Twenty years looking
chance in the eye

It all started long before ProbaMax - and really, long before me. In the 1990s, my father set out to analyze lottery draws. Not out of superstition, but out of engineering curiosity. He wrote his first programs with the tools of the time: GW-BASIC, then VBA in Excel. Heavy macros, frequency tables calculated by hand, nested loops on machines that took minutes to run.

It was not magic. It was rigor. The idea that data - even in the apparent disorder of chance - deserves to be examined seriously and methodically.

I grew up in that environment. At six, while others played outside, I watched my father code - and I wanted to try. Our family machine was an Amstrad 1512, and the language was GW-BASIC. My first lines of code drew colored lines on screen - and somewhere in the loop, one sentence kept appearing: "Sorry Mom". I had done something silly and had been sent to my room. In the late 80s, though, no one was worried about screen addiction yet - and my room had the Amstrad in it. Today, when a parent punishes a child, they remove console, internet, and phone. Back then, punishment ended with a GW-BASIC program and a coded apology. Not a "Hello World." An apology program at six years old. The tone was set.

I completed a software engineering degree. I earned a perfect 20/20 in Software Engineering - the top grade, in the subject that mattered most to me. That 20/20 came from a special exam project: a Tetris fully rebuilt in Java, initially designed for a non-profit hospital multimedia room. The idea was great on paper. In practice, I never delivered it - in 2003, games built by full studios were obviously far more polished than my learning project, and I knew it. But the examiner graded the code. And the code was solid.

Meanwhile, the probabilistic analysis project inherited from my father kept running in my mind. Languages changed. The core ideas stayed.

Since 2003, the project evolved alongside my career, my life, and the technology stack. What started in VBA became Java, then Python. Databases replaced Excel sheets. Probabilistic models became sharper. The combination-generation algorithm was rewritten many times - each version more rigorous than the last, each rewrite driven by dissatisfaction with what already existed.

If I am honest, though, this obsession is much older than 2003. Back then my father played 200 francs per draw - a serious amount. My uncle did too. I grew up in a home where lottery was both analytical subject and regular practice, where results were tracked and streaks discussed. It was not compulsive gambling - it was a deep ritual-like passion. I watched, absorbed, asked questions. The obsession did not begin with my first program. It was already there.

The guiding thread never changed: analyze historical draw data honestly, without promising the impossible, while building tools that respect both mathematics and the intelligence of those using them.

ProbaMax as it exists today is not something I could have built alone. With David, my friend and fellow developer, the project reached another level. We have long shared the same conviction: probabilistic analysis tools should exist beyond local scripts and hacked spreadsheets. They deserve a real platform, polished interfaces, and an experience accessible to everyone - not just developers who can read Python.

I bring product vision, algorithms, and a long-standing obsession rooted in family history. David brings system architecture, robustness, and the engineering precision that makes a platform last. Together, we built what neither of us could have built alone.

ProbaMax is the result of all that: a family legacy turned into a rigorous platform, decades of algorithms distilled into interactive studies, and a passion for analysis made accessible to anyone who plays the lottery and wants to understand what they are doing.

The platform combines in-depth statistical studies on historical EuroMillions, French Loto, and other lotteries with a combination generator powered by a proprietary algorithm I have refined for more than twenty years. Every feature is designed to stay honest: I show you the data, explain probabilities, and promise nothing else.