Methodology — Pillar VI of VII
Conviction, in a number that can't be edited
All the argument — four pillars of evidence, a weighted bear case — lands as one number: the Gladiator Score, 0 to 100. It is published before the outcome is known, and it never changes afterwards. Here is how it is built, and why the lock matters more than the number.
How the score is built
The four research pillars — fundamentals quality, valuation, momentum in context, balance-sheet risk — each contribute evidence, and the bear case pushes back against all of it. What survives that collision becomes the Gladiator Score: a 0-to-100 expression of how much conviction this desk has in the pick. Higher means the evidence stacked up further and the bear case knocked less of it down.
How much each input counts is part of the methodology we apply consistently from stock to stock — what we will not do is dress the mix up as a magic formula. Judgement is involved, and pretending otherwise would be the most dishonest sentence on this site. So instead of hiding the judgement, we disclose it: when an analyst overrules what the evidence framework alone would say, the override note is published with the pick, in plain sight, with the reasoning. You see the number and the human thumb on the scale, whenever there is one.
Two things the score is not. It is not a price target — it measures our conviction in the argument, not a predicted destination. And it is not a portfolio instruction: how any position fits an individual investor's circumstances is a different question, and not one a published number can answer.
Why it matters
Numbers force honesty in a way adjectives never will. "We like it" can never be wrong — it can always be reinterpreted later, softened, or quietly forgotten. A 74, stamped on a dated page, can be checked against what actually happened. That checkability is the entire point. We publish a number not because conviction is precisely measurable, but because a number is accountable and a mood is not.
Fixing the score at publication is the other half of the mechanism. A score that could be revised after the fact would drift, every time, toward whatever makes the record look best — not through villainy, just through the ordinary human preference for being right. Locking it removes the temptation to grade our own homework. The score you see on a pick today is the score members saw the day it published, whatever has happened to the stock since.
What would change our mind
Conviction changes — that is not a flaw in the system, it is the reason the system has a publication rule.
- When new evidence changes our conviction, we publish a new note with a new score, dated, with the reasoning
- The original score stays exactly as published — the record shows the sequence, not a retouched summary of it
- Re-scoring is triggered by evidence: results, bear-case tripwires, balance-sheet changes — never by price movement alone, and never by how the record is looking that month
- Downgrades get the same prominence as upgrades; there is no quiet channel for bad news
The score is not static across time. It is honest across time — those are different promises, and we are making the second one.
Figure VI — From evidence to a locked number.
Four streams of evidence go in, the bear case pushes back, and what survives becomes the score. The padlock is the part that matters: from the moment of publication the number belongs to the record, not to us. The dial shows an illustrative example — 74 out of 100 — not a live pick.
General information only — not personal financial advice. Research is provided to wholesale clients within the meaning of ss 708 and 761G of the Corporations Act.