You guys are neglecting to consider mathematics. If only 1% of cop thrillers are good and only 1% of remakes are good, then only .01% of remade cop thrillers will be good.
That's a misuse of statistics. If you take it the other way, you get "If 99% of cop thrillers are bad and 99% of remakes are bad, then only 98.01% of remade cop thrillers will be bad." And if you add the two results, you only get 98.02%, not 100. So obviously the mathematical assumption there is incorrect. It's not multiplicative in that way. (Not to mention that by Sturgeon's Law, 10% of everything would be good, not 1%.)
What you're forgetting is that "cop thrillers" and "remakes" are both subsets of the larger group defined by Sturgeon's Law, namely "everything." Look at it this way: let's say that 5% of Americans are redheads (Wikipedia says 2-6%, but we'll pick 5% for convenience). That means that 5% of Ohioans are redheads, and it means that 5% of American women are redheads; but it does
not mean that only 2.5% of Ohioan women are redheads. The
number of redheaded Ohioan women will be half the number of redheaded Ohioans -- but the number of
all Ohioan women will be half the number of all Ohioans, so the
percentage -- the ratio of the two numbers -- remains the same, since you're reducing both the numerator and the denominator by the same amount.
So if only 10% of cop thrillers are good and 10% of remakes are good, then we still get the result that 10% of remade cop thrillers are good. They're all just different-sized slices of Sturgeon's "everything" pie, and the percentage that's good remains consistent however large the slice, because that's how percentages work. They're ratios, not absolute quantities.
This is what we're saying. Citing statistics
in isolation, without considering their relationship to the broader context or how they actually come about, is meaningless. Statistics aren't about single entities in isolation, they're about how different things
relate to one another, about how they fit into larger patterns. You can't just spout numbers out of context and make them mean whatever you want them to mean, any more than you can pull a few random words out of a text and ignore their context. ("Look, this soliloquy in
Hamlet contains the words 'slings,' 'arrows,' and 'bodkin,' so obviously it's a discussion of handheld weaponry and combat tactics.") You have to understand where those numbers come from and how they fit into the specific pattern being evaluated.