A problem with the theory of natural selection- Diminishing effectiveness

Each female housefly can lay up to 500 eggs in several batches of 75 to 150 eggs over a three to four day period. As many as 10 to 12 generations of housefly may occur annually in temperate regions, while more than 20 generations may occur in subtropical and tropical regions.

That’s a lot of flies born every year and the common housefly has been around for about 50 million years.

And that’s only the housefly, not all the other small flying insects that have been around millions of years.

One female Spider may produce as many as 3,000 eggs in a series of several sacs over a period of time.

So also an awful lot of spiders and web spinning spiders are at least 110 million years old.

With all those spiders and all those flies being born for all those millions of years, why haven’t, through natural selection and over millions of years, flies evolved to see and avoid or escape spider webs?

It seems logical that, if natural selection is true, the methods of a particular predator should, over time, become less and less effective the more prey they kill; that over millions of years the prey should evolve, through natural selection, to instinctively know and avoid the traps and snares of the creatures that prey upon them forcing the predators to have to evolve.

And yet the 50 million year old housefly (aswell as other flying insects, some older than the housefly)  is still getting caught by the 110 million year old Spider Web design.

Note: I’m not saying that Natural Selection doesn’t exist, I’m simply asking, if it does exist, why do flies still get caught in spider webs? Why are the same spider webs still as effective today as they were 110 million years ago?