not checking if they've won (and this doesn't increase the return of someone who does check - the money just goes back into the fund).

RoundTower

they really don't write if you win a small one? I used to have some of these, wonder what I missed out on.

RoundTower

TommyGunne wrote: »

AKAIK, claims that aren't collected within a certain time period go back to An Post, except the large prizes, where An Post must contact the owner.

OK according to the site it's just untrue

An Post wrote:

How do I know if I've won a prize?
All prize winners are contacted following each draw. In order to maintain confidentiality and security, correspondence is sent in plain envelopes. For prizes worth 250 or less a cheque is enclosed. For any prizes worth more than 250 a form is enclosed in order for the investor to claim their prize.

To ensure that you can be contacted to claim your prize, it is important that you inform us - by signed mail or fax - if you change address.


What happens to unclaimed prizes?
Very few prizes are not claimed, but if someone does not claim their prize within 6 months of the draw date, the prize is entered onto our unclaimed prizes database. The prize is then held until the holder comes forward to claim it.

Michael Collins

RoundTower wrote: »

OK according to the site it's just untrue

Yeah, I think the confusion is because it used to be the case that for prizes under a certain value you were not notified so the owness was on you. But that's obviously changed since. And they actually send you out the check for small values now which is nice.

Although 3% return isn't really that good, even these days you can get banks offering over 5% AER (some with withdrawl delays and deposit limit but I'd say there's some without those clauses too).