The Realised Value Engine™
Right player.
Wrong team.
The most expensive mistake isn’t a bad signing.
It’s a good one who doesn’t fit.
1,904 players · 98 clubs · 186,592 scored
02 / What it computes
A player is only as good as the system.
The same player thrives in one team and fails in another. The same draft pick transforms one roster and is wasted on the next. The player has not changed. The system has. Dextant scores how well any athlete fits any system, and works out what that fit is worth.
Fig.02 / The engine
Input
System + player
Engine
Sealed engine
Output
Realised value, 0–1
What goes in, and what comes out. How it computes stays inside.
We publish results, not methods.
03 / The method
Declare a system.
Score the fit. Find the optimum.
01
Declare
Describe how a system plays: what it demands of the people inside it.
02
Decompose
Every candidate is scored against it in detail, not as a single number.
03
Optimise
The engine finds the system where the athlete's value is highest.
Clubs see the score and the reasons. The computation never leaves our servers.
04 / Live
One engine.
Every athlete.
Every system.
Give the engine two systems and the same roster ranks differently in each. Built first in football, where value is priced in public every week, the method reads any sport that keeps a record. The value a system captures is computed, not argued.
Fig.03 / System re-sort
- 1P-07
- 2P-19
- 3P-31
- 4P-15
- 5P-23
- 6P-11
- 7P-02
- 8P-04
- 9P-08
- 10P-27
- 1P-11
- 2P-02
- 3P-07
- 4P-27
- 5P-23
- 6P-15
- 7P-31
- 8P-04
- 9P-19
- 10P-08
Fig.03 · re-rank under a second declared system. 8 of 10 positions move. Illustrative.
05 / What it reads
Fit is not one number.
A player is not good or bad. They are strong at some things and weak at others, and a system leans on some of them and not the rest. Every sport asks for a different mix. The engine reads a candidate across the parts that matter to the system in front of them, not as a single number.
Fig.04 / Fit decomposition (illustrative)
06 / Applications
Every roster decision
is a bet on fit.
Transfers, drafts, line-ups, contracts. Billions move on whether talent fits a system. Dextant prices that bet before it is made.
A transfer prices the player. It should price the fit. Dextant scores every candidate against the way the buying side actually plays, so the fee is spent on output the system can use, not on a reputation built somewhere else.
A draft pick is a bet that a young athlete will translate. Translation depends on the system they land in. Dextant ranks a class by what each prospect would add to the roster as it is, not in the abstract.
Most rosters are collections. A system is a set of demands. Dextant reads the gap between the two: where the current squad leaves the system short, and which profile closes it.
Extend, sell, or let it run down. The call depends on what the athlete is worth to this system, now. Dextant puts a number on realised value so the decision is priced, not argued.
07 / Method and limits
What we hold back,
and what we do not.
The scoring method is a trade secret. The limits are not. It runs in one sport so far. The engine is bounded, and its bounds are stated to the people we work with. What it does not yet do is named, not hidden.
Not this
Dextant does not predict injuries. It is not a scouting database.
The null result
We tested whether the engine’s reads predict what happens when a player is absent. They do not. 50.9% against a 53.4% matched baseline, across 3,289 tested cases. Teams re-route. We publish that.
What it refuses
It will not compare candidates across roles asked to do different things. It will not estimate what it cannot observe. It will not print a score beside a named, identifiable player in public. A blank is a real output, and rival tools never return one.
The record
What has been tested, on what, and what it showed, is written down and dated. It is not published. It opens to a technical reviewer under confidentiality. Validation.
The engine, its scoring method and its outputs are proprietary and protected as a trade secret.
Appendix / Definitions
- System
- How a team plays. What it asks of the people inside it.
- Fit
- How well a candidate meets what the system asks.
- Realised value
- How much of a candidate a system turns into performance.
- Optimum
- The system that realises the most value from a candidate.
Figures
- Fig.01Demonstration surface
- Fig.02The engine
- Fig.03System re-sort
- Fig.04Fit decomposition
- Fig.05Section through the value surface, long axis
- Fig.06Section through the value surface, transverse
08 / Access
The value is already there.
Dextant computes where it is highest. Access is by introduction.
Who
Adrian Goransch
Co-founder.
UEFA B licensed coach.
Knows this market from the inside.
Oliver Cormack
Co-founder.
Built Dextant end to end.
Company in formation · 2026