This should give FirstAscent a stroke...

[bb113]

Predicting the direction of something is 50/50 in a random-walk model, when did you prove this earlier? In a model with causation you generally can predict the direction and magnitude with some level of certainty (and you should be able to generate some error bars as well.)

This is confused. The actual direction of a 100% stochastic process is 50/50. The actual direction of a deterministic process is obviously not 50/50. However, when there are various unknown initial parameters and relationships between parameters it is 50/50 from the perspective of the investigator.

OK, but the point was, this method is for analyzing 100% stochastic systems, not deterministic ones, it has been misapplied. I also have some doubts about that second sentence, on the surface it seems unlikely that you can factor in the massive influences we understand on the climate and still be stuck at 50/50. I don't need to know the initial point, exact impact angle, or velocity for a baseball to know that if it is hit it will likely go forward, but there is a small chance it might go backwards or straight up.

Ok, I see the problem. That is just not true. What is your source for this?

Hmm, I must have misunderstood something. This was from that blog post from yesterday, the one explaining how they had invalid tests being performed and they were failing to reject the presence of the unit root. I'll have to look at it again.

maybe it was a stochastic variable, not system or I'm making another semantic error. You obviously know a bit about it, care to share with the class?

The key is that you can have a process that is part deterministic and also part stochastic. You can also have different types of stochastic processes (stationary and nonstationary):

http://i49.tinypic.com/r9lvdf.jpg


In order of left to right:
1) y= random number from normal distribution of
mean= 0
standard deviation=1

2) "Random Walk": y= random number from normal distribution of
mean= previous y
standard deviation=1

3)  y= .01*x +
random number from normal distribution of
mean= 0
standard deviation=1

4)  y= .01*x +
random number from normal distribution of
mean= previous y
standard deviation=1


As I said, I am not clear on what relationship this has to the cointigration analysis. I am just saying that the debunking for it you provided doesn't make sense to me.

[myrkul]

I'm a little insulted that you think you know me so well. I'm not having a misunderstanding in ANY of those areas, and I've been defending YOUR position as well I believe, which causes me to be a bit amazed that you feel I need to go back to basics. Especially after you chimed in with a list of 3 things to add to my list (of which 2 were actually covered in the list at least partially.)

...

You might want to reconsider your way of interacting with folks, I'm pretty unimpressed.

FirstAscent, making new friends already.

[scrybe]

As I said, I am not clear on what relationship this has to the cointigration analysis. I am just saying that the debunking for it you provided doesn't make sense to me.

Me neither at this point, don't feel bad

Thanks for the feedback, I'll come back at this later and see if I can figure out what happened.

[FirstAscent]

Predicting the direction of something is 50/50 in a random-walk model, when did you prove this earlier? In a model with causation you generally can predict the direction and magnitude with some level of certainty (and you should be able to generate some error bars as well.)

This is confused. The actual direction of a 100% stochastic process is 50/50. The actual direction of a deterministic process is obviously not 50/50. However, when there are various unknown initial parameters and relationships between parameters it is 50/50 from the perspective of the investigator.

OK, but the point was, this method is for analyzing 100% stochastic systems, not deterministic ones, it has been misapplied. I also have some doubts about that second sentence, on the surface it seems unlikely that you can factor in the massive influences we understand on the climate and still be stuck at 50/50. I don't need to know the initial point, exact impact angle, or velocity for a baseball to know that if it is hit it will likely go forward, but there is a small chance it might go backwards or straight up.

Ok, I see the problem. That is just not true. What is your source for this?

Your diligence is semi-admirable, but may I suggest something? As I said earlier, you essentially lack common sense. I don't mean common sense as in you can't fix yourself a sandwich, but as in, you don't understand climate science at the general level. Interested laymen know much more than you. Your nose is stuck in spreadsheets, but you have no general understanding of the forces at work, the dynamic interactions, etc. Sort of like someone who has no real 'feel' for hitting a baseball.

Learn about the following:

- Ice ages and their causes
- Ice albedo feedback loops
- Current fieldwork on glacier melting
- Ice core analysis, tree ring analysis
- Climate change induced species migration
- Sea level rise and its causes
- The changing of precipitation patterns
- Political blockades
- Consensus view
- Prior EPA success stories

And lastly, and perhaps most importantly, I think you discount scientists doing field work out in the real world too much. I sensed this when you tried to make light of the article which summarized the effects of climate change on the migration of species. It's akin to an armchair mountaineer analyzing the decisions of a team pushing a new route on an 8,000 meter peak in the Himalaya. You're not seeing the things that field workers are seeing - years of study allow them to intuit the truth in ways you're not familiar with.

I'm a little insulted that you think you know me so well. I'm not having a misunderstanding in ANY of those areas, and I've been defending YOUR position as well I believe, which causes me to be a bit amazed that you feel I need to go back to basics. Especially after you chimed in with a list of 3 things to add to my list (of which 2 were actually covered in the list at least partially.)

My misunderstanding is in statistical math, specifically econometrics, which I have almost no experience with. I have been following IPCC reports, US govt reports, I followed the ozone hole scare, I followed CFC regulation and it's results, and just this morning I was looking at pictures of ice floe coverage and a couple polar bears on a blog. This is not something that I do every day, or systematically, but it has fascinated me for years and I feed my fascinations as much science as they can stomach.

You might want to reconsider your way of interacting with folks, I'm pretty unimpressed.

My post was clearly addressed to bitcoinbitcoin113, someone I have been interacting with for at least a year, and that was my summary of a year's interaction with him. Thanks for your opinion though.

[bb113]

Predicting the direction of something is 50/50 in a random-walk model, when did you prove this earlier? In a model with causation you generally can predict the direction and magnitude with some level of certainty (and you should be able to generate some error bars as well.)

This is confused. The actual direction of a 100% stochastic process is 50/50. The actual direction of a deterministic process is obviously not 50/50. However, when there are various unknown initial parameters and relationships between parameters it is 50/50 from the perspective of the investigator.

OK, but the point was, this method is for analyzing 100% stochastic systems, not deterministic ones, it has been misapplied. I also have some doubts about that second sentence, on the surface it seems unlikely that you can factor in the massive influences we understand on the climate and still be stuck at 50/50. I don't need to know the initial point, exact impact angle, or velocity for a baseball to know that if it is hit it will likely go forward, but there is a small chance it might go backwards or straight up.

Ok, I see the problem. That is just not true. What is your source for this?

Your diligence is semi-admirable, but may I suggest something? As I said earlier, you essentially lack common sense. I don't mean common sense as in you can't fix yourself a sandwich, but as in, you don't understand climate science at the general level. Interested laymen know much more than you. Your nose is stuck in spreadsheets, but you have no general understanding of the forces at work, the dynamic interactions, etc. Sort of like someone who has no real 'feel' for hitting a baseball.

Learn about the following:

- Ice ages and their causes
- Ice albedo feedback loops
- Current fieldwork on glacier melting
- Ice core analysis, tree ring analysis
- Climate change induced species migration
- Sea level rise and its causes
- The changing of precipitation patterns
- Political blockades
- Consensus view
- Prior EPA success stories

And lastly, and perhaps most importantly, I think you discount scientists doing field work out in the real world too much. I sensed this when you tried to make light of the article which summarized the effects of climate change on the migration of species. It's akin to an armchair mountaineer analyzing the decisions of a team pushing a new route on an 8,000 meter peak in the Himalaya. You're not seeing the things that field workers are seeing - years of study allow them to intuit the truth in ways you're not familiar with.

I'm a little insulted that you think you know me so well. I'm not having a misunderstanding in ANY of those areas, and I've been defending YOUR position as well I believe, which causes me to be a bit amazed that you feel I need to go back to basics. Especially after you chimed in with a list of 3 things to add to my list (of which 2 were actually covered in the list at least partially.)

My misunderstanding is in statistical math, specifically econometrics, which I have almost no experience with. I have been following IPCC reports, US govt reports, I followed the ozone hole scare, I followed CFC regulation and it's results, and just this morning I was looking at pictures of ice floe coverage and a couple polar bears on a blog. This is not something that I do every day, or systematically, but it has fascinated me for years and I feed my fascinations as much science as they can stomach.

You might want to reconsider your way of interacting with folks, I'm pretty unimpressed.

My post was clearly addressed to bitcoinbitcoin113, someone I have been interacting with for at least a year, and that was my summary of a year's interaction with him. Thanks for your opinion though.

This is fair enough (except I have never fixed myself a sandwhich... don't like them). You can sum it up more precisely by saying that the researchers in the field use a different prior probability than I do due to various background knowledge that I lack.

[scrybe]

My post was clearly addressed to bitcoinbitcoin113, someone I have been interacting with for at least a year, and that was my summary of a year's interaction with him. Thanks for your opinion though.

Apparently I misunderstood your point when you mentioned my name in the subsequent post to the one I quoted. Since he was pointing out a mistake in my logic related to math, I didn't connect that you would be taking him down for some completely different reason.

My apologies, and I hope you can understand the reaction.

[Rassah]

So... what if we do something like cut all oil subsidies and stop military support/protection for oil wells in troubled areas of the world, making gas prices in US possibly go up to $6 a gallon, and stopped subsidies for highways and roads, making them depend on tolls and other means to raise money for upkeep, thus making driving itself very expensive as well? I.e. get the government out of that particular part of transportation. Would that help?

[myrkul]

So... what if we do something like cut all oil subsidies and stop military support/protection for oil wells in troubled areas of the world, making gas prices in US possibly go up to $6 a gallon, and stopped subsidies for highways and roads, making them depend on tolls and other means to raise money for upkeep, thus making driving itself very expensive as well? I.e. get the government out of that particular part of transportation. Would that help?

Never happen. Too sensible.