Premium Bond Winner ?

SnowMan

https://forums.moneysavingexpert.com/discussion/comment/82190237#Comment_82190237

And if you add a custom list in excel which just includes your area (exactly as it appears in the spreadsheet) then each month you can sort on area, choosing as order that custom item and your area will then appear at the top.

strawb_shortcake

£75 on full holding, my luck really has been dismal the last year.

My Son (17) has just bought £300 his first draw is December so will see how he does, he tends to have some luck.

SnowMan

No win on 1K holding.

Last wins £100 in July 2024 and £100 in September 2024 are now a distant memory

Podseas

Relative has won £1725 so far this year on FH, so according to above graph very close to the mean.

SnowMan, are you generating your charts through mathematical analysis or are you running a monte carlo simulation?

Marpau

£100 on FH

vienna28

I started with a £4,700 amount in August
2022. I always auto reinvest any prizes. After a £25 prize in November my
total is now £5,600.

My yearly wins:

2022 = £25

2023 = £100

2024 = £175

2025 = £325

2026 = £275

lascem

£125 from a FH (£25 and 2 x £50). Rolling 12-month total is £1750 (also YTD). 3.5% return. I was thinking of cashing in a few months back, but in light of the budget will stick with PBs for a while.

£150 for OH on his FH (£1850 YTD).

SnowMan

https://forums.moneysavingexpert.com/discussion/comment/82190325#Comment_82190325

I set myself the challenge of working out the discrete probability distribution of win amounts in excel as it seemed to me this was doable without using any significant approximation or simulation.

I am working out very close to exact probabilities for the win amounts at every win amount in an excel spreadsheet. I'm not using simulation at all. The only approximations I make are trivial, such as assuming that the returns of any bond are independent from other bonds of that holder in a given month, and that a single bond can win more than one prize in a month. And then I ignore multiple win combinations that are so unlikely that they are more than one thousand times less likely than winning a million even over 5+ years.

Once I take out all the probabilities I've calculated of winning amounts below about 300K, I am left with the almost exact probability of winning a million, which is a good check everything is working well and there are no floating point errors. The amounts agree the MSE calculator percentages exactly to the 1 decimal place rounding they use, although the MSE calculator only works on the current prize distribution. It is pretty clear to me that the MSE calculator isn't using simulation and is super accurate.

My approach is to first calculate probability of win numbers (amounts) from combinations of £25 and £50 prizes in a two dimensional grid with multiples of £50 amount wins in the columns and amounts of multiples of £25 representing the combined £25 and £50 win amount in the rows. For the combined win amount of £200, the first possibility is no £50 win and eight £25 wins, and in that case in the cell representing the £200 combined win amount row and 0 x £50 column we multiply the binomial probability of eight £25 wins by the binomial probability of no £50 win. And the second possibility is one £50 win and six £25 wins, and in that case in the combined £200 win amount row and 1 x £50 column we multiply the binomial probability of six £25 wins by the binomial probability of one £50 win. And so on. And the row total is then the probability of a £200 win by any combination.

The probabilities of say six £25 wins, or one £50 win are binomial probabilities and excel handles those binomial probabilities well even though the probabilities are small. And you only need to calculate those binomial probabilities once for each £25 win number and each £50 win number, as for example winning £200 through eight £25 wins, and winning £250 through one £50 win and eight £25 wins use the same binomial probability for the £25 wins as both involve eight wins.

Then I add in win amounts from £100 prizes again using the same grid template. This time the columns are the numbers (amounts) of £100 wins and the rows are the amounts of combined £25, £50 and £100 wins. And so in the cell representing the combined £200 win amount row and 1 x £100 column we multiply the binomial probability of one £100 win by the probability of combined winnings of £100 from £25 and £50 wins. This time the combined £25 and £50 winnings probability is not a new binomial probability but instead the probability already calculated in the £25 and £50 sheet. And the row total is the probability of a £200 win by any combination of £25, £50 and £100 wins.

Then I add in £500 prizes, then £1,000 prizes and so on using the same template of the two dimensional grid I used for the first £25 and £50 combining. You could combine them in a different order but working up with increasing win amounts works well enough and is most systematic. As you add in higher win amounts it becomes increasingly easy to do this. For example when adding in 100K wins you only need 5 columns at most as winning six 100K prizes is ridiculously unlikely. Doing it this way is an efficient but highly structured way of evaluating the different combinations.

It all recalculates in excel in a split second when I adjust the number of months or the holding amount or draw month.

pmets

https://forums.moneysavingexpert.com/discussion/comment/82190195#Comment_82190195

I share similar luck, £950 so far this calendar year on a FH, £75 in last draw

Dugal

On a FH £125 this month from 2 prizes giving £1625 for the last 12 months which is bang on the median from the excellent data from @SnowMan . I have had a FH since January 2021 but have yet to win a prize over £100.