Now the probability of a long-term outcome is actually somewhat predictable.

So even though in one generation there's a roughly equally likely for

an allele to go up or down in frequency, that the frequency of A may go up, may go

down, the long-term loss or fixation of allele is actually more predictable.

Again, if you start at dead center here you're equally likely to

hit either of the two ends.

Think of that blindfolded person.

If the frequency is much less than 0.5 then it's likely for

the allele to be lost.

You're likely to run into that zero end.

If it's much greater than 0.5, you're more likely for that allele to be fixed.

Well there's a very simple relationship you might have already anticipated.

The probability of eventual fixation.

And again, this is over the long-term, not one generation.

Probability of eventual fixation of allele A will equal the allele frequency of A.

So if your allele frequency is 0.9,

you're probability of hitting that 1.0 end is 90%.

The probability of hitting the 0.0 end is 10% or 1 minus your allele frequency.

So let's look at this.

Here's four sample runs of that program I showed you before AlleleA1.

Our starting allele frequency is 0.75.

You are towards this end of the overall cage.

Again, it's be-bopping around there, three out of four times it ends up hitting 100%.

One out of four times, it actually be-bops far enough down that it hits 0%.

I actually didn't manipulate the data.

I actually just ran this as the same four sample runs and

this is what actually what it came out with.

So as you can see,

this probability of eventual fixation is equal to the allele frequency.

Now notice there is still some chance the allele will be lost even if it starts out

very abundant.

And you see that it's happened in this case.

It started at 75%, it be-bopped around and

after a few sudden losses it did actually go to zero.

So this does happen, but this frequency tells you the probability of it.

Now let's turn this around a little bit.

Let's think of it in the context of species versus population.

Everything we've looked at so far has been thinking of what's happening in

one population and focusing just on that.

Now what happens if we're looking at a whole species that has many

isolated populations?

Let's imagine we're looking at for example, the Galapagos land snail,

which is found on various of the Galapagos Islands.

Here's a thought question for you.

If there are four populations of Galapagos land snails and

each individual population is started with an allele frequency of 0.75,

what would the allele frequencies be in the populations many years later?

Well you probably can guess that some are going to fix, some are going to be lost.

In fact, 75% of them will fix, 25% will be lost.

So is there's four populations we expect on average, three to be fixed,

one to be lost.

The second one is something I want you to actually answer here in video question.

What would the average allele frequency across all populations

be many years later?

Let's measure the four populations had approximately the average size.

They eventually do their things, some are fixed, some are lost.

In the species as a whole, averaging what's happened to each

of the four isolated populations, there's no migration between the populations or

from anywhere else, what will the average allele frequency of A be?