Bill Nye vs. Ken Ham Debate Catch-All

LarryC wrote:
Well, under your definition the results of rolling a six-sided die are non-random. That's not usually how it's described.

Which of the six numbers are rolled is random. Whether or not you get a 7 is not.

The number you do get is still random though. That there is a limited set of possible results does not make the result within that set non-random.

Stengah wrote:
LarryC wrote:
Well, under your definition the results of rolling a six-sided die are non-random. That's not usually how it's described.

Which of the six numbers are rolled is random. Whether or not you get a 7 is not.

The number you do get is still random though. That there is a limited set of possible results does not make the result within that set non-random.

Well, in addition to that, the results population-wide should be predictable. The individual result is random, just as subatomic behavior is random. But Ohm's law and gas laws are not random. I work with these probabilities - I use both statistics and parameters. I deal with outcomes en masse as well as on an individual basis.

In that mutations as a process of evolution result in a particular set of results considered across thousands and hundreds of thousands of samples, then they are somewhat predictable. Any individual result is random, but the distribution across millions of samples is not.

LarryC wrote:
Stengah wrote:
LarryC wrote:
Well, under your definition the results of rolling a six-sided die are non-random. That's not usually how it's described.

Which of the six numbers are rolled is random. Whether or not you get a 7 is not.

The number you do get is still random though. That there is a limited set of possible results does not make the result within that set non-random.

Well, in addition to that, the results population-wide should be predictable. The individual result is random, just as subatomic behavior is random. But Ohm's law and gas laws are not random. I work with these probabilities - I use both statistics and parameters. I deal with outcomes en masse as well as on an individual basis.

In that mutations as a process of evolution result in a particular set of results, then they are somewhat predictable. Any individual result is random, but the distribution across millions of samples is not.

Population wide results are somewhat predictable, in that if you expose a population to a particular selection pressure, you can reasonably expect that the population will adapt to that pressure, but it's random how they'll adapt to it, and it's possible that they won't ever adapt to it. A specific mutation occurring is not guaranteed though, so it's still random.

IMAGE(http://happynicetimepeople.com/wp-content/uploads/2013/11/pi-is-three.jpg)

@LarryC
You know what random means, though, right?

LarryC wrote:

But Ohm's law and gas laws are not random.

However, genetic interactions across populations do not follow the same patterns as do collisions of molecules, distribution of heat across a surface and the like.

Chumpy_McChump wrote:

@LarryC
You know what random means, though, right?

I'm starting to wonder about that.

Robear wrote:
LarryC wrote:

But Ohm's law and gas laws are not random.

However, genetic interactions across populations do not follow the same patterns as do collisions of molecules, distribution of heat across a surface and the like.

Neither do Ohm's law and Boyle's law follow the same pattern. They follow different patterns.

What I'm saying is that when a gambler gains a lot of money on a casino purely through chance, that's random. When the casino industry posts profits at the end of a fiscal year - that's not.

Stengah wrote:
LarryC wrote:
Stengah wrote:
LarryC wrote:
Well, under your definition the results of rolling a six-sided die are non-random. That's not usually how it's described.

Which of the six numbers are rolled is random. Whether or not you get a 7 is not.

The number you do get is still random though. That there is a limited set of possible results does not make the result within that set non-random.

Well, in addition to that, the results population-wide should be predictable. The individual result is random, just as subatomic behavior is random. But Ohm's law and gas laws are not random. I work with these probabilities - I use both statistics and parameters. I deal with outcomes en masse as well as on an individual basis.

In that mutations as a process of evolution result in a particular set of results, then they are somewhat predictable. Any individual result is random, but the distribution across millions of samples is not.

Population wide results are somewhat predictable, in that if you expose a population to a particular selection pressure, you can reasonably expect that the population will adapt to that pressure, but it's random how they'll adapt to it, and it's possible that they won't ever adapt to it. A specific mutation occurring is not guaranteed though, so it's still random.

You're talking about different things here, actually. A population can adapt to a selection pressure without the benefit of chance mutations. If the population already has a phenotype that responds favorably to the pressure, then the population adapts by increasing the incidence of that phenotype - and that "how" response is predictable.

Incidence of mutations is randomized, but can be predictable across a population if we knew more about genomes and DNA molecules. I suppose it depends on what you mean by "random." You have a non-guaranteed chance of getting the office every day. You could get into an accident and die. But I don't know that it's correct to say "I randomly survived today and got to work."

Is it?

LarryC wrote:

You're talking about different things here, actually. A population can adapt to a selection pressure without the benefit of chance mutations. If the population already has a phenotype that responds favorably to the pressure, then the population adapts by increasing the incidence of that phenotype - and that "how" response is predictable.

It's the initial existence of the trait that responds favorably and it's spread through the population that's random. Once it's there, yes, you can reasonably predict that it will spread if you introduce a pressure that specifically selects for it.

Incidence of mutations is randomized, but can be predictable across a population if we knew more about genomes and DNA molecules. I suppose it depends on what you mean by "random." You have a non-guaranteed chance of getting the office every day. You could get into an accident and die. But I don't know that it's correct to say "I randomly survived today and got to work."

Is it?

That'd depend. Narrowly avoiding an accident or getting into an accident and coming out unscathed would be randomly surviving. The accident itself and how it affected you is random. The accident not occurring in the first place is the default, so it not occurring is not random.

To tie it back to mutation, the mutation occurring in the first place is random (unless an outside force is specifically attempting to force a mutation to happen), as is the exact nature of that mutation (within a limited set of possibilities). Just as you won't turn into a dinosaur by being in a car accident, you're not going to become Spiderman by being bit by an irradiated spider.

You're inching back to my original point, Larry. Mutations are random in their effects within a population, but selection pressures are powerfully non-random.

LarryC wrote:

What I'm saying is that when a gambler gains a lot of money on a casino purely through chance, that's random. When the casino industry posts profits at the end of a fiscal year - that's not.

The composition of winners and losers in a population follows a random distribution curve if the games are actually random (mutations are, individually, you accepted that). So that's the first part - mutations are like winning and losing gamblers as a population. "Winning" and "losing" and "breaking even" are *effects* of playing games, so what we are discussing here are effects. Mutations can be read as "playing the game". And everyone is playing all the time.

Note that your point seems to be that some mutations are favored by conditions and the rules that govern molecular actions, and those are non-random. That's analogous to the different odds of winning in each *type* of game. Some yield more losers than others. And some mutations are expected more than others. But that does not make their *effects* across the population non-random, because most mutations are bad, just as most gamblers are losers. While cells have many ways to reduce or fix bad mutations (many gamblers avoid low-payoff games), they can't deal with all of them (some players will *always* play low-payoff games). So many or most of the non-random *effects* of mutations are deleterious and fall out of the population; they don't affect the random nature of useful mutations across the population to a significant degree in most situations. (There are enough winners in the system that players keep playing, but only in extreme edge cases do we see a cluster of big winners together.)

Selection pressure is the casino's tweaking of the game algorithms to ensure a profit across the population of players. It regulates the number of people willing to play the games (the number of members of the population) after taking the effects of mutations into account.

So while the success or failure of any player is random within the confines of the game rules (the genetic mutations of any individual and their effects are random within the limits of the available gene pool), and the overall distribution of winners within the population follows a random distribution curve, the bias towards taking profits (selection pressure) is non-random in it's effects.

Stengah wrote:

Quote:

Incidence of mutations is randomized, but can be predictable across a population if we knew more about genomes and DNA molecules. I suppose it depends on what you mean by "random." You have a non-guaranteed chance of getting the office every day. You could get into an accident and die. But I don't know that it's correct to say "I randomly survived today and got to work."

Is it?

That'd depend. Narrowly avoiding an accident or getting into an accident and coming out unscathed would be randomly surviving. The accident itself and how it affected you is random. The accident not occurring in the first place is the default, so it not occurring is not random.

Huh. Consider this. We have a population of 100,000 people; of that population, let's say 3 will get into an accident each day. Now, some days it will be 5, and other days 7, and some days zero, but most days will be around 3. That's a bell curve distribution with 3 in the center.

Now, we can look at any individual who *had* an accident and say "Wow, that was just back luck, there was ice on the bridge", or "Wow, that guy had 12 DUIs and hit someone after a night at the bar, that's totally predictable". But we can't look at the population and say "This guy's getting in an accident today". Even though we *know* some of them are pre-disposed to it.

Knowing that some behaviors lead to accidents does *not* make the accident rate across the population follow a non-random distribution curve.

It's also the case that we can't predict on any particular day whether someone will get into an accident. So while their particular accident and it's timing may be *influenced* by non-random factors, it's specific occurrence will fit the random distribution curve in the vast majority of cases.

Note here that "accidents" are beneficial mutations.

LarryC wrote:

A population can adapt to a selection pressure without the benefit of chance mutations. If the population already has a phenotype that responds favorably to the pressure, then the population adapts by increasing the incidence of that phenotype - and that "how" response is predictable.

Selection pressure is non-random. Check.

Incidence of mutations is randomized, but can be predictable across a population if we knew more about genomes and DNA molecules.

And mutations are random. Even though we may eventually learn how to predict them better, they are functionally random at this time.

These two points are the ones I've been making. Whew!

Robear:

I was trying to clarify it that way - that you were equating random with "unknown." I suppose that's a handy shortcut if you're not really into it, but to me it's a lot like "God did it." It's a dead end that presupposes certain things are simply unknowable. The objective of investigation and experiments is to find out things we don't already know. If I were to accept that mutations are simply random, then I would not investigate them further. To be willing to investigate the mechanisms of mutations involves throwing away that assumption.

Which is not to say that the assumption isn't useful - only that it's limiting. I think we've moved beyond it already, tbh.

"I'd like to point out A, B, and C."
"Yeah, but you're forgetting B and C. This is B and this is C."
"I haven't forgotten B and C. Here's some more examples of B and C."
"Look, I'm just trying to point out B."
"I'm already talking about B. Here's some more examples of B."
"You haven't said anything about A though."

LarryC wrote:

If I were to accept that mutations are simply random, then I would not investigate them further. To be willing to investigate the mechanisms of mutations involves throwing away that assumption.

I wasn't saying that they are random in that sense; that was the point of the cite on the previous page. Sigh. I feel like all we've done here is to put people off of the general topic by quibbling over details.

When I asked if you were equating random with unknown, you could've just said yes. Shrug. It's not like people can't skip over posts.

Robear wrote:

I wasn't saying that they are random in that sense; that was the point of the cite on the previous page. Sigh. I feel like all we've done here is to put people off of the general topic by quibbling over details.

I definitely hit the point of "well, time to start skimming the posts instead of really reading them with interest." Seems to be a pattern for some of these threads.

LarryC wrote:

When I asked if you were equating random with unknown, you could've just said yes.

So you didn't understand when he explained his yes again and again in much more detail and specific terms or you decided that it didn't have any bearing on a reality where a 'random outcome' is, by definition, 'unknown?'

Not unknowable. Unknown.

Edit: A little more 'raw' than I would like, but that exchange was a pretty frustrating detour.

Edit2: I think I've identified the nugget that was rubbing me wrong. That method of redefining concepts which is why we even have this thread. The concept of random is pretty well understood and we are quite aware of how very few things are 'truly' random. We lack total perspective and have an annoying tendency to demand permanence in our realities. So once an event has resolved we tend to apply hindsight and assume that since we can track the resolution down to causes that a future manifestation is less random than before. It stops being random once we know how it happens. That's not how it works. Mr. Ham wants to redefine Science because the way it works doesn't fit his explanations for the past. By changing the parameters for Science he tries to invalidate all the work done under the 'wrong' version of Science, even if all the work is still applicable even if they were operating under the 'delusion' of being scientists. If something is random that means it cannot be accurately predicted. It can be guessed. It can be speculated. You can wager. There may even be odds. But if it is random then you can only be correct in hindsight. Once that manifestation resolves and stops being random and part of 'historical science.'
altered to try clarifying that clunky statement. I really need to stop wading into these things during work hours.

Rezzy:

It appears that you have a different understanding of the term as well - definitely different from how Robear and I resolved it. I think that justifies nailing down exactly what we mean and how "random" is meant in mutations. As far as I understood it, it is not what you meant, though I could be wrong about Robear.

This is not so much a redefinition as it is a clarification tangent. Indeed, there's actually nothing wrong with getting your statements all lined up. As much as I don't like Ham's goals, his procedural methods gain a lot more respect from me than what seems to me to be muddy reasoning and outright fallacious reasoning that happens from time to time, even around here.

One of the first things a scientist looks for in a paper is "Definition of Terms." There is a good reason why that is.

LarryC wrote:

One of the first things a scientist looks for in a paper is "Definition of Terms." There is a good reason why that is.

As illustrated here. Sadly, not as much rigor can be expected from forum posts. Sometimes it might be best to take things within the context they're given and move on instead of dragging the thread through iteration after iteration of "You mean *this*?!" Robear's meaning was clear to me from the start. It's painful to see him restating it patiently in so many different ways. Now we're moving on to Rezzy?

LouZiffer:

Rezzy's definition and Robear's clarified definition are different. I'm pointing out that I'm not the only one who reads "random" that way.

To put things back on track, maybe, Ken Ham would like to draw line in application that we don't see in the data. There are mutations in populations, and there are selections across populations, and nowhere in that is there a limit to the changes which can occur over time. It's just not there.

As I pointed out earlier, there are other issues with his claims that would show differently in areas like microbiology.

EDIT: Robear has the right idea. Removing the dead horse and holstering the stick.

Rezzy wrote:

EDIT: Robear has the right idea. Removing the dead horse and holstering the stick.

That was very random of you.

Quintin_Stone wrote:
Rezzy wrote:

EDIT: Robear has the right idea. Removing the dead horse and holstering the stick.

That was very random of you.

Don't holster that stick just yet, Rezzy.

I'm going to flip a coin. Insert the coin into a parboiled potato. Insert the potato into a potato-gun mounted onto a platform that can both tilt and rotate a full 360 degrees on every axis. Around the edges of the platform I will mount between 0 and 100 fireworks tethered to the platform with elastic string carefully selected to have just enough tensile strength to not break when the firework launches. The ropes will have been placed in a pile which was lightly sprinkled between 0 and 5 times from one of two bottles. One bottle contains a weak acid the other bottle contains either a strong acid or water. The elastic rope reacts with the weak acid, the mooring points of the rope react to the water. Both react to the strong acid. The platform is then suspended from an autonomous balloon using unregulated solar cells to determine the altitude of the assembly. The fuses of the fireworks are then attached to an ignition device that reacts to thresholds on one of three criteria: Airspeed. Moisture. Or ambient decibel levels. The potato gun is then attached to an ignition device timed to fire 10 seconds after the fireworks trigger. I then select the top 10 Reddit posts. Divide the number of total posts in each by the number of blue cars I see that day. Take that number and open the CNN homepage in a text only browser. I count down the number of characters equal to the Reddit derived number and then select the immediately following paragraph. I paste that paragraph into google maps and select the first address it suggests. I mail the assembly there and instruct the recipient to release it from the nearest window leading outside as soon as it is delivered.
Then I stand in my backyard with a catcher's mitt and wait.

If it is heads then I will dirty my beatin' stick again, if it is tails then Greedo shot first. You weren't there!

Rezzy wrote:

If it is heads then I will dirty my beatin' stick again, if it is tails then Greedo shot first. You weren't there!

You know that a coin is statistically going to land on heads slightly more often then tails.

kazar wrote:

You know that a coin is statistically going to land on heads slightly more often then tails.

Even when inside a parboiled potato? I'll need to see the evidence.

A new article about the work Richard Lenski has been doing with bacterial evolution (previously mentioned in the Post a picture, argue with me thread, though this one's probably a better home for it).

After testing 10 trillion ancestral cells from early generations, he got no growth. But when he tested cells from the 20,000th generation on, he began to get results, eventually finding 19 mutants that could use citrate as a power source. The results showed that the citrate-eating mutation was most likely not the result of a single mutation, but one enabled by multiple changes over 20,000 generations.

In further testing to determine if the new bacteria were different enough to qualify as a new species, Lenski’s researchers found that beyond changes to the genes responsible for glucose and citrate consumption, other changes had occurred in the organism that had made it less fit to survive in a glucose-only environment,

“We find they are getting less fit in the ancestral niche over time,” Lenski said. “I would argue that citrate users are — or are becoming — a new species.”

Well, now we know the real reason Ken Ham debated Bill Nye.

Ham's been trying to build a 510 foot (300 cubit) replica of Noah's Ark in Kentucky and turn it into a biblically-themed theme park. He needed at least $73 million to break ground on the first phase in May and only managed to raise $14 million from his fellow Creationists.

Ham claimed that the debate helped spur interest in the project and helped him meet his funding goals.

The reality, however, is a bit different. Apparently he got the city of Williamsburg, KY, where the park will be built, to issue between $55 and $62 million in unrated bonds back in December to fund the project.