CORRELATION DOES NOT IMPLY CAUSATION*

>scientific rigor
L0Lno fgt pls

Wait a mo, user is preparing a considered scientific response and will post again shortly.

>correlation dub fun imply causation
>u can't know NUFFIN
>fuckin librals

Not a libtard, but correlation is always evidence of causation unless you knew in advance that the outcomes were going to correlate.

Global warming has other arguments than just correlation.

I disagree with OP on everything except for global warming meme. So does everyone else.

You realize there are computer models of the effect of human activity on the climate right? "Switch" human activity off and you get average temperatures and CO2 concentration, switch it on and you get increased temperatures and CO2 concentration very close to actual measured values.
You're right on account of evolution and nonexistence of God though.

This is not even good bait. Morons.

Why would they execute useful idiots?
A bit of an idealistic opinion of govt m80

>correlation is always evidence of causation
not a statistically significant line of reasoning.

you don't know much statistics, do you?

>If these things are so important and they are so sure of them why can't they prove them with scientific rigor?

We didn't prove causation, we simply falsified every other competing theory until those were the only ones left.

I've always wondered, is evolution falsifiable? How would you go about proving evolution wrong?

Everything (news, data, memory...) should be questioned; choosing only to accept what is reasonable but with continued uncertainty.

I accept it is reasonable that no one can prove negative or positive: of anything or anyone.

All of us have only opinion based on data: such as "santa" that requires at least one more brain circuit to discredit the bad data.

its falsifiable, but everything seems to fit.
I'm not sure how to put it to the test, because everything seems to fit so well. Creationists have been trying for centuries. Darwin's own book is a massive, intricate, and detailed argument of his case, because he knew creationists wouldn't bite. now we have even more evidence.

Exactly my point. You can't come up with a single scenario that evolution wouldn't be able to explain away. It's not a falsifiable theory and therefore not scientific.

Just like using God to explain everything. There isn't a single scenario you can come up with that can't be explained by God.

Both theories are just as valid in my opinion, and neither are scientific. I'm not saying evolution is false, it most certainly exists, it's been observed. But using it to explain the origin of man isn't as scientific as everyone touts it to be.

You just used your own "muh correlation" counter-argument.

Go back to >>>/8gag/

what does correlation and causation have to do with the nonexistence of God?

there's no substantial proof for it, period.

>I've always wondered, is evolution falsifiable? How would you go about proving evolution wrong?
A rabbit fossil from the precambrium would do the job nicely. So would pigs that started to grow wings in the next hundred thousand years (without human meddling).

>every1 is dum but me :^)

>humans caused the nonexistence of God
doesn't even makes sense.

Exactly. You pretty much summed it up

Actually there's nothing inherently impossible about pigs evolving wings if the environment favors it
Though if a pig gave birth to a pig with functioning wings that would be hard evidence against evolution

>Actually there's nothing inherently impossible about pigs evolving wings if the environment favors it
True -- but there is a lot wrong with it happening in this short a timeframe. Complex machinery takes ages to emerge due to evolution; that is a true and hard prediction of evolution as a theory. There is math on time limits involved, and that math says that it's absolutely not going to happen in a hundred thousand years. Such an observation (in the absence of human meddling) would disprove evolution as we know it outright.

Let Co represent "Thing B correlates with thing A"
Let Ca represent "Thing A causes thing B"

P(Co|Ca) = 1 # if thing A causes thing B then B will always correlate with A

P(Ca|Co) = P(Co|Ca)*P(Ca)/P(Co)
=> P(Ca|Co) = P(Ca)/P(Co)

Now, if you know in advance that B will correlate with A, P(Co) is 1 and P(Ca|Co) = P(Ca) i.e the probability that A causes B did not change when you observe B correlate with A. Otherwise, P(Co) is less than 1 and P(Ca|Co) > P(Ca) i.e. the probability that A causes B has increased when you observed B correlate with A.

>you don't know much statistics, do you?
Do you?

I don't think anyone has pointed this out yet, but correlation does imply causation, it just doesn't definitively mean causation. Right?

you're an idiot.
to falsify it you simply find evidence that contradicts it.
currently there is no such evidence.
therefore, it remains the dominant theory.

>P(Co|Ca) = 1 # if thing A causes thing B then B will always correlate with A
False. This would be true if A was the only thing which affects B. A lack of correlation is completely possible even if A causes B, because other factors might make B not occur, and thus no correlation is observed.

I was thinking "cause" as in logical implication, but if that's what we want to look at fine:

Let Co represent "Thing B correlates with thing A"
Let Ca represent "Thing A positively effects B but other factors might make B not occur"

If you don't know in advance that B will correlate with A, then B is merely more likely to correlate with A than with some unknown thing:

P(Co|Ca) > P(Co)
=> P(Co|Ca)/P(Co) > 1
=> P(Co|Ca)*P(Ca)/P(Co) > P(Ca)
=> P(Ca|Co) > P(Ca)

Same result. Happy?

>P(Co|Ca) > P(Co)
This is also false. P(Co|Ca) can be equal to P(Co) if P(Co|-Ca) = 0

But nice try.

dis bait lel

my gott, pure ideology!

this thread definitely needs more cute feet

If P(Co|!Ca) = 0 then

P(Co) = P(Co|Ca)*P(Ca) + P(Co|!Ca)*P(!Ca)
=> P(Co) = P(Co|Ca)*P(Ca)
=> P(Co|Ca)*P(Ca)/P(Co) = 1
=> P(Ca|Co) = 1

So again P(Ca|Co) > P(Ca), unless you already knew absolutely that there was a causal relationship.

So I will amend my original statement: correlation is always evidence of causation unless you knew in advance that the outcomes were going to correlate OR YOU KNEW IN ADVANCE THAT THERE WAS CAUSATION

Happy?

Evolution is a fact. You can observe it happening. Just look around if you have a cat or dog.

Take two groups of males, one has males with big penises (whatever the criteria for that is), other one has males with penises smaller than that.
Give everyone an IQ test.
Group with big penises has a bigger average IQ.
So, either:
>correlation doesn't imply causation
or
>OMG GUISE STUDIES SHOW THAT HAVING A BIG PENIS MEANS THAT YOU HAVE BETTER CHANCES AT HAVING A HIGHER IQ HEHEHEHEHE

dumbass

All scientific evidence is correlative.

The job of the scientist is to construct their experiments such that causation is the only reasonable explanation for observed correlation.

>Microevolution is a fact. You can observe it happening.
ftfy

>drumpf voters

all distinctions between micro- and macroevolution are arbitrary

Honestly this looks like a Reddit poster who went on /pol/ for the first time, or someone who is doing a bad impression. People take all of the shit they say jokingly serious and vice versa.

>basing your """theory""" on deductive """"""reasoning""""""
ayyyyyyy lmaaooooooooooo

causation is the invention of man

>prove them with scientific rigor
>prove ... rigor
Lrn2science fgt pls

Why don't you have your magical ghost *appear*, so that memes and shit end

everyone that has ever died drank water at some point in their life, therefore water kills people

epic

I made a mistake. I should have said that P(Co|Ca) can be equal to P(Co) if P(Co|Ca) = P(Co|-Ca). This in essence is the idea behind correlation does not imply causation.

So your still wrong, you keep assuming what you're trying to prove.

...

It has nothing to do with knowing in advance. Certain things are necessarily going to be correlated without causation, regardless of whether you know they will correlate. For example it just so haired that everyone who pays taxes dies. This does not mean taxes causes death because their lack of causation had no effect on their chance of correlating.

Another issue with your math is that it appears to argue that both A causes B and B causes A in every case where they are correlated, since correlation is transitive.

source or bullshit on the computer model

"correlation does not imply causation" is a definite example of mount stupid, because hardly anyone who says this phrase seems to understand that correlation, when other variables are either controlled for or blocked for ( i.e.e spread out randomly and evenly over a large enough number of data) does provide strong evidece of causation.

If liberal arts are considered a general education requirement in USA universities then some experimental design and statistics DEFINITELY should be

Actually there is. Pigs arent built for flying. If they evolved to fly they would have to xhange so much you might aswell not call it a pig.

Thats only if u use evolution like an idiot. There is plenty falsifiable about evolution. No theory is immune to adhoc ammendment

...

>CORRELATION DOES NOT IMPLY CAUSATION*
>*except for evolution, human caused global warming, and the nonexistence of God
None of those imply causation from correlation.

Showing DNA having zero relation to ancestry would have been a great way to falsify common ancestry.

These people are out there, and they vote. I'm so extremely happy they'll be the ones losing their healthcare.

>>>>/pol
>>>/x

I can't decide which board is worse, /pol/ or /x/. On one hand, /x/ doesn't seem too concerned with supporting evidence. They're basically telling stories, a type of digital folklore. Now, /pol/, being one collective mass of idiocy, demagogy, and cuckolding (they claim to hate Jewish people but vote for an extremely pro-Israel candidate who is going to fill his cabinet with Jewish folks) likes to pretend they understand things.

First, something can be evidence without being conclusive evidence. Second, this would surely be a case where you knew in advance that the things would correlate.

Just like the guy above you are giving a bad example because presumably you know in advance that tax paying and dying are going to correlate. Pick an example that actually gives what I ask for: a correlation where you don't know in advance that the things will correlate or that there is a causal relationship.

Or if you are somebody who doesn't know that tax paying and dying are going to correlate, then you are simply failing to adjust your beliefs properly if you do not increase your belief in causation when you observe the correlation.

>Another issue with your math is that it appears to argue that both A causes B and B causes A in every case where they are correlated, since correlation is transitive.
Why is that an issue?

>I should have said that P(Co|Ca) can be equal to P(Co) if P(Co|Ca) = P(Co|-Ca)
But P(Co|Ca) = P(Co|-Ca), which is the same as saying P(Co) = P(Co|Ca), is not consistent with what Ca represents.

>Let Ca represent "Thing A positively effects B but other factors might make B not occur"

Because those other factors might make B not occur the chance that B will correlate with A doesn't increase to 100%, but you are saying that the chance B will correlate with A literally doesn't increase at all.

Let's consider a concrete example. Flipping a light switch positively effects the lightbulb lighting up but other factors might make the lightbulb not light up (e.g. a mouse could have chewed up the wiring, the bulb could have burned out on you, etc.). Saying that P(Co) = P(Co|Ca) in this case is saying that if you knew nothing at all about the light switch and the light bulb and then learned this fact about the relationship between the switch and the bulb, your assignment of likelihood that the light bulb will light up when the switch is flipped should not increase at all.

Do you actually agree with that?

but what stretches the space? what is the link between *thing* (particles) and force? because force acts on things, and things don't just magically act differently on their own.

Love it when SimpletonScience rewrites history to hide the utter failure of Climate "Science" predictions.

Here are the actual predictions of James Hansen, former director of NASA GISS. Which show an utter failure; Scenario C is where there's a huge cutback in CO2 production (best fit, but this didn't happen with CO2 production); Scenario A doesn't fit the data at all. Yet that's what happened in terms of CO2. An increase in anthropogenic CO2.

And why does SimpletonPseudoScience start the graph in 1990? Oh yeah, because the divergence of temperatures was already beginning, and SkS wanted to hide it. To cover up this fraudulent behavior SkS and Gavin Schmidt said much baseline is wrong. Even Though They Changed the Baseline! The UN IPCC diagram clearly shows the baseline to be 1979. Pic related, from UN IPCC AR4, Fig 10-26.

Source: ipcc.ch/publications_and_data/ar4/wg1/en/figure-10-26.html
Graph In the lower left hand corner of the page, enlarged with updated data.hide that. Pic related. The actual predictions from the IPCC AR4, with added instrumental data (and enlarged).

>much baseline is wrong.
"muh the baseline is wrong."

SkS and Schmidt have been thoroughly debunked on their bogus "muh baseline" excuse.

Gavin Schmidt gets roasted:
climateaudit.org/2016/04/19/gavin-schmidt-and-reference-period-trickery/

Nuttercelli destroyed:
wattsupwiththat.com/2016/02/20/the-guardians-dana-nuccitelli-uses-pseudo-science-to-libel-dr-john-christy/

>nb4 denier blog. Don't care. Try facts and logic instead of ad hominem. If you can.

>But P(Co|Ca) = P(Co|-Ca), which is the same as saying P(Co) = P(Co|Ca), is not consistent with what Ca represents.
It's not consistent with what you incorrectly believe Ca represents.

>Because those other factors might make B not occur the chance that B will correlate with A doesn't increase to 100%, but you are saying that the chance B will correlate with A literally doesn't increase at all.
Wrong. I'm saying it doesn't necessarily increase. You assumed it does.

>Flipping a light switch positively effects the lightbulb lighting up but other factors might make the lightbulb not light up (e.g. a mouse could have chewed up the wiring, the bulb could have burned out on you, etc.). Saying that P(Co) = P(Co|Ca) in this case is saying that if you knew nothing at all about the light switch and the light bulb and then learned this fact about the relationship between the switch and the bulb, your assignment of likelihood that the light bulb will light up when the switch is flipped should not increase at all.
You are getting very confused here. Learning of the causal relationship would make P(Ca) = 1, which would mean P(Co) = P(Co|Ca)(1)+P(Co|-Ca)(0) = P(Co|Ca). Before you know of the causal relationship, whether P(Co) = P(Co|-Ca) is dependent on several factors which your example does not elucidate.

>Just like the guy above you are giving a bad example because presumably you know in advance that tax paying and dying are going to correlate.
I said they just happen to correlate. This has no bearing on the argument, which is that it is possible for causation to have no effect on correlation. You assumed this is never the case, so only one example is necessary to disprove your assumption.

And try to remember what we're arguing about, "correlation does not imply causation." The correlation is already known.

>Love it when SimpletonScience rewrites history to hide the utter failure of Climate "Science" predictions.
So you aren't going to respond to the fact that IPCC projections from 1990 are accurate? Yeah that's what I thought.

Instead you focus on Hansen's 1988 predictions which used a high climate sensitivity and you claim that scenario A describes true emissions when it doesn't. In fact, none of the scenarios describe the emissions that actually occurred. Climatologists don't make predictions, they make projections based on certain variables like solar activity, GHG emissions, and volcanic activity. Judging a model based on the wrong variables is misleading.

>And why does SimpletonPseudoScience start the graph in 1990?
Because that's when the projection was made.

>The UN IPCC diagram clearly shows the baseline to be 1979. Pic related, from UN IPCC AR4, Fig 10-26.
How exactly did you determine the baseline from that blown up graph? You didn't, you just made up the baseline. And the data you overlayed can't even stay inside the blown up line.

>climateaudit.org/2016/04/19/gavin-schmidt-and-reference-period-trickery/
So this argues that the baseline doesn't matter since the only point is to compare trends. If that's the case then why did the original graph not have trendlines? The point was to create divergence between the data and the models by adding the heat from an anomalous year to the models. So this doesn't respond to the criticism.

>wattsupwiththat.com/2016/02/20/the-guardians-dana-nuccitelli-uses-pseudo-science-to-libel-dr-john-christy/
This makes the same non-response as above. It then responds to the fact that the graph shows no error bars with the non sequitur that the IPCC 1990 predictions are inaccurate, which we already know from is false. It then responds to the fact radiosonde data is improperly averaged by talking about... satellite data. It then responds to the fact that the graph uses data from high in the troposphere by saying that the IPCC says things about the upper troposphere. Monckton fails to justify the use of high troposphere data to judge surface temperature projections because he gets caught up in a tangent repeating one of his denier memes. He then makes the same misrepresentations about IPCC projections that you made by using false scenarios as strawmen. All in all, a bunch of debunked memes and non-sequiturs that make Monckton look senile.

>>Love it when SimpletonScience rewrites history to hide the utter failure of Climate "Science" predictions.
>So you aren't going to respond to the fact that IPCC projections from 1990 are accurate? Yeah that's what I thought.
There not. see the gif here: This is how pathetic the warmists are. They first had a 1979 baseline which failed miserably. So in preparation for IPCC AR5 they used a 1990 baseline. But those failed miserably. So that added an incredible amount of variance to the models; making the 'predictions' go all over the place and thus guaranteeing unfalsifiability. Finally, SimpletonScience tells a bogus story here: Which is debunked by the IPCCs own graphs,

>Wrong. I'm saying it doesn't necessarily increase. You assumed it does.
Can you describe a concrete example where it wouldn't?

I gave an example that models the situation and asked you if you agree with it. Do you?

>You are getting very confused here. Learning of the causal relationship would make P(Ca) = 1, which would mean P(Co) = P(Co|Ca)(1)+P(Co|-Ca)(0) = P(Co|Ca)
No, seriously, you're the one who is confused. In the expression "P(Co) = P(Co|Ca)", "P(Co)" is the prior probability of Co. As in prior to learning anything about Ca. The expression is saying that the probability of Co before learning about Ca is the same as the probability of Co conditional on Ca being true.

What you've done above is move the situation forward in time to a point after we've learned that Ca is true and then derive the tautology that after we know that Ca is true, the probability of Co is the same as the probability of Co conditional on Ca being true. Well duh. Obviously once you know something, conditioning on it again isn't going to change anything.

>Before you know of the causal relationship, whether P(Co) = P(Co|-Ca) is dependent on several factors which your example does not elucidate.
Like what? Again, can you give a concrete example of a situation in which two things A and B have a causal relationship, yet learning about this relationship should not increase the probability of B correlating with A?

>Wrong. I'm saying it doesn't necessarily increase. You assumed it does.
Can you describe a concrete example where it wouldn't?

I gave an example that models the situation and asked you if you agree with it. Do you?

>You are getting very confused here. Learning of the causal relationship would make P(Ca) = 1, which would mean P(Co) = P(Co|Ca)(1)+P(Co|-Ca)(0) = P(Co|Ca)
No, seriously, you're the one who is confused. In the expression "P(Co) = P(Co|Ca)", "P(Co)" is the prior probability of Co. As in prior to learning anything about Ca. The expression is saying that the probability of Co before learning about Ca is the same as the probability of Co conditional on Ca being true.

What you've done above is move the situation forward in time to a point after we've learned that Ca is true and then derive the tautology that after we know that Ca is true, the probability of Co is the same as the probability of Co conditional on Ca being true. Well duh. Obviously once you know something conditioning on it again isn't going to change anything.

>Before you know of the causal relationship, whether P(Co) = P(Co|-Ca) is dependent on several factors which your example does not elucidate.
Like what? Again, can you give a concrete example of a situation in which two things A and B have a causal relationship, yet learning about this relationship should not increase the probability of B correlating with A?

(Other than ones where, like I've said, you already have definite knowledge of the correlation)

>>And why does SimpletonPseudoScience start the graph in 1990?
>Because that's when the projection was made.
Wrong, that's what the projection was changed to, after the utter failure of the original prediction from UN IPCC AR4 which starts at 1979. "hindsight" is not "prediction."

>>The UN IPCC diagram clearly shows the baseline to be 1979. Pic related, from UN IPCC AR4, Fig 10-26.
>How exactly did you determine the baseline from that blown up graph? You didn't, you just made up the baseline.

You idiot, the zero point of that graph is at about 1979. The starting value of temperature anomaly predictions is, of course, zero. That makes the baseline 1979.

>And the data you overlayed can't even stay inside the blown up line.
Gosh, what is smoothing.

So Nice to talk to an autist.

First this Second, evolution defined broadly as "change of allele frequency over time" is observable in anything with a short enough lifespan. MRSA is probably the most commonly cited such example, we saw it happen. It's also a fair proof for natural selection, because we saw the mutations happen and we saw them proliferate because of an advantageous trait. Evolution describes nothing beyond that which is already observable with the right technology.

>Can you describe a concrete example where it wouldn't?
I already did. Can you just prove mathematically that P(Co) < P(Co|Ca) instead of making examples? you can't prove a blanket statement like this with examples.

>I gave an example that models the situation and asked you if you agree with it. Do you?
I already explained that your example lacks the information to calculate the relevant probabilities, so I neither agree nor disagree. You also make the same mistake of representing my argument as "it won't increase" when it should be "it doesn't have to increase."

>No, seriously, you're the one who is confused. In the expression "P(Co) = P(Co|Ca)", "P(Co)" is the prior probability of Co. As in prior to learning anything about Ca.
No, P(Co) is simply the probability of A correlating with B. Nowhere did we say anything about it being prior, that would depend totally on context. And I already talked about both cases of before and after learning about the causation, so this seems like an irrelevant point.

>What you've done above is move the situation forward in time to a point after we've learned that Ca is true and then derive the tautology that after we know that Ca is true, the probability of Co is the same as the probability of Co conditional on Ca being true. Well duh. Obviously once you know something, conditioning on it again isn't going to change anything.
So what? I don't see what the point is in repeating what I said.

>Like what? Again, can you give a concrete example of a situation in which two things A and B have a causal relationship, yet learning about this relationship should not increase the probability of B correlating with A?
I already described the situation completely, which is whenever P(Co|Ca) = P(Co|-Ca). An example of this would be when P(Co|Ca) = 1/2 and P(Co|-Ca) = 1/2. If you would like to prove this is impossible, go ahead. Until then, your argument is non-mathematical.

>Instead you focus on Hansen's 1988 predictions which used a high climate sensitivity and you claim that scenario A describes true emissions when it doesn't. In fact, none of the scenarios describe the emissions that actually occurred.

Now you're just flat out lying. CO2 output has GROWN almost every year in the past 50 years or so. That's scenario A. Pic related; CO2 emission GROWTH per year.

Data Source: wri.org/resources/data_sets

P.S. The fact that you'll just brazenly make up crap to defend an unfalsifiable belief system is a sure sign of a paid shill.

>Wrong, that's what the projection was changed to, after the utter failure of the original prediction from UN IPCC AR4 which starts at 1979. "hindsight" is not "prediction."
You just contradicted yourself in two sentences. AR4 was in 2007. So is hindsight a prediction or not? Did the "prediction" start at 1979 or 2007?

>You idiot, the zero point of that graph is at about 1979. The starting value of temperature anomaly predictions is, of course, zero. That makes the baseline 1979.
You idiot, "the zero point" or starting value of the graph is not the baseline. A baseline is normally the average of a range of temperatures. It will cross the data at some point, but where it cross the data is not necessarily the baseline. And you can't even tell where it crosses the data, since you blew up a tiny graph. So either you purposefully made up the baseline being 1979, or you have no idea what you're talking about. Which is it?

>Gosh, what is smoothing.
Smoothing wouldn't make some part more extreme and others less extreme. Nice deflection dumbass.

>This makes the same non-response as above. It then responds to the fact that the graph shows no error bars with the non sequitur that the IPCC 1990 predictions are inaccurate, which we already know from is false

You "know" from a post-hoc rewrite by Simpleton Science? Sorry buddy, the IPCC said it themselves here
Stop referencing that dishonest clod John Crook. He has no credibility

JOHN COOK DEBUNKED:
motls.blogspot.com/2010/03/john-cook-skeptical-science.html

JOHN COOK LIES
hiizuru.wordpress.com/2014/01/27/john-cook-is-a-filthy-liar/
www.forbes.com/ sites/ jamestaylor/ 2013/ 05/ 30/ global-warming-alarmists-caught-doctoring-97-percent-consensus-claims /
wattsupwiththat .com/2012/02/03/monckton-responds-to-skeptical-science/
impactofcc.blogspot.com/2013/05/john-cook-et-al-willfully-lie.html
populartechnology.net/2012/03/truth-about-skeptical-science.html

>Smoothing wouldn't make some part more extreme and others less extreme. Nice deflection dumbass.
You're the dumbass. I'm talking about the smoothness of the originally graphed data vs. The higher variability of the (less smoothed) overlaid data.

You really are an autist.

>Now you're just flat out lying. CO2 output has GROWN almost every year in the past 50 years or so. That's scenario A.
so the scenarios only describe CO2 emmissions? No, it's about all GHG forcings! So again we must ask, did you know that and only mentioned CO2 to be misleading, or did you not know that and are you just making arguments from rank ignorance?

But yeah, I'm clearly the liar since you utterly misunderstand what Hansen's scenarios are.

>P.S. The fact that you'll just brazenly make up crap to defend an unfalsifiable belief system is a sure sign of a paid shill.
The pot calling the kettle black, and I just proved it.

>You just contradicted yourself in two sentences. AR4 was in 2007. So is hindsight a prediction or not? Did the "prediction" start at 1979 or 2007?
Autist Boy, the graph, of course, starts with hindsight and ends with a projection. Stop getting your panties in a bunch. And leave the basement more often.

>I'm talking about the smoothness of the originally graphed data vs. The higher variability of the (less smoothed) overlaid data.
So am I, dumbass. The original data is more extreme that the overlaid data in some places and less extreme in others.

>Autist Boy, the graph, of course, starts with hindsight and ends with a projection.
>Wrong, that's what the projection was changed to, after the utter failure of the original prediction from UN IPCC AR4 which starts at 1979.
Arguing with yourself is a clear sign of confused thinking. It's hard to keep track of the facts when you keep making shit up, like the prediction "starting at 1979." Why even mention 1979?

Duh, I even said I agreed with that. Using evolution as an explanation for the origin of man is not observable and I thought not falsifiable.

However
Has a point about finding a fossil before it's time. I wonder if scientists would admit evolution isn't the source of life if they found such a fossil.

>I already did.
No you didn't. Concrete as in a plausible real world example. Like I gave with the light switch example.

>Can you just prove mathematically that P(Co) < P(Co|Ca) instead of making examples? you can't prove a blanket statement like this with examples.
The problem is our dispute isn't entirely mathematical. It's like if we were modelling a train that moves at one hundred miles an hour and so I wrote down "trainSpeed = 100" and you asked me to prove mathematically that trainSpeed = 100 rather than 200 or 7.

>I already explained that your example lacks the information to calculate the relevant probabilities, so I neither agree nor disagree.
Please name the missing information. What would that information have to be so that you would agree with what I wrote?

>You also make the same mistake of representing my argument as "it won't increase" when it should be "it doesn't have to increase."
Fair enough. But again, can you give a real world example where it would not increase?

>No, P(Co) is simply the probability of A correlating with B. Nowhere did we say anything about it being prior, that would depend totally on context.
What? Firstly, there is no such thing as a probability without context. Probabilities change as you learn information about the world. That's what we're talking about: how the probability of a causal relationship between two things changes as you learn about their correlation. So secondly, of course P(Co) is the prior, with P(Co|Ca) being the posterior. You're manipulating the symbols of conditional probability, but you seem to not understand what they represent:

en.wikipedia.org/wiki/Conditional_probability

>So what? I don't see what the point is in repeating what I said.
So what you wrote there doesn't apply to what I had said in my light switch example because what you wrote was based on a different state of knowledge than the one at hand.

...

do you realize that what you're saying makes no sense whatsoever?
conclusions from the AR4 models were published in 2007. they're pretty accurate through 2009, including a long backcast and a brief assessment of forecast, pic related.
but you claimed that AR4 failed badly after 1979. and when it was pointed out to you that AR4 didn't make any predictions about the interval you spoke of, your response was to call the guy who pointed it out names. presumably this is because you don't actually have any evidence to back up your claim on account of you just spouting whatever bullshit sounds good to you. the models are actually pretty good, for the most part, but that's never stopped deniers from hysterically screaming otherwise.

>No you didn't. Concrete as in a plausible real world example. Like I gave with the light switch example.
OK, two black boxes are tested in a lab. one has a switch which causes a light on the box to turn on. The other has a switch which is not connected to a light on the box. It is observed that 1/2 of the time the first box's switch is turned on, the light turns on. It is observed that 1/2 the time the second box's switch is turned on, the light turns on. Happy?

>The problem is our dispute isn't entirely mathematical. It's like if we were modelling a train that moves at one hundred miles an hour and so I wrote down "trainSpeed = 100" and you asked me to prove mathematically that trainSpeed = 100 rather than 200 or 7.
No it isn't. This isn't an empirical matter. You made a mathematical claim that you have yet to prove. That's it.

>Please name the missing information.
What is P(Co|Ca) and what is P(Co|-Ca)? Or give information that would allow us to calculate these.

>What? Firstly, there is no such thing as a probability without context.
Not what I said. I said whether it describes the probability before or after learning about the causation is dependent on context.

>That's what we're talking about: how the probability of a causal relationship between two things changes as you learn about their correlation.
I already described what would happen before and after given the information. You keep ignoring this and then using the lack of what you're ignoring to claim I don't understand conditional probability. This is laughable.

>So what you wrote there doesn't apply to what I had said in my light switch example because what you wrote was based on a different state of knowledge than the one at hand.
*sigh* you made a claim here about whether P(Co) would increase after knowing about the correlation, so clearly it applies. If you are going to say that something increases, you have to know what it was before and after, don't you?

>>Now you're just flat out lying. CO2 output has GROWN almost every year in the past 50 years or so. That's scenario A.
>so the scenarios only describe CO2 emmissions? No, it's about all GHG forcings! So again we must ask, did you know that and only mentioned CO2 to be misleading, or did you not know that and are you just making arguments from rank ignorance?
>But yeah, I'm clearly the liar since you utterly misunderstand what Hansen's scenarios are

> Look at my post-hoc "fit the data" emmissions values I got from realbogusclimate.org
> realclimate.org/index.php/archives/2007/05/hansens-1988-projections/

Not impressed by your shilling or your shill sites. What did Hansen actually say?
pubs.giss.nasa.gov/docs/1988/1988_Hansen_ha02700w.pdf

"Scenarios A assumes the growth rates of trace gas emissions typical of the 1970s and 1980s will continue indefinitely"

And see attached graph for the influence of other gases than Co2 (from Hansen). The only significant trace gas that has had a big decrease is CFCs. But even SkS admits they have little effect. skepticalscience.com/CFCs-global-warming.htm

On the other hand, coal use went through the roof via China. So no buddy, Scenario A is the most accurate description. You've really got to stop drinking the SkS/RealFakeClimate kool-aid.

Let me try another approach. Let's say we have two things, A and B. In situation 1, A has a causal relationship with B, and A happens. In situation 2, A doesn't have a causal relationship with B, and A happens.

If that's all we know about the world, do you agree that in situation 1, B has a higher chance of happening?