This year, we had to revise a club’s plan because the players available to its coaches had changed.
The games were still on the calendar. The staff still had to prepare. An assessment built around the previous roster could no longer answer the question in front of them.
We had to reconsider catching coverage, pitching allocation and how to build an offense from the players who remained available.
That is a useful moment to introduce into a baseball technology demonstration.
Run the demo again. This time, remove the catcher.
Keep the same fixture list. Respect the registration deadline. Give the staff the actual remaining pitching options. Then show what your recommendation changes, what it sacrifices and why.
I would learn considerably more from that than from another slide announcing a revolution in baseball analytics.
It also explains how I see the year we have spent building Statbase.
Even though the club’s identity is confidential, I can describe the work without publishing its players, opponents or specific competitive circumstances. What matters here is the sequence: an initial assessment became a preparation plan; new information forced changes; the results forced a review; that review became work for the following season.
Our tools have to be useful along that sequence. A feature list gives very little indication of whether they are.
Before the roster changed, we had already examined what the club could reasonably expect from it. That meant bringing previous and current performance together, accounting for competitive context, and separating established contributions from small samples. A promising rate over a few appearances needed a different degree of confidence from a role supported by sustained workload.
That work relies on the player histories we organize in StatIndex. This year, we have continued building out those records and the tools for examining them: career and season views, comparisons, league percentiles, recent form and scouting analysis.
For the club, those were the beginnings of an answer. The next step was to examine how the individual contributions fitted together.
PVI helped frame performance value in its league context. PFI added game impact and recent form. TPM supplied a team-strength view. Roster analysis then examined where the production and workload were concentrated, and what the options looked like behind the leading players.
That last part matters when a name disappears from the lineup.
A team average can remain respectable while a particular role becomes very difficult to fill. The cost can spread: covering one absence changes the defensive arrangement, which changes the available lineup, which changes how much run support the pitching plan may require.
There are various baseball & baseball related software, which are already being advertised with automatic lineup optimization based on statistical averages and WAR. That is a sporting claim worth demonstrating under these conditions.
Mathematically and statistically speaking, adding individual WAR values produces the same total when the same hitters are reordered. Any system using that information to recommend a batting order therefore needs to show what further objective and relationships distinguish the alternatives. It also needs to respect who can play, where they can play and what the staff needs across the schedule.
“Optimized” is an obligation to explain the objective. It is a very expensive word to use when you mean “sorted.”
In our club work, the task extended across games. Committing a pitcher to one opportunity affected the resources available for the next. A lineup that looked stronger in isolation could create a catching or workload problem elsewhere in the series.
We used StatPredict, running on the metrics and inputs behind our analysis, to examine those alternatives. Opponent preparation supplied the other side: likely pitching matchups, offensive strengths, contact tendencies where supported, and the parts of the game that deserved closer attention.
The resulting work included different pitching and lineup scenarios, coaching briefs, game cards and proposed decision triggers. We examined when to spend a scarce resource, when to preserve it, and what new information should cause the staff to reassess.
Then availability changed, and we went back through the plan.
We assessed the contribution being lost, the internal alternatives and the constraints on bringing someone else in. The staff needed choices they could actually implement. A recruitment recommendation for a player they could not register would have left the immediate problem untouched.
A list of better players is easy to produce. A workable plan for the players a coach actually has is where the analysis earns its place.
As the season continued, we updated the assessment against what happened. We examined the games, the remaining possibilities and the consequences of the resources already used.
We also had to revisit our own assumptions. Some of the early expectations were too generous. The review examined where individual performances, roster depth and the competitive environment had been misread, and which weaknesses required a different approach.
I consider that part of the product. If we publish an assessment, the work includes returning when the evidence makes the original answer uncomfortable.
Our public forecasting work put that discipline through different tests. Before the playoffs had resolved, StatPredict’s leading final pairings were Bonn–Regensburg in Germany and Neptunus–Amsterdam in the Netherlands. Both became the actual finals. The engine had to account for different rosters, competition structures and paths through a bracket.
In France, combining the French D1 and Challenge de France cohorts in our September 7 internal audit gives 68 correct winner calls from 93 completed games: 73.1%. That scorecard brings together league forecasts and rolling cup forecasts; the final-pairing results are recorded separately from game-level accuracy.
Those forecasts gave us something to return to when the games were over: where the model held up, where it missed, and which assumptions deserved another look. That habit of checking the answer matters just as much when the question comes privately from a coaching staff.
Publishing the forecast before the game gives the claim a deadline.
The same principle applies to the club engagement: preparing the analysis, making a recommendation and establishing that it improved an outcome are separate achievements. The coaching decisions remained with the staff. We had to be precise about what our work could demonstrate.
Eventually, the review moved into the following season. We prepared a blueprint around role priorities, player-specific development, recruitment criteria and checkpoints for assessing progress.
That changed the recruitment question. We could describe the contribution a signing needed to provide, the depth it needed to protect and the opportunity the club could offer in return.
It is the same two-sided question behind Fit Match and the Find My Club route. A player’s evidence must make sense for the destination, and the proposed role must make sense for the player. Cross-league comparisons need their assumptions and uncertainty visible; a screening estimate starts an investigation, while availability, eligibility, budget and direct conversations determine what is feasible.
Audit My Club, which we have been building, follows the organizational side of that sequence: diagnose the roster, agree the priority with authorized staff, build a meaningful recruitment brief, then review potential options. The club’s context belongs in the calculation from the beginning.
The following-season plan also required decisions about the players already there. Development priorities in the INDEX connect an assessment to specific practice suggestions and benchmarks a coach can review. A player can work toward a defined role, with a clearer basis for the next evaluation.
The Baseball CV helps carry that work beyond the immediate conversation. Performance records, charts, analysis, club experience, confirmed national-team selections and relevant video can travel together, with a route back to the live profile. A prospective coach gets something to investigate, and the athlete gains a more coherent account of their progress.
This is where I want the word “visibility” to become concrete. An imported profile, a person actively using a platform and a genuine recruiting opportunity are three different things. Organizations should expect providers to distinguish them when presenting their numbers.
Our current public footprint covers eleven leagues, more than 4,000 player profiles and 124 clubs. Those are coverage figures. They establish the scope of the records we work with. The usefulness of what we build from them has to be demonstrated through the work.
Some of the questions raised by player assessment also reach beyond those records. A coach may need to inspect how a pitch was delivered, where it crossed the plate or whether a pattern persists across games. A season line cannot supply a measurement the source never recorded.
That boundary is one reason Diamond Vision belongs in the same ecosystem.
Available to selected early-access participants, its working video-review experience connects pitch replay with ball tracking, location charts and model-inferred estimates of release, movement and plate crossing. Player views bring observations together across games, including estimated pitch mix and release patterns.
A coach can return to the footage behind an observation and examine it in context. The estimates carry their own requirements for footage quality, calibration and validation. Building the dependable connection between that evidence and the wider INDEX remains continuing work.
When the evidence is missing, the next job is to obtain it or define what can reasonably be estimated. Adding another decimal place will not do either.
The club engagement showed how many connected questions one organization can face. A federation faces related questions across an entire competition and over several seasons.
It may need to follow players through changing clubs, understand where younger athletes receive meaningful playing time, or identify people worth examining for a national-team camp, with eligibility checked separately. Our public youth-roster and national-team analyses have already explored those uses of the records.
Making the underlying information reusable also gives member clubs a better starting point. The scorekeepers and federation staff have already done substantial work to record the sport. A technology partner should extend its usefulness to the next coach, athlete and season.
That is a sensible basis for an institutional evaluation. Give providers an actual task. Include incomplete records, a small sample and a change in availability. Ask them to explain their conclusions, revise the recommendation and show where the staff still needs observation or better information.
A federation can lend a company credibility. It should first make that company demonstrate something useful to its coaches.
League dossiers, research reports, player cards, season awards and public stories help put the resulting work in front of people. They can bring attention to a performance and lead a reader toward the underlying evidence. That gives clubs and federations material with which to recognize athletes and explain their competitions to supporters and partners.
But the person who follows the link should arrive somewhere useful. There should be a record to examine, a comparison to investigate, or a better question to bring to a coach. That is how the public-facing work connects back to development, scouting and preparation.
This year has moved Statbase further along that path. The INDEX and our reports already provide working tools and evidence. Other capabilities are available through early access or scoped analytical work; club workflows and the connections between video and the wider platform continue to develop. We need to be clear about those stages when agreeing an engagement.
For a club, the potential value begins with scarce resources: staff time, pitching, playing opportunities and the recruitment budget. For a federation, it includes making the records it already holds useful across more clubs and more seasons. A pilot should have a specific job and a result against which to judge it.
I am putting my own money, years of work and professional reputation into building something that can keep answering those questions. The scope has grown because the next problem keeps arriving: a changed roster, a different opponent, a player ready for another role, a gap in the evidence.
That is the story behind the ecosystem we have spent the year building.
The first assessment mattered. So did the revised plan. So did the review that changed our thinking about the next season.
Bring the real roster. Change the assumptions. Make us show the work.
