Product Name: Black Box AI Bets – £100 to £10,000 from Saturday 31st October
All orders are protected by SSL encryption – the highest industry standard for online security from trusted vendors.
Black Box AI Bets – £100 to £10,000 from Saturday 31st October is backed with a 60 Day No Questions Asked Money Back Guarantee. If within the first 60 days of receipt you are not satisfied with Wake Up Lean™, you can request a refund by sending an email to the address given inside the product and we will immediately refund your entire purchase price, with no questions asked.
Description:
Two full months of this season went into the box before a single bet came out. From Saturday 31st October the bets land before 10:30am UK every morning, with the model’s read on every game beside them. Data goes in. Predictions and bets come out. You see everything it concludes – and never how it got there.
A new football model, built on more capable AI and fourteen seasons across twenty-two divisions, then kept quiet for two full months of this season while it learned every team’s new shape. On Saturday it starts betting in public, from £100.
£37 once. Rest of the season. 30-day money-back guarantee.
My name’s Stephen James. I’ve been betting on football for twenty years and building models to do it better for a good few of those. I had some early involvement in building the Hivemind Model, but I didn’t carry on with that project – I wanted to do my own thing. This is the story of what that turned into, and why you haven’t heard about it until now.
Midway through last season, something changed. The AI tools I build with suddenly got a lot more capable – not a bit better, a different class. The research that used to take me hours per fixture could be handed to AI agents that did it in minutes, at a depth no human keeps up with. Formations, line-ups, the finer details of a team’s last five games, who’s carrying a knock, how many days since the European trip. Every detail, for every game, every morning. I looked at that and thought: everything I’d built before was built for the old tools. What would I build with these?
So, in the background, I built it. I collected a crazy amount of back data – fourteen seasons of results across twenty-two divisions, every top flight that matters in Europe, the second tiers in the big five, England down to League Two, Scotland down to the fourth – and layered in team strength ratings, expected goals, the bookmakers’ closing prices, rest days and fixture congestion. Fifty-five things the model knows about a game before a ball is kicked. The model itself took about a week to build. The testing took months, and it hasn’t stopped. It’s been evolving all season.
Everything I’d learnt from the models before it went in. Then more. And I have to say it plainly: it has not disappointed.
Then I did the hard part. I waited. I could have launched this in August. I didn’t, because August lies – new managers, new signings, promoted sides nobody has a read on, a window still open. If I’d started on the first weekend there’d have been more upsets and more losing bets, and I knew it. So I fed it every result, every single day, and sent nothing. Two full months. Until every team in every division had found its level, and the model – and I – understood how they function this season. The model is built on more than a decade of results, but it’s those two months that drive the accuracy now.
I tracked every bet it put up the whole way through. Honest answer on how that went: the early weeks were patchy, as the data was thin and the season was still settling. Then from about week five the growth became consistent, with very few losers. It feels – and this is the only way I can put it – like the model now understands the teams in this season.
What surprised me most is how right it is when it’s right. It doesn’t just get the winner – it gets the scoreline. I’ve watched it call 3-2 at 15/1. There will always be losers – a couple of this season’s upsets no model and no human could have seen – but the pattern is consistency on the ordinary bets and the odd outlandish one that pays for a month.
That’s what starts on Saturday 31st October. A bank of £100. The model’s bets every morning before 10:30am UK – match odds, both teams to score, over 1.5, 2.5 and 3.5, and correct scores at very small stakes. Every bet on the public record before kick-off, every result the next morning, every penny accounted for. The target is £10,000 by the end of the UK season. You get what to bet, at what odds, for how much. And every day you also get the Daily Output – the model’s predictions and the research behind them for every game it ran that morning, not just the ones that became bets. What stays inside the box is how it gets there, and how the formula decides which of those predictions are worth money. That’s the product, and you don’t need it to follow the bets.
Fifty of you can come with me. That’s not a marketing number: if hundreds of people copy the same bets at the same prices, the bookmakers spot the trend and the prices go. Fifty keeps us under the radar. Doors close Friday at midnight so everyone starts on the same morning I do.
Sold already? Skip to the order box
Not a demo, not a stock image, not somebody else’s screenshot. This is my Betfair account on 25th October 2026 – the football profit and loss for the last thirty days, and the balance it’s sitting at.
That’s my betting, at my stakes. The £100 bank that starts on Saturday is a clean slate so you can follow it to the penny from day one – but you should know the person sending you bets backs football with his own money, and has the account to show for it.
Steady bets most days, the odd big one when the box is sure, every stake a slice of the live bank so it compounds without anyone chasing. How that works is below. This is where it’s going.
Pacing is illustrative. Banks don’t grow in straight lines, and ours will have losing days you’ll see in full. The target is the target, not a promise.
Every result in 22 divisions going back 14 seasons. Closing prices, Elo ratings, xG and xGA. Formations, line-ups, injuries, rest days and European midweeks. And two months of this season, fed in every day.
The bets. Fixture, market, odds and stake – match odds, BTTS, the goals lines and correct scores – before 10:30am UK, almost every day. And the Daily Output: the model’s prediction and research for every game it ran, bets or not. What never comes out is how it got there.
Every morning the model is fed the day’s fixtures across the divisions it’s built on – the top two tiers in England, Germany, Spain, Italy and France, England’s third and fourth, all four Scottish divisions, the top flights in Belgium, the Netherlands, Portugal, Turkey and Greece, plus the European nights. For every game, AI agents pull together what a serious punter would look at and a great deal they’d never have time for.
Seven factors carry most of the weight: team strength ratings that update match by match, venue form with home and away kept apart, expected goals for and against over the season and the last five, the bookmakers’ closing price, rest days and fixture congestion with European games counted, attacking and defensive shape from shot data, and each league’s own home advantage and goal rates. Seven more adjust it – goal-market habits at the venue, manager changes, injuries and line-ups weighed by position, what’s at stake, rotation around European midweeks, head-to-head at the ground when there’s a proper sample, and style.
From all of that it prices every market, not just the winner – home, draw, away, both teams to score, over and under 1.5, 2.5 and 3.5 – with a fair price to hold against the bookmaker’s. Where the two fall out, that’s a mismatch, and mismatches are what I capitalise on. It also predicts the scoreline properly, a full goals distribution per game corrected so it doesn’t lazily default to 1-1, which is where the 15/1 correct scores come from. And it marks its own homework: every prediction is logged and settled when the result comes in, the ratings refit every run, the weights re-estimated monthly. The version you follow in April is sharper than the one you join in October.
No preamble, no “we fancy”, no three paragraphs about a press conference. The bets first – the games the model has picked out and the bets built from them, with the price to take and the stake in pounds from the live bank. Some days one bet. Big Saturdays and European nights, a few. Some days the box says nothing, and that’s a bet too.
Illustration of the format – not real selections.
Then there’s the Daily Output, and it’s part of the package. Every day the model runs every fixture in its twenty-two divisions plus the European games, and every day you get the compiled output – the predicted score for each game, the probabilities on every market, the fair prices against the bookmaker’s, the expected-goals line, and the research the agents pulled on form, line-ups, rest and congestion. Not just for the games that became bets. For everything.
Not a mock-up and not a sample built for this page – this is the real Daily Output from Jattvibeday morning, 25th October, as it was produced. One of these lands every morning. The Card tab is every game the model ran, with the predicted score, the probabilities, the fair prices and the research. The Bets tab beside it is the day’s exact bets and stakes from the live bank – the part you act on. Both come together, before 10:30am UK.
Now note the accuracy. Those three predictions were published on Jattvibeday morning. Here’s what happened on Jattvibeday afternoon.
Three from three on the result, one exact score, and the other two among the five scorelines it rated most likely. One day’s output, not a season’s record – but it’s the output you’d have had in your inbox before 10:30 that morning, and it’s what you’ll get every day from Saturday.
What the Daily Output doesn’t show you is how the model got to those numbers, or why the formula turned three of them into bets and left forty alone. That’s the box.
Your bit takes five minutes. Open the email, place the bets at the price listed or better with any bookmaker or the exchange, close your phone. No in-play, no cash-out, no watching the box. Results and the bank update land next morning. Miss a bet here and there? It happens to all of us – I’ll try to make up for it along the way.
The model gives me the price it thinks every outcome deserves, and shows me where the bookmaker disagrees. The formula – the part I spent the two months on – decides what to do with that. It runs at two speeds.
Singles where one price is plainly wrong. Doubles where two line up. Sensible stakes, a few percent of the bank, most days. This is what gets the bank into profit and keeps it there.
Where the model’s scoreline read is strong and the price is big – 6/1, 8/1, 15/1. Staked very, very small. Most lose. The ones that land pay for weeks. In testing, these are what turned steady into significant.
Three things hold it together. Profit first: the early weeks are about getting the bank above £100 with the low-risk bets, and once it’s in profit I’m betting with profit, so the box gets more room. Stakes step up with the bank: every stake is a slice of the live bank, so the bets grow as it grows and nobody chases – the spreadsheet does the sums. And the bank is never in jeopardy: I’m cautious by default, I take the risk when the model says the risk is worth it, and no single day’s bets can do real damage.
Where it’s been tested. The model was backtested on ten seasons of results, every prediction scored against what actually happened. The formula was then run through the two months I waited – every bet built and tracked, none sent to anyone. Consistent growth from about week five, very few losers, and in the last few weeks the growth has been significant. Those figures will sit on this page as the live record builds; I’d rather show you a bank than quote a test.
The model is free to think whatever it likes about a game. The service is not free to bet however it likes. Underneath the clever stuff sits a set of hard guardrails that no amount of “confidence” can override. They exist because every one of them cost money before it became a rule.
Nothing above 6.0 at a normal stake, ever. The correct scores at 8/1 and 15/1 are in the plan – at a sliver of the bank that can never hurt it.
Five without a win and the box walks past, however tempting the price.
Every bet is sized to the live bank. Singles carry the most, doubles less, correct scores least. No chasing, no “all in”.
Winners and losers go on the record the next morning, to the penny. If you ever catch a loser missing, email me and I’ll put it right.
You see everything the model concludes – the Daily Output gives you its prediction for every game. What you don’t see is how it gets there, or how the bets are chosen from it. You’re entitled to ask why, so here’s the straight answer. First, it’s the product. The data pipeline, the factor weights, the rulebook, the formula and the way it learns from its own results took the best part of a year to get right, and printing it under every bet would be handing it to anyone with a copy-and-paste button.
Second, it makes you a worse punter. I told you about the double I glossed over with the model’s numbers right in front of me. Show a member why a bet was chosen and a fair number of them start “improving” the bets – skip the one they don’t fancy, double the one they do, back a Daily Output prediction the formula deliberately left alone. Their record stops being the box’s record and becomes their own, and their own was the thing they joined to escape.
Third, you can judge it without seeing inside. A black box is only scary if nobody checks what comes out of it. Every bet here is published before kick-off, settled at the price I told you, and added to a bank anyone can see. The record is the window. Look through that, not the lid.
The daily Black Box email, before 10:30am UK almost every day of the week including the European nights, with every bet, its odds and its stake – across twenty-two divisions and the European competitions. The Daily Output with it: the model’s predictions, probabilities, fair prices and research for every game it ran that day, whether it became a bet or not – all of it, compiled, every day, as part of the package. The public £100 bank, updated daily to the penny, losers included, for you to copy exactly or scale to whatever you start with. Every model upgrade through the season, free, with two months of this season already inside it. And me on support, same-day replies, no ticket queue.
Then there’s the spreadsheet, which is yours to download and keep. It does every calculation for you: the stake for each bet from the live bank, the running profit, where you stand against the target. Type in the result and it moves. On its own it’s worth the price – it’ll size any bet you ever place properly, for any service, for as long as you bet.
Take one of the 50 places
The bank starts at £100 on Saturday 31st October. From then on this block updates every day with yesterday’s bets and the running total, to the penny. Join now and your record is the same as ours.
If you’ve followed tipsters before, you’ve probably been hurt – promised the earth, sent three losers and a shrug. I’ve been on the receiving end of that too, which is why this isn’t built on promises. It’s built on a bank you can watch. You’ve seen the Daily Output above and what it did on Jattvibeday. You’ve seen my own account. From Saturday you’ll see every bet before kick-off and every result the next morning, to the penny.
So judge it on what it does, not on what I say. Join, and for the first thirty days you don’t have to place a single bet if you don’t want to. Open the Daily Output each morning, watch the bank move, see the results land. If it isn’t making money, claim your refund and you’ve lost nothing – not a penny, no questions. If it is, you’re one of fifty people in from the first morning, following a bank that will never be smaller than it is on Saturday.
That’s the whole offer. Thirty days of watching it work, at no risk to you. You decide what the upside is worth.
Join, and treat the first month as a trial. You can follow the bets from day one or you can sit on your hands and just watch – open the Daily Output every morning, see what it predicted, see what happened, watch the £100 bank move on the public record. You’re not risking a penny of betting money unless you choose to.
If at any point inside 30 days you look at the bank and it isn’t making money, email me or ClickBank and you get every penny of the £37 back. No form, no “but why”, no quibble.
I’ll be honest about why I can offer that so easily. Once you’ve watched the box for a couple of weeks – the predictions landing, the scorelines it calls, the bank doing what it’s supposed to – the chances of you bothering to ask for a refund are close to nil. You’ll have seen how powerful this is for yourself, and by then you’ll be following it, not watching it. The guarantee is there so you can find that out with nothing on the line.
Every bet of the £100 to £10,000 bank build, from Saturday 31st October to the end of the season
Once. Rest of the season. No monthly, no rebill, no hidden extras.
About a fiver a month for seven months of bets – or one losing acca, if we’re honest about where the money usually goes.
Secure checkout through ClickBank. One payment – nothing rebills. 30-day money-back guarantee.
Anything at all – a question before you join, a missing email, a price you couldn’t get – write to support@blackboxaibets.com. It goes to my phone. I reply the same day, usually within the hour.
Results shown on this site are recorded at the prices advised in the daily email and may differ from the prices you obtain. The £10,000 figure is a target, not a guarantee, and past or paper-traded results are not a guarantee of future results. The model’s selections are opinions generated by software and are provided for information and entertainment only. You are responsible for your own bets. 18+ only.
© 2026 Black Box AI Bets. All rights reserved.
All orders are protected by SSL encryption – the highest industry standard for online security from trusted vendors.
Black Box AI Bets – £100 to £10,000 from Saturday 31st October is backed with a 60 Day No Questions Asked Money Back Guarantee. If within the first 60 days of receipt you are not satisfied with Wake Up Lean™, you can request a refund by sending an email to the address given inside the product and we will immediately refund your entire purchase price, with no questions asked.
