I don't think space is expanding.

Status
Not open for further replies.
Mike Helland said:
Here's what I have so far:

(junk snipped)
Click to expand...

What you have so far is completely wrong, as has been amply demonstrated by a number of posters. I am quite sure that the only reason that you continue posting this crap is adequately explained by the Dunning Kruger effect. You do not appear to have the intellectual capacity to understand that you are wrong, never mind why you are wrong. Your ideas will never, ever, be taken seriously by anyone working in the field of cosmology, physics generally, or even lay persons with a reasonable knowledge of mathematics. You have convinced no one here in a small sub-forum of a backwater internet message board. Your chances of convincing anyone, anywhere, ever, (with the possible exception of Pixie of key) that you have a workable idea that should be considered rather than laughed at are somewhere between nil and zero.
 
Mike Helland said:
My computer is better at discrete time steps than spacetime continuums.
Click to expand...

Using discrete time steps to numerically approximate an integral is very standard practice for problems that cannot be solved or aren't worth solving analytically. But you're doing it wrong. Like, not a bit wrong, but completely wrong. As in, you don't have a clue about how numerical integration works, or even what integration is.

There is some evidence that redshifts are quantized.

Make of that what you will.
Click to expand...

I could make a hat, or a broach, or a pterodactyl...

But I can't make your fruitcake ideas work.
 
Ziggurat said:
No, Mike, it didn't. You thought it did because you're doing the math wrong. And now you're just proving that in addition to being clueless about physics, you're clueless about calculus.
Click to expand...

Ignore the QM then.

I used the same idea on classical light and it happens too:

https://mikehelland.github.io/hubbles-law/other/reflection3.htm

That's nice. It's also irrelevant, because (again) you're still doing the math wrong.
Click to expand...

The exact same technique, based on my hypothesis, applied to two different models, changes the result to the answer I expected.

What are the odds of that being a mistake?

Here's a hint: dt isn't what you think it is. When doing a numerical approximation of the integral, you have to choose a value for your time step, it is not determined by any clock. And if you have chosen a good value for your time step, then making it smaller won't appreciably change the answer. If it does appreciably change the answer, then you done ****** up, boy.
Click to expand...

I get what you're saying.

If I run the model at timestep = 1 and timestep = 1234 I should get the same results.

And I do.

But this technique does something different.

It uses a timestep of 1/2 for z=1 photons, and a timestep of 1 for z=0 photons.

The time step is different before and after the mirror.

Using only Fermat's least time principle, and counting time individually for photons based on their redshift, the results work out in my favor.

And then I also get the favorable results when doing the same thing in Feynman's solution.

Pretty wild coincidence.
 
Last edited:
Here's a little exercise for you, mike. Use numerical integration to calculate how far a drop object falls in a uniform gravitational field as a function of time. This is an easy problem because we have an exact analytic solution to compare the numerical solution to. But try doing the numerical solution: x = integral(v(t) dt). Try different time step sizes, see what happens.

Maybe you'll figure out what you're doing wrong. I doubt you will, though.
 
Mike Helland said:
Ignore the QM then.

I used the same idea on classical light and it happens too:

https://mikehelland.github.io/hubbles-law/other/reflection3.htm
Click to expand...

Am I supposed to be impressed that you've found multiple ways of screwing up?

The exact same technique, based on my hypothesis, applied to two different models, changes the result to the answer I expected.

What are the odds of that being a mistake?
Click to expand...

100%. What did I tell you before? You are always wrong. It's like a super power.

You introduce fudge factors to things you don't understand, and if you get the right answer then you think those fudge factors mean something. But they don't. They never did.

I get what you're saying.

If I run the model at timestep = 1 and timestep = 1234 I should get the same results.
Click to expand...

No, you don't get what I'm saying, at all. You don't have a god damn clue about numerical integration.

It uses a timestep of 1/2 for z=1 photons, and a timestep of 1 for z=0 photons.
Click to expand...

That doesn't make any sense. That isn't how time steps work in numerical integration.
 
Ziggurat said:
100%. What did I tell you before? You are always wrong. It's like a super power.
Click to expand...

You're being irrational.

If you use the "master clock" in the least time principle to determine which path took the least time, when there's a speed change at the mirror, it'll be a photon that changes speed sooner rather than later.

If you use the photons' clocks, and if photons are moving slower when their clock is running slower, the lower speed before the mirror is canceled out by the lower clock speed, resulting in a reflection at a "normal" angle.
 
Mike Helland said:
Right, one of the terms in Feynman's sum is dt.
Click to expand...

Feynman's sum is part of Feynman's math, which is part of the math that yield's Snell's law and a bunch of other things that you either deny, or which contradict your theory.

Where's your math, the math that yields not only everything we've observed about QED, but also everything you need to show that space is not expanding?

You're writing programs to represent models that you haven't even developed yet. The representational programs should be the very last step, not the first and only step. You don't have a model, and you won't get a model by applying bits and pieces of other people's models, cargo-cult style. Feynman isn't going to do your work for you. You have to do it yourself.

Where's your math?
 
Last edited:
theprestige said:
Feynman's sum is part of Feynman's math, which is part of the math that yield's Snell's law and a bunch of other things that you either deny, or which contradict your theory.

Where's your math, the math that yields not only everything we've observed about QED, but also everything you need to show that space is not expanding?
Click to expand...

https://mikehelland.github.io/hubbles-law/other/reflection3.htm

https://mikehelland.github.io/hubbles-law/other/reflection_nm.htm

Right click. View page source.
 
theprestige said:
Feynman's sum is part of Feynman's math
Click to expand...

What Mike did wasn't Feynman's math. Mike did it wrong. Everyone is telling Mike that. Everyone is telling Mike the same reason why it's wrong. Because everyone with a clue about calculus can recognize his mistake instantly.
 
Last edited:
Mike Helland said:
So you're saying, don't include the timestep in the sum?

And then I should get Snell's Law.

And I told you, yeah, that's what happens. That's what I had first.

Then I added the redshifts.

[qimg]http://www.internationalskeptics.com/forums/imagehosting/thum_762186056142463bf5.png[/qimg]
Click to expand...

Snell's Law is what's observed in reality, though. Models that predict something other than what's observed are obviously wrong. Adjusting your model until it predicts something other than what's observed seems like a stupid thing to do. Why do you keep doing it?
 
Ziggurat said:
What you did wasn't Feynman's math. You did it wrong. Everyone is telling you that. Everyone is telling you the same reason why it's wrong. Because everyone with a clue about calculus can recognize your mistake instantly.
Click to expand...

Maybe it would make more sense if you saw the control version next to the hypothesis:

https://mikehelland.github.io/hubbles-law/other/reflection_nm.htm

Make sure to hit refresh. There should be 3 demos there.

The top one has photons always moving at c.

The middle one is photon's moving at 0.5c, then reflecting at c.

The bottom one uses the photon clock idea.

*edit* set the boundary to 2500 to see the dials. github's not refreshing pages right now.
 
Last edited:
theprestige said:
Plugging values into an equation isn't a model. It's just algebra. The math you're looking for lies on the other side of differential calculus, not one step away from arithmetic.
Click to expand...

Ok. Well.

Are you looking for this then:

dt' = dt / (1 + z)

?

This little bit of code will produce redshifts for an expanding universe that is accelerating:


Code: 
const dt = 1
const c = 1
const H = 25000000000

for (var photon of photons) {

	var dt_p = dt / Math.pow(1 + photon.x / H, 2)

	photon.clock += dt_p
    	
    photon.x += photon.dx * dt_p
    photon.y += photon.dy * dt_p

}

Units:

c in lightyears per yer
H in lightyears
dt in years
x in lightyears

The += operator just means take what value I had, and add to it. So:

x += 1

And

x = x + 1

do the same thing.

This is the hypothesis, and this same solution produces the right (IMO) angles for photons in a vacuum at traveling at less than c.

What would be the way to express that model, in your view?
 
Mike Helland said:
Maybe it would make more sense if you saw the control version next to the hypothesis:
Click to expand...

No, Mike. It isn't an issue of me not understanding. It's that you're doing it all wrong. That isn't the way any of it works. You don't actually understand the quantities you are working with.
 
Ziggurat said:
No, Mike. It isn't an issue of me not understanding. It's that you're doing it all wrong. That isn't the way any of it works. You don't actually understand the quantities you are working with.
Click to expand...

Here.

Top is v=c, middle is v=0.5c to v=c at the mirror.

Left shows the top has finished and middle is part way done.

Right shows the final results.

Top and middle produce the results you expect. So I'm doing it right.

feynmanpathintegral_ab.png


The bottom creates a wave within the dials that's unique to both of them (edit: but the final result is the top pattern).
 
Last edited:
Mike Helland said:
Here.

Top is v=c, middle is v=0.5c to v=c at the mirror.

Left shows the top has finished and middle is part way done.

Right shows the final results.

Top and middle produce the results you expect. So I'm doing it right.
Click to expand...

No, Mike, you are not doing it right. Repeating the assertion won't make it so. You really don't understand what dt is, at all.
 
Mike Helland said:
Ok. Well.

Are you looking for this then:

dt' = dt / (1 + z)

?

This little bit of code will produce redshifts for an expanding universe that is accelerating:


Code: 
const dt = 1
const c = 1
const H = 25000000000

for (var photon of photons) {

	var dt_p = dt / Math.pow(1 + photon.x / H, 2)

	photon.clock += dt_p
    	
    photon.x += photon.dx * dt_p
    photon.y += photon.dy * dt_p

}

Units:

c in lightyears per yer
H in lightyears
dt in years
x in lightyears

The += operator just means take what value I had, and add to it. So:

x += 1

And

x = x + 1

do the same thing.

This is the hypothesis, and this same solution produces the right (IMO) angles for photons in a vacuum at traveling at less than c.

What would be the way to express that model, in your view?
Click to expand...

Anybody can write code that draws pretty pictures. I'm asking you for your mathematical model of reality, that yields reality-describing code as a by-product.

(No, not reality as you need it to be in order for your hypothesis to be true. Reality as it is actually observed, with which all true hypotheses are consistent.)
 
Status
Not open for further replies.

Back
Top Bottom