Strategy · Published 8 September 2026 · 8 min read
How I made 96 units in my first five weeks.
Firstly, hello. I am Matt and I created LastMinutesAI. I have been betting on the last minutes for years using my models, and I finally decided to turn it into an app and share it with you. This article will try to explain how I used my knowledge of the markets and the app to help you try to start off a successful last minute betting system.
I bet one thing more than anything else: total corners, over the line, in a match that is still alive. Across 386 settled bets between 4 August and 8 September 2026 I finished 96.21 units up. This will explain how I have been using my app and an honest account of what I have kept doing and what I am still testing.
The main bet
A match is level or within one goal, the clock is somewhere between 75 and 85 minutes, check the Asian corner line, and if the value and the odds beat the fair price then that is the bet.
The app shows these for you, which is the part that makes it easy to run. Open a corners market that the bookmaker is offering you. This will show a variety of markets, but we care about Asian corners.
There are two ways to back the same thing here. The main line is over 8.5 at a fair price of 2.79. Under it sits the Asian 8.0 at 1.86, which gives your stake back if the match ends on exactly eight corners. The half line is my default because it has paid better; I take the .0 when I want that protection. Either way I only bet when the bookmaker is longer than our fair price. My stake comes from experience of these markets and the value the app says I am getting. First though, the corner markets themselves.
The corner markets
Three different corner markets turn up in this window and they behave very differently. If you have only ever bet goals, these are worth understanding before you bet any of them.
Total corners, over and under
The easy one. It counts every corner in the match by both teams. "Over 8.5 corners" wins if the match finishes with nine or more. Because the line ends in .5 there is no way to draw: you win, or you lose. These can be useful when you want to back more corners than the Asian line offers.
Asian corner lines
Asian lines are the same bet except the bookmaker sets one value as the line, and these can either be .5 or .0. A .5 behaves the same as a total corner bet. A .0 Asian line is different.
- Whole line, for example over 8.0. Nine or more and you win, seven or fewer and you lose. Exactly eight and the bet is pushed, so your stake comes back. It is the insurance version, and you pay for it with a shorter price: in that screenshot 8.5 is 2.79 while 8.0 is 1.86.
Those 26 pushes in my record are whole Asian lines landing exactly on the number. Which line you take is a real choice, and so far my numbers point one way.
Choosing your line
My own tracker, 4 August to 8 September 2026. So far the insurance has cost me more in price than it has saved in pushes.
Where to find them: most in-play sections keep corners under their own tab, with the Asian lines a level deeper, sometimes behind an "Asian" or "alternative lines" label. Not every bookmaker prices them in-play, and those that do vary in how deep the lines go, so it is worth shopping around.
Corner races
A corner race asks which team reaches a given number of corners first. Race to 7 means exactly that: the first side to win seven corners takes it, and if neither side gets there by full time, the "neither" option wins.
Why this bet works late
A side chasing the game commits players forward. That produces crosses, blocks and scrambles, and those produce corners. It is the outcome closest to the pressure you can see.
The reason I favour corners over goals is that a corner is a cheaper event. A goal can change the outcome of a match; a corner only needs the attacking to continue. When a team is camped in their defensive third, the team would much rather concede a corner than a goal. That does not mean goals are a bad bet, I still bet them plenty, as we will get to.
The five entry rules
If I were to give you five rules based on my performance and experimentation over the last five weeks, my best performing markets followed all five of these.
The core position
All figures from my own tracker, 4 August to 8 September 2026.
Where I have stayed consistent
The spine has not moved since week one: total corners, over, live match, two corners to find, at a price that beats our line. What changed is how much of my book it takes up.
Corners totals were 47 percent of my bets in the first two weeks, 52 percent across late August, and 61 percent in the first week of September. Corner races fell from 20 percent of my volume to 9. Nothing forced that. I kept a record, looked at it, and drifted toward what was working.
Stakes have stayed in a tight band throughout, with a median of 2 units and no increase after a losing day. 93 percent of my corner bets were in games level or within a goal. And I have taken the price only when it beat our line, which is what produces the 14.4 percent average edge the tracker reports.
What I am still experimenting with
I have been genuinely experimental, so I am not going to pretend otherwise with a list of things I never touch. These are live, with results attached, and some are costing me.
Open experiments
Open experiments as at 8 September 2026, all figures from my own tracker.
The pattern I have noticed in myself is that the experimental bets cluster on quieter nights. When there are ten qualifying matches, I can find my bread and butter. Remember that experimenting with different methods is fine, but you should have a core strategy.
The second bet: one more goal
About a quarter of my volume is total goals, over, in a match that needs exactly one more goal.
The pattern is narrow. Of my 83 goals-over bets, 76 needed exactly one more goal, placed around the 82nd minute at an average price of 2.90, returning 7.9 percent. Half the return of the corners position at a longer price, so a thinner edge. Same match states, same reason: a side pushing for an equaliser finds one more often than the pre-match numbers suggest.
What I stake
A median of 2 units, a range from 0.25 to 5, and no increase after a losing day. Size follows conviction in the setup, never the result of the last bet.
The core corners position averages 2.40 units. Goals average 1.50, because they are longer and swingier. Races average 1.22. That ordering does real work: the bet I understand best gets the most money, the experiments get about half. Sizing decides how expensive a lesson is before you know it is one, which is why I keep experimenting at small stakes rather than not at all.
What it returned
386 settled bets, 96.21 units up on 761 staked, a 12.6 percent return, winning 46 percent at an average price of 2.77. I lost more bets than I won and finished well ahead, which is the whole idea.
At an average price of 2.77 a bet needs to land about 36 percent of the time to break even. Landing 46 percent leaves the margin. It does not come from picking winners, because I do not pick more winners than losers. It comes from only taking prices that beat the line.
That corners line is the clearest evidence I have. 93 wins against 89 losses is as close to even as a market gets, and it still produced 54.13 units. The profit is in the price, not the prediction.
Before you copy any of it
The rules above are specific enough to follow, but variance means results are not guaranteed.
Look again at the curve at the top. For the first two and a half weeks I was up about 14 units, betting in ones and twos. Then on 22 August I had 59 settled bets and 26.44 units in a day, late August added two more big days, and 6 September added 19.75 from 50 bets. In between, a five-day stretch where the line went backwards. Most people who bet do not finish ahead, and this is one person's five weeks rather than a guide to what you should expect.
This app is not a tipster app. It is a tool to help you develop your own strategies, and hopefully you are now a bit clearer on how I have used it and excited to try developing your own. If you made it this far you should join our community Telegram channel so you can discuss and compare other strategies.
When betting stops being fun, or you find yourself chasing, that is the signal to stop. Support is at begambleaware.org.
I built LastMinutesAI and I use it myself during live matches. These guides are how I bet, not advice on what you should do. Every number comes from my own tracker, on the dates shown.