This may happen on a case by case basis. But will it be true statistically. Of 100 eagles with good eyesight and 100 eagles with poor eyesight, which group is more likely to have more children, and even more importantly, more grandchildren.
The probabilities of natural selection are statistical. There can always be individual cases that go against the statistical norm, but they don't generally impact the overall picture.
In many cases the choice of mate is not random. Assortative selection of mates is one of the signs and one of the causes of speciation. Assortative selection of mates is often based on signals of fitness. One study showed that in frogs, the strength and lenght of its mating call was a proxy for fitness. Females preferentially chose males with the strongest and longest mating calls.
However, even if mate selection in eagles is random, natural selection will still move the population toward better eyesight because those with poorer eyesight will be statistically less likely to survive to or through their full reproductive years, and hence, statistically likely to have fewer offspring.
And with a double dose of undesirable genes, the offspring of such matings are even less likely to find a mate and to produce viable offspring themselves.
3) Another random factor, is whether or not the top choices actually live. They have a higher likelyhood of surviving, but can just as easily die at the hands of a skilled predator. The odds may be higher that they live, but you're still betting on random odds.
How do you track a higher likelihood of surviving? Through the statisical documentation of who survives and who doesn't. May I recommend again Jonathan Weiner's
The Beak of the Finch. This is exactly what Peter and Rosemary Grant have done over the last 25 years.
Yes, on a case-by-case basis, some accident can end the life of the most fit individual in the population befoer it has a chance to reproduce, but this does not affect the overall big picture, which is based on all cases, not this particular one.
4) Then you have to consider all the random ways a healthy creature can die: predation, being in the wrong place at the wrong time and catching a disease, a list of natural disasters such as floods, famine, a long list of harsh whether possibilities like hurricanes, snow, extreme heat or cold..............there are far to many random occurances to consider.
No, all that needs to be considered is whether the organisms with a particular trait have a greater likelihood of leaving descendants than those without that trait. Various natural disasters will affect both the fit and the less fit. The question is whether they affect them randomly or non-randomly. A hurricane may affect them randomly. Extreme cold may not. Some may have a greater capacity to tolerate cold than others. These will have a greater likelihood of surviving and transmitting their genes to the next generation.
Natural selection is a completely random process, of passing off randomly occuring genes.
How can natural selection be random when it is the preferential passing on of a certain gene or set of genes?
True. But this would be like trying to pic a pure red marble in a mixed bag of other marbles of different colors, patterns, and color combinations. Health on an organism can increase it's chances of having it's genes passed on; but that's like just adding more red marbles to the mixed bag---your chances of getting the red one increase, but it's still random chance to pick the red one.
I am not following your analogy. Who is picking? And why? You are admitting that the chance of getting a red one increases. Why does it increase? It is not just because there are more of them. It is because there are proportionately more of them.
If you begin with a bag of 100 marbles,99 mixed colour and 1 red, the proportion is 99:1
If the red marble reproduces a red marble and all the mixed colour marbles each produce a mixed colour marble, you now have a bag of 200 marbles, 198 mixed colour and 2 red. The proportion is still 99:1. Even though there has been a doubling in the number of red marbles there is no more probability of picking a red one than at the beginning.
But if the red marble reproduces a red marble, and only 90% of the 99 mixed colour marbles reproduce a mixed colour marble (i.e. 89 new mixed colour marbles) our total population is now 190 with 188 mixed colour marbles and 2 red marbles. The proportion is now 94:1, so there is a greater likelihood of picking one of the two red marbles.
Lets try a bag of 100 marbles in which 10 instead of only 1 is red. Let 90% of the red marbles reproduce so that after reproduction there is a total of 19 red marbles. Let 85% of the 90 mixed colour marbles reproduce so that after selection there are 90 + 76 = 166 mixed colour marbles. Now our proportion of mixed colour to red marbles is 166:19 or approximately 8.7:1
Now it may be that the one red marble that did not reproduce was the purest, reddest, most fit of all the red marbles. Its failure to reproduce did not stop the other 9 from reproducing.
There is no getting around the fact that evolution is a process that affects species not individuals. Therefore it must be understood statistically. It is not the particular effect of particular disasters on particular fit individuals that is relevant, but the question of whether survival correlates with fitness statistically.
Natural selection is completely random, and is another reason why evolution is hard to believe.
If there is a correlation between a characteristic and survival it is not random. If that characteristic becomes fixed in the species, it is not random.