I floated half heartedly a theory yesterday that a half-arsed attempt to fix up mistakes with the data MAY explain a lot of what we are seeing. I'm sort of saying the same thing today in a slightly bigger font. The font may get bigger over time.
Let us consider the circumstantial evidence.
- Everything about Hugo voting and nomination in 2023 was late and difficult to use. Access to voting was spotty. We do not need to assume there were difficulties because we all saw that there were difficulties. Some of these were due to circumstances in particular the known difficulty of data exchange across the internet borders with China and the more complex nature of works in two different languages.
- Less objectively, I had the impression from the US based members of the con committee that they were less interested in the Hugo Awards and more interested in the complexity of what was a genuinely impressive in-person convention in a newly built venue. Pulling that off in time would have been a challenge in itself and other elements of the Worldcon would have been lower priority as a result.
- We know that the finalist results were late. We know that a set of finalists was sort of published but that it had multiple errors in it and had to be fixed up. We know therefore that this was a process that was both delayed and rushed and had QA and version control issues.
- We know that the published nomination total was larger than the final vote total before any stats were published. This looked like an anomaly. It remains odd. (Yes, we can imagine reasons for it but it remains weird).
- Both the final voting figures and the nomination figures were delayed being published. Both sets of data were not ready to be seen by voters in the immediate aftermath of the awards. That definitely means the stats were not in a clean state for public consumption - extra QA and tidying needed to be done which implies that the level of QA that had been done prior to the release of the finalists and the release of the winners was "only just" good enough for that purpose i.e. they got the nomination stats to a point where they could say "here are the finalists" (only just) and then later the winners were in a similar state.
- There were admitted errors in the nomination stats when released (Turing Food Court)
- The EPH totals are inconsistent with the vote totals (Marshall's observation). This is a definite error. It might just be a typo in the published total but even that implies poor QA of the data even at this point.
- I am 90% sure there are other cases of long list nominees having copy-and-paste errors based on the way EPH votes spread. I think some Chinese-language nominees have been accidentally listed as English-language nominees.
- Many people have suggested that Dave McCarty's responses are more than just him being unpleasant but that he is providing cover for other members of the committee. This is usually framed in terms of sinister aspects of CCP censorship which I remain sceptical of but who knows. However, Dave is covering up for mistakes made by a team that he should have been supervising more closely but didn't - well, heck, that makes a bunch of sense to me. It is the sort of mix of self-interest and self-perceived altruism (defending one's team) that humans do
Cons: This does not explain the "cliff". To some extent, the accidental release of the incorrect finalist creates one issue for the hypothesis - people at the time noticed Babel wasn't on the list, which means if it was accidentally left off initially the Hugo team did get a big hint. The theory requires a whole bunch of things to have happened and a series of attempts to deal with them with quick fixes. Maybe data was lost and so maybe that explains the Cliff in some categories.
I don't like that it requires a whole bunch of different things to have happened but then ALL of the hypotheses/conspiracy theories I've seen need a whole bunch of things to have happened and THERE REALLY IS A WHOLE BUNCH of observable issues here that don't add up to a whole (incorrect totals, the cliffs, the delays, the obfuscation and, of course, the ineligibles). What I do like is that we have a single disaster movie-like underlying cause. A ship accidentally steers near some rocks and tries to fix its course and that makes things worse etc.
Except...surely this would be less damaging for the Chengdu Worldcon to admit to than all the other things people are currently imagining? Heck, if people thought I was involved in some weird corruption involving Communist Party censorship of Neil Gaiman, I'd be loudly announcing "No, no, I'm just a complete fuck-up and messed up all the data!". Even more so given that the "Communist Party censorship of Neil Gaiman" is looking not bat-shit crazy as a theory at the moment.
No comments:
Post a Comment