Thread: Science project Oil and shutting up the tree huggers for good.

Originally Posted by Cebelius

*laughing* My god the repressed memories are just bubbling to the surface...

"The research is far from conclusive," is my end game, no bullshit answer. People who claim it is tend to have an agenda. (People who claim it doesn't may ALSO have an agenda, but in this particular case while I disagree with the premise that the science is settled, my reasons for doing so include both self-interested opinions AND self-evident facts) Certain branches of science ARE fairly settled. Newtonian physics works in the macro world. Fine we get it. Mathematics hasn't changed much in the last several centuries either. We've added to it, but none of the basic functions have been demonstrated to be incorrect or inaccurate. THAT is settled science. Chemistry is settled science. You can APPLY settled sciences. We can use chemistry to make bombs and aspirin. We can use physics to predict ballistic behavior and combine that with math to make sure bullet drop in our video games is accurate; etc.

Cosmology and Climatology are FAR from settled sciences. When the meteorologists MAKE it rain instead of predict the rain with increasing but still questionable accuracy, then I'll rethink whether or not climate science is settled. Taking a system as complex as THE WORLD and telling me you have an accurate model for what will happen NEXT WEEK is something they can't do with more than generally acceptable accuracy. Telling me I have to eat my dog and ditch my truck, wear hemp clothes and go live on a commune to save the environment and have oxygen for our children to breathe going into the next generation is.... amusing. Perhaps an interesting thought. Certainly not an actionable one.

You know your going to burn in hell for this right?

Originally Posted by Cebelius

*laughing* My god the repressed memories are just bubbling to the surface...

"The research is far from conclusive," is my end game, no bullshit answer. People who claim it is tend to have an agenda. (People who claim it doesn't may ALSO have an agenda, but in this particular case while I disagree with the premise that the science is settled, my reasons for doing so include both self-interested opinions AND self-evident facts) Certain branches of science ARE fairly settled. Newtonian physics works in the macro world. Fine we get it. Mathematics hasn't changed much in the last several centuries either. We've added to it, but none of the basic functions have been demonstrated to be incorrect or inaccurate. THAT is settled science. Chemistry is settled science. You can APPLY settled sciences. We can use chemistry to make bombs and aspirin. We can use physics to predict ballistic behavior and combine that with math to make sure bullet drop in our video games is accurate; etc.

And so you have to ask yourself: what's the appropriate response to uncertainty?

Your response is: "Until there is conclusive proof that the environment is changing in dangerous ways, and that coordinated human activity has an effect, then fuck all that. Moreover, before considering any change in behaviour, I demand that the proof be easily understood by everyone without them having to learn intermediate math or any science more complicated than they were taught in 9th grade."

You go on about agendas. Sure, there are agendas all over the place. I'm sure Proctor and Gamble are happy to charge people a premium for laundry detergent packaged in a white-and-green bottle. I'm also sure that Enviro-politicos aren't above playing on fear to increase donations. The agendas of capitalists and nut-jobs don't affect what's happening. The Earth doesn't care how much profit or political chaos are created in the next 20 years. Why do you want to lump the agendas in with researchers? It may be the case that someone is insisting that you eat your dog, but I promise you that if that person even exists they aren't a climate scientist.

The main thrust of your posts seems to be your desire to maximize your fuel consumption, a complete lack of guilt, and how well you sleep after eating beef. Are you sure you want to focus the discussion on agendas?

Just to be clear; I drive a car that isn't particularly fuel-efficient, and I eat meat. My consumption patterns respond mostly to prices and personal desires, just like most everyone's. But I also think that CAFE standards haven't been a bad thing. I'm not, in principle, against doing things now that hedge against long-term dangers.

Originally Posted by Cebelius

Cosmology and Climatology are FAR from settled sciences. When the meteorologists MAKE it rain instead of predict the rain with increasing but still questionable accuracy, then I'll rethink whether or not climate science is settled. Taking a system as complex as THE WORLD and telling me you have an accurate model for what will happen NEXT WEEK is something they can't do with more than generally acceptable accuracy. Telling me I have to eat my dog and ditch my truck, wear hemp clothes and go live on a commune to save the environment and have oxygen for our children to breathe going into the next generation is.... amusing. Perhaps an interesting thought. Certainly not an actionable one.

My investment club trading in straw-man futures is looking better. There are... let's see... at least 6 in that paragraph alone.

Among biologists, chemists, physicists, meteorologists, geologists, and climate scientists - what percentage do you think wear hemp and live on communes?

Your dismissal of meteorology is weird. An inability to make precise predictions in the short term (whether it will rain on your house next Thursday) has very little to do with the ability to make long-term predictions. That's simply the nature of complex systems and random functions.

For example, I can't predict exactly where the roulette ball will fall next spin, but I can tell you with exacting precision what the house take will be over a large number of spins. You tell me how much is getting bet on roulette, and I'll tell you how much the house will make. The interesting thing about these systems is that my ability to predict any single outcome never changes, but my ability to predict aggregate outcomes and long-term net effects improves with time.

Cheers,


AetheLove

Originally Posted by AetheLove

For example, I can't predict exactly where the roulette ball will fall next spin, but I can tell you with exacting precision what the house take will be over a large number of spins. You tell me how much is getting bet on roulette, and I'll tell you how much the house will make. The interesting thing about these systems is that my ability to predict any single outcome never changes, but my ability to predict aggregate outcomes and long-term net effects improves with time.

Cheers,


AetheLove

I doubt that you really could due to the movements of players, increasing of bets, lowering of bets, differing of placements of bets, and so on. If everyone were to place black or red then you stand a good chance. In either case i wouldnt call it exacting precision.

Besides the long term model for the discussion hasnt been right yet. Its kind of like reading nostradamis's predictions. Some can see it and others cant. If it were done with exacting precision then there wouldnt be an issue to begin with.

Originally Posted by AetheLove

This isn't some speculative thing where I'm making a school-yard bet that I can do it.

I've done it. Roulette is easy. The only real variation is whether there is one or two house slots (whether there is a double-zero).

It's not like BlackJack, where the expected value of the game over time is subject to how well an individual plays.

Anyway, even if I couldn't calculate it ahead of time (and I can), I could simulate it (using a method called, not coincidentally, a Monte Carlo simulation). Look up the the Law of Large Numbers. Convergence is a beautiful thing, DG.

Cheers,


AetheLove

I read what you mentioned but it doesnt seem to account for all the possible variables. It seems more to me to be about predicting the average amounts of time the house would win over a specific player. That i can agree with being possible but not the possibility of 1 person to say 20 beting various amounts. You could possibly come up with a number close to how many times the house can win but guessing the houses take? Not likely without knowing what that tables does on an average basis. At that point its more about having the average already and not coming up with it.

Originally Posted by AetheLove

Is it the mindset of a marine to only grudgingly give ground even in the face of a clearly untenable tactical situation?

Modeling roulette is trivial.

Without the zero (or double-zero), the payouts for all bets on a roulette table are fair (no house advantage). If you bet on a single number...

- chance of winning is 1-in-36
- payout if you win is 36 (your original bet plus 35 times that bet)

All other bets are effectively constructed from that one. So, an "odd" bet:

- chance of winning is 1-in-2 (50%)
- payout if you win is 2 (your original bet, plus that amount again)

Red/Black, High/Low are the same as Odd/Even.

In what sense are those "constructed" from the single-number bet? An $18 bet on "Odd" is effectively the same thing as a $1 bet on each of the 18 odd numbers. In each case the chance of winning is 18-in-36 (of 36 possible outcomes, 18 are winners, and 18 are losers).

There are at least two ways to bet thirds of the wheel (dozens):

- chance of winning is 1-in-3
- payout if you win is 3

These can also be constructed from the single-number bet.

You can bet groups of 3 (chance to win 1-in-12, payout 12), or groups of 6 (1-in-6, payout 6), etc.

That's all for a "fair" wheel - 36 numbers.

To make this game worth their while, the house adds one or two extra slots - zero and double-zero. Let's consider the case of adding two zeros.

The effect of these two extra slots is to reduce your chances of winning any bet. A single-number bet pays the same, but the odds of winning have been reduced to 1-in-38 (from 1-in-36). Having added the zeros, the house also gives you a way to bet on them - but the payout relative to the odds of winning such a bet are exactly the same as any other bet (with one exception).

What does this mean in money? For a $1 bet on a single number, your odds of winning are 1-in-38 and the payout if you win is the same as before - $35 plus your $1 bet.

So: win $35 1/38 times, lose a dollar 37/38 times .... 35/38 - 37/38 = -2/38 = -0.05263157...

That's $0.0526, which is 5.26 cents. Over time, for every dollar you bet on a double-zero wheel, you will lose five and a quarter cents.

Since all the bets have this same ratio of chance-to-win vs. payoff, all the bets have this same expected value.

The exception I mentioned above is a bet on the set {0, 00, 1, 2, and 3}. In this case the payoff is worse than all the other bets (I have no idea why, maybe there's some historical reason). That bet pays $6 on a $1, and wins 5-in-38 times.

So: win $6 5/38 times, lose a dollar 33/38 times .... 30/38 - 33/38 = -3/38 = -0.07894736...

That's almost 8 cents. This bet sucks more, but unless a significant number of people play this particular (and worse) bet, the long-run transfer of money to the house won't be very different. Also, casinos keep track of how their patrons bet and are likely to have an excellent idea of how frequently this one bet gets made.

All of that is for a wheel with two zeros. A single-zero wheel has a lower expected value for the house; instead of 2/38 it's 1/38.

Over time, every dollar bet will earn the house 5.26 cents. The only variable is how long people sit and play the game. The house earns a nickle every time a dollar is bet. The house wants to keep people feeding their dollars into the process until they have no more dollars left.

There are other variations on roulette. All of them are completely predictable.

An interesting note: in the case of a biased wheel, the bias will eventually be noted by anyone who is keeping track of the history of the ball. Casinos keep track. The easy way to deal with this problem is to re-adjust (or replace) the wheel, but if they wanted to they could leave the wheel biased, let everyone know what the bias was, and adjust all the payoffs to reflect the changes in the actual odds and still maintain the 0.0526 advantage.

Cheers,


AetheLove

No the mindset of a marine in a tactical situation is you never give any ground. If they get said ground it will be over your dead body.

I still dont see where it can account for multiple players making numerous different size bets. With 36 numbers on the board what if 30 numbers are being played by 30 different people and none of which are making the same size bet spin after spin? I clearly see how you can judge 1 person making bets.

I can even see how you could fairly accurately predict the house winnings with a constant number of people all never verring from the size of the bet. However we can set up an example and I dont think any amount of math can make those adjustments in real life. I can see how we could make it look predictable on paper though. As i love to tell architechs and interior designers just because it works on paper doesnt make it pratical in real life.

If it really was this simple there would be alot more gambling rich millionaires from the table.

Originally Posted by deathgodusmc

No the mindset of a marine in a tactical situation is you never give any ground. If they get said ground it will be over your dead body.

Good to know. That's a particularly nasty sort of stubbornness. I guess it's foolish for me to type out the math when your dead body is a necessary part of the proof.

Originally Posted by deathgodusmc

I still dont see where it can account for multiple players making numerous different size bets. With 36 numbers on the board what if 30 numbers are being played by 30 different people and none of which are making the same size bet spin after spin? I clearly see how you can judge 1 person making bets.

I can even see how you could fairly accurately predict the house winnings with a constant number of people all never verring from the size of the bet. However we can set up an example and I dont think any amount of math can make those adjustments in real life. I can see how we could make it look predictable on paper though. As i love to tell architechs and interior designers just because it works on paper doesnt make it pratical in real life.

Then at this point the issue is: what's it worth to you to learn? I used get between $10 and $60 an hour to tutor math (or statistics, or economics, or formal logic, or some sections of philosophy, or a few others) to undergrads. I got more for grad students and athletes. Last I looked, and depending on the school, an instructor might get something like $1500 to $2800 per class per semester (seems rates haven't gone up since last I did it).

If you and at least 19 others put up $100 each, I'd teach a Forum course in applied probability and statistics that focused on casino games and lotteries.

A WAY better thing for you (if you're interested, and have time) is to check out the many free web-based courses. MIT puts up a lot of their course work. Khan academy is supposed to be pretty good too.

Originally Posted by deathgodusmc

If it really was this simple there would be alot more gambling rich millionaires from the table.

Almost correct. It really is that simple, and there are gambling millionaires - they're the ones who own the table.

Seriously, from a casino's perspective, over time (let's say, a month), a roulette wheel works like this:

All the money bet on roulette in a month goes into a huge pile. The casino takes 5.26% of the pile (it's actually a little higher, because of that one weird bet). Everything that's left in the pile gets distributed back to the people who put the money into the pile. Some get more than they put in. Most get less. There are an uncountable number of ways that what's left over can get redistributed. The casino doesn't care which way it is; except that it's good PR if every now and then someone gets back a lot more than they put in. That helps encourage future play.

That's true as long as there are enough dollars going through the machine. With enough repetitions, the process converges.

A nickel per dollar. One hundred thousand dollars yields one hundred thousand nickels. As many people betting, in as many ways as they like, in different variations every time. Doesn't matter. A nickel per dollar. The casino's incentive is to maximize the amount of money processed by the wheel.

Cheers,


AetheLove

n.b.: I know fuck-all about managing a casino floor. The mix of games, the mix of minimums and maximums, the countless strategies to keep people playing; I know these things exist and have a real effect on the profitability of the room, but I don't know more than anyone else who's sat at watched the zombies play.

Originally Posted by AetheLove

Good to know. That's a particularly nasty sort of stubbornness. I guess it's foolish for me to type out the math when your dead body is a necessary part of the proof.




Then at this point the issue is: what's it worth to you to learn? I used get between $10 and $60 an hour to tutor math (or statistics, or economics, or formal logic, or some sections of philosophy, or a few others) to undergrads. I got more for grad students and athletes. Last I looked, and depending on the school, an instructor might get something like $1500 to $2800 per class per semester (seems rates haven't gone up since last I did it).

If you and at least 19 others put up $100 each, I'd teach a Forum course in applied probability and statistics that focused on casino games and lotteries.

A WAY better thing for you (if you're interested, and have time) is to check out the many free web-based courses. MIT puts up a lot of their course work. Khan academy is supposed to be pretty good too.




Almost correct. It really is that simple, and there are gambling millionaires - they're the ones who own the table.

Seriously, from a casino's perspective, over time (let's say, a month), a roulette wheel works like this:

All the money bet on roulette in a month goes into a huge pile. The casino takes 5.26% of the pile (it's actually a little higher, because of that one weird bet). Everything that's left in the pile gets distributed back to the people who put the money into the pile. Some get more than they put in. Most get less. There are an uncountable number of ways that what's left over can get redistributed. The casino doesn't care which way it is; except that it's good PR if every now and then someone gets back a lot more than they put in. That helps encourage future play.

That's true as long as there are enough dollars going through the machine. With enough repetitions, the process converges.

A nickel per dollar. One hundred thousand dollars yields one hundred thousand nickels. As many people betting, in as many ways as they like, in different variations every time. Doesn't matter. A nickel per dollar. The casino's incentive is to maximize the amount of money processed by the wheel.

Cheers,


AetheLove

n.b.: I know fuck-all about managing a casino floor. The mix of games, the mix of minimums and maximums, the countless strategies to keep people playing; I know these things exist and have a real effect on the profitability of the room, but I don't know more than anyone else who's sat at watched the zombies play.

First let me say what someone makes to teach doesnt mean everything all falls together perfectly enough for a exacting percision to any equation to be made. Let me get real with you for a second. When i am installing floors i average $125 dollars an hour. Now on some jobs i average higher and on some i average lower. The reason i cant give you an exact number is the same exact reason why you cant give me an exact answer for what the house take is at a roulette wheel. The unknown variables. You can postulate what the answer will be but the more variables you add the farther from exact you will be.

Do i think its possible to hit a ballpark figure? Sure why wouldn't i agree with that. Do i think you can give me with exacting percision what it will be? Hell no. Not when your talking about variable after variable being added to the equation. My issue with your statement had nothing to do with being able to hit a ballpark figure within a certain range. It was your choice of wording.

Originally Posted by deathgodusmc

First let me say what someone makes to teach doesnt mean everything all falls together perfectly enough for a exacting percision to any equation to be made. Let me get real with you for a second. When i am installing floors i average $125 dollars an hour. Now on some jobs i average higher and on some i average lower. The reason i cant give you an exact number is the same exact reason why you cant give me an exact answer for what the house take is at a roulette wheel. The unknown variables. You can postulate what the answer will be but the more variables you add the farther from exact you will be.

I mentioned the figures as a way to express how inexpensive it is, not how important statisticians are or that the money they get is a measure of how powerful their knowledge is. You can get sufficient instruction to learn this stuff for free. You can get one-on-one help for cheap. Real-world examples are everywhere. Roulette makes a good example because the number of variables are small, and player choice (or skill) has no effect. It sits easily in the border between a theoretical random variable and a tangible, physical, thing.

I deny the equivalence of your two examples. The unknown in the roulette example is how much total money is being wagered. That's it. For a single spin of the wheel or a small number of bets, it's impossible to know the outcome. For a large number of bets and spins, it's easy.

That's why I chose it as a counter-example to the criticism that if meteorologists can't predict the weather a week from now, then climate predictions of the future are worthless. There are many, many, examples in nature and in laboratories of processes where the state of the system is hard to predict in the short term and easy to predict in the long run.

Originally Posted by deathgodusmc

Do i think its possible to hit a ballpark figure? Sure why wouldn't i agree with that. Do i think you can give me with exacting percision what it will be? Hell no. Not when your talking about variable after variable being added to the equation. My issue with your statement had nothing to do with being able to hit a ballpark figure within a certain range. It was your choice of wording.

Rozencrantz and Guildenstern are dead.

Cheers,


AetheLove