Champions are not born on the stage —
they are honed before it, in a thousand rounds no one sees.
Qatar Debate trains judges and coaches through capacity-building programmes — twelve intensive hours, five editions, forty participants from fifty institutions. Then they return to their universities, and the standard lives only in their memory. Saql stays with them: a continuous Arabic adjudication reference, assessing as they were trained to assess.
This was built specifically for Qatar Debate Center — not as a public offering, but as a concept for how your programme can extend across the full year.
Twelve hours of training. Then a year with no reference.
Your championships span Doha to Chicago, New York, Illinois, London and Kuala Lumpur. Six universities in the national championship alone; a national team facing eighty countries at the World Championship. The training team: one, based in Doha.
When the capacity-building programme ends, the trained judge returns to their university carrying what they memorised in three days. No reference to return to when they doubt, no standard to measure against when they disagree with a colleague, no drill to anchor what they learned before the next championship.
The gap is not in the training — the training is excellent. The gap is in what comes after.
How Saql adjudicates
An example from the first exploratory run: one test debate, assessed by a human judge and by the engine with the same anchors, indicator by indicator for each argument. This is not the transcript — it is what the judge sees, and what the two assessments are compared on:
حجتا المناظرة التجريبية: «المنع يرفع التركيز» و«الهاتف أداة تعليمية»
The two arguments of the test debate, quoted as recorded in Arabic.
Before any score, the judge sees the behavioural anchor for each indicator. This is the panel for claim scope (AC5); anchor texts are quoted verbatim from the approved Arabic rubric:
هل يتناسب نطاق الادعاء ودرجة اليقين والشروط مع قوة ما يدعمه؟
- 0ادعاء مطلق، والدعم لا يبرر إلا جزءًا صغيرًا منه.
- 1النطاق أوسع بوضوح من الدعم، مع تحفظ شكلي فقط.
- 2متناسب في جانب وغير متناسب في آخر.
- 3متناسب عمومًا، مع تجاوز ثانوي.
- 4النطاق والشروط مطابقة لما يحمله الدعم.
قاعدة اللغة الحذرة وحدها لا ترفع الدرجة، والادعاء غير المقيَّد المدعوم بالكامل يأخذ 4.
Illustrative rendering of the system interface.
Each assessment is shown as a heatmap: rows are the debate's arguments, columns the indicators. This is how the two arguments came out in the human and the engine assessments:
| AC1 | AC2 | AC3 | AC4 | AC5 | AC6 | RQ2 | RQ3 | RQ4 | RQ5 | CR | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| المنع يرفع التركيز | 4 | 2 | 2 | 0 | 1 | 0 | 1 | 1 | 4 | 1 | 4 |
| الهاتف أداة تعليمية | 3 | 0 | 0 | — | 2 | 0 | 2 | — | 4 | 2 | 3 |
| AC1 | AC2 | AC3 | AC4 | AC5 | AC6 | RQ2 | RQ3 | RQ4 | RQ5 | CR | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| المنع يرفع التركيز | 4 | 2 | 2 | 1 | 0 | 0 | 2 | 1 | 4 | 1 | 3 |
| الهاتف أداة تعليمية | 3 | 0 | 0 | 0 | 1 | 0 | 2 | 0 | 4 | 1 | 3 |
Illustrative rendering of the system interface, with data from the first exploratory run.
When the two assessors disagree on a score, the disagreement is shown in one card with both scores and each side's rationale (rationales quoted verbatim in Arabic):
«نطاق الادعاء أوسع بوضوح من الدعم: كل المدارس مقابل تجربة مدرسة واحدة.»
«طُرح الادعاء كما هو مطلقًا: «يجب منع» الهواتف «داخل المدارس» دون تحديد نطاق أو شروط أو درجة يقين، فهو يشمل كل المدارس وكل…»
The anchor applied by both: the AC5 panel above.
Illustrative rendering of the system interface, with data from the first exploratory run.
Both assessors saw the same flaw — a claim stretched over limited evidence — and differed only in its severity. This is what makes every disagreement diagnosable: either one of them applied the anchor imprecisely, and the rationale shows it, or the anchor itself needs tuning.
In this run, the human and engine assessments matched exactly in 62% of pairs, and 94% fell within one point. An exploratory result from a single debate, not a calibrated measurement — and the decision, in every case, is the judge's.
Human judgment is always final. Saql organises, documents and trains — it crowns no one.
Three forms Saql takes
Judges' Reference
A trained judge dissects a real debate, then compares their assessment to Saql's methodological evaluation — seeing where their judgement diverged and why. Practice through application, without requiring an expert present.
The Permanent Opponent
A team in Missouri or Kuala Lumpur submits a motion, Saql builds the full opposing case in Arabic, debates them, and delivers an immediate assessment with their weakest point. No travel, no waiting for the annual visit.
The Sealed Archive
Every championship, every motion, every adjudication — sealed in permanent, indelible documentation. An archive of contemporary Arabic debate, accumulating in the Centre's name, generation after generation.
What it extends for the Centre
Permanent presence, not episodic
Your presence today is tied to championships and visits. Saql makes you present in every team's daily practice, in every country you have reached.
Consistency in adjudication
Dozens of judges, dozens of temperaments — and a single methodological draft each reviews and decides against. Championship quality rises without touching the judge's standing.
Arabic-language learners in particular
The teams debating in Arabic as they learn it — whose participation was the defining moment of your last championship — are the most in need of continuous classical Arabic practice, and the least served by it.
Leadership no one else holds
The first AI adjudication training infrastructure in Classical Arabic — from Doha. A language for which these tools were not built, and you are best placed to build them.
From al-mirāʾ — the disciplined disputation our predecessors knew — to a Saql worthy of our tools.
Built by Abd Al-Rahman Yazouri, a developer from Gaza —
who built Fasil to restore to Arab debate its dignity, and to argument its weight.
"Truth emerges where minds meet."
What do you think?
Is Saql worth extending your capacity-building programme across the full year?
abd@quotefather.com