An official box score gives you a record of what happened on a baseball field.
It does not automatically validate the rating, scouting grade or recruitment recommendation your software builds from it.
The federation recorded the hit. Responsibility for the interpretation belongs to you.
I care about that distinction because I have put a substantial part of my life into building on the other side of it.
I started Statbase during COVID. I was a sports data analyst, a former baseball player, and someone who wanted to do something useful with those two parts of my background. For a while, it was a hobby project. I maintained it, updated it and moved it forward when I had time.
The foundations were already there. For years, European baseball had been building an extensive public record of itself: results, box scores, rosters and season statistics, spread across federation websites, competition pages and archives.
Behind those records were scorekeepers, volunteers and federation staff doing work that deserves recognition. There were also enthusiasts following the updates, connecting the dots and producing their own research. Anyone building in this space today owes something to that groundwork.
What interested me was how much more we could learn from it.
Those archives contained years of player development, changing roles, workloads and team trajectories. With enough care, they could help answer questions that mattered well beyond the final score.
I saw a gap between the richness of that record and the tools available to the people it described. A young athlete could have several seasons of publicly recorded performances and still struggle to turn them into a coherent account of their development. A coach could have access to the statistics and still face substantial work before they became useful for a decision.
That possibility stayed with me as I maintained Statbase.
The turning point came during the Paris Olympics in 2024, after meeting another sports analyst. I came away with the uncomfortable feeling that we were still facing too many of the same problems. Los Angeles 2028 was already approaching, and I wanted us to build something useful before another opportunity passed.
That conversation marked the beginning of Statbase’s transformation into the much larger undertaking it is today.
We fund it entirely ourselves. The total commitment, in money and time, is not something I find particularly relaxing to think about. We have gone all-in because we believe in the future of baseball in Europe. As former players, we want younger athletes to have tools we would have valued decades ago.
We saw an opportunity worth committing our own money and years of work to. Building something dependable on top of that shared groundwork became the responsibility we chose to take on.
That explains our commitment. The quality of our analysis still has to be demonstrated.
A recent meeting with a federation representative prompted this article. It made me think again about what a technology provider should be able to defend before asking for the sport’s institutional trust.
My answer starts with a distinction that deserves considerably more attention.
“The data is official” answers a question about the source. It leaves the questions about your model unanswered.
Once you transform those records into an evaluation, you introduce decisions. You choose the inputs, weights, comparison groups and minimum samples. You decide what to do with missing information. You decide whether a number describes past production, current form or a claim about future ability.
The source website cannot take responsibility for those choices.
Call a number, for example, WAR (fWAR, bWAR, rWAR or whatever philosophy you prefer – we have none of it), and you should be able to explain the replacement-level baseline and the assumptions behind it. Claim to project performance in a stronger league, and you should be able to show how that projection has been tested. Describe a lineup as optimized, and you should be able to explain the objective and constraints.
These are ordinary questions about the product being sold.
Familiar baseball terminology does not answer them. A precise-looking number does not answer them.
At Statbase, even establishing that the imported records agree with the source requires substantial work.
Our internal season-close audit for Belgium and Croatia covered 207 games and 16,645 plate appearances. We reconciled 9,734 player-level counting fields against official game boxes, checked pitching workload in outs, resolved identities and listed scheduled entries excluded because they lacked final box scores.
Those figures have a precise meaning. They describe checks on the consistency of our imported records.
They do not prove that every source entry was correct. They do not validate every player evaluation we might build afterward.
Our audit still labels local run-value calibration for those competitions provisional. That qualification belongs beside the results. Agreement with a box score cannot be stretched into evidence for an unrelated modelling claim.
We cannot borrow more credibility from our own audit than it earned. Neither can anyone else.
A proprietary model can retain its code and still explain its purpose, assumptions, validation and limitations. Coaches do not need your entire repository to ask why a player received a particular assessment. A federation does not need to own your algorithm to expect evidence that it is suitable for the proposed use.
“Proprietary” should describe how a company protects its work. It should never become an escape route from scrutiny.
There is a human reason to insist on this.
Behind a player profile is someone training, competing and trying to create an opportunity. An evaluation may influence who receives attention, who gets another look and what a coach decides to work on. The people affected should be able to understand the basis of that judgement.
European baseball’s limited resources make that responsibility more pressing. Clubs have little time to waste. Federations have competing demands on their budgets. Players have short windows in which to develop and be noticed.
A small market cannot become an excuse for small standards.
I want Statbase to earn trust through work that can withstand difficult questions. That includes questions that reveal a weakness, force a correction or show that a claim needs to be narrower.
Our investment does not exempt us. Our intentions do not exempt us. Our enthusiasm for the game does not exempt us.
I started this project because I believed we could give athletes something useful. That remains the reason to keep building—and to take responsibility for what we publish about them.
If “the data is official” is your strongest defence, you have not yet defended your analysis.
The figures above refer to Statbase’s internal Belgium–Croatia season audit, verified on 6 October 2026, against the official Belgian D1 and Croatian Open League records.