Possible vs. Likely

On Nov 27, 10:15 am, hersheyh <hershe...@xxxxxxxxx> wrote:

Again, you confuse what is possible with what is likely . . .

And you think it "likely" that there is only one possible "gap"
involved in going from your "hypothetical unnamed undescribed"
starting point to the final teleologic "system" you name?

You simply can't stop using the word "possible". You just did it
again here. Anything is possible. There are many "possible" gap
distances. Which of these possibilities is most likely without
knowing the answer ahead of time? That's the question here.

You simply assume that the minimum possible gap distance is likely
without actually doing any statistical analysis to support this
assumption - nor have you presented any actual demonstration or direct
observational evidence. All you really have are bald assertions and
just-so stories. That's not science.

It
isn't enough to present what is possible Howard. You have to actually
show that what you're presenting is even remotely likely in a given
span of time . . .

Yes. Which is exactly what the evidence shows for humans and
chimpanzees.

Name a unique functional difference between humans and chimps that
isn't just quantitative but qualitative . . . Then, once you have
done this, show the minimum number of genetic differences needed to
produce this novel functional difference.

The problem here, of course is that no one knows anything about such
functional differences between humans and chimps - at least not
adequately at this point. That is why you use this example all the
time. You're conveniently hiding behind a lack of information. Try
dealing with examples that we know more about - like subcellular
systems.

You're argument is not based on any form of statistical analysis
whatsoever when it comes to known differences in levels of functional
complexity. You're argument is based entirely on what you know could
happen but have no idea how likely it is to happen. That's not
science Howard. That's wishful thinking.

< snip >

You
never even think about what is
likely - only about what is possible.
That's not science Howard.

Your entire argument is equivalent to claiming that it was almost
impossible for Barack Obama to be elected president of the U.S. in
2008 because the likely probability of any individual being elected
president is one over the number of people that meet the legal
requirements (the size of total sequence space that meets the minimum
functional and structural requirements for being president). That is
exactly the argument your math presents, only substituting "total
sequence space that meets the minimum sequence requirements for some
named function that already exists"). It is an ahistorical argument
based on a false premise.

Someone within the population is guaranteed to win the presidency
ahead of time within a certain period of time. Finding a target
within sequence space is not guaranteed to happen in a given span of
time for a given population. The locations of the targets are
unknown. The search algorithm is completely random. The odds of
finding a target in a given span of time depend entirely upon one
thing, one thing that you never even consider: The odds that a
potential target within an unknown location will be within a given
distance of a known starting point.

Even with predicting who will be president, knowing the likely pool
from which the president will be drawn does improve the odds of a
successful pick, but the odds are still there. There are non-winner
options. You don't consider any odds of success vs. failure
whatsoever in your "theory". You don't use any odds analysis at all.
You simply argue that eventually it will be known, with a 1:2 odds of
success, who will be president.

Of course, if you actually know that the starting point is so close to
the target ahead of time, you can get very very accurate. Using this
analogy you actually claim that the odds of finding novel targets in
sequence space in a given span of time can be known to be as high as
50%? - regardless of the level of functional complexity under
consideration?

Do you not understand how strained this argument really is? Again,
you are in fact arguing that what is possible is what is likely
again. In reality, you have no idea as to the actual odds of success
without actually knowing the location of both the starting point AND
the target points in sequence space ahead of time. You don't know the
location of the target point(s) ahead of time like you know the target
point of the presidency ahead of time (or the actual final two
candidates ahead of time).

Determining the odds of success when it comes to finding targets in
sequence space involves at least some understanding of the likely
ratio and distribution of potential targets vs. non-targets in
sequence space at various sizes of sequence space. That's it. Yet,
you don't even consider such odds. You simply state that it is
possible. What about considering the likelihood of what is possible
Howard? That's the science part of the problem.

You're problem is that you assume that the ratios remain essentially
the same regardless of the size of sequence space. You simply assume
that because it is possible to have the minimum gap distance of one at
any size of sequence space that this is also likely. That's not
true. What is possible is not necessarily what is likely. You need
to actually consider the odds of what is likely. You don't do this.


Real analysis of real historical events that have already happened do
not, as you do, assume that all the precursors were equally likely to
exist. It looks for *evidence* of precursors that would made the
event that actually happened possible. In the world of the evolution
of protein systems, that means looking for evidence of useful
subfunctions that parts of the current modern function could do
without major change in function or structure.

You have to consider the odds that you will actually find something
close enough Howard. You can search all you want, but the odds that
you will find something depend upon the nature of sequence space.
And, the search itself demonstrates my very point. Finding targets
within one or two mutations of each other gets exponentially more and
more difficult at higher and higher levels of functional complexity.
This is a fact. Like 3-letter words, lower-level targets are commonly
within one or two character changes of each other. This is not as
often the case for meaningfully beneficial 100-character sequences,
and exponentially less so for 1000-character sequences . . . etc.

You can't simply search without any idea as to what you're likely to
find. That sort of search is based on nothing but wishful thinking -
not science.

Your position can't be falsified, even in
theory, because everything is possible.
Science deals with estimates
of likelihood - not imagined possibilities.

And for historical events, it looks for evidence of how that event
actually happened by *possible* mechanisms,

We are talking about what is likely here, not about what is merely
possible. What is the likelihood that your proposed mechanism did the
job Howard? You don't even consider that question. You simply argue
that "it's possible" like a mantra. That's not science.

not the probability of
recurrence based on a process that is unlikely to be the one that was
actually used, such as the idea that Obama became president by a
random lottery pick from the total sequence space of total sequence
space of legally possible candidates.

As I've told you before, the gap distance in real life is ALWAYS
smaller than the size of total sequence space - always. The pool of
likely starting positions is much smaller than the total space of
potential starting points. This doesn't remove the need for odds
analysis - of how many mutations it will likely take to find a novel
target when the location of potential targets is not already known
ahead of time. All that is known is that they are almost certainly
closer than the maximum possible distance.

Hint: The *real* historical process of Obama becoming president
involved a series of events, some of which were due to chance (such as
the Jack Ryann Senate candidacy imploding in a sex scandal), some of
which involved the character and qualifications of the candidate. The
probability of the last step (or gap) in the process was essentially 1
out of 2. There are ways of identifying evidence for the steps in the
historical process. None of them involve assuming that Obama became
president in a single step gap-crossing with a probability due to
division by the total number of legally possible candidates for
president.

We are talking about a situation where the actual gap distance is not
already known ahead of time Howard. We are talking about predicting
the gap distance. What gap distance is likely? That's the question
here. In order to determine this without actually knowing ahead of
time, you need to do some real science Howard. If you already know,
you don't need science. Since you do NOT already know what the actual
gap distances were, you DO need science. That means you need to start
using some real statistical analysis Howard - some real numbers. You
need to go beyond your mantra that "It's possible" to "It's likely"
based on some real mathematical analysis. I know that's tough for
someone who has spent his whole life telling just-so stories, but
until to go beyond this, what you're doing is not real science.

Sean Pitman
www.DetectingDesign.com

.

Relevant Pages

  • Re: Possible vs. Likely
    ... *You* are assuming a single gap. ... what was the probability that a President would be shot by a ... requirements (the size of total sequence space that meets the minimum ... finding a target in a given span of time depend entirely upon one ...
    (talk.origins)
  • Re: Humpty Dumpty sat on a wall . . .
    ... acid and protein system, which as far as I know no abiogenesis ... specificity requirement comes into play. ... target system that requires at least 1000 highly specified residues? ... considering the size of the sequence space at this level is 1e130. ...
    (talk.origins)
  • Re: Humpty Dumpty sat on a wall . . .
    ... If protein X has 1000 extremely secified ... target system that requires at least 1000 highly specified residues? ... based on gap distance that are in fact known from what currently ... considering the size of the sequence space at this level is 1e130. ...
    (talk.origins)
  • Re: Howard Hersheys Challenge of Sean Pitmans Assumptions
    ... bull pucky about proteins if you don't have any science worth talking ... claiming that the other guy has the problem and not Sean. ... What does the actual data tell you about your bogus notions of protein ... that possible if what you claim about sequence space is at all true? ...
    (talk.origins)
  • Re: Estimated Ratio of Viable Proteins
    ... numerology, bullshit handwaving, apologetics, but not science. ... of viable sequences in sequence space. ... My positions are based, to the extent that there is evidence, on ... You simply assume that the mechanism of RM/NS ...
    (talk.origins)