Does evolution have a chance?

[Tomk80]

You need to revise your high school probability here. How do you determine the probability of two sixes in the roll of a two dice. Note these events are said to be independent.

Enough mathematics for the night. Time for bed. Good night all.

But the two wings aren't comparable to two sixes. As Jet Black outlined, they look similar, but are very different.

A comparison would be rolling two dices and calculating the probability of rolling 8. There are 5 ways of rolling 8 on two dice. 6 on dice one and 2 one dice two, 5 on dice one and 3 on dice 2, 4 on both dice, 3 on dice one and 5 on dice two and 2 on dice one and 6 one dice two. All these outcomes are similar, they are not the same.

In the same way, a bird's and a bat's wings are similar, but they are not the same. So we would need to add up the probabilities, not multiply them. If you know of any case where the exact same protein or structure has evolved independently (as opposed to two different proteins or structures that look similar), I would be very interested to hear of it. I know of none.

It's not I who should review my mathematics, it's you who should review your biology.

 

[Jet Black]

Please demosntrate what you are saying mathematically.

you're the probability whizz here - I'm sure you can cope. it's quite a simple calculation, but I have to go home now. consider it tonight's homework for you.

 

[Loudmouth]

1. Determine the probability of a specific point mutation occurring that meets the requirements of NDT.

Given the current human mutation rate, a population of 40 million per generation could concievably cover every possible point mutation every generation. I would say that the probability is 1.

2. Determine the probability of that point mutation surviving and becoming fixed in the population.

Mutations are not required to be fixed. This probability is not needed. For instance, hemoglobin S is a mutated version of hemoglobin C (the sickle cell allele). HemS prevents the malarial parasite from killing you. As you can guess, HemC has not been fixed in the population, but this wasn't required for the next step (HemS) to occur.

Also, speciation might push an allele towards fixation without it needing to be fixed in the original population.

3. Determine the number of viable point mutations needed to give a realistic probability of a single step of evolution occurring.

This is completely unknowable, given that each environment and each individual define what is beneficial.

4. Assuming that number of point mutations, and the number of steps required for a change of species to occur, determine the number of possible paths for evolution to take.

Given that there are approximately 5 million species or so, and the current biodiversity only represents the tip of the iceberg when previous species are considered, I would say that the possible paths are quite numerous.

5. Given that number of paths, what is the probability of convergence based on the number of convergent paths observed to exist in nature.

Genetic convergence would be extremely rare. Morphological convergence would be much more likely. For instance, if we were able to recover the complete Tasmanian wolf genome we would find that it is much closer to kangaroos than it is to North American wolves despite the morphological convergence seen between the two species. Even with a limited Tasmanian wolf genome, it can be shown that North American wolves are more closely related to humans than they are to the Tasmanian wolf.

In this thread we have confined our discussions to steps 4 and 5 of the calculation.

Which are impossible to calculate. Instead, creationists give us big numbers and hope that no one asks to see their math.

 

[Alchemist]

I think your attacking the argument from a bad angle. The chances of wings forming exactly the way they are now are small, no matter what time period you put on it. However, there are several flaws that you should be focusing on.

Of course, the analogy is completely flawed to begin with. My point was that assuming the analogy was correct, Micaiah's math is still wrong. In his post, Micaiah claims that evolution of a wing would take 1x10^2997 years. But as a six could be rolled 3855 times in succession the first time it was attempted, there is probability the evolution of the wing would take only 5.4 hours.

In other words, Micaiah is claiming the wing will take 6.16x10^3001 times longer to evolve that it actually could. Even if the analogy is useless, I think that's a quite important observation.

Peace,
Nick

 

[Alchemist]

This is a good example of the kind of miscommunication that is occuring. I was trying to give a mental picture of the kind of numbers we're talking about...

The point is though, Micaiah, your numbers are rubbish. If you roll a die 3855 times, there is a probability that you will roll a six every time. Yes, the probability of it happening is only 1x10^3000. But it can still happen. And if it does, then according to your analogy, wings could evolve seperately twice in 5.4 hours.

...not claiming this is how long the wing takes to evolve.

What...? That huge number you posted was how long you claim a wing would take to evolve. It has everything to do with it.

 

[Micaiah]

Has Tomk80 worked out how to do probaility yet? Take two dice. The chance of throwing a 6 AND a 3 is 1/6 x 1/6. The probabality of of throwing a 6 OR a 3 is 1/6+1/6.

If my memory serves me okay the rule is stated as follows:

For two independent events A and B, the probabaility of event A AND event B is:

P(A and B) = P(A) x P(B)

The probabilty of an event A OR an event B is:

P(A or B)= P(A) + P(B)

Alchemist, show me where I stated this.

That huge number you posted was how long you claim a wing would take to evolve.

 

[ebia]

Has Tomk80 worked out how to do probaility yet? Take two dice. The chance of throwing a 6 AND a 3 is 1/6 x 1/6.

No it's not. The chance of throwing a 6 on the red die and 3 on the blue die (say) is 1/6 x 1/6. The chance of throwing a 6 and a 3 is twice that because there are two possible favourable outcomes (6,3) and (3,6)

The probabality of of throwing a 6 OR a 3 is 1/6+1/6.

Wrong again.
The chance of rolling a 6 on the red OR 3 on the blue is
1/6 + (5/6 * 1/6) = 11/36
or, more simply
1 - (5/6 * 5/6) = 11/36

(unless you meant XOR, in which case it's (1/6 * 5/6) + (5/6 * 1/6))

If the 'order' is interchangable (which is the implication of the question), then the probability is
1 - (2/3 * 2/3) = 5/9 = 20/36

If my memory serves me okay the rule is stated as follows:

For two independent events A and B, the probabaility of event A AND event B is:

P(A and B) = P(A) x P(B)

Correct, but be careful how you apply that.

The probabilty of an event A OR an event B is:

P(A or B)= P(A) + P(B)

Wrong.

Easily demonstrated - roll 7 dice looking for at least one 6. The answer cannot be 7/6.

If you want to base an argument on probability, you really need at least a 1st year University understanding of it. A badly flawed high-school understanding is not sufficient.

 

[Jet Black]

has miciah worked out the probability of him being born yet?

 

[Jet Black]

miciah, in his search for convergence and having failed at demonstrating how the pax6 gene helps his argument, suggested other features, and the one he seemed most interested in was echolocation, particularly in bats and dolphins. To start with, all that is convergent about these is that both use sound in order to detect the location and other properties of objects. The question then, is how similar are the methods that they actually use. the answer, is not very.

All microbats echolocate, the sound they make is a chirp of varying frequency from high to low(interestingly humans use this in very sophisticated radar banks, it's called frequency modulation). Dolphins however use clicks.
As bats get close to their target, the reflections will be louder and louder. to compensate for this, the hearing systems of the bat become less sensitive by reducing the tension on the tympanic membrane and adjustment of the angle at which the stapes meets the oval window. As Dolphins get closer to their target, they face the same problem, however in their case, they modulate the volume of their clicks.
Bats sense the returned sound through their ears,
Dolphins sense the returned sound through their lower jaw.

So already we can see that there are a large number of differences in these echolocation methods, so there is little claim for convergence, especially given that many blind people can echolocate too.

so all we really have in the case of echolocation that is similar between bats and dolphins, is the modification of an already existing system (hearing) in order to be more fine tuned. The adaptations that both have undergone are completely different, particularly when you look at the auditory systems. I have already mentioned a couple of ways the bats fine tune their auditory systems though adjustment of membrane tensions and bone positions, but dolphins can't do this. The dolphin ear is a very small hole and basically irrelevant in dolphin hearing. Dolphins detect the sound through a fluid filled cavity in the jawbone, which transmits the sounds to the auditory bulla. this whole area is quite different from terrestrial mammals. The auditory bulla is a bone case that is isolated from the rest of the skull by fibrous tissue. So the actual hearing systems are very different between dolphins and bats also (interestingly, we see intermediates between the hearing systems of terrestrial and aquatic cetaceans as we examine the fossil record)

so where is the claim of convergance? that both use sound to find where they are?

 

[Micaiah]

No it's not. The chance of throwing a 6 on the red die and 3 on the blue die (say) is 1/6 x 1/6. The chance of throwing a 6 and a 3 is twice that because there are two possible favourable outcomes (6,3) and (3,6)


Wrong again.
The chance of rolling a 6 on the red OR 3 on the blue is
1/6 + (5/6 * 1/6) = 11/36
or, more simply
1 - (5/6 * 5/6) = 11/36

(unless you meant XOR, in which case it's (1/6 * 5/6) + (5/6 * 1/6))

If the 'order' is interchangable (which is the implication of the question), then the probability is
1 - (2/3 * 2/3) = 5/9 = 20/36



Correct, but be careful how you apply that.


Wrong.

Easily demonstrated - roll 7 dice looking for at least one 6. The answer cannot be 7/6.

If you want to base an argument on probability, you really need at least a 1st year University understanding of it. A badly flawed high-school understanding is not sufficient.

Okay, I stand corrected. Thankyou. I didn't quite state the examples as intended. The scenarios of interest are as stated below:

The probability of a specified series of consecutive mutations occuring, or the convergent evolution of three animals (ie two animals evolving independently, and then a third animal evolving independently) can be modelled with consecutive throws of the dice. eg. what is the probability of a three followed by a six:

= 1/6 x 1/6

The probabality of a mutation occuring given several locations on the DNA at which it can occur can be modelled by throwing a dice once and then determining the probability of say a 3 or a 6. In this case, the probability is calculated as:

= 1/6 + 1/6.

In this thread, I was trying to show that if you have three animals that evolve independently, then the probability of this happening is determined by multiplting the the probability of a single animal evolving twice, not by adding the probability twice. So the probability of convergence evolution of three animals occurring independently is (10^3000)^2.

 

[Micaiah]

Jet Black, you are correct. I am serching for clear examples of convergence. Ideally what I'm looking for are two traits in an animal that are the same and where that can be shown genetically. For example two animals with identical genes that cannot be explained in terms of the evolutionists basal genes. I'm not saying the other examples don't demonstrate the improbabaility of convergence, just that an example such as that would be more clear cut.

A couple of questions to evolutionists:

1. Do you know of any cases where this has occured?

2. How do evolutionists explain this?

 

[Jet Black]

I'm not saying the other examples don't demonstrate the improbabaility of convergence...

no, but then at no point have you ever demonstrated how these thenotypical cases, where two organisms have developed a similar sort of lifestyle or a vaguely similar sensory apparatus are actually a problem. You see you are taking the current situation as it is, and then saying "hey this situation is extremely improbable" but so what? all possible situations are extremely improbable, but there are a heck of a lot of improbable situations, and so the probability of one of these highly improbable situations occuring is 1. demonstrating a single one of them is improbable is kind of pointless.

1. Do you know of any cases where this has occured?

no

2. How do evolutionists explain this?

explain what? something we have never even seen? how would one explain something that one has never seen to occur and for which there are absolutely no details to study.

 

[ebia]

Okay, I stand corrected. Thankyou. I didn't quite state the examples as intended. The scenarios of interest are as stated below:

The probability of a specified series of consecutive mutations occuring, or the convergent evolution of three animals (ie two animals evolving independently, and then a third animal evolving independently) can be modelled with consecutive throws of the dice. eg. what is the probability of a three followed by a six:

= 1/6 x 1/6

Yes and No.

You've got a lot more than two (or three) animals (dice) available to produce the two or three sixes you are after. It's also not at all clear what constitutes a 'six', and you are deciding what is a favourable outcome after the event, which throws the whole thing out.

The probabality of a mutation occuring given several locations on the DNA at which it can occur can be modelled by throwing a dice once and then determining the probability of say a 3 or a 6. In this case, the probability is calculated as:

= 1/6 + 1/6.

That does not seem like a sensible modeling of the question you claim to be modeling. Two dice, with either producing a six considered a favourable outcome would seem a more reasonable model unless you think the DNA can only produce exactly one mutation.

In this thread, I was trying to show that if you have three animals that evolve independently, then the probability of this happening is determined by multiplting the the probability of a single animal evolving twice, not by adding the probability twice.

As you are deciding what to look at after the event, it is neither.

As others have pointed out, there are no instances of independent evolution producing the same event, only examples of superficially similar events. But suppose for a minute that there were.

Your 'model' then is that you have rolled a large number of dice (lets say 10), with a large number of faces (lets say 100), and three have come up the same. Not that two have come up 100 (or six), but that two have come up the same. The probability of that is not (1/100)^2. It's a damn sight more effort to calculate than that EVEN IF YOU SPECIFIED WHAT YOU WERE LOOKING FOR BEFORE YOU ROLLED THE DICE. But you did not, which completely screws your calculations anyway.

To give a well known example, if you put 30 people in a room, what is the probability that two of them share a birthday? It's not (1/365). The chance of 3 of them sharing a birthday is not (1/365)^2

If you are happy to look for any "co-incidence", then it approaches 1.

 

[Micaiah]

Yes and No.

You've got a lot more than two (or three) animals (dice) available to produce the two or three sixes you are after. It's also not at all clear what constitutes a 'six', and you are deciding what is a favourable outcome after the event, which throws the whole thing out.

The number of animals in the population is considered in the first part of the calculation. Remember, you are looking through a keyhole.

See last comment.

That does not seem like a sensible modeling of the question you claim to be modeling. Two dice, with either producing a six considered a favourable outcome would seem a more reasonable model unless you think the DNA can only produce exactly one mutation.

See last comment.

As you are deciding what to look at after the event, it is neither.

Irrelevant

As others have pointed out, there are no instances of independent evolution producing the same event, only examples of superficially similar events. But suppose for a minute that there were.

We'll see.

Your 'model' then is that you have rolled a large number of dice (lets say 10), with a large number of faces (lets say 100), and three have come up the same. Not that two have come up 100 (or six), but that two have come up the same. The probability of that is not (1/100)^2. It's a damn sight more effort to calculate than that EVEN IF YOU SPECIFIED WHAT YOU WERE LOOKING FOR BEFORE YOU ROLLED THE DICE. But you did not, which completely screws your calculations anyway.

Carefully read over the comments I've made in this thread about the model and the way the mathematics is related to the model. You have not grasped that yet.

 

[ebia]

The number of animals in the population is considered in the first part of the calculation. Remember, you are looking through a keyhole.

Then you better spell out what calculation you think you ARE making, because it is wrong somewhere but I'm not going to trawl through your posts trying to piece together what calculation you are trying to make to figure out where.

Irrelevant

No it is NOT. When you decide what consistutes a favourable outcome is always extremely relevent in probability. You clearly have no understanding of what you are talking about.

 

[Jet Black]

Carefully read over the comments I've made in this thread about the model and the way the mathematics is related to the model. You have not grasped that yet.

I think the problem is not that we have not grasped it yet, but that you are unable to make this mathematical model of yours clear.


Have you noticed how everybody here has exactly the same objection to your calculation? I would assume that at least a few of those objections are independently derived rather than copied from other posters (mine certainly was, so unless everybody else copied their objection from me there should be other people who came up with the idea on their own)

 

[Jet Black]

The way I grasp what Miciah is saying is as follows.

Starting from an evolutionary assumption of a common ancestor between two species that do not at this stage have a certain characteristic, Miciah then looks at the modern species that do have this characteristic, but have independently derived it. An example of this is the echolocation, since the common ancestor between a dolphin and a bat most likely did not echolocate (though it could hear). Then he is saying, what are the odds of the dolphin and the bat both developing echolocation as their primary method of hunting/not bumping into things and determining that these odds are very small. We can do the same for the eye really, although the problem with that is that the argument centred around the Pax6 gene, which exists in more basal organisms than those with eyes.

does everybody else interpret Miciah's argument in the same way that I have?

 

[Karl - Liberal Backslider]

It's rather hard to calculate the odds of bats and dolphins both evolving echolocation independently because we don't know what the odds of either group doing so were.

Given that both inherited from their more basal common ancestor the ability to vocalise, and stereophonic hearing apparatus, it may not be that unlikely at all. Even humans have a very basic echolocational ability, not that it's good enough for us to use for anything.

 

[ebia]

The way I grasp what Miciah is saying is as follows.

Starting from an evolutionary assumption of a common ancestor between two species that do not at this stage have a certain characteristic, Miciah then looks at the modern species that do have this characteristic, but have independently derived it. An example of this is the echolocation, since the common ancestor between a dolphin and a bat most likely did not echolocate (though it could hear). Then he is saying, what are the odds of the dolphin and the bat both developing echolocation as their primary method of hunting/not bumping into things and determining that these odds are very small. We can do the same for the eye really, although the problem with that is that the argument centred around the Pax6 gene, which exists in more basal organisms than those with eyes.

does everybody else interpret Miciah's argument in the same way that I have?

Pretty much.

Seems unconvincing as an argument to me, but (being honest) it's the blatantly incorrect maths that's really annoying.

 

[Micaiah]

It's rather hard to calculate the odds of bats and dolphins both evolving echolocation independently because we don't know what the odds of either group doing so were.

Estimates can and have been made.

Given that both inherited from their more basal common ancestor the ability to vocalise, and stereophonic hearing apparatus, it may not be that unlikely at all. Even humans have a very basic echolocational ability, not that it's good enough for us to use for anything.

This may be an easier approach. Any responses?

I am serching for clear examples of convergence. Ideally what I'm looking for are two traits in an animal that are the same and where that can be shown genetically. For example two animals with identical genes that cannot be explained in terms of the evolutionists basal genes. I'm not saying the other examples don't demonstrate the improbabaility of convergence, just that an example such as that would be more clear cut.

A couple of questions to evolutionists:

1. Do you know of any cases where this has occured?

2. How do evolutionists explain this?