Well, I’m happy to point out that this is a misconception. Differential reproductive success is what results in evolutionary change.
1 Like
Keep in mind that the science is about populations, not about individuals. And survival of populations does depend on reproduction.
It has to do with equilibrium which is the rate at which new mutations and back mutations are occurring.
That’s false. You might credibly say that you are confused by what I write, but to claim that as my motivation is just dishonest.
And you are confused because you are trying (not hard enough) to learn the material while simultaneously taking a strong position that is not supported by the evidence.
3 Likes
Yes, I’ve got that.
Yes, I’ve got that too.
OK - but I still don’t understand what you’re trying to point out exactly.
I was being a little playful. Of course you don’t try to confuse me on purpose.
If substitution an A for a G is a deleterious nearly neutral mutation then changing the G back to an A would be a beneficial nearly neutral mutation. They would balance each other out and stop a steady decline in fitness, if that even happens due to accumulating slightly deleterious mutations.
1 Like
A very good observation, I think. ONE side of this is differential fitness, and being better in some way at surviving to reproduce. The OTHER side of this to not be the unfortunate misfit that gets “selected out”. Not all misfits will be equal, and some might have an advantage relative to other misfits allowing them to survive.
I’m not basing this on biology, but rather on my tinkering with Genetic Algorithms; a sort of search method that is patterns on mutation and selection. What I found was the speed at which searches would find the best solutions varied a lot depending on the rule I used to remove misfits from the population. Part of this was population size, with larger populations having more opportunity to improve quickly.
The other rule that seemed to help, was to allow “new” misfits to remain in the population for a few generations, giving them opportunity for a second mutation. This increased the diversity of the population, and helped the search avoid getting stuck in dead ends.
Long story short, diversity in a population is good, so long as the entire population are not horrible misfits. 
I agree Not sure how realistic your algorithms are, but the general statement makes sense.
Not at all realistic in the biological sense, but the same principle should still apply. I think you are correct: both fit and misfit matter.
I think if I were to make my program sufficiently complex, those different ways of being (mis)fit would drive something like speciation. Speculation tho - I’m not likely to try it. 
I like being correct. 
It happens so rarely in this forum… You’re going to make an ogre give you a big hug.
ah…but that’s the creationist model I think
Only in the sense that I would have created the simulation, and all the souls I created would be deleted as soon as I got bored with it. 
More seriously, there are limits to such simulations. They are great for exploring how certain possibilities can arise given the assumptions, but ultimately what you get from it is only as good as the assumption you start with. That’s a big part of the argument with Carter and Sanford, for example.
I agree. Current evolutionary models do the same thing.
I disagree. Current evolutionary models are tested empirically. Creationist models are only rarely pseudo-tested with imaginary or cherry-picked data.
3 Likes
To be fair, it might depend on which model is being considered. The kind of model I might construct will be full of holes, and I don’t want to overstate my opinion to Valerie. Models based on empirical evidence are going to be MUCH stronger.
I don’t know just what assumptions Carter and Sanford are using for their Mendel’s Accountant models, but I am highly suspicious their claims. A search on Google Scholar shows me that it’s mostly people publishing in Creationist journals who are using this software, and very few in mainstream journals. I guess I could dig into the code, but that seems like a lot of work for a foregone conclusion (also, I bet someone else has already done this. I just need the right search to find it).
1 Like
Well THAT didn’t take long!
https://www.reddit.com/r/DebateEvolution/comments/g4le0v/is_mendels_accountant_really_that_flawed/
3 Likes
I would suggest that the motivated reader should ignore Mendel’s Account and use SLiM instead. It requires coding skills, but it’s extremely flexible and unlike MA it is actively being developed, tested, and used by a large number of researchers. SLiM is also incredibly well documented.
See the following papers (each with more than 100 citations according to Google Scholar):
I had the priviledge of attending a 1 week workshop led by the main developer, Ben Haller, about a year ago. He’s super helpful and incredibly responsive to questions or problems.
5 Likes