To outline this argument, I'll use the simplest example that can point out the problems in this methodology - Death Star acceleration. As one of the simpler examples, it's also one of the less dramatic, because the naïve arguments in question diverge most dramatically from a truly empirical approach in the more complex examples.
We have one particular commonly iterated argument that boils down to a claim of 100g acceleration based on the series of positions of the Death Star on a Rebel diagram. Solely, in other words, on a particular interpretation of a single small detail of the VFX.
As usual, there are minor problems of scaling, and the Death Star at one of these positions happens to be a faded icon sitting on the edge of a dark line across the diagram, i.e., an apparent blind zone. The figure also seems predicated on Yavin being noticably larger than Jupiter, when we consider the distances claimed (100,000 km from 15 minutes to 5 minutes, and 200,000 km from 5 to firing position).
But it's reasonable enough on first inspection, errors aside.
So what happens when we plug this 100g back into the original scenario, looking at the Rebel diagram? We get a very strange starting position for the Death Star, particularly when we consider an Imperial diagram that also shows the approach.
It becomes clear that something is wrong with the claim that the Death Star has 100g. The more evidence we examine - dialogue speaking of orbiting at maximum velocity, and when we dig into the mathematics of the Death Star's attack on Yavin, we can formalize an explicit mathematical proof demonstrating that the Death Star cannot have 100g of acceleration*. Place the Death Star anywhere that fits with the diagrams and dialogue, and 100g results in Yavin being destroyed before Luke reaches it.
Can we fit a lower acceleration into the Rebel diagram? Sure. Remember first that the 5 minutes position is a "last known" position and you're fine.
So let's summarize the technique that was used here to arrive at the faulty figure.
- Find a small detail of VFX.
- Interpret this detail in a fashion that maximizes or minimizes whatever you're looking to maximize or minimize.
- Make sure #2 seems to be the simplest possible explanation. After all, simple explanations sell well.
- Don't even bother looking at the rest of the explicit visual evidence. After all, since the VFX are completely and utterly accurate (per the suspension-of-disbelief philosophy of analysis) and the simplest possible explanation is the best one (per a misinterpretation of Occam's Razor**), your argument lays forth the best possible estimation, and is necessarily correct in its details.
- Likewise, ignore qualitative evidence - the only thing that matters is numbers, dialogue can be freely ignored.
- Above all, do not try to fit your conclusion back into the movie or TV series.
Point #6 bears particular emphasis. One of the most easily used counterarguments in the VS debate is reductio ad absurdum. In its most useful form, this means taking the conclusions arrived at, and seeing what would have happened onscreen were the figures in question true.
*The messy details are an exercise left to the reader, who surely can manage it from all the figures being thrown around on the various articles. I could write it out here, but this post is long enough already.
**Occam's Razor, properly cited, states that the simplest explanation that fits all of the evidence is usually the best one.