Playtesting a Standard Model Game

[Qaanol]

TTG, I don’t understand why you are still arguing this point. It is a basic matter of the relativity of simultaneity: “Where an event occurs in a single place–for example, a car crash–all observers will agree that both cars arrived at the point of impact at the same time.”

If two things differ in space, or differ in time, then observers will not agree on their relative locations and times. But if two things are identical in space, and identical in time, then all observers will agree they are the same event. Because they are the same event.

It seems a lot of the arguments here would be far more widely reaching than intended. Suppose we aren't concerned about three reference frames anymore, just the usual two. (In the following examples, the relative speed v=.866c - set to be such that gamma is two. Things contract to half their length.) The raptor is standing still on an 'x' on the ground. 100 feet away from that 'x' is a stationary wall. Some questions:

1. The raptor starts running towards the wall at velocity v. How far away should that wall be in the raptor's frame (and thus drawn on the screen)? (a) 100 feet away still, or (b) contracted to 50 feet?

b

2. Barely after leaving the 'x', the raptor comes to a full stop. Where does the wall appear to be now? (a) 200 feet away, (b) 100 feet away

b

3. The raptor is standing still on the 'x' again, and remains still. On the wall there is a cannon, aimed at the raptor. It fires a bullet directly towards our hero at speed v. Where should the bullet first appear in the raptor's frame? (a) 100 feet away, or (b) contracted to 50 feet away.

a
The raptor looks at the bullet and says “The cannon and the x are at rest with respect to me, and are 100 feet apart. That bullet is 100 feet away.”
The bullet looks at the raptor and says “The cannon and the x are moving with respect to me, and are 50 feet apart. That raptor is 50 feet away.”

Why do they differ? Because the bullet and raptor are in different locations and different reference frames, so the two points of interest (cannon and x) cannot be in both their frames of reference (they happen to be in the raptor’s frame).

You can read about this effect, where two observers on course for a head-on collision disagree about how far apart they are, in the context of muon decay here (the paragraph after the block quote). The saving grace is that one observer sees the distance as contracted, whereas the other observer sees the time as dilated, in counteracting amounts.

4. There is a pane of bulletproof glass directly in front of the cannon. So, just moments after being launched (and traveling a negligible distance) the bullet stops again. The bullet, the glass, the raptor, and the cannon are all in the same reference frame. Where is the bullet now? (a) 200 feet away, (b) 100 feet away

b
This is the question you want to think about, because it clearly shows the game is currently wrong. If the game were right, then at the moment of firing the cannon, in the frame of reference of the raptor, the bullet would teleport (or at least go faster than light) to where it first appears away from the cannon, then at the moment of striking the bulletproof glass it would teleport (or go FTL) back to the end of the muzzle. That is clearly impossible, since it requires superluminal velocity.

5. A T-Rex is standing still next to the cannon through all these experiments. Zero feet away from the cannon. Where would the bullet first appear to him?

Right at the front of the cannon. And since the T-Rex and the Raptor are in the same frame of reference (rest with respect to the room), they agree on all observations.

In fact, we can let the raptor have a t-rex buddy, and have them communicate a plan ahead of time so that they both move together, always remaining in the same reference frame, 100 feet apart. They can run by the cannon, with the t-rex brushing the mouth of the cannon at the instant a bullet is fired, so the t-rex gets hit. Every observer agrees on this fact, because the t-rex and the bullet are in the same place at the same time.

However, the raptor 100 feet away is moving so the bullets appear in game some distance away from the cannon. And the t-rex is moving at the same velocity as the raptor, so the raptor and t-rex agree on all observations (with some delay from light-speed, but they do agree). That means either a bullet far away from the t-rex struck the t-rex, or the game is displaying things wrong.

...photons from the bullet as it punctures paper target 2,048, will follow identical paths at the same speed, so they will reach the dinosaur simultaneously (according to all observers, though they may disagree on the specific time) from the same direction (according to all observers, though they may disagree on the specific direction) so the raptor will see the bullet and the target in the same location at the same time, for every single one of the paper targets filling the space along a straight line (in proper rest frame of targets) between the gun and the target.

You're argument seems to be that because of identical the photon paths, all observers will see the events overlap. That isn't the case. Check out http://www.spacetimetravel.org/ to learn about the weird warpy world of seen relativity. They have good descriptions of how when we view a non-co-moving reference frame, it will bend and curve. A bending path of bullets wouldn't line up with a straight path no matter what one did. I think there are some Newtonian assumptions underlying your argument as it stands.

I don’t know what you think that link says, because I don’t see any description there. If you could provide a more specific link to the actual description they provide, or quote it here, that would be helpful. In any case, my argument is entirely correct.

Imagine a pipe or tube of arbitrarily-small diameter, at rest in some reference frame. Let some high-speed particles be fired down this pipe from one end to the other. The particles start inside the tube, they end inside the tube, and at no point do they puncture the walls of the tube. The particles are always inside the tube (quantum indeterminacy notwithstanding).

All observers agree that the particles were fired from one end of the tube to the other, and stayed inside the tube the whole time. No observer sees the particles hitting or passing through the walls of the tube.

Now, for any arbitrary observer, the path through space taken by the high-speed particles is the same as the path through space occupied by the pipe. Observers might disagree on what that path looks like in space and how it curves, but every observer agrees that, whatever the path looks like, both the tube and the particles follow it. Because the particles are inside the tube.

[JWalker]

I'd just like to say that while Velocity Raptor was quite fun to play, the physics is kind of off. Not only are the bullets treated incorrectly as other posters have said, but the length contraction is depicted incorrectly as well. Its a common misconception with special relativity that moving objects appear shorter, but that is actually not quite true! In fact, the length contraction everybody learns about is not directly observable. You would actually see contraction, elongation or just rotation (the so called Terrell rotation) of the object, depending on the direction of the relative velocity, and this is quite different from the usual length contraction. This is due to the different signal propagation times from different parts of the object.

Here is a nice animation showing this effect: http://faraday.physics.utoronto.ca/PVB/ ... sible.html

[douglasm]

I'd just like to say that while Velocity Raptor was quite fun to play, the physics is kind of off. Not only are the bullets treated incorrectly as other posters have said, but the length contraction is depicted incorrectly as well. Its a common misconception with special relativity that moving objects appear shorter, but that is actually not quite true! In fact, the length contraction everybody learns about is not directly observable. You would actually see contraction, elongation or just rotation (the so called Terrell rotation) of the object, depending on the direction of the relative velocity, and this is quite different from the usual length contraction. This is due to the different signal propagation times from different parts of the object.

Here is a nice animation showing this effect: http://faraday.physics.utoronto.ca/PVB/ ... sible.html

Unless you've gotten to level 25 (the "With your Own Eyes" set of levels), signal propagation time is deliberately omitted. If you haven't tried those levels yet, withhold judgment on that issue until you have. I'm not saying he got it right in those levels - enough weird stuff and advanced math is needed to deal with it that I have no idea whether the game's right or not - but anything before then is not claimed to include the detail you're nitpicking about in the first place.

The bullets are still definitely way off, though.

[gmalivuk]

Yeah, I think signal propagation time is handled correctly, it's just intentionally not part of the game until pretty high levels, at which point everything gets quite wobbly and difficult and I eventually gave up because I'm just not leet enough to guide a velociraptor at relativistic speeds around an ice rink with guns shooting at me, I guess.

[JWalker]

Unless you've gotten to level 25 (the "With your Own Eyes" set of levels), signal propagation time is deliberately omitted. If you haven't tried those levels yet, withhold judgment on that issue until you have. I'm not saying he got it right in those levels - enough weird stuff and advanced math is needed to deal with it that I have no idea whether the game's right or not - but anything before then is not claimed to include the detail you're nitpicking about in the first place.

The bullets are still definitely way off, though.

I only played until level 20 or so, but now I'm going to have to go back and try the later levels, as its a great idea for a game. Sorry for not checking thoroughly enough!

[TestTubeGames]

...the vertical component of their velocity is nowhere near high enough. I know the bullets' already high horizontal speed means the relativistic velocity addition formula would reduce that effect, but I looked up that formula and the reduction is nowhere near enough to match what I see on screen. I measured the bullets as moving at 2.4 mph, or .8c, by matching velocity with them. Yet even when I'm moving vertically at speeds even higher than that, the bullets are barely creeping along at all vertically.

First off, I modified the program as you requested, so feel free to give it another whirl. Be free! Free without the bounds of 'walls!' Free to run as fast as your little raptor can go. Also, it's funny how alike people can behave and act. Just the other day, when you all got me thinking about this again, I found myself with a straightedge across the screen watching if the bullets rose or fell. It worked (though I did feel like the proverbial person using whiteout on the screen )

I'm hoping the problem you're having will be resolved with the free-er version of the level. Reading what you wrote, I agree with your reasoning. Even given relativistic velocity addition, though the vertical component of velocity of the bullets should be diminished, it shouldn't be zero. And if you go fast enough, it should be obvious (even to those of us not using rulers), though I'll have to recheck the math. I think the big problem is we weren't able to consistently go fast enough (without crashing into a wall). Take a look at the new version, and see if it is better. I am quite sure the bullets are qualitatively doing what they should (not appearing on the expected Newtonian path), but if you've found any quantitative error at all, I'd be grateful! As I said, I'll keep thinking about this one, too...

...at which point everything gets quite wobbly and difficult and I eventually gave up because I'm just not leet enough to guide a velociraptor at relativistic speeds around an ice rink with guns shooting at me, I guess.

I think there's a rad tagline hiding somewhere in there

No observer sees the particles hitting or passing through the walls of the tube.

Qaanol, I'll repeat again: think of it as an illusion. I am not suggesting anything physical happens to the bullets, where they pass through walls, or particles are bursting out pipes. Just as passengers in a train speeding past don't feel squished, we can't take everything we perceive in relativity too literally.

I notice your answer to Question 3 is based on the raptor deducing the location rather than perceiving it. Why not calculate it, like for the very similar Question 1? In fact, some calculations will point out the contradiction inherent in the positions you took:

Suppose there are two bullets fired from that cannon. One, followed shortly by the second. And in the frame of the bullets, they (the two bullets) are 100 feet apart. Surely you agree that in the raptor's (and room's) frame the bullets would seem a mere 50 feet apart? Basic relativity.

Now, the moment (raptor's frame) the second bullet is fired, where would the raptor note the first bullet? By your answer to Question 3, you claim the second one should seem to be at the cannon (to all observers). Thus the second bullet is 100 feet away... the raptor measures a 50 foot difference... meaning the original bullet is now 50 feet away.

And now, both bullets are stopped abruptly and simultaneously (according to the frame of the bullets). Perhaps by some more of that bulletproof glass. Since they were both stopped at the same time (according to their own frame), the distance between the bullets - 100 feet - is preserved. (Their velocities matched each other at all times - according to their own frame - they couldn't drift together or apart). Again, no funny business here.

Where are the bullets now? Well, by your answer to Question 4, I can tell you'd agree with me that the second bullet would be found right at the cannon, 100 feet from the raptor. It was just fired, after all! And we know the first bullet is 100 feet in front... which places it at the feet of the raptor. Just before we stopped the bullets, that bullet was 50 feet away. Now it is right at the raptor! It just jumped 50 feet! The very type of behavior you want to say is impossible. On the contrary, that type of behavior (on the other end, with the perceived location of the FAR bullet moving) is necessary. The answer to Question 3, in perfect analogy to Question 1, is (b).

[Sizik]

Simultaneity depends on your reference frame. Let's say you have a "smart" bullet which observes when the two bullets collide. If you put it so that both bullets are equidistant from it (like if it was right in between them), then if it sees an event to each bullet at the same time, it can be sure that they actually did happen at the same time, since light has a constant velocity. Suppose it sees both bullets stop simultaneously. It flashes, signalling a simultaneous event.

Let's look at this from the raptor's point of view. It sees the bullets stop and the smart bullet flash. But then the raptor ponders: If both bullets stopped at the same time, and the smart bullet is moving, wouldn't it "see" one bullet stop before the other one? Since the smart bullet is moving towards the first bullet and away from the second, the light from the first bullet stopping should reach it before the light from the second bullet stopping. Yet it flashed, so it must have seen both at the same time. How can this be?

Then the raptor remembers. He saw the second bullet stop first! Then, a few seconds later, the first bullet stopped. There was enough time between the stoppings that the light from the second bullet had a head start, allowing it to reach the smart bullet at the same time as the light from the first bullet. Thus, while to the bullets, their distance stays at 100 ft. and they stop at the same time, to the raptor, their distance starts at 50 ft., increases to 100 ft. as one stops while the other keeps moving, and then stays at 100 ft. as the other bullet stops.

tl;dr The raptor sees the bullets stop at different times, accounting for the distance discrepancy.

[douglasm]

...the vertical component of their velocity is nowhere near high enough. I know the bullets' already high horizontal speed means the relativistic velocity addition formula would reduce that effect, but I looked up that formula and the reduction is nowhere near enough to match what I see on screen. I measured the bullets as moving at 2.4 mph, or .8c, by matching velocity with them. Yet even when I'm moving vertically at speeds even higher than that, the bullets are barely creeping along at all vertically.

First off, I modified the program as you requested, so feel free to give it another whirl. Be free! Free without the bounds of 'walls!' Free to run as fast as your little raptor can go. Also, it's funny how alike people can behave and act. Just the other day, when you all got me thinking about this again, I found myself with a straightedge across the screen watching if the bullets rose or fell. It worked (though I did feel like the proverbial person using whiteout on the screen )

I'm hoping the problem you're having will be resolved with the free-er version of the level. Reading what you wrote, I agree with your reasoning. Even given relativistic velocity addition, though the vertical component of velocity of the bullets should be diminished, it shouldn't be zero. And if you go fast enough, it should be obvious (even to those of us not using rulers), though I'll have to recheck the math. I think the big problem is we weren't able to consistently go fast enough (without crashing into a wall). Take a look at the new version, and see if it is better. I am quite sure the bullets are qualitatively doing what they should (not appearing on the expected Newtonian path), but if you've found any quantitative error at all, I'd be grateful! As I said, I'll keep thinking about this one, too...

Much better, now I can actually get up to high speeds and still have room to coast and observe. It's interesting how the room seems to stay mostly motionless as I rocket off into the great beyond as long as I continue accelerating, only to zoom off behind the instant I release the arrow key after achieving 2.99999 mph. It doesn't have the infinite line of cannons or the "hide the room" button I asked for, but those are much less important.

Anyway, I have observed the bullets while traveling straight down at 2.98 mph, and even at that enormous velocity they are still moving almost exactly horizontal. At that speed, vertical velocity should be overwhelming their normal horizontal velocity, not the other way around, yet any vertical speed they might have is still at the "need a ruler on the screen to even notice" level.

[TestTubeGames]

tl;dr The raptor sees the bullets stop at different times, accounting for the distance discrepancy.

I completely agree.

And treating the closer bullet as the 'x', and the far bullet as the 'wall'... we have a complete equivalence between Questions 1 and 3. If the moment (in the raptor frame) the close bullet ('x') hits the raptor... where will the far bullet ('wall') seem to be for the raptor. Then both bullets stop -simultaneously in the bullet frame- preserving distance (just as the 'x' remains the same distance from the 'wall' in their rest frame). The jump in apparent position of the far bullet is no more surprising than the jump in apparent position of the wall.

[Sizik]

And treating the closer bullet as the 'x', and the far bullet as the 'wall'... we have a complete equivalence between Questions 1 and 3.

Nope. In question 1, the raptor's accelerating, and in question 3, the 'wall' and 'x' are accelerating. Big difference.

If the moment (in the raptor frame) the close bullet ('x') hits the raptor... where will the far bullet ('wall') seem to be for the raptor. Then both bullets stop -simultaneously in the bullet frame- preserving distance (just as the 'x' remains the same distance from the 'wall' in their rest frame). The jump in apparent position of the far bullet is no more surprising than the jump in apparent position of the wall.

The moment the close bullet hits the raptor, the far bullet will be 100 ft away, since it will have stopped already to ensure the bullets stop simultaneously in their former reference frame.

[Twistar]

You're misunderstanding/applying length contraction. Length contraction doesn't mean "things that are moving appear half as far away as they would if they were still", it means "things that are moving appear half as long as they would if they were still." Thus, in question 1, since the room is seen as moving, it appears half as long, so the opposite wall is only 50 feet away. The bullets in question 3, however, would only look like they were squashed to half their size, not look like they suddenly appear 50 feet from the cannon.

I want to reiterate this point. This is where the confusion comes from. If the raptor and gun are 1000 feet apart and then the gun fires a bullet the bullet will be 1000 feet from the raptor, NOT 500 feet from the raptor. Think about it this way, moving rulers are length contracted. However, the ruler that the raptor uses to measure the position of the bullet is in the raptors frame and thus it isn't length contracted. If the raptor tries to use a ruler to measure the distance between two bullets that ruler will be moving, thus raptor sees the bullets closers together than they see each other.

edit: Oh and I have another unrelated question: levels 12 and 13 with the color distortion. Are these levels perfectly calibrated so that there is a speed when all of the colors will be the same? It seems like it would be very difficult to get all three colors to line up in a case like level 12. I would need to be moving with a very specific two dimensional velocity vector, but it doesn't seem difficult to achieve this at all in the game. Am I missing something, or am I just better at sweeping a 2d parameter space than I think?

[lightvector]

tl;dr The raptor sees the bullets stop at different times, accounting for the distance discrepancy.

I completely agree.

And treating the closer bullet as the 'x', and the far bullet as the 'wall'... we have a complete equivalence between Questions 1 and 3. If the moment (in the raptor frame) the close bullet ('x') hits the raptor... where will the far bullet ('wall') seem to be for the raptor. Then both bullets stop -simultaneously in the bullet frame- preserving distance (just as the 'x' remains the same distance from the 'wall' in their rest frame). The jump in apparent position of the far bullet is no more surprising than the jump in apparent position of the wall.

Jumps in the apparent position of an object can only occur when you switch reference frames, and can never occur simply because the object itself switched velocities.

Situation A: Raptor suddenly starts running with velocity v towards cannon.
Raptor sees wall and cannon jump forward from 100 ft away to 50 ft away because the raptor accelerates and changes his reference frame from one that is stationary wrt the floor to one that is moving at velocity v relative to floor.

Situation B: Cannon fires bullet at velocity v towards raptor.
Raptor sees the bullet appear 100 ft away, at the location of the cannon. The raptor does not accelerate, so its reference frame does not change, so it sees no jump in the position of anything.
Meanwhile, the bullet sees raptor 50 ft away after being fired. Prior to being fired, the bullet saw the raptor 100 ft away. The jump in the position of the raptor from the perspective of the bullet occurs because the bullet accelerated, so its reference frame changed.

Situation C: Cannon fires two bullets such that in the reference frame of the moving bullets, they will be 100 ft apart in the moving bullet frame. Both bullets are timed to explode simultaneously in the moving bullet frame just after the second bullet is fired.
In the moving bullet frame, the distance to the raptor from the cannon is only 50 ft and the bullets are 100 ft apart when they are in flight, with the second bullet exploding at the cannon. So what happens is the first bullet is fired and flies over the raptor's shoulder and then when it is 50 ft behind the raptor, the second bullet is fired and then both bullets explode, 100 ft apart.
In the raptor's frame, the distance to the cannon is 100 ft, but the bullets are only 50 ft apart when they are both in flight. So what the raptor sees is that the first bullet is fired, travels 50 ft, then the second bullet is fired and immediately explodes, then the first bullet continues flying 50 ft, passes over its shoulder, flies another 100 ft, then explodes. The raptor observes no jump in the position of anything because it remains stationary the whole time - its reference frame is fixed. Instead, simultaneity is seriously broken.

(I think I got this right, although there might be an error here)

So, to reiterate: If the raptor is moving at a constant velocity so that its reference frame is fixed, then it cannot observe objects jump in position. So the raptor must observe the bullet coming from the cannon as it is fired. The game is wrong.

[TestTubeGames]

The game is wrong.

After thinking about this for a bit, I think I know where you went wrong - instead of including Thomas Precession, you left it out.

And that was indeed the problem. Yikes!

Thank you all for being patient with my confusion/delusion! I now see where my reasoning went awry. I had been thinking the weird rotation was an effect of Thomas Rotation... but it was the other way around. After doing the two length contractions I discussed... you need to add in a rotation. As douglasm thought, I had left it out. Putting it in fixes everything. (And to make matters worse, there was a small error in a step of the calculations... so it didn't seem like rotating the frame could get everything lined up... until I found that other error this evening. Again, thank you douglasm, your ruler-screen measurements made me take a closer look at that code.)

Now it works like butter. Thank you all. I was so very wrong (and embarassed this was overlooked til now!) Nothing worse than an intuition-builder that builds the wrong intuition. The model level I put up has been fixed, and I'll fix the game tomorrow.

Thanks for playing, and thanks for being clever, perceptive folk...

Oh and I have another unrelated question: levels 12 and 13 with the color distortion. Are these levels perfectly calibrated so that there is a speed when all of the colors will be the same?

Yes, they are. Level 13 especially took a lot of trial-and-error to design right. It ends up not being as hard as it sounds to get such levels to work, though, since the color-match doesn't need to be perfect. I think they just need to be in around a 10nm range.

[Sizik]

Now it works like butter. Thank you all. I was so very wrong (and embarassed this was overlooked til now!) Nothing worse than an intuition-builder that builds the wrong intuition. The model level I put up has been fixed, and I'll fix the game tomorrow.

Looks a lot better now!

[douglasm]

Excellent, that matches the behavior I'd expect from the bullets perfectly now. Nice to know my first guess turned out to be right after all.

Incidentally, this has turned the model level into an interesting demonstration of time dilation and the resolution of the twin paradox - if I move up at high speed, the bullets are almost frozen in their horizontal trek; when I then turn around, during the period of acceleration towards the bullets (deliberately not using the space bar or shift keys) they seem to move at an extremely boosted speed, and then slow back down again when I release the arrow key. On both the outbound and inbound trips the bullets are moving very slowly, as if time is progressing slowly for them; during the turnaround acceleration, however, time is sped up dramatically for the bullets, and the magnitude of this increase depends on how far away you are when you do it. Thus, which twin is the younger one in the paradox depends on which one does all the accelerating.

In fact, how perfectly this matches up to the twin paradox now should serve as extra conclusive evidence that you got it right.

Don't forget to update (or maybe just remove) the bullets question in your Relativity 101 page.

And now that that's all resolved, maybe we can all get back to the game this thread was actually started for.

[Qaanol]

The game is wrong.

After thinking about this for a bit, I think I know where you went wrong - instead of including Thomas Precession, you left it out.

And that was indeed the problem. Yikes!

Thank you all for being patient with my confusion/delusion! I now see where my reasoning went awry. I had been thinking the weird rotation was an effect of Thomas Rotation... but it was the other way around. After doing the two length contractions I discussed... you need to add in a rotation. As douglasm thought, I had left it out. Putting it in fixes everything. (And to make matters worse, there was a small error in a step of the calculations... so it didn't seem like rotating the frame could get everything lined up... until I found that other error this evening. Again, thank you douglasm, your ruler-screen measurements made me take a closer look at that code.)

Now it works like butter. Thank you all. I was so very wrong (and embarassed this was overlooked til now!) Nothing worse than an intuition-builder that builds the wrong intuition. The model level I put up has been fixed, and I'll fix the game tomorrow.

Thanks for playing, and thanks for being clever, perceptive folk...

Oh and I have another unrelated question: levels 12 and 13 with the color distortion. Are these levels perfectly calibrated so that there is a speed when all of the colors will be the same?

Yes, they are. Level 13 especially took a lot of trial-and-error to design right. It ends up not being as hard as it sounds to get such levels to work, though, since the color-match doesn't need to be perfect. I think they just need to be in around a 10nm range.

Freakin’. Awesome.

I’ve counted as many as 11 bullets per cannon on screen “simultaneously” and as few as 2, depending on my velocity. This is excellent, well done.

[eSOANEM]

Now it works like butter. Thank you all. I was so very wrong (and embarassed this was overlooked til now!) Nothing worse than an intuition-builder that builds the wrong intuition. The [url=http://testtube

Thanks.

The measured view certainly seems right now, but the seen view still seems off, the bullets are appearing from behind the cannons (which for reasons discussed previously can't be right and then they're following paths other than the straight lines marked on the floor) even when stationary. I'm not sure if this is the same issue, but it does look a lot like how it did in the seen view before the change so I'm guessing it is.

[douglasm]

Oh yeah, the seen view, I forgot to check that. Hmm, looking at the not-yet-fixed main game, that looks close enough to the same phenomenon that it might be a case of separate calculation code paths and Thomas Precession being fixed only in one of them. I don't think that would explain it for when the raptor is at rest, though, which makes me think there's probably some other error involved.

Two things being in the same place at the same time is something that all views should agree on in all reference frames, so until the seen view shows the bullets appearing at the cannons and following the lines there is still work to be done.

[TestTubeGames]

The measured view certainly seems right now, but the seen view still seems off...

The reason for the apparent shift in the 'seen' view stems from the finite propagation of light. With that factor included, not every observer will see the same thing. Depending on proximity, some will have seen (say) a bullet already fired, and others not so much.

This leads to the strange bending of space you see in the seen view. Ignoring the bullets for the moment, and running around a level like 25: the apparent position of objects you are running toward get pulled towards your axis of motion, and apparent position of objects you are running away from get pushed away from it. The result is that straight lines get curvy. (And they get curvy in a different manner depending on where you are, so different observers will 'see' different results.)

There is an explanation of the cause of this curvy-ness (aberration) here, though be forewarned there is math and if you want the full context you'll need to read a few pages. I link to this particular page because near the bottom, figure 10 displays what I am talking about. That site also has videos of the effect in 3 spatial dimensions, such as this one.

If you accept the curvy-ness of level 25, then you can think of the bullets' path in the same way. If you are stationary, the room has nice straight lines for walls. But the bullets aren't stationary, so the 'straight lines' in their frame will have to be bent and curvy. And a curvy line won't match up point-for-point with a straight line.

Again, I think this does not conflict with a lot of the arguments that have been presented against the 'measured' view, because we shouldn't be expecting each observer to 'see' the same thing in the first place.

[douglasm]

Again, I think this does not conflict with a lot of the arguments that have been presented against the 'measured' view, because we shouldn't be expecting each observer to 'see' the same thing in the first place.

Different observers will see different things, yes, but all observers should see the bullets and the lines curving together. Facts of the form "at the time of event A, object B was present at the location of A" are universal constants independent of velocity, reference frame, and signal propagation. The bullets being on the lines can be expressed in that form, so all views should show the bullets on the lines.

I think I have a pretty good idea of how and why your calculations are resulting in something different from that, and on first glance the logic I think you're using seems sound. I'm sure there's some additional correction mechanism that you left out, similar to Thomas Precession, but what it is is not immediately obvious. I'll do some thinking and google searching to try to figure it out later.

[TestTubeGames]

I think I have a pretty good idea of how and why your calculations are resulting in something different from that, and on first glance the logic I think you're using seems sound. I'm sure there's some additional correction mechanism that you left out, similar to Thomas Precession, but what it is is not immediately obvious. I'll do some thinking and google searching to try to figure it out later.

I'm on cloud nine after fixing the 'measured' bullets. And the more I look at the seen view, I have a sneaking suspicion you may be right here, too. For instance, in this video (from the same site I was referencing before), the moving dice clearly fly above the stationary dice. I'm not yet sure how to reconcile that in a consistent manner.

The 'seen' view, as it works now, is just a series of rotations of all the points around the raptor (just as in the figure 10 before). Maybe each of those needs to be modified by a Thomas-precession-eque rotation so it all lines up. (The Thomas Precession is added in by rotations-about-the-raptor, too, which may be more than just coincidence). I'll try to see if I can find the right modification, but if anyone thinks of anything in the meantime, let me know.

[Qaanol]

Well, one option, and I don’t know how well this would work for practical calculation, is to treat everything as being part of a 3D space (like Minkowski space but missing a spatial dimension, so ++- signature). Then look at where things intersect the boundary of the past light-cone of the raptor, since those are exactly the locations where the raptor can see things “right now”.

[Twistar]

Yeah so intuitively speaking.. In the 'seen' mode if the raptor is standing still it seems like he should see the bullets travelling in a simple straight line. There is happening to screw with distances while he is standing still. There will maybe be some weird spacing effects depending on how far away the bullets are due to time propagation discrepancies, but no matter when the light ray from some bullet reaches the raptor it looks as far away as it should.
It's very possible I'm misunderstanding this bit though. I haven't thought about it very much.

Also, on the test round I can do the following: Let a row of bullets hit the wall and then run to the left and those bullets appear again. If the level were bigger I could speed up and slow down and make the bullets disappear and reappear over and over again. Is this right? I feel like no matter how my velocity changes I shouldn't see things (even if they are far away) running backwards in time (i.e. the glass reconstructing itself after I have seen it be hit by a bullet.) I understand that in my first frame I see the bullet hit the wall and then I accelerate into a different frame where timing is different so maybe the bullet hasn't yet hit in that frame... but in moving from one to the other should I see events reverse themselves? Is this some non-special relativity effect that has to do with accelerating from one reference frame to another?

[TestTubeGames]

I think I've figured out the problem with the 'seen' view. The website I've linked to does a great job explaining how the angle of the image changes. And since they are drawing a 3D world, that's all you need. Distance is implied. But in my 2D birds-eye-view world, I needed distance, of course. I calculated it out... but I'm almost certain now that's the step that went awry. So 'seen' is not right at the moment for either the bullets, or the room. It approximates it, and the angles are right... but not the distances. Yet.

So the last problem was with angles. This one is with distances. Hopefully we've gotten them all!

I'll give it a shot recoding this problem. Might take me a bit longer than for the measured view, but, hey, if I hurry maybe I will have fixed two problems in 24 hours. A productive day!

[eSOANEM]

Well, one option, and I don’t know how well this would work for practical calculation, is to treat everything as being part of a 3D space (like Minkowski space but missing a spatial dimension, so ++- signature). Then look at where things intersect the boundary of the past light-cone of the raptor, since those are exactly the locations where the raptor can see things “right now”.

I have no idea how hard this would be to program, but this would guarantee giving the correct appearance for the seen view because it is exactly what would be happening in real life. It also would mean that you wouldn't have to think/program it in terms of specific effects as thomas precession etc. would all drop out necessarily.

[TestTubeGames]

Well, one option, and I don’t know how well this would work for practical calculation, is to treat everything as being part of a 3D space (like Minkowski space but missing a spatial dimension, so ++- signature). Then look at where things intersect the boundary of the past light-cone of the raptor, since those are exactly the locations where the raptor can see things “right now”.

I have no idea how hard this would be to program, but this would guarantee giving the correct appearance for the seen view because it is exactly what would be happening in real life. It also would mean that you wouldn't have to think/program it in terms of specific effects as thomas precession etc. would all drop out necessarily.

Actually my first attempt was something kind of along these lines. I figured I'd store all the info about the location of all the particles at all past times. And perhaps I was doing it wrong (and I'm not the best programmer), but it ended up lagging a lot either from memory storage reasons or the computation. I did get it "working", but shortly thereafter stumbled across the much simpler math presented on spacetimetravel.org. I'm hoping I can avoid going back to that method, but you're right, it would give the right answer.

[douglasm]

Also, on the test round I can do the following: Let a row of bullets hit the wall and then run to the left and those bullets appear again. If the level were bigger I could speed up and slow down and make the bullets disappear and reappear over and over again. Is this right? I feel like no matter how my velocity changes I shouldn't see things (even if they are far away) running backwards in time (i.e. the glass reconstructing itself after I have seen it be hit by a bullet.) I understand that in my first frame I see the bullet hit the wall and then I accelerate into a different frame where timing is different so maybe the bullet hasn't yet hit in that frame... but in moving from one to the other should I see events reverse themselves? Is this some non-special relativity effect that has to do with accelerating from one reference frame to another?

It has to do with accelerating frames. If you are accelerating, then in your frame time runs faster ahead of you and slower behind, in direct proportion to distance from you. With sufficient acceleration and distance, this can cause time to run backwards in areas that you are accelerating away from. However, that extreme of this effect should only be apparent in measured view; signal propagation time will always work out such that you would never actually perceive such a reversal.

Elaborating on this a little more, one of the most important principles in the development of general relativity is that gravity and acceleration behave almost identically. Acceleration even has its own equivalent of black holes. Looking at the measured view, I think the point where everything is standing perfectly still in that view is the point where in a proper "seen" view you would have an event horizon that you cannot see past until you stop accelerating because the light coming from that area can never quite catch up with you. Implementing that particular detail might require a special effort to bring in some things from general relativity, but then again it might not, I really don't know - as far as I know it could be a natural result of constantly recalculating standard special relativity details with a series of different inertial frames, in which case it would automatically appear if special relativity rules are implemented correctly.

[Qaanol]

Also, on the test round I can do the following: Let a row of bullets hit the wall and then run to the left and those bullets appear again. If the level were bigger I could speed up and slow down and make the bullets disappear and reappear over and over again. Is this right? I feel like no matter how my velocity changes I shouldn't see things (even if they are far away) running backwards in time (i.e. the glass reconstructing itself after I have seen it be hit by a bullet.) I understand that in my first frame I see the bullet hit the wall and then I accelerate into a different frame where timing is different so maybe the bullet hasn't yet hit in that frame... but in moving from one to the other should I see events reverse themselves? Is this some non-special relativity effect that has to do with accelerating from one reference frame to another?

Good eye. Unless the raptor moves faster than light, it should never see the same event twice (in special relativity—general relativity is another matter). Once light from a given bullet-hits-target event has reached and passed the raptor, the only way for the raptor to see it again is by catching up to that same light, hence traveling faster than light, which is impossible.

But as mentioned above, that only applies to what the raptor actually sees. The raptor could certainly measure time-reversal elsewhere during acceleration, so the measured view seems plausibly correct. The seen view of course has not yet been fixed, but once it is, we can expect no time-reversal to be observed.

[TestTubeGames]

These are two of my favorite emergent properties of the game (ones that surprised me when they just appeared).

Qaanol and douglasm have addressed the time-oddity well. In the measured view, if you run around right, time can seem to go backwards at certain distant locations. The rub is given propagation times, you'll never actually 'see' such an event. I particularly enjoy watching the bullets go back into the cannons. Some levels in the game have grids of clocks, which, though less fun than bullets, is perhaps more illuminating.

And douglasm, as far as the black hole comparison goes, I think you're right on. If you accelerate through space, there is an event horizon, behind which (new) light won't be able to reach you. Before I added walls into the game, I spent a bunch of time just running as fast as I could, and watching the world shrink down, down, down but stop moving away. Eventually (since this isn't reality), the game's calculations break down. But, there it is. douglasm, you mention incorporating this into the seen view. I'm curious how it will look once I finish fixing it up, too. So far it's looking like we'll have a similar phenomenon, without having to add anything in by hand.

[TestTubeGames]

Got it! The seen view is working now. You can check it out again in the game. It also mostly works in the special level I'd made, though I didn't get a chance to properly set the cannon's location, so the bullets seem shifted there. You'll know what I mean when you take a look. But it all works right in the legit game.

It turns out the problem was indeed that the distances were set incorrectly. After some trig, and a bunch of debugging, it is fixed. Thanks again, all, for pointing me in the right direction on that. Here's hoping this means the game is all physically-accurate now.

[douglasm]

Looks good. In addition to bullets now following the lines, I tried the time reversal/twin paradox thing again in seen view, and I was indeed properly unable to get the bullets to visibly reverse course. I also couldn't get the super-speed thing from twin paradox acceleration to happen, but on reflection I think that's correct too - it's something that should only happen in measured view.

Add that to the approaching bullets appearing to be hyper-fast while receding ones appear to move slowly, both of which are correct and I don't think were happening before, and I think I can confidently say seen view is now correct too.

And now that we've finally finished (I think) picking apart and reassembling an entirely different game from the one this thread was "supposed" to be about, maybe we can get back to that Standard Model game you mentioned... Too bad I only have a computer, not an iPhone or other "app"-running device. I suppose I might be able to find an emulator for one, though.

[Twistar]

haha! Seen view is crazy now! Way harder. Thanks for the explanation on the measured time reversal. I had never heard about this. Also, thanks putting this game together, TTG, it's really great!

Anyways, I have an app-running device and would try the standard model game.

[Yakk]

Go up/down at high speeds in XKCD text with "center" view.

After everything leaves your screen, start going in the opposite direction (don't cheat with space).

Funky space warping fun.

[TestTubeGames]

Go up/down at high speeds in XKCD text with "center" view.

After everything leaves your screen, start going in the opposite direction (don't cheat with space).

Yup. I first noticed that when I was redo-ing the 'seen' view, and thought I had messed everything up! But it turns out, that just happens because the level-drawing algorithm assumes you're within the original walls. If you go past that, it paves the floor with funkiness. I just left it because there are always walls in the game, and hey, it looks even more mind-trippingly rad.

And now that we've finally finished (I think) picking apart and reassembling an entirely different game...

Woohoo! Again, thanks to everyone for pointing out those problems. Ah yes, there is another game I'm supposed to be working on...

Twistar, I'll get you hooked up with a beta copy of the new game. Thanks to everyone who has already given it - Agent Higgs - a play. And for those of you without iPhones, fear not, there are more flash games coming in the future, too (likely even some version of Agent Higgs).

[Qaanol]

The seen view still has a problem whereby the bullets don’t appear to start at the cannon—even when I am standing perfectly still, the bullet come into existing a significant distance ahead of the cannons.

[Yakk]

Anyhow, I'd be interested in testing out the standard model game.

[douglasm]

The seen view still has a problem whereby the bullets don’t appear to start at the cannon—even when I am standing perfectly still, the bullet come into existing a significant distance ahead of the cannons.

As he said, that's specific to the special test level. There's some location data that needs to be corrected which he didn't have time to fix after doing everything else. The actual game has any such data fully corrected, though, so look at level 28 of the main game to see it truly working right.

[Qaanol]

Ah, thanks.

Back on the original topic of the standard model game being beta tested, are the beta-testers allowed to talk about it here? How about posting screenshots?

[TestTubeGames]

Back on the original topic of the standard model game being beta tested, are the beta-testers allowed to talk about it here? How about posting screenshots?

Most definitely! Feel free to post whatever you'd like.

And a little background about the game (since one could be forgiven for forgetting the origins of this thread), it is a puzzle game based on particle physics. There is a big difference between it and Velocity Raptor, in that this new one (Agent Higgs), is not meant to strictly be a simulation. Velocity Raptor was a game based off a (now accurate ) relativity simulation. With Higgs, you move the particles around and they *roughly* follow the rules they should. Two muons, for instance, will repel. A particle and its antiparticle will annihilate. Those kinds of things. My big hope was to make engaging puzzles, where people would learn the basics of the standard model just by playing.

I'm planning to have the game released by June. And after that, there is an increasingly good chance I'll make a (slightly modified) flash version.

[Qaanol]

All right then, questions for other beta testers:

Have you gotten a perfect score on Neutrinos #30? I am stuck 1 away. I just solved it! (Excellent level design TTG!)

Also, on what levels have you done better than the nominal perfect score? On Neutrinos #19 I beat it by 1, and on Neutrinos #34 I beat it by 2.

@TTG: I really like how tauons behave.

Edit: …and the new version of the beta came out, so the perfect scores got updated.