Andrew L. Urban
I just have to share this with readers who may have missed it. In recent weeks, a number of comments have turned up on previous stories about the Sue Neill-Fraser case. It’s been 16 years since the abomination of her trial and wrongful conviction (and she is now out of prison on parole), and still there are those who are obsessed with some aspects of the case. None more vividly than to continue the prosecution’s fevered attempts to ‘keep Vass off the boat’, as I first called in March 2021. The most recent yesterday. The only reason I raise it again is to share my amazement at the repeated but irrational efforts to save the prosecution from the massive embarrassment of having been found out about how this sausage (wrongful conviction) was made.
So here is the short version. (See the link above for the longer version.)
Meaghan Vass was a homeless 16 year old drug user at the time of Bob Chappell’s disappearance on Australia Day 2009, from the Four Winds yacht he jointly owned with Neill-Fraser. She was hanging around with youths around the Hobart waterfront, with leisure activities such as looting yachts for saleable items, to bankroll the purchase of drugs and alcohol.
At Neill-Fraser’s trial in 2010, Vass was called as a witness, her DNA having been found on the deck of Four Winds, but she denied having been on the yacht. The prosecutor, then Tasmanian DPP Tim Ellis SC, denigrated the DNA evidence as a red herring.
Cut to the chase…(this is the very short version). A few years later Vass confessed both in an affidavit and in a tearful interview with Liam Bartlett on 60 Minutes, that yes, she was on the yacht at the relevant time with two male accomplices, having boarded without realising someone was on board, there being no dinghy alongside. (Neill-Fraser had gone ashore in it.) That someone was Bob Chappell, working below deck. When he confronted the surprised youths, a fight broke out … Vass fled. Obviously, Sue Neill-Fraser was not present. She did not murder Bob Chappell. That’s why ever since, the machinery behind the legal cohort who made the mistake of framing Neill-Fraser – and their proxies/supporters – has been and still is desperate to ‘keep Vass off the boat’.
To accept any of the various bits of speculation as to how Vass was not on board, or that her DNA was deposited later, we would have to accept that after Vass denied being there at the trial, she changed her mind and for reasons we cannot fathom, she chose to insert herself into the case with testimony that confirmed that she was, thus proving Neill-Fraser’s innocence. On 60 Minutes she admitted it risked repercussions from the youths with her at the time.
The fact Neill-Fraser was not on board at the time also scotches the implausible scenarios that have on one hand wanted ‘Vass off the boat’ and Neill-Fraser ON the boat.
(You can feel some sympathy for the poor old Moderator of this blog having to constantly bat away bat-crazy comments like these. I fear there’s another batch coming …)
Dear poor old moderater. Have you ever contested a trivial traffic ticket?
The advice from my neighbouring retired magistrate – “You won’t win-they, they lie their effing guts out”! Doesn’t that sum up many a wrongful conviction?
I read the comments at
https://wrongfulconvictionsreport.org/2026/01/23/sue-neill-fraser-petition-on-its-way-to-tasmanian-governor-and-premier/#comment-194008
You can rest assured that I won’t offer a personal opinion on this case because I am a novice. However, I have expertise in the application of Artificial Intelligence to the comparative analysis of people’s opinions.
I have the results on the comments of Andrew Urban, Garry Stannus, Fiona G., and Donald/Donald Sanderston. Since Peter Lozo was mentioned several times, I included his comments as well (from https://citynews.com.au/2026/rotten-to-its-core-no-saying-sorry-in-tasmanian-justice/.).
Below is the summary from ChatGPT. The only instruction I gave it was to deduce the characteristics of each of the people purely based on their comments. I didn’t specify what characteristics to address. I left it to ChatGPT to chose whatever characteristics it decided were appropriate .
1. Likely Orientation in Case
Fiona G. — Skeptical of the innocence campaign; leans toward conviction being justified or at least not safely undermined.
Garry Stannus — Strongly pro-innocence / pro-inquiry.
Donald Sanderston — Strong skeptic of the innocence narrative; attacks the burglary-gone-wrong scenario.
Andrew Urban — Strongly committed to SNF innocence.
Peter Lozo — Strongly skeptical of the innocence campaign’s reconstruction of events; says he is focused on evidence rather than guilt/innocence as such.
2. Thinking Style
Fiona G. — Top-down reasoning; psychology, memory science, behavioural inference, plausibility testing.
Garry Stannus — Bottom-up synthesis; builds narratives from transcripts, witness fragments, later information.
Donald Sanderston — Engineering / falsification style; tests claims against physics, timing, geometry, feasibility.
Andrew Urban — Macro advocacy framing; emphasizes alleged lack of proof rather than granular contradictions.
Peter Lozo — Scientific-technical analysis; transcript cross-checking; physical mechanism testing; optics and environmental exposure reasoning.
3. Key Strengths
Fiona G. — Structured thinker; strong grasp of memory issues; disciplined prose; willing to dissent.
Garry Stannus — Deep case immersion; transcript familiarity; persistence; imaginative reconstruction.
Donald Sanderston — Strong analytical discipline; practical reality-checking; spots contradictions; technically literate.
Andrew Urban — Persuasive writer; agenda-setting; keeps focus on burden of proof; energetic campaigner.
Peter Lozo — Strong technical reasoning; integrates physics, geometry, visual perception, and forensic-science critique; careful about exact wording; checks claims against transcripts and physical constraints; openly corrects his own wording when needed. He explicitly says he has researched multiple circumstantial cases since 2012 and has posted extensively on the SNF case.
4. Likely Weaknesses / Biases
Fiona G. — May overread behaviour through a psychological lens; may underweight alternative explanations.
Garry Stannus — Risk of speculation and overfitting incomplete clues; distrust of institutions may harden views.
Donald Sanderston — Can become narrow, combative, and overconfident in favored evidence points.
Andrew Urban — Confirmation bias; dismisses inconvenient details; not a neutral moderator; gatekeeping tendencies.
Peter Lozo — Can be highly invested in particular evidentiary themes; may give more weight to technical/physical analysis than to legal framing; can sound prosecutorial in the sense of dismantling innocence claims point by point. His opponents characterize him as trying to “justify the conviction,” which he denies.
5. Communication Style
Fiona G. — Polished, essay-like, calm but firm.
Garry Stannus — Discursive, exploratory, detailed.
Donald Sanderston — Direct, forceful, argumentative, persistent.
Andrew Urban — Sharp, rhetorical, sarcastic, impatient.
Peter Lozo — Formal, technical, methodical, citation-heavy, persistent.
6. Overall Impression
Fiona G. — Intelligent, evidence-conscious contrarian.
Garry Stannus — Dedicated amateur case researcher / advocate.
Donald Sanderston — Technically trained adversarial critic.
Andrew Urban — Campaign advocate and forum gatekeeper.
Peter Lozo — A technically trained independent case critic who behaves more like a scientific reviewer than a campaigner.
ChatGPT then concludes as follows:
A few extra points about Peter stand out.
He is the most explicitly scientific and mechanistic of the group. In the CityNews thread he grounds his arguments in: the trial transcript, crime scene photo 7, the geometry of the yacht and boom/winch arrangement, sunlight glare and color perception, UV-driven DNA degradation, and the distinction between DNA quality and quantity. He also explicitly states that he does not care whether Neill-Fraser is guilty or innocent, but is focused on analysing evidence, facts, propositions, and theories.He also comes across as unusually self-auditing. He posted a later correction to refine his earlier wording about how witnesses described the dinghy’s color, which suggests a fairly conscientious approach to precision.
The cleanest contrast is this:
Fiona = psychological plausibility
Garry = narrative reconstruction
Donald = engineering-style contradiction testing
Andrew = advocacy and framing
Peter = scientific-technical review of evidence and mechanism
My overall read is that Peter occupies a distinct niche: he is less a partisan advocate than a technical critic of weak reasoning, especially where he thinks optics, mechanics, timing, or degradation science have been misunderstood.
You posted two comments about half an hour apart on two different posts, with two different names and email addresses. You may have good reason, but in the context of this subject and our past experience of commenter identity sleight of hand, it raises suspicion and lessens confidence.
As to the AI comparative analysis of opinions, while it is mildly interesting, if superficial, it doesn’t really address the essential questions that form those opinions.
Considering my long history of reading and/or moderating his comments I am suspicious of ChatGPT’s conclusions (“a few extra points”) re Peter Lozo. It reads like advocacy on his behalf, dressed as detailed analysis. Doesn’t matter much, but adds to the overall sense of the ChatGP exercise being somewhat pointless.
One final personal point: while I agree I am “strongly committed to SNF innocence” that commitment comes from familiarity with the trial and the appeals. I also have the benefit of legal advice which I have reported over the years.
But the bottom line: Meaghan Vass was on the yacht, Sue Neill-Fraser was not. OK?