John Derbyshire: “I will not do my homework!”

I’ve run into the same problem repeatedly. People want to argue against “ID”, then admit they’ve never read anything first-hand on it. Or they do something like call you “Dembsky” unwittingly showing the same.

To me, it is like a Shakespeare critic never having read any Shakespeare, only reading Cliffs Notes.

If you’re gonna critique ID, then at very least learn the arguments first-hand. It isn’t that hard, there are lots of ID resources online and in print.

I’ve noticed that the lottery ticket rebuttal (someone has to win right? RIGHT??!!!) is often used by creditionaled persons — who one would think would know better — to rebut I.D.

I picked up pretty quickly long ago that what you were getting at was an attempt to specifically address, and go beyond, discussions of simple probablilty.

I don’t consider myself a genius. I just consider those who can’t see this to be as dumb as bricks.

NOTE: for those of you who can’t see what Dembski is getting at and might be reading this: “dumb as bricks” is meant figuratively, not literally.

I find it interesting that most materialists (evolutionists) will hardly ever address any of the conclusive evidence for design. Yet, when they bring up some suggestive similarities as evidence, ID proponents investigate the matter with rigor. It should be noted the evidence for design keeps growing as each piece of suggestive evidence is dealt with and put in the proper ID perspective. Evolutionists don’t seem to grasp the necessity for empirical validation in science.

I’m tempted to praise Dr. Dembski for such a great analysis. However, it seems awfully silly coming from me, someone with hardly any education at all. It’s not worth much. So, for whatever it’s worth, thank you Dr. Dembski.

I listened to the AEI debate and among the silly objections offered by Derbyshire, he made what I think must be an extraordinarily embarassing flip-flop. And actually, I was a little disappointed that neither West nor Gilder called him on it…

During his presentation, he responded to the typical Christian testimony of how Christianity gave purpose to life, etc. by proudly challenging with the question “Yes, but is it REALLY TRUE?” And as a Christian myself, I do agree with him on the importance of that question, and yes, I do think it’s “really true.”

Derbyshire compared Christianity with Plato’s “noble lie” and really made a big, hairy deal out of the importance of the truth of a given proposition as opposed to how it made people feel or how comfortable people were with it.

But then he turned right around and essentially defended Darwinism on the basis that “scientists are happy with it.” Suddenly the TRUTH of the proposition didn’t seem all that important to Derbyshire. Suddenly he was more concerned with how it made scientists feel, or with how comfortable scientists were with it.

This was a truly amazing about-face. It revealed to me that Darwinism itself can be compared to the “noble lie”… at least from a naturalist’s perspective. Obviously, as lies go, I don’t think that one’s so “noble” myself.

And there’s nothing new under the sun.

“To me, it is like a Shakespeare critic never having read any Shakespeare, only reading Cliffs Notes.”

How about the ‘reams of evidence’ canard? Search on “evolution” and “reams of evidence” and you get 3,940 hits. Read a few of them and laugh (or cry). Reminds one of pronouncements made at the Dover trial. Sadly, Judge Jones bought it.

bornagain77: “Evolutionists don’t seem to grasp the necessity for empirical validation in science.”

Of course not. They demand it everywhere from IDists, but are incapable of providing any themselves. Yet they all clamor constantly that “there are mountains of evidence” for NDE!! I’ve yet to see any after 30 years of debating them.

“scientists [they always imply by this that ID scientists are not real scientists] are happy with it” !!!

Interestingly, Michael Chrichton, comments. “Whenever you hear the consensus of scientists agrees on something or other, reach for your wallet, because you’re being had.”

One thing that I think is confusing, and makes probabilities difficult to understand, (for me at least) is that saying that an event has a “one in one million” chance of happening is NOT the same (correct me if I’m wrong) as saying that it WILL happen once in one million times. Isn’t that right?

Seems to me that this expression of chance is frequently used as some sort of “proof” that the result they want would happen eventually. While I’m the furthest thing from a mathematician, this strikes me as being quite incorrect.

If anyone would care to elaborate, I’d be very interested.

Note the similarity between this situation and the situation in Galileo’s time. The folks opposing Galileo insisted that he was wrong and yet patently refused to look through his telescope and consider the evidence for themselves. Their worldview precluded Galileo’s argument so there was nothing more to discuss. Today, materialism and naturalism preclude design so there is no need to bother with the specifics of the case. It is ruled out of bounds by definition.

The Scubaredneck

TRoutMac -

When a probability is given, it is only meaningful over a long range of measurements. For example, even a 1 in 10 chance may not happen for 15 or 20 trials without arousing too much suspicion, but the longer you go, the more in doubt you are about the original estimate of probability. As you do more and more trials, the average of all successes divided by all attempts is you experimental probability. Some people argue that theoretical probability of one time events is essentially meaningless (from a mathematical standpoint).

Still, given enough time, all sorts of weird things will happen. If you have a goal of flipping 10 heads in a row, it will happen eventually. An estimate can be give to the chances (1 in 1024), but you can’t gather up 1023 of your closest friends and have them all flip 10 coins and be certain it will happen.

The reverse is true, of course, as well. Even if something is a 1 out of a million chance, it might happen to one person twice in a row the first two times they try it. It is unexpected, but not impossible.

“I’ve noticed that the lottery ticket rebuttal (someone has to win right? RIGHT??!!!) is often used by creditionaled persons — who one would think would know better — to rebut I.D.”

Yet even a lottery itself have to be design well for them to work(draw people in). If you set the odds too high (for example 1 win in 1000 years) then there will not be any winners at a reasonable time, Thus with time people see it as nothing but a scam and stop buying tickets. If the odds are too low you don’t have any big time winners to advertise and draw more people into the lottery.

If you have a goal of flipping 10 heads in a row, it will happen eventually.

That depends, on the probabilistic resources.

An estimate can be give to the chances (1 in 1024), but you can’t gather up 1023 of your closest friends and have them all flip 10 coins and be certain it will happen.

No, but you can calculate the probability that it will happen ;).

1024 people are equipped with 10 coins, with each coin having a “heads” side and a “tails” side.

All 1024 people toss their ten coins and record how many heads and how many tails appeared amongs theuir ten coins.

What is the probability that at least one of the 1024 people would record ten heads as the result of their tosses, assuming all are being honest in their recording to the acutual toss and that the chances of seeing a heads on each toss was equivalent to the chances of seeing a tails?

Maybe someone should point out a good refutation of the lottery example since it is a common argument used by Darwinists with the dare that “you cannot refute it.”

Have any of the ID proponents provided a good answer to it?

Eric wrote “For example, even a 1 in 10 chance may not happen for 15 or 20 trials without arousing too much suspicion”

Well, I think we’re on the same page then. ‘Cuz I was thinking that an event that has a probability of 1 in one million doesn’t actually NEED to happen in the first million. I mean, even if you try 10 million times, it’s still conceivable that this event didn’t happen in any of those. And yet, it MIGHT have happened on the first try, or it might happen on the 10,000,001 try.

But it seems like when the Darwinists talk about probabilities, they seem to be looking at that one in a million as some sort of “guarantee” that it’ll happen AT LEAST ONCE in that million tries. Now, I may be overstating this somewhat just to illustrate my point, but that’s the way it looks to me sometimes.

Think of it this way… If you were able to make your one million tries ONE MILLION times (1 mil x 1 mil) what would be the odds that ONE of those sets of one million tries would actually match up with your probability estimate of one in a million? I’m thinking not very good… because now you’ve specified a TARGET and that changes everything. Weird.

I know, I’m WAY outta my league here, and I’m sure it shows. I gave up on math a LONG time ago.

Has anyone read “Why Intelligent Design Fails”?

I read the chapters critiquing Dembski , but did not spent much time thinking through.

It seems that the arguments presented in the book are Derbyshire’s kind.

I also have not read Dembski myself. Now I know he has factored in “probabilistic resources”.

I am a bit scared of Probability mathematics now. The last time I touched it was 1982.

Maybe someone should point out a good refutation of the lottery example since it is a common argument used by Darwinists with the dare that “you cannot refute it.”

Try looking at it this way:

That someone will win a lottery is a near certainty.

Life forming from dust stirred up on your walk to the store is, well, not.

Why is someone winning a lottery a near certainty? Because a lottery is designed that way. In fact, if a particular lottery never has a winner you will be wise to assume — heh heh — it is designed that way.

So to correlate the winning lottery ticket to the inevitability of the existence of life one would have to presume life to be . . . . ?

Now, here’s something else to ponder based on the hole in one probability.

With thousands of golfers playing thousands of holes you can count on an occasional hole in one.

BUT suppose every golfer had some muscle condidtion that prevented him from hitting the ball more than 10 yards? And suppose all were blind? And suppose all were facing the wrong way? Even with a trillion billion trillion golfers playing on a trillion billion trillion Sundays could you ever except a hole in one with those conditions?

The existence of physical impossibility is reality.

It is impossible for life to exist by chance. It is inevitable for life to exist with the right Designer.

TRoutMac asks,

an event has a “one in one million” chance of happening is NOT the same (correct me if I’m wrong) as saying that it WILL happen once in one million times. Isn’t that right?

Yes, that is correct. Be aware of the following fallacy of probabilistic reasoning. What is the probability of throwing a 6 on an unloaded die? It is 1/6. If you throw the die and a 6 turns up, what is the probability of that 6? If you say 1, you have fallen for the fallacy. The probability is 1/6 no matter what the outcome. This fallacy lies at the core of many Darwinian assertions about probability and chance.

“Maybe someone should point out a good refutation of the lottery example since it is a common argument used by Darwinists with the dare that “you cannot refute it.”

I think that the fallacy of the argument can be explained by this comparison:

1- Real lottery: n-digit series (let’s say n=7) number and 10^n= ten million tickets sold (at the best). There is a Pr=1 that one person will win the lottery.

2- Difficult lottery: N-digit series number (let’s say N=16) and ten million tickets sold.
What’s the prob. that one person will win? It’s a depressing:

Pr=10^n/10^N=10^-9= 1/1,000,000,000

And if N would be 100?

Muy interestante:

On the Lottery Argument:

1] Context:

Probability thinking and linked statistics are loaded with all sorts of subtleties and technicalities. They are also, very, very important — and are often of great consequence on vital matters.

A set-up for tricksters, who commonly reap a rich harvest. Thus, Darrell Huff’s classic expose, “How to Lie with Statistics.” And, full of traps for the unwary or those who let their attention falter for a moment.

2] Condom roulette and winnable lotteries:

Consider a certain condom that fails in use 1 of 10 times on average. If someone uses such a condom 10 times in a high risk context, what are the odds he has been “protected” all 10 times?

ANS: First time, 1 - .1 = 0.9. Assuming the tries are independent [and they may not be - bad technique . . .], odds of protection all ten times are therefore [0.9]^10 ~ 35%. [That's like playing so-called "Russian Roulette" with four loaded chambers out of six!]

So increasing exposure does shift overall odds. The idea that if we have a million tries at a 1 in a million chance, we are “bound” to come up a winner, is strictly wrong, but is fairly close to an important point — multiplying opportunities shortens overall odds. Thus, too, we see that a lottery with a well designed number of tickets and size of winning number — number of digits — is reasonably winnable.

But, are all “lotteries” winnable?

No, and here is why:

3] Dembski’s upper probability bound:

Say a certain lottery has tickets with 150 decimal digits, and that there is just one winning ticket. Then, other things being equal the odds of winning are 1 in 10 for the first digit, 1 in 10 for the second, etc, so overall it is 1 in 10^150. A lottery of this type, of this size or bigger is arguably unwinnable within the scope of the observable universe. (BTW, in binary digits, this is about 500 bits.] Here’s why, as Dan Peterson summarises:

Dembski has formulated what he calls the “universal probability bound.” This is a number beyond which, under any circumstances, the probability of an event occurring is so small that we can say it was not the result of chance, but of design. He calculates this number by multiplying the number of elementary particles in the known universe (10^80) by the maximum number of alterations in the quantum states of matter per second (10^45) by the number of seconds between creation and when the universe undergoes heat death or collapses back on itself (10^25). The universal probability bound thus equals 10^150, and represents all of the possible events that can ever occur in the history of the universe. If an event is less likely than 1 in 10^150, therefore, we are quite justified in saying it did not result from chance but from design. Invoking billions of years of evolution to explain improbable occurrences does not help Darwinism if the odds exceed the universal probability bound.

Of course, key informational macromolecules such as DNA and proteins etc, by far exceed this storing capacity, and we need dozens to hundreds in the compass of a cell for viable life to exist.

That is why it is a very good empirical observation to note that there is a functionally specified complex information content in life that so far exceeds the reasonable probability resources of the known universe, that it is well-warranted to infer that life as we see it is designed to meet a specification. For agents routinely produce systems using intelligence, which are well beyond the Dembski bound. [This post is an example.]

4] Thermodynamics-scale odds . . .

Can we beat odds like that or worse?

Logically and physically yes, but there comes a point where the odds of winning are so low that we simply do not reasonably expect a win. (Cf my case study thot expt from a discussion with Pixie on that.)

Thermodynamics provides many a case in point, as Dr Granville Sewell often points out. For instance, there is a possibility that every oxygen molecule by random chance in the room where you are will all rush to one end, and so you could dies of lack of oxygen. But he odds of that are so low relative to the changes of a more or less even distribution that we will never observe it in the gamut of the observed universe. [And to smuggle in the assumption that there is a vastly wider universe as a whole that reduces the overall odds, is of course speculation, not science . . .]

Okay, trust that helps

GEM of TKI

I posed the question about the refutation of the lottery because I was interested if anyone who has written about ID has a good response to it. I had not seen one so was actually looking for a cite or a series of discussions about it.

The typical materialist response is that what you see is not the only thing you could possibly get but what just happened to happen. So while any specific rock formation is of incredibly low probability we witness zillions of incredibly low probabilistic formations all over the universe. Similarly, while the actual molecules that eventually led to life are also incredibly low probability what we witness is just the actual formation that emerged and zillions of others were possible but the one that we witnessed is the one that happened.

This is supposedly what CSI is supposed to eliminate but we had a very long thread a few months ago without anyone providing an easily understandable discussion of what CSI is. I don’t want to start this again but it ran for over 200 comments and I do not believe anything was resolved.

I doubt there is anything written that would convince John Derbyshire that the lottery example is not valid since we here were unable to find a clear understandable explanation.

Personally, I intuitively understand it and believe the lottery example nonsense but what I was looking for was well thought out arguments against it that the typical person who reads about this topic could understand.

You can show that all the proteins of length 40 or 60 depending upon how you do the calculations exhaust all the matter in the universe so the assembly of anything by chance is infinitely small. The answer to this is that there is incredibly large number of functional possibilities for life and the combinations that emerged are just one of this very, very large set. Just like the specific rock formation is of incredibly low probability but some formation emerged.

That is why the argument of the shooting the arrow at the wall fails because there is not just one target but zillions of targets on the wall and it would be difficult for the archer to miss one of them.

That is why the argument of the shooting the arrow at the wall fails because there is not just one target but zillions of targets on the wall and it would be difficult for the archer to miss one of them.

I think I see the point you are missing. There are never zillions of targets. There is only just one. There may be zillions of potential targets, however. How do you differentiate between hitting a target and random shooting?

You say you have an intuitive understanding of CSI.

Here is something to consider:

Is it possible to function in society without making assumptions in which you discriminate between design and chance?

What is the process by which you make these assumptions?

Now try to articulate and quantify that process and apply it to this debate.

jerry wrote: “I doubt there is anything written that would convince John Derbyshire that the lottery example is not valid since we here were unable to find a clear understandable explanation”

Well, I think it all boils down to this, with either the lottery example OR the golfing example:

Someone wins the lottery (or gets a hole in one) because people are TRYING to win the lottery. I would say that the Darwinist who uses the lottery example to defend chance has just defended design… in particular, the design of the participantts. People buy lottery tickets because they CHOOSE to… they INTEND to. And they do so because they want to be filthy rich. If no one played the lottery, there’d be no intent, no “design” towards winning it, and you’re probability would fall off dramatically (as if it’s not already very low)

Likewise, A hole-in-one is something that every golfer strives for… it’s the whole object of the game; to get the ball into the cup with the fewest strokes. That someone eventually gets a hole-in-one reflects their “design” (read: intent)

Right now I’m not playing golf. In fact, I DON’T play golf. So what are the odds that in the course of NOT playing golf, I’ll hit a hole-in-one? He, he.

I guess if Derbyshire wants to impress me with examples of people winning the lottery or getting a hole-in-one, he’ll have to find examples of that where the person that won the lottery hadn’t bought a ticket. Or the person that got the hole-in-one wasn’t playing golf. Otherwise, it’s just another example of design.

Am I right?

Hey jerry, others,

Allow me to throw my thoughts into the mix.

I have just had a week long email exchange with a Darwinist friend on this very topic.

I posed him a question, to see if he could help me resolve it. I said, can low probabilities ever be used to rule out the “chance” occurrence of an event? We said yes, we do it in stats. We prespecify our rejection region, then reject the chance hypothesis.

But then I asked how we could do that consistently, when unlikely events of arbitrarily low probability happen all the time?

Take for example the quote mentioned earlier:

If an event is less likely than 1 in 10^150, therefore, we are quite justified in saying it did not result from chance but from design.

Now, let’s say I flip a fair coin 501 times and write down the resulting binary string. Was that specific string the result of chance? If so, we have an example of > 500 bits which is the result of chance.

So we kept talking, and eventually came to the conclusion that we need both specification and low probability (viz. classic ID), but I couldn’t come up with a good reason why the specification should “prevent” certain outcomes. (I couldn’t say it was the low probability that prevented them, since lower probability events could happen at any time.)

But somehow, macroscopically describable, low description length events of low probability do not happen by chance. I can’t say why not, I just see that they don’t.

Atom,

We had a very similar discussion before on a long thread in February though the concept of lottery was not really mentioned

http://www.uncommondescent.com.....me-online/

At the end great_ape, kairosfocus, and gpuccio were trying to illuminate the problem but it stopped when the thread essentially ran off the list of threads and got too long. No one after 200 comments had defined CSI to everyone’s satisfaction.

As far as I could see no one could answer the lottery example which is not just that someone has to win the lottery. That is really a bad choice of words and the real issue is that there could be an enormous number of possible starting points for life all of incredibly low probability and the one that happened or “won the lottery” was just one of these starting points. So this lottery was really a lottery of starting points and one of the starting points is the one that we observe. If time had marched on then some other starting point may have arisen.

In fact several researchers propose that life had more than one origin and not all may have been DNA based but the DNA based one is the one that won out. There is no evidence for this but because life exists, they say there must be an explanation. The use of low probabilities does not eliminate all potential starting points and the one that appeared was the one that won the this “lottery.”

I just placed a comment up and it did not appear. I assume it is caught in the filter, probably because of the link in it. Could a moderator check it out when they have time to see there is nothing wrong with it. And then remove this comment.

Thank you.

Since my comment is stuck in the filter, I recomposed it without the link.

Atom,

We had a very similar discussion before on a long thread in February though the concept of lottery was not really mentioned

Search for Michael Egnor Responds. There was a long thread in February. The actual link seems to put a comment into the spam filter.

At the end great_ape, kairosfocus, and gpuccio were trying to illuminate the problem but it stopped when the thread essentially ran off the list of threads and got too long. No one after almost 200 comments had defined CSI to everyone’s satisfaction.

As far as I could see no one has ever answered the lottery example which is not just that someone has to win the lottery. That is really a bad choice of words because it doesn’t define what the lottery is. The real issue or lottery is that there could be an enormous number of possible starting points for life all of them of incredibly low probability and the one that happened or “won the lottery” was just one of these starting points. So this lottery was really a lottery of starting points and one of the starting points is the one that we observed 3.5 billion years ago as cells fossilized in ancient rocks. If time had marched on then some other starting point may have arisen. There is of course other lotteries along the way after the first lottery. These are the nature of multi-celled organisms, the various phyla of the Cambrian Explosion, flight, legs, 4 chambered hearts, etc.

For the origin of life, several researchers propose that life had more than one origin and not all may have been DNA based but the DNA based one is the one that won out. There is no evidence for this but because life exists, they say there must be an explanation or a lottery winner. The use of low probabilities does not eliminate the possibilities that one of all potential starting points didn’t happen just that the one that appeared was the one that won the this “lottery” of low probability events.

jerry wrote: “For the origin of life, several researchers propose that life had more than one origin and not all may have been DNA based but the DNA based one is the one that won out. There is no evidence for this but because life exists, they say there must be an explanation or a lottery winner.”

Yes, some do take this tack. However, they’ve just fallen into a trap because, just as the multiverse theory (which holds that there really is an infinite number of universes but ours just happened to land on the magic combination) that theory is untestable.

They’re free to assert this, of course, but they shouldn’t expect to retain any credibility since they badmouth ID as being “untestable.”

TRoutMac says:

They’re free to assert this, of course, but they shouldn’t expect to retain any credibility since they badmouth ID as being “untestable.”

Is is testable, to some degree. (No, we can’t tell what actually happened, I realize that, but hear me out here…)

So, if the hypothesis is that there are multiple ways for life to start, we can test this by reproducing what we believe to be the initial conditions and seeing what arises. Possibly, we can do thought experiments regarding what combinations of proteins, etc. are self-replicating, etc. - but I know that many don’t consider those very good evidence.

The point is that, if we can reproduce initial conditions and we find that self-replicating substances can form from there, we’ve shows that it is possible. Here were are not predicting / testing that it did happen, only that it could.

Note: these experiments may be extremely difficult to do and take a long time to get results. I know that frustrates a lot of people when scientists don’t give up on a theory just because it hasn’t borne fruit yet, but there are many areas of science (particle physics, for example) that are playing the same game of experimentation lagging theory by a wide margin.

Oh, and for what it’s worth, there are a couple of proposed tests for a multiverse concept as well - although they would be very difficult to do.

Atom: So we kept talking, and eventually came to the conclusion that we need both specification and low probability (viz. classic ID), but I couldn’t come up with a good reason why the specification should “prevent” certain outcomes. (I couldn’t say it was the low probability that prevented them, since lower probability events could happen at any time.)

It might to think of it this way.

Consider writing down any sequence of 100 heads or tails. It doesn’t matter if it follows a “pattern” or not. By writing it down you have specified the sequence. No matter which sequence it was, the probability that 100 coin tosses will get that sequence is exactly the same.

OR, have someone toss the coins first, and then try to guess (i.e. specify) the sequence that was tossed without looking (also excluding any ESP or any other knowledgable “help”). The probability that the sequence tossed is the one you picked is still the same low probability.

NOW, contrast this with tossing the coin 100 times and getting any old sequence of heads and tails. What is the probability that you will get something or other. Quite high! In fact, it is a certainty.

If you want to say more specifically, “What is the probability that I will get some arrangement or other that is not such-and-such kind of pattern?”, then just

1) define “such-and-such kind of pattern,
2) calculate the probability for getting that kind of pattern (usually very low), and
3) subtract the answer from 100% (or from one, for zero to 1 probabilities).

The result is that it will still usually be very quite likely that you will get some sequence that is not-that-pattern, i.e. not a certainty but still quite high.

BOTTOM LINE:

Specification does not need to mean “regular pattern”, and specification does not “prevent” anything. However, whether a regular pattern or not, a specified sequence can be extremely unlikely.

The mistake is to think that any-old-result is just as unlikely as the specified result. It would be if and only if that particular sequence is specified. If you are willing to take whatever turns up without specifying independently, that is not unlikely at all!

It isn’t the orderliness or symmetry or regularity of a pattern, per se. Its specifying vs. not specifying.

ericB wrote: “It isn’t the orderliness or symmetry or regularity of a pattern, per se. Its specifying vs. not specifying.”

Right. For example, if you randomly grab Scrabble letters one after the other, it’s actually pretty likely that you’ll frequently generate a meaningful word like CAT or DOG. That’s not that impressive and obviously is totally explainable by reference to chance. But if you choose (specify) a sequence like “CAT” BEFORE you start grabbing the letters, then the likelihood that you will pick that particular sequence will drop significantly. And that’s just a three-letter sequence.

TRoutMac: One thing that I think is confusing, and makes probabilities difficult to understand, (for me at least) is that saying that an event has a “one in one million” chance of happening is NOT the same (correct me if I’m wrong) as saying that it WILL happen once in one million times. Isn’t that right?

That is correct. About this question, others have already provided helpful information. I just want to point out that it is not hard to see why and to see the difference exactly.

If you want to see the probability that some unlikely event will happen at least once in some number of independent attempts, it is actually easiest to

1) figure the odds that it won’t happen at all over all those attempts, and then

2) subtract this from 100%

EXAMPLE: If you have a 10% chance of success (1 in 10) for a single try, then you have a 90% chance of failure on each try. Two attempts is 90% x 90% = 81% chance of not succeeding on any try. N attempts is (90%)^N, so ten tries is (90%)^10 = 34.9% chance of not succeeding on any of those attempts.

So, even with a chance of 1 in 10, over ten attempts you still only have about 100% - 34.9% = 65.1% chance of success — NOT 100%.

Likewise, trying for heads with coin flips (1 in 2 chance), there is a 50% x 50% = 25% chance that even after two flips you will still not get any heads, so only a 100%-25% = 75% chance you will get at least one head — again NOT 100%.

This is where I don’t understand how all this probability discussion applies to ID/evolution:

For example, if you randomly grab Scrabble letters one after the other, it’s actually pretty likely that you’ll frequently generate a meaningful word like CAT or DOG. That’s not that impressive and obviously is totally explainable by reference to chance. But if you choose (specify) a sequence like “CAT” BEFORE you start grabbing the letters, then the likelihood that you will pick that particular sequence will drop significantly. And that’s just a three-letter sequence.

Isn’t that the point of critics of ID attacks on evolution? You are specifying the “pattern” after it’s already been observed in nature. Since evolution does not have a “goal” or a specific end point, is it fair to use probabilities calculated after the fact?

Moreover, since evolution isn’t acting all at once, it’s not a good analogy to claim it’s like throwing out letters and forming an 8-letter word. The analogy would have to account for a simple beginning and then some methodology for mimicking natural selection, et. al. (or whatever term you wish to use for the evolutionary process).

I understand that ericB, thanks for the discussion.

Imagine we set up a coin flipper to spit out 501 random bits. We let it run, and it spits out a 501 bit string that doesn’t match any pattern we know of. Our coin flipper is fair, and we know the string was generated randomly . Obviously the universe and laws of nature allowed that particular string of low probability to occur.

Now we get ready to run the flipper again, over and over until the end of the cosmos. But before we do, we write down a string of 501 bits on paper.

Question: Will the coin flipper be able to produce this specific bit string, given from now until the end of the cosmos, by random chance?

If yes, then we just produced a CSI string by chance, since I independently pre-specified my string. (It is meaningful to me, but to a stranger may appear as random as any other string.)

If not, then something prevents this particular 501 bit length string from occuring. Remember, our flipper can produce many 501 bit length strings, of equally low probability. The only difference with this string and my first one was that I took the time to specify this one. So what in my specification act changes things?

I guess an easier way to ask would be:

What keeps CSI strings from forming by chance in random processes? Low-Probability?

If Low-Probability, why doesn’t Low-Probability stop even more unlikely unspecified events from occuring all the time?

If Specification, what about specifying changes what is allowed by nature?

” For example, if you randomly grab Scrabble … But if you choose (specify) a sequence like “CAT” BEFORE you start grabbing the letters, then the likelihood that you will pick that particular sequence will drop significantly. And that’s just a three-letter sequence.”

Or even better yet what are the odds of throwing all the letters on a table and spell “cat” with the C standing perfect on top of A and A on top of T on their ends. Here you have the law of gravity working against you.

Okay . . .

“Once more unto the breach, dear friends . . .”

I see the issue has now “evolved” to the question of both complexity and specification. I will make a few remarks on that, but first, on the issue . . .

1] The problem of selective hyperskepticism

In Western thinking, we often meet those who imagine that by default the objection must prevail. For instance, we typically hear a quote from Sagan; “extraordinary claims require extraordinary evidence.”

They are wrong, and wrong based on the issue of consistency. For, properly, claims only require ADEQUATE not extraordinary evidence — on pain of inconsistency between standards on what we accept and those on what we reject. And, in the empirical world of science, our evidence and arguments are provisional, so we look on which explanation is best or most credible relative to accounting for the material facts, being coherent and being explanatorily powerful but bot simplistic or ad hoc.

Otherwise, we are simply begging the question, and may be guilty of the most blatant inconsistencies.

2] Fisherian Hyp testing and inference to design

This immediately exposes the problem on rejecting inference to design on observing CSI.

For, routinely, we characterise estimated distributions for events, and when they are sufficiently far out into the tails [usually at 1 in 20 or 1 in 100 levels] we accept that chance — the usual null hypothesis — is an inadequate explanation and revert to agency or natural regularity depending on the case in view. [Of course we run risks of errors: accepting chance when we should reject it, or rejecting it when we should accept it; but since when is the risk of error a new, or even an avoidable, thing?]

As someone in that Feb thread said, as I recall, Dembski’s 1 in 10^150 is just about as conservative a rejection region as he has ever seen. In short, only the most extraordinary cases will be rejected, relative to the chance null hyp!

So, let us see the selective hyperskepticism at work here for what it is.

3] Now, what of [F]CSI and the explanatory filter?

Here we look at two criteria, having first insisted on contingency so that natural regularities do not determine the outcome:

[1] complexity, in the sense of being beyond 500 or so bits of information storing capacity,

[2] specification, in the sense of fitting with an independently known pattern; in the cases of interest, a FUNCTIONAL pattern. (Sure, any 500 coin set is equi-probable, but when someone tells you he just tossed and lo THTH .. TH appeared, no-one will believe him! For excellent reason. Just as Moshe didn’t believe Aaaron when he said they just tossed the gold in the fire and the golden calf “just came out” . . . the first recorded inference to design!]

In short, it is like having miles of a wall, with just one fly on it in a 100 yard stretch. then, bang, a bullet hits the fly. A lucky shot, or an aimed shot? [And why do you plunk for the "lucky shot" explanation?]

4] Relevant cases:

In the case of bio-information, DNA ranges 500k to in excess of 3 billion storage units, each capable of storing 2 bits of information. Even in bacteria, cutting out below about 360 k of storage units, destroys bio-function.

But, 4^360,000 ~ 4.0 *10^216,741 bits. That’s many orders of magnitude beyond 500 bits!

Then too, the functionality of the DNA chain’s stored information is quite easily observed: do you have a viable life form that can feed, move, reproduce etc in appropriate environments?

So, how did we hit this lonely fly on the wall, given the obvious deep isolation of the functional states in the overall configuration space? [And cf here the way that statistics is deployed in thermodynamics to ground the 2nd law of thermodynamics. My nanobots and microjets example in the always linked through my name will help see this. In short I am pointing to selective hyperskepticism at work, again. Probability hurdles are as real as potential walls!]

–> Of course, introducing the ideas that we can only infer to material entities in science simply begs the question.

–> Similarly, trying to speculate on a vastly wider unobserved cosmos than what we see is a resort to speculative metaphysics, and is ad hoc to try to rescue a hyp that is preferred but otherwise in deep trouble on accounting for observed facts.

GEM of TKI

PS: Onlookers, look at my always linked through my handle for updated details, esp in the appendix on thermodynamics, after a follow-up debate with Pixie in my own blog.

THat filter strikes again.

Trying:

Okay . . .

“Once more unto the breach, dear friends . . .”

I see the issue has now “evolved” to the question of both complexity and specification. I will make a few remarks on that, but first, on the issue . . .

1] The problem of selective hyperskepticism

In Western thinking, we often meet those who imagine that by default the objection must prevail. For instance, we typically hear a quote from Sagan; “extraordinary claims require extraordinary evidence.”

They are wrong, and wrong based on the issue of consistency. For, properly, claims only require ADEQUATE not extraordinary evidence — on pain of inconsistency between standards on what we accept and those on what we reject. And, in the empirical world of science, our evidence and arguments are provisional, so we look on which explanation is best or most credible relative to accounting for the material facts, being coherent and being explanatorily powerful but bot simplistic or ad hoc.

Otherwise, we are simply begging the question, and may be guilty of the most blatant inconsistencies.

2] Fisherian Hyp testing and inference to design

This immediately exposes the problem on rejecting inference to design on observing CSI.

For, routinely, we characterise estimated distributions for events, and when they are sufficiently far out into the tails [usually at 1 in 20 or 1 in 100 levels] we accept that chance — the usual null hypothesis — is an inadequate explanation and revert to agency or natural regularity depending on the case in view. [Of course we run risks of errors: accepting chance when we should reject it, or rejecting it when we should accept it; but since when is the risk of error a new, or even an avoidable, thing?]

As someone in that Feb thread said, as I recall, Dembski’s 1 in 10^150 is just about as conservative a rejection region as he has ever seen. In short, only the most extraordinary cases will be rejected, relative to the chance null hyp!

So, let us see the selective hyperskepticism at work here for what it is.

Pause . . .

GEM of TKI

Continuing . . .

3] Now, what of [F]CSI and the explanatory filter?

Here we look at two criteria, having first insisted on contingency so that natural regularities do not determine the outcome:

[1] complexity, in the sense of being beyond 500 or so bits of information storing capacity,

[2] specification, in the sense of fitting with an independently known pattern; in the cases of interest, a FUNCTIONAL pattern. (Sure, any 500 coin set is equi-probable, but when someone tells you he just tossed and lo THTH .. TH appeared, no-one will believe him! For excellent reason. Just as Moshe didn’t believe Aaaron when he said they just tossed the gold in the fire and the golden calf “just came out” . . . the first recorded inference to design!]

In short, it is like having miles of a wall, with just one fly on it in a 100 yard stretch. then, bang, a bullet hits the fly. A lucky shot, or an aimed shot? [And why do you plunk for the "lucky shot" explanation?]

4] Relevant cases:

In the case of bio-information, DNA ranges 500k to in excess of 3 billion storage units, each capable of storing 2 bits of information. Even in bacteria, cutting out below about 360 k of storage units, destroys bio-function.

But, 4^360,000 ~ 4.0 *10^216,741 bits. That’s many orders of magnitude beyond 500 bits!

Then too, the functionality of the DNA chain’s stored information is quite easily observed: do you have a viable life form that can feed, move, reproduce etc in appropriate environments?

So, how did we hit this lonely fly on the wall, given the obvious deep isolation of the functional states in the overall configuration space? [And cf here the way that statistics is deployed in thermodynamics to ground the 2nd law of thermodynamics. My nanobots and microjets example in the always linked through my name will help see this. In short I am pointing to selective hyperskepticism at work, again. Probability hurdles are as real as potential walls!]

–> Of course, introducing the ideas that we can only infer to material entities in science simply begs the question.

–> Similarly, trying to speculate on a vastly wider unobserved cosmos than what we see is a resort to speculative metaphysics, and is ad hoc to try to rescue a hyp that is preferred but otherwise in deep trouble on accounting for observed facts.

GEM of TKI

PS: Onlookers, look at my always linked through my handle for updated details, esp in the appendix on thermodynamics, after a follow-up debate with Pixie in my own blog.

Try again part 2:

3] Now, what of [F]CSI and the explanatory filter?

Here we look at two criteria, having first insisted on contingency so that natural regularities do not determine the outcome:

[1] complexity, in the sense of being beyond 500 or so bits of information storing capacity,

[2] specification, in the sense of fitting with an independently known pattern; in the cases of interest, a FUNCTIONAL pattern. (Sure, any 500 coin set is equi-probable, but when someone tells you he just tossed and lo THTH .. TH appeared, no-one will believe him! For excellent reason. Just as Moshe didn’t believe Aaaron when he said they just tossed the gold in the fire and the golden calf “just came out” . . . the first recorded inference to design!]

In short, it is like having miles of a wall, with just one fly on it in a 100 yard stretch. then, bang, a bullet hits the fly. A lucky shot, or an aimed shot? [And why do you plunk for the "lucky shot" explanation?]

pause 2 . . .

Thread locks up on continuation . . . maybe a new spam filter will help, folks?

GEM of TKI

The real issue or lottery is that there could be an enormous number of possible starting points for life all of them of incredibly low probability and the one that happened or “won the lottery” was just one of these starting points.

Jerry, I appreciate your posts and demands for ever clearer explanations. They only help improve our articulation.

It really comes down to what is the most reasonable explanation. It’s theoretically possible for wind and rain to etch the alphabet on a rock, but if you found the alphabet etched on a rock would you really, really believe it was done by wind and rain?

The materialists arguing that life came by known natural causes is akin to arguing that wind and rain etched not just an alphabet but the works of Shakespeare on a rock.

It gets to the point where you have to shake your head, understand they are fools, then go on to more practical things.

kariosfocus — GREAT POST.

or half-post anyway

Hey kairosfocus,

What does “GEM of TKI” mean? It sounds like a Hip-Hop crew shout out:

“Yo, this is the G.E.M., representing TKI to the fullest. Tha Killa Instinct crew remains from 2007 until the Heat Death!”

lol. Seriously though, is that just your initials and the shire you come from?

An addition/correction to a previous post of mine when I responded to this (in part):

But if you choose (specify) a sequence like “CAT” BEFORE you start grabbing the letters, then the likelihood that you will pick that particular sequence will drop significantly.

I forgot to point out that this is not technically true. Picking the sequence does not change the probability of that sequence. It does change the probability of a “success” in your trials, but only because you changed what defined a success.
In the first case, your specified set was “any three letter word” as opposed to “that” three letter word.

Again, these probabilities are only really meaningful when the set of desired outcomes is specified BEFORE the trials. To specify them AFTER the trial has run vastly confuses the discussion. Besides, it is not an analogy to how evolution works, since evolution is not completely random. Using pure probability on the end result only is not a realistic “test” of the theory.

Just as Moshe didn’t believe Aaaron when he said they just tossed the gold in the fire and the golden calf “just came out” . . . the first recorded inference to design!

LOL!

Kairosfocus,

Thank you for your posts, I always appreciate your insights.

Just so you know, I am not trying to undermine ID with my questions regarding low-probability and specification. They are honest questions, places I feel IDers can give clearer answers.

The answers given so far go towards answering them. So I appreciate them.

tribune7,

You do not have to convince me that the process could have happened by accident, chance, law or whatever non-intelligent process someone names. But I find the ID answers to the materialists objections/claims often vague and simplistic. While I generally understand the concept of CSI, I have yet to see a convincing explanation of it that is not picked apart.

The materialist will argue that it all happened in steps over deep time because that is the way they handle the low probabilities. They will sometimes even assert it is a certainity given enough time. They haven’t any proof of there ever existing any steps but they dare you to prove it couldn’t have happened this way. ID should focus much of its attention on refuting the step approach. That is the best way I believe to refute the lottery cop-out instead of quoting just absurdly low probabilities.

Also I once listened to a shrill student challenging ID by saying where is your proof when she said someone like Darwin was a real scientist who meticulously gathered empirical data to support his hypotheses. The answer was look at the complexity of the cell when the answer should have emphasized that Darwin had no empirical evidence for his hypotheses and that exactly zero new species have been confirmed to have arrived on the planet by natural selection.

The main argument for Darwin these days is common descent when even if you accept common descent there is no evidence that it ever happened in a gradual fashion nor is any other mechanism indicated.

I have yet to see a convincing explanation of it that is not picked apart.

Just because you can’t convince someone doesn’t mean you aren’t right and haven’t made a reasonable case.

People are self-delusional. The materialist is self-delusional. You can never convince them. Just because your opponent says you haven’t won the argument doesn’t mean you haven’t.

A good sign is when they start resorting to ridicule.

30 percent of the Democratic Party believes that President Bush knew beforehand of the attacks on the World Trade Center.

You can never convince most of them otherwise.

Debating a materialist is like debating one of those people.

Hi Folks . . .

OOPS: From it won’t post to it multiple posts! [Pardon . . . BTW, Trib 7, how do you get those neat smileys to post at UD?]

Next, thanks for the many kind words.

Atom, FYQI: GEM is the acrostic formed by my initials, and TKI is my organisation. I hail from, live and work in the Caribbean — currently volcano island, now with a suspiciously quiet ol’ smokie . . .

Now on a few follow up points:

1] Jerry & Trib 7: The real issue or lottery is that there could be an enormous number of possible starting points for life all of them of incredibly low probability and the one that happened or “won the lottery” was just one of these starting points

The real issue is isolation of the islands of functionality.

As John Leslie pointed out long ago now, whether or no there are many regions of the wall with flies, and even regions positively carpeted ans swarming with then, when we see quite isolated regions beyond the reasonable reach of a random walk or random targeting, then the wondrousness of hitting the target emerges. [In short, LOCALLY isolated islands of function is all we need to defeat the chance hyp.]

And, whatever other life technologies or architectures are possible — maybe even non-physical [dare I say "spiritual?!] — the fact is that in general small random disturbances of the observed bio-functional molecules typically destroy function.

Further, the required information to get to the cluster of functional molecules for life is way, way beyond the sort of 500 or so bit limit on chance increments we have discussed. In short, even having got to life, body plan innovation by RM + NS is seriously problematic. Also, the OOL problem is so intractable that Shapiro recently panned the whole RNA world hyp in Sci Am; with issues of getting to information (which he appar does not see equally apply to his own metabolism first model).

Of course, Leslie was in the main talking about the quasi-infinite array of sub cosmi issue,a nd was in effect highlighting that since local changes in many many key parameters make for a radically life-hostile cosmos, the finetuning issue does not go away so easily.

2] Eric: Picking the sequence does not change the probability of that sequence. It does change the probability of a “success” in your trials, but only because you changed what defined a success.

Go to the head of the class!

The vital point relevant to both origin and body plan level diversification of life, is that “success” is independently “functionally specified.

Next, it is empirically, highly complex and integrated in such a way as to in many cases be credibly irreducibly complex, which strains the capacity of variations and co-opting of parts for other purposes.

Third, it is so information rich [in the Shannon sense of data storing capacity] that it is utterly unlikely for something that meets the three observed constraints to happen by a chance-dominated process in the gamut of the observed cosmos. (The same extends in effect to the formation of a life-habitable cosmos. For details cf my always linked and freshly updated after the exchanges with Pixie and Dave Scott.)

But, routinely, agents produce such entities — even this post is a case in point — i.e we KNOW a credible source for FSCI/CSI.

Therefore on inference to best explanation, the cosmos we observe, the habitable planet we inhabit, and the life [including life that requires mind and morals . . . i.e us] we see on that planet all most credibly due to intelligent agency.

So strong is this case, that it is only by selective hyperskepticism linked to institutional dominance of evolutionary materialism and associated methodological naturalism that blocks this from being he new paradigm.

But, that is coming.

GEM of TKI

BTW, Trib 7, how do you get those neat smileys to post at UD?

colon +dash + right parenthesis

Following up . . .

Had to do family deliveries.

3] Jerry: While I generally understand the concept of CSI, I have yet to see a convincing explanation of it that is not picked apart.

Of course the first problem with that,is just what Trib 7 pointed out: selective hyperskepticism can easily demand an undue degree of “proof” on an inconvenient claim, whilst on other claims that are acceptable tot he agenda, a much lower standard of evidence will do.

But also, let us note:

a] CSI is NOT — repeat, “NOT” — an ID-originated concept. CSI predates ID and in fact is part of the set of emerging ideas on the challenge of the origin of life that helped trigger the emergence of the design school of thought in the ea