Have a burning desire to win Lotto? - AI is here to help!

[randomx]

Yeah , the odds are buckleys and none for sure but l use to think well they're also those same odds for the one that wins it every wk to so yaknow, we've all got as gooda chance as anyone else is the way l look at it.

Mind you, l don't even play it ha ha, haven't for yrs , thinking l might start again though, why not.

 

[red750]

I always say, it only takes one game to win it. The guy that won the $50mill this week won with one game, lost with three. If you are meant to win it you will. Buying 50 games won't help if you're not meant to win it. You might pick up a small consolation prize, but probably wouldn't cover the cost of the ticket.

[red750]

A Melbourne dad almost missed the news he'd just won $50 million in Powerball after mistakenly deleting the email telling him about it.

The Victorian man, who was watching the AFL on the couch with his son on Thursday night, ignored the first message from The Lott because he assumed it was junk mail.

'I actually saw the first email and didn't even open it,' he said.

'I thought it was spam and deleted it.'

But a second email made him curious enough to check his ticket on the app - and that's when the reality of his win hit.

'Unbelievable. Fifty million dollars - that's insane,' he said.

'I'm just sitting here watching the footy with my son. When my wife gets home, I've got some news for her.

'She probably won't believe me, but when I show her the ticket, I think she'll start shaking and crying.'

The man held the only division one winning entry in Powerball draw 1556, drawn Thursday 12 March 2026, taking home the entire $50 million jackpot.

The winning ticket was just a $6.30 four-game entry bought through The Lott app.

'I only ever play the smallest entry. You don't need a big one to win,' he said.

[rgmwa]

What would you do with $50 million? Probably try and avoid all your new friends for a start. 

[red750]

This may sound odd, but I don't need much. Firstly I would make sure my 3 kids are comfortably well off. Then, considering the current and likely future economic conditions, I would give assistance to my siblings and Irene's siblings. I have 4 cousins, and would give them each a couple of thousand. I would make donations to a few charities.

 

I don't think I have many years to go, so I don't want a large bank balance to pay probate on. I have money put aside for my cremation and have enough to meet my expenses. so I don't need a big bank balance.

 

 

 

 

 

[rgmwa]

That sounds pretty reasonable. I’d do much the same in terms of the family and I’m not likely to live long enough to spend the rest. I’ve got two cars (one was my wife’s) that date from 2002 and 2008 respectively, but they’re going fine at the moment so I don’t even need a new car. What else to spend the other $49m or so on? It’s too much money for most people. As long as you’ve got enough to live pretty well, having a lot more is not going to do much for you. 

[onetrack]

3 hours ago, rgmwa said:

What would you do with $50 million? Probably try and avoid all your new friends for a start. 

Well, for a start, I wouldn't stop sending out the begging letters. You never know when the $50M might run out. 😄 

[pmccarthy]

I would pay people to finish all my half finished jobs, including vehicle projects.

[octave]

13 hours ago, pmccarthy said:

I would pay people to finish all my half finished jobs,

I wouldn't, which is not necessarily a plus.

We built our own house in 1990, and when we sold it, we were forced to install architraves and skirting boards and even get a completion certificate from the council so that we could sell it. I have that combination of the drive to do everything myself and a wonderful ability to complete 85% of the job.

[willedoo]

1 hour ago, octave said:

I wouldn't, which is not necessarily a plus.

We built our own house in 1990, and when we sold it, we were forced to install architraves and skirting boards and even get a completion certificate from the council so that we could sell it. I have that combination of the drive to do everything myself and a wonderful ability to complete 85% of the job.

octave, I built my own place in 1996 when I was in my early 40's and I'm thankful I did it then and not now. Looking at some of those beams I got up there on my own gets me trying to remember what it was like to be young and physically capable. The deepest stump hole was up to almost the top of my head and too tight for a jackhammer, so I levelled out a punching pad in the rock base with a crowbar. I used it like a percussion drill, letting the hand pressure off just as it hit the rock so it would twist a bit like a drill and chip a little bit out at a time. It was a bit claustrophobic down there. I dug all the stump holes with a shovel, crowbar and jackhammer. As far as raising bearers and beams, it's amazing what you can do with pulleys and temporary gin poles. To raise and place the poles in the holes I had pulleys and ropes set up at different points so I could tie the rope onto the 4WD and just drive away and watch the pole stand up in the rear vision mirror. It was hard work but a lot of fun and I'm glad I did it. All the 4x2 wall studs, king posts and collar ties, the 6x2 rafters and verandah joists and the 10x2 ridge board are all recycled ironbark from when they pulled down the old Maroochydore RSL. I bought a whole truckload of it from a demolition mate. Great seasoned timber but as hard as your mother in law's heart. You can't drive a clout into it without pre drilling.

[red750]

Many many years ago, my dad built a large building by himself. I don't know much about it, because I was about 5 or 6 at the time. Willi's story reminded me of what I had heard about it.

 

We lived on a farm in a valley which was acquired by the Victorian State Rivers and Water Supply Commission in about 1949. They were planning to build a dam in the valley, and pump the water into Melbourne's water supply. Our entire farm ended up under water. The building was to house the pumping station. The support columns were cut from trees surrounding the site. We didn't have a tractor in those days so they were hauled into place with horses.

 

[randomx]

Good old horsey power 😁

 

We we renting a place in NSW back when and the farmer would grab me for a lift with something now and then.

He was only 87 so you know, just a young fella haha. Well one day he asked for a hand with a gate post. This bloody thing was half a damn tree tell you what it must've been about in old school 7'x 18" or so.

We go over to the spot the post is just laying next to the hole and bugger me before l had time to even get a hold on the thing he had it about half a mtr of the ground and still lifting.

l was only about 30 so not wanting to be put to shame by an 87yr old 🥲 , l quickly got a hold and started lifting too.

l swear though l hardly helped , he was only a small man too but l could feel the post lifting and bang into the hole like it was nothing.

 

 

 

 

 

 

 

 

 

[facthunter]

The Odds are very much stacked against you. The HOPE you might win keeps you buying tickets. Without a ticket you won't ever win..  How much of the take that is Paid out%, gives you an indication of how fast you will lose your money. Gambling is a Big drain on any society. People you would think would never succumb to it, lose everything quite commonly.  It's EVIL. Nev

[Jerry_Atrick]

Of course, no AI or other software can change the odds of winning. But I would argue odds and probability are statistically two different things. 

 

The lotto is a totally random selection of 6 or so numbers from 45 (depending which lottery it is you are playing). That gives 1 in 14m-ish odds. We'll leave the system entries out. 

 

If you buy a ticket - you have exactly 1 in 14m-ish chances of winning.  That is the odds regardless. Assuming you have randomly selected the numbers or say used a child's birthday week in, week out, you are likely to have the same probability as odds. 

 

But if you use some sort of historical simulation - especially something like a path-dependent Monte Carlo simulation which assumes randomness in the probability chain, you may detect certain patterns of numbers which come up more often than others. The more data sampling points you have, the more likely these patterns will repeat. However, below a certain threshold, the simulation is meaningless, and you would have to introduce variables into your simulation that you thought were material - that could be simply the combinations of numbers, slight imperfections in the balls that result in certain balls being in the scooping place more often relative to others, the order in which the balls are selected may also change that, the humidity even, etc etc. The point is, with enough historical data, you won't need to observe all those other minutiae factors as the patterns will tell their own story...

 

And you may find, it can better predict the numbers likely to be drawn... and although your odds are exactly the same because it is a totally random draw, your probability of striking it lucky may increase. Of course, because the draw is totally random, it is not fool proof..

 

But it may say something like, for each number, if it is drawn first, this is the likely set of drawn numbers, and if the number drawn second is x, this is the likely set of following drawn numbers, etc. 

 

 

 

 

 

 

Edited yesterday at 11:31 AM by Jerry_Atrick

[pmccarthy]

Jerry that makes no sense to me. The outcome is either random or it is not. The next number is not dependent on previous results.

Edited 11 hours ago by pmccarthy

[rgmwa]

22 hours ago, Jerry_Atrick said:

But if you use some sort of historical simulation - especially something like a path-dependent Monte Carlo simulation which assumes randomness in the probability chain, you may detect certain patterns of numbers which come up more often than others. The more data sampling points you have, the more likely these patterns will repeat. However, below a certain threshold, the simulation is meaningless, and you would have to introduce variables into your simulation that you thought were material - that could be simply the combinations of numbers, slight imperfections in the balls that result in certain balls being in the scooping place more often relative to others, the order in which the balls are selected may also change that, the humidity even, etc etc. The point is, with enough historical data, you won't need to observe all those other minutiae factors as the patterns will tell their own story...

I looked at reams of historical draws years ago and noticed apparent variations in the frequency with which certain numbers came up. So I tried writing programs to analyse the data and improve my odds, while knowing that in the long run all numbers had an equal probability of being drawn, assuming there was no bias in the system. In effect, I assumed there was some kind of short to medium term bias. However, even if there was it became clear that the odds were still so high I might as well have stuck to random numbers.

[Jerry_Atrick]

1 hour ago, pmccarthy said:

Jerry that makes no sense to me. The outcome is either random or it is not. The next number is not dependent on previous results.

The outcome is totally random - yes. As I mentioned, a historical simulation may show a more or less universal distribution of drawn numbers, in which case the odds and the probability would be the same for all practical purposes. However, a historical simulation may show certain combinations of number drawn more often than others material enough to be a pattern or series of patterns emerging that increases the probability of sets of numbers being drawn. The draw is stall random, but the probability - and that is all it is - a probable and no guaranteed - set of drawn numbers may become higher than any old random selection. This could be because of say manufacturing imperfections of the balls where minutely different weight distributions result on some balls being more likely to be scooped than others.. And it could be that the probability if subsequent balls being drawn changes depends on what ball/s were previously drawn for the same reasons. In order to prove it, you would need a lot of data, including that of the factors that we define as determinative to the probability - but without those factors, we can only rely on the historical number draws.

 

2 hours ago, rgmwa said:

I looked at reams of historical draws years ago and noticed apparent variations in the frequency with which certain numbers came up. So I tried writing programs to analyse the data and improve my odds, while knowing that in the long run all numbers had an equal probability of being drawn, assuming there was no bias in the system. In effect, I assumed there was some kind of short to medium term bias. However, even if there was it became clear that the odds were still so high I might as well have stuck to random numbers.

Given the permutations and combinations, coming to a single set of numbers likely to be drawn using a historical simulation would improve the probability - but only marginally at best. These simulations predict probable outcomes over various scenarios and normally a confidence level is provided. And it will be a probability map. So for example, it may give a probability for each number of that number being drawn first. For each number, it will give the probability of every other number being drawn second. Then for each number drawn second of the number that was drawn first, it will come up with a probability of each number not drawn yet to be drawn third, and so on.  Taking this approach, there are 130m nodes in the map for each number drawn first. So, if 1 is drawn first, there will be a map of 130m possible permutations and combinations. Now multiply by 6 and that is how big the probability map will be.

You can use a binary coefficient approach to reduce it, but you attenuate the probability calculation as a result because we don't know the factors that are causing some numbers to be picked more than others.

 

Anyway, for a 6/45 lottery draw, the probability map reaches something like 900m nodes, each with a confidence factor. Of course, if we could pick our numbers just before each number is drawn, we could really improve our probability of winning. But we can't, so the only way to really improve the probability is to try and predict what the first through 6th number is going to be dependent on what we think the first number is going to be. So, in defining our confidence level, we have to look for any other correlations. As we have virtually no other known factors, the only thing I can think of is to look at time based cycles and add that to the algorithm to predict the different potential numbers to draw. And that is the rub. Coming up with one most likely to draw is not really going to to reduce it much; intuitively, I would reckon you need to put around 200k minimum bets on it spread across the different scenarios. And, it is still not guaranteed. 

 

I have no idea what the cost of a singe entry is, or 200k entries are against the normal prize value. So, it may be futile.. buit the above is a glimpse of algo trading (well, that would get smashed in the markets.. so I would not recommend it). 

 

Now I am rambling as I have had my first wine in 6 months - a Jim Barry McLaren Vale.. and very nice. What impressed me is my local Majestic Wines knew exactly which aisle I was heading for when I entered the shop... 

 

I digress.

 

Anyway, chasing the single winning entry will probably not yield too much more likelihood than the basic odds.. it is how you play the probability map.

[Jerry_Atrick]

BTW, you can replace drawing balls in an entry with this bit of python code:

 

import secrets

def draw_numbers():

  return sorted(secrets.SystemRandom().sample(range(1, 45), 6))

print(draw_numbers())

 

 

That is it.. 

Edited 1 hour ago by Jerry_Atrick