01
Understand
Find the real problem — not just its initial wording.
VIC·TORI
We create a form of human–AI thinking that can be tested against reality.
If you have a problem, we will solve it.
VIKTOR
TORI
AT WORK
BRING US A REAL PROBLEM
We solve it this way.
We start with the problem itself — not with a preferred tool, model or technology.
02
Reconstruct
Reveal the structure, constraints, signals and relationships that matter.
03
Hypothesize
Build competing explanations and possible solutions instead of committing to the first plausible one.
04
Build
Turn the strongest structure into the simplest working mechanism: algorithm, model, code or another concrete solution.
05
Test against reality
Measure what actually works. Keep what survives. Change what does not.
Then we continue until the problem is solved.Bring us: CONTACT
PUBLIC LIVE EXPERIMENT · ENG0
16 future matches.
Decisions frozen before kick-off.
Snapshot prepared
03 Sep 2026 · 18:44 EEST
Public track record is now live.
03 Sep 2026 · 18:44 EEST
Public track record is now live.
8 SINGLE
5 ANTI / double chance
3 PASS
8 RESOLVED
8 PENDING
SYSTEM is our final decision. OUTSIDE is a snapshot of public previews/models and market-facing narratives. STABILITY shows whether the ensemble keeps its direction when the current container is deliberately perturbed. Outside view is context, not ground truth.
04 SEPFINAL · 0–2
Ipswich — Liverpool
X2ANTI-H
OUTSIDEA
STABILITYUNSTABLE
Pair H↔A · T=K=A · REALITY A · ANTI-H SURVIVED · OUTSIDE EXACT
05 SEPFINAL · 1–1
Brentford — Sunderland
PASSPASS
OUTSIDEH
STABILITYPAIR STABLE A
T=K=H · Pair=A · REALITY D · PASS PRESERVED
05 SEPFINAL · 1–1
Brighton — Leeds
DSINGLE@3.45
OUTSIDEH
STABILITYSTABLE 3/3
System and outside view disagree · REALITY D · EXACT
05 SEPFINAL · 2–3
Fulham — Crystal Palace
HSINGLE@2.25
OUTSIDEH
STABILITYSTABLE 3/3
REALITY A · H FAILED · CAS+SOLVER=A SURVIVED
05 SEPFINAL · 2–2
Newcastle — Bournemouth
HSINGLE@2.19
OUTSIDEH
STABILITYSTABLE 3/3
REALITY D · H FAILED · STRUCTURE MISS
05 SEPFINAL · 0–0
Nottingham — Tottenham
12ANTI-D
OUTSIDEH / A
STABILITYUNSTABLE
Pair H→D→H · REALITY D · ANTI-D FAILED
06 SEPFINAL · 2–1
Arsenal — Chelsea
HSINGLE@1.68
OUTSIDEH
STABILITYSTABLE 3/3
REALITY H · EXACT H · Arsenal came from 0–1 down to win 2–1.
06 SEPFINAL · 2–2
Everton — Man Utd
ASINGLE@2.20
OUTSIDEA / MIXED
STABILITYSTABLE 3/3
REALITY D · A FAILED · ⚠ Equalised at 90+6′ (final minute of stoppage time).
12 SEP
Aston Villa — Nottingham
HSINGLE@1.86
OUTSIDEH
STABILITYSTABLE 3/3
12 SEP
Bournemouth — Brentford
12ANTI-D
OUTSIDEH
STABILITYUNSTABLE
Pair H→D→H
12 SEP
Crystal Palace — Ipswich
1XANTI-A
OUTSIDEH
STABILITYUNSTABLE
Pair D→D→H
12 SEP
Liverpool — Fulham
PASSPASS
OUTSIDEH
STABILITYUNSTABLE
Three competing internal directions
12 SEP
Sunderland — Arsenal
ASINGLE@1.45
OUTSIDEA
STABILITYSTABLE 3/3
12 SEP
Tottenham — Everton
PASSPASS
OUTSIDEMIXED
STABILITYUNSTABLE
Pair D→D→H
13 SEP
Man Utd — Man City
DSINGLE@3.50
OUTSIDEA
STABILITYSTABLE 3/3
Stable system decision against market favorite
14 SEP
Leeds — Newcastle
12ANTI-D
OUTSIDEH / D
STABILITYUNSTABLE
Pair H↔D
AFTER REALITY · PROTOCOL FROZEN BEFORE RESULTS
Reality Debrief.
The prediction is only the first half. After each match we compare what the system saw, what the outside field said, what action was allowed — and what reality returned.
Evaluation protocol fixed
03 Sep 2026 · before first kick-off
Results will fill this page; the rules will not move.
03 Sep 2026 · before first kick-off
Results will fill this page; the rules will not move.
01
Before
Frozen public snapshot. No result is known to the system.
02
Decision
SINGLE, ANTI or PASS — each is a different claim and is judged differently.
03
Outside
Public previews, models and market-facing narratives remain visible as an independent field.
04
Reality
Actual result. No explanation can replace it.
05
Debrief
What survived, what failed, and what the next system state should learn.
NO POST-HOC REWRITE. The original decision, outside view and stability state stay in the record. A loss is data, not something to edit away.
THREE DIFFERENT CLAIMS · THREE DIFFERENT TESTS
SINGLE
We acted on one outcome.
After the match we ask whether the chosen direction survived reality, what the financial outcome was, and whether stability or disagreement with the outside field carried useful information.
ANTI
We excluded one outcome.
ANTI is not a hidden exact-score prediction. Its claim is weaker and different: one outcome was not admissible. It is judged by whether that excluded outcome occurred, separately from SINGLE.
PASS
We refused to force a decision.
PASS is a decision about admissibility before the result, not a failed prediction. The later winner does not automatically make PASS wrong. We ask whether the pre-match internal conflict justified no action.
OUTSIDE FIELD ANALYSIS
Another measurement.
Not another oracle.
OUTSIDE is a snapshot of previews, public models, market-facing narratives and consensus available before kick-off. We keep it visible because agreement, conflict and ambiguity between the system and the outside field may themselves contain information.
Where did the outside field see reality better than the system?
Where did the system survive against the dominant outside direction?
Was disagreement itself informative — or only noise?
When did agreement add confidence, and when was it merely shared error?
Outside field · full frozen snapshot.
ALL 16 MATCHES · BEFORE RESULTSThis is the complete comparison field, not a selection of convenient examples. For ANTI, “compatible” means the outside direction does not contradict the excluded outcome. PASS remains a refusal to authorize action, so no directional agreement is claimed.
| Date | Match | System | Outside | Relation | Stability |
|---|---|---|---|---|---|
| 04 SEP | Ipswich — Liverpool | X2ANTI-H | A | COMPATIBLE | UNSTABLE |
| 05 SEP | Brentford — Sunderland | PASSNO ACTION | H | NO ACTION | PAIR STABLE A |
| 05 SEP | Brighton — Leeds | DSINGLE | H | CONFLICT | STABLE 3/3 |
| 05 SEP | Fulham — Crystal Palace | HSINGLE | H | AGREE | STABLE 3/3 |
| 05 SEP | Newcastle — Bournemouth | HSINGLE | H | AGREE | STABLE 3/3 |
| 05 SEP | Nottingham — Tottenham | 12ANTI-D | H / A | COMPATIBLE | UNSTABLE |
| 06 SEP | Arsenal — Chelsea | HSINGLE | H | AGREE | STABLE 3/3 |
| 06 SEP | Everton — Man Utd | ASINGLE | A / MIXED | MIXED | STABLE 3/3 |
| 12 SEP | Aston Villa — Nottingham | HSINGLE | H | AGREE | STABLE 3/3 |
| 12 SEP | Bournemouth — Brentford | 12ANTI-D | H | COMPATIBLE | UNSTABLE |
| 12 SEP | Crystal Palace — Ipswich | 1XANTI-A | H | COMPATIBLE | UNSTABLE |
| 12 SEP | Liverpool — Fulham | PASSNO ACTION | H | NO ACTION | UNSTABLE |
| 12 SEP | Sunderland — Arsenal | ASINGLE | A | AGREE | STABLE 3/3 |
| 12 SEP | Tottenham — Everton | PASSNO ACTION | MIXED | NO ACTION | UNSTABLE |
| 13 SEP | Man Utd — Man City | DSINGLE | A | CONFLICT | STABLE 3/3 |
| 14 SEP | Leeds — Newcastle | 12ANTI-D | H / D | MIXED | UNSTABLE |
PASS is part of the experiment.
3 REFUSALS TO FORCE A DECISION05 SEP · PASS
Brentford — Sunderland
The Pair direction was stable on A, but independent pre-match evidence pointed strongly the other way. Stability alone was not enough to authorize action.
Frozen reason: a stable but isolated direction is not automatically an admissible bet.
12 SEP · PASS
Liverpool — Fulham
Three competing internal directions were present. Selecting one would have hidden the conflict rather than resolved it.
Frozen reason: unresolved internal competition made action inadmissible.
12 SEP · PASS
Tottenham — Everton
The ensemble was unstable and the independent evidence did not collapse to one direction. The system refused to manufacture certainty.
Frozen reason: multiple live hypotheses remained; no single direction earned authority.
Full Reality Debrief.
8 RESOLVED · 8 PENDING · RESULTS FILLED IN PLACEAfter each final whistle, the original columns remain unchanged. Only REALITY and DEBRIEF are added. This keeps wins, losses, abstentions and outside-field errors in one audit trail.
| Date | Match | Decision | Outside | Stability | Reality | Debrief |
|---|---|---|---|---|---|---|
| 04 SEP | Ipswich — Liverpool | X2ANTI-H | A | UNSTABLE | A · 0–2 | ANTI-H survived reality. Outside A was exact. Instability limited the strength of the claim rather than invalidating the exclusion. |
| 05 SEP | Brentford — Sunderland | PASS | H | PAIR STABLE A | D · 1–1 | PASS preserved. Stable Pair=A was not enough to authorize action; reality landed on D. |
| 05 SEP | Brighton — Leeds | DSINGLE | H | STABLE 3/3 | D · 1–1 | Exact D. Stable disagreement with outside H carried information in this case. |
| 05 SEP | Fulham — Crystal Palace | HSINGLE | H | STABLE 3/3 | A · 2–3 | H was a real match trajectory, but CAS+Solver=A survived reality. Stable Pair H should not silence aligned corrective opposition. |
| 05 SEP | Newcastle — Bournemouth | HSINGLE | H | STABLE 3/3 | D · 2–2 | Structure miss. The match trajectory was closer to A/D; Newcastle recovered the draw rather than confirming H. |
| 05 SEP | Nottingham — Tottenham | 12ANTI-D | H / A | UNSTABLE | D · 0–0 | ANTI-D failed. D was not merely an unstable intermediate state; it was the realized outcome. |
| 06 SEP | Arsenal — Chelsea | HSINGLE | H | STABLE 3/3 | H · 2–1 | Exact H. Arsenal came from 0–1 down to win 2–1. System, outside H and STABLE 3/3 aligned, and reality confirmed the consensus. |
| 06 SEP | Everton — Man Utd | ASINGLE | A / MIXED | STABLE 3/3 | D · 2–2 | A failed as the final outcome. Man Utd led twice; Everton equalised at 90+6′, the final minute of stoppage time. Stable A captured a real match trajectory, while OUTSIDE MIXED remained a meaningful warning. |
| 12 SEP | Aston Villa — Nottingham | HSINGLE | H | STABLE 3/3 | PENDING | PENDING |
| 12 SEP | Bournemouth — Brentford | 12ANTI-D | H | UNSTABLE | PENDING | PENDING |
| 12 SEP | Crystal Palace — Ipswich | 1XANTI-A | H | UNSTABLE | PENDING | PENDING |
| 12 SEP | Liverpool — Fulham | PASS | H | UNSTABLE | PENDING | PENDING |
| 12 SEP | Sunderland — Arsenal | ASINGLE | A | STABLE 3/3 | PENDING | PENDING |
| 12 SEP | Tottenham — Everton | PASS | MIXED | UNSTABLE | PENDING | PENDING |
| 13 SEP | Man Utd — Man City | DSINGLE | A | STABLE 3/3 | PENDING | PENDING |
| 14 SEP | Leeds — Newcastle | 12ANTI-D | H / D | UNSTABLE | PENDING | PENDING |
Three highlighted questions.
SELECTED BEFORE RESULTS · NOT THE WHOLE ANALYSIS05 SEP
SINGLE · SYSTEM ≠ OUTSIDE
Brighton — Leeds
SYSTEMD · SINGLE
OUTSIDEH
STABILITYSTABLE 3/3
After reality: did a stable disagreement carry information, or did stability preserve the wrong decision?
RESOLVED · D EXACT
12 SEP
ANTI · EXCLUSION CLAIM
Crystal Palace — Ipswich
SYSTEM1X · ANTI-A
OUTSIDEH
STABILITYUNSTABLE
After reality: was excluding A justified even though the ensemble direction itself was unstable?
PENDING
12 SEP
PASS · ACTION NOT ADMISSIBLE
Liverpool — Fulham
SYSTEMPASS
OUTSIDEH
STABILITYUNSTABLE
After reality: was abstention better than forcing one direction when three internal directions competed?
PENDING
The goal is not to rescue the system after a loss. The goal is to learn when the system is allowed to act.
Contact
Viktor
If you're ready for a direct conversation. Sometimes more direct than expected.
Tori
If you want to see the problem from another angle. Sometimes a completely different one.