“Science” has no methodical way to produce innovation, invention and discovery. They’re coincidental at best. “Science” claims way too much credit for things.
The "scientific method" is just rebranded 'trial and error,' a technique that long predates the terminology "science."
"Computer models" just help an author visualize their assumptions. They don't prove anything, they just regurgitate the original assumptions.
A mathematical formula resulting in a prediction does not explain why and can always just be a coincidental.
Correlation does not imply causation.
It's a bias of its own to look at a category of entities and decide there's an inherent relationship between the entities because there are differing percentages of them. This often results in folks deciding everything needs to be 50-50 for no reason. For instance, deciding that because there are 8 female cheerleaders and 2 male cheerleaders in a group, there's discrimination against males. This neglects all kinds of other information about the situation like maybe most males just don't want to be cheerleaders. And by looking at it from the sex of the cheerleaders in the first place, you've already decided that there's a relationship; your test (counting) hasn't done anything.
Saying a coin flip has a 50% probability because it can be 1 of 2 outcomes is just not helpful. There's the motion of the throw, the strength of the throw, the wind while it's in the air, the weight of the coin, what side the coin started on, whether or not the coin has been altered, whether the coin will be caught or fall to the ground and surely other factors. You could accurately calculate the result of the toss with 100% accuracy prior to the throw with enough time and effort. It's actually even easier on a computer as random number generators aren't true random number generators and can be manipulated.