Search found 950 matches

by mfb
Sun Feb 14, 2016 11:13 am UTC
Forum: Science
Topic: Problems in medical research:Tracking outcome switching
Replies: 90
Views: 11686

Re: Problems in medical research:Tracking outcome switching

The fact that clinical trials are lengthy and expensive puts a great deal of pressure on the project to at least provide SOME kind of useful data. A null result is useful. I could write a lengthy post, but I think that is the main point. As long as null results are not considered as equally valid a...
by mfb
Sun Feb 14, 2016 10:58 am UTC
Forum: Science
Topic: "Miracles" of science idea for a XKCD panel
Replies: 6
Views: 2340

Re: "Miracles" of science idea for a XKCD panel

This article is relevant to the topic: http://www.astronomy.ohio-state.edu/~pogge/Ast162/Unit5/gps.html It explains how GPS would be totally broken if we did not know about general and special relativity. The article gets it completely wrong. The numbers there would be true if you would compare the...
by mfb
Sun Feb 14, 2016 10:51 am UTC
Forum: Mathematics
Topic: Determining a probable range for intraday stock prices
Replies: 3
Views: 1855

Re: Determining a probable range for intraday stock prices

Pretty crap shooty, and using the statistics you know is a non-starter, because they tend to rely on some assumptions of normalcy. Central limit theorem does not always apply. The stock market will have things like Bear Stearns and Uber (more often than tails of Gaussians would have it). No one sai...
by mfb
Sun Feb 14, 2016 10:47 am UTC
Forum: Mathematics
Topic: Combinatorics : Castle, rooms and doors
Replies: 14
Views: 3121

Re: Combinatorics : Castle, rooms and doors

I get 86 for 2x2. Starting from 54 for the 3-room"L", 11 have two open doors towards the fourth room (->11 options), 11 have no open door (->11 options), and 32 have one open door (->2*32 options), for a total of 86 options. Finding that number of 11 was not trivial, so I don't see an easy...
by mfb
Thu Feb 11, 2016 10:30 pm UTC
Forum: Logic Puzzles
Topic: Unknown Digits
Replies: 6
Views: 2442

Re: Unknown Digits

Are there any constraints on the values of Q, R and so on?
by mfb
Thu Feb 11, 2016 10:27 pm UTC
Forum: Logic Puzzles
Topic: Escape the Frictionless Circle
Replies: 156
Views: 64141

Re: Escape the Frictionless Circle

With actual Earth: perturbations from Moon, Sun and other sources are sufficient to give some non-zero acceleration. Just wait. With an idealized, rotating sphere and no other objects in the universe, if you are not very close to one of the poles: Jump up repeatedly. Coriolis force will eventually g...
by mfb
Sat Feb 06, 2016 12:47 pm UTC
Forum: Individual XKCD Comic Threads
Topic: 0893: "65 Years"
Replies: 221
Views: 85950

Re: 0893: "65 Years"

Image

Edgar Mitchell, 1930-2016 :(
by mfb
Sat Feb 06, 2016 12:38 pm UTC
Forum: Mathematics
Topic: Least numbers discarded when dealing cards
Replies: 30
Views: 6523

Re: Least numbers discarded when dealing cards

It is optimal for every finite run. It is not optimal if you run the algorithm multiple times, and if your set of random input numbers (e.g. the 3-sided die) is larger than the set of outputs (coin flip). You can avoid this if you store additional information at the end of a run (here: a binary info...
by mfb
Sat Jan 09, 2016 2:51 pm UTC
Forum: Mathematics
Topic: Game-tree complexity of Connect Four
Replies: 7
Views: 3573

Re: Game-tree complexity of Connect Four

A hard upper bound: If every player could randomly pick one out of 42 spots and the whole board is filled (don't stop with a winner), we get 42! possible games. For all 7 columns we can check the order of the moves, there are 6! permutations and only one leads to a valid game. Therefore, there are 4...
by mfb
Sat Jan 09, 2016 2:42 pm UTC
Forum: Mathematics
Topic: Least numbers discarded when dealing cards
Replies: 30
Views: 6523

Re: Least numbers discarded when dealing cards

It would lead to an odd poker feeling if you always get red cards. You can use unimportant random information to mix that: use the next to last card of the previous game that the player saw, or something similar. It does not have to be fully random as it does not influence game play, but it would ma...
by mfb
Sat Sep 19, 2015 2:52 pm UTC
Forum: What If?
Topic: What-If 0140: "Proton Earth, Electron Moon"
Replies: 50
Views: 42764

Re: What-If 0140: Proton Earth, Electron Moon

Relativistic mass in comment 4, really? That model is nearly as outdated as the Bohr model of the atom. I don't think it is valid to use classical electrodynamics to calculate the total energy of the moon and then apply GR to find a black hole that has the same energy. To contain 10^52 elementary ch...
by mfb
Wed Sep 02, 2015 10:18 pm UTC
Forum: Individual XKCD Comic Threads
Topic: 1572: "xkcd Survey"
Replies: 366
Views: 127678

Re: 1572: "xkcd Survey"

Also, two of my responses I'm going to share, since I didn't sign any confidentiality thing. Sharing the two means you share all. I'm quite sure you will be the only one to pick exactly those five words. Where should we post predictions on possible trends in the data? I missed that option in the su...
by mfb
Wed Sep 02, 2015 10:00 pm UTC
Forum: News & Articles
Topic: The Thread To Remind Me We're Living In The Future
Replies: 1922
Views: 385898

Re: The Thread To Remind Me We're Living In The Future

it is implied since it is pretty much impossible for human drivers to be less, or indeed, any more, idiotic that they are. . Traffic is so different in different countries - with huge differences in the level idiocy. I guess in some countries the current google cars would never be able to cross lar...
by mfb
Mon Aug 10, 2015 10:22 pm UTC
Forum: Science
Topic: Puzzling artifacts and reasonable reasoning
Replies: 79
Views: 14436

Re: Puzzling artifacts and reasonable reasoning

Like others here, I'd probably go with coincidence being the most probable explanation. In fact, I'd say coincidence is probably much more likely than it being a hoax. There's lots of perfectly plausible reasons why an ancient civilization might have wanted rocks with masses in ratios of ~1:5:10, a...
by mfb
Mon Aug 10, 2015 10:15 pm UTC
Forum: Individual XKCD Comic Threads
Topic: 1562: "I in Team"
Replies: 43
Views: 11150

Re: 1562: "I in Team"

There are no "vowels" in the alphabet! Not even in "the alphabet".
by mfb
Sat Aug 08, 2015 5:46 pm UTC
Forum: Mathematics
Topic: Factorial expressed as sum of consecutive numbers
Replies: 25
Views: 4489

Re: Factorial expressed as sum of consecutive numbers

Well, we don't have to sum them, we have a closed formula for the sum. Still, (10^500)! has O(10^500) factors, there is no way to sort them into two groups. We could not even store the result on computers.
by mfb
Sat Aug 08, 2015 5:20 pm UTC
Forum: Science
Topic: A million years
Replies: 59
Views: 9151

Re: A million years

I think we have to interpret the diagram as "at 1 GHz, the radiation is as intense as we would expect from a 3 K blackbody radiation", "at 100 MHz, the radiation is as intense as we would expect from a 900 K blackbody radiation" and so on. Gamma rays are hard to focus. Lasers or ...
by mfb
Sat Aug 08, 2015 5:14 pm UTC
Forum: Science
Topic: Puzzling artifacts and reasonable reasoning
Replies: 79
Views: 14436

Re: Puzzling artifacts and reasonable reasoning

Here's a related question. Let's say we did find something (perhaps a plaque from a crashed alien probe) with the exact masses of all the elementary particles, but with a mass for the Higgs Boson that's noticeably different from the 126 GeV/c 2 we're currently estimating. Would such a discovery be ...
by mfb
Tue Aug 04, 2015 11:11 pm UTC
Forum: Science
Topic: A million years
Replies: 59
Views: 9151

Re: A million years

Forget radioactive decays as signal. There is significant cosmic background, even our whole nuclear inventory [i]on Pluto[i] would probably be impossible to distinguish from background on Earth. What about a nuclear explosion? Electronics is needed anyway. The nuke also needs chemical explosives tha...
by mfb
Tue Aug 04, 2015 10:38 pm UTC
Forum: Science
Topic: Puzzling artifacts and reasonable reasoning
Replies: 79
Views: 14436

Re: Puzzling artifacts and reasonable reasoning

Unless I miss my guess, it's 1 in 10/1.55 * 10/0.98 * 10/0.73 * 10/0.13 * 10/0.13 or 1/5,624,212 1 in 500 You only have four independent variables, and one independent one. Removing the quark masses, see asdfex. 1 in 500 looks quite reasonable. We can reduce this even more if we assume the two larg...
by mfb
Tue Aug 04, 2015 10:02 pm UTC
Forum: Mathematics
Topic: I don't seem to understand my own brain teaser, please help!
Replies: 22
Views: 5077

Re: I don't seem to understand my own brain teaser, please h

By the Kelly criterion, you should only bet 1/10 of your time each time (if you get that option). Betting everything is clearly far away from optimal. If you get this option, you can play 10^10 times, or as long as you like (but ask for a suicide option in case 1 million years don't turn out to be s...
by mfb
Sun Jul 05, 2015 9:47 am UTC
Forum: Mathematics
Topic: Mancala game and factorization of big numbers
Replies: 7
Views: 1625

Re: Mancala game and factorization of big numbers

Having of the order of sqrt(n) pits is not efficient, it is completely impractical. Typical factorization problems involve factors that exceed the total number of atoms in the whole universe.
by mfb
Sun Jul 05, 2015 9:26 am UTC
Forum: Mathematics
Topic: Factorial expressed as sum of consecutive numbers
Replies: 25
Views: 4489

Re: Factorial expressed as sum of consecutive numbers

Does it imply that the length will be always odd? 7! and 8! are counterexamples. 14! has an odd result: a=99*2^12 = 405504 is the ideal length, while b=105*2^12 = 430080 leads to a=405405 (note the difference!). Concerning complexity: if we convert everything to logarithms, finding the best product...
by mfb
Sun Jul 05, 2015 8:08 am UTC
Forum: Science
Topic: Mass and Weight Question
Replies: 6
Views: 1216

Re: Mass and Weight Question

Still about 11 km/s. The usual approximation that the other mass won't accelerate during the process will be completely wrong if you are "the other mass".
by mfb
Thu Jun 18, 2015 10:38 pm UTC
Forum: Mathematics
Topic: How many integers between 100 and 999 with repeated digits
Replies: 5
Views: 4354

Re: How many integers between 100 and 999 with repeated digi

How many integers do not have repeating digits?
9 choices for the first digit (no zero), 9 for the second (no repetition), 8 for the third => 9*9*8=648

648+252=900 as expected.
by mfb
Thu Jun 18, 2015 10:34 pm UTC
Forum: Mathematics
Topic: standard deviation on number of events.
Replies: 15
Views: 3414

Re: standard deviation on number of events.

Depending on the situation, the whole approach could be flawed. Are the times between your events independent of each other? If they are: fine, see above. If they are not, you are probably overestimating (or underestimating) the uncertainty. Let's consider an extreme example: the events actually com...
by mfb
Sun May 10, 2015 12:34 am UTC
Forum: Mathematics
Topic: Powers of 2 and Prime Numbers
Replies: 13
Views: 10211

Re: Powers of 2 and Prime Numbers

That gives you a better approximation for small n but it is still wrong.
f(n) = log2(2^n / n) = n - log2(n) is not linear.
by mfb
Tue Apr 14, 2015 10:45 pm UTC
Forum: Science
Topic: What is the smallest object that has gravity?
Replies: 65
Views: 19052

Re: What is the smallest object that has gravity?

@Quizatzhaderac: It is easy to show that gravity acts both on neutrons and protons+electrons. The other direction - showing that small objects do attract other objects - is harder to show. Lab experiments are down to distances below 1mm, and there is no reason to expect a minimal mass (those experim...
by mfb
Tue Apr 14, 2015 10:41 pm UTC
Forum: What If?
Topic: What-If 0136: "Spiders vs. the Sun"
Replies: 89
Views: 43061

Re: What-If 0136: "Spiders vs. the Sun"

I'm disappointed he didn't go into the gravitational pull of all the world's spiders gathered together in a ball (loosely, densely, and compressed into neutronium) right next to you, which would increase their gravitational pull due to proximity. I was expecting the same. Fine, we can calculate it....
by mfb
Tue Feb 24, 2015 8:15 pm UTC
Forum: Individual XKCD Comic Threads
Topic: 1490: "Atoms"
Replies: 92
Views: 19265

Re: 1490: "Atoms"

It is easy to know that the sample contains beryllium. Every macroscopic sample contains all the more common elements.
by mfb
Fri Feb 06, 2015 9:54 pm UTC
Forum: Individual XKCD Comic Threads
Topic: 1478: "P-Values"
Replies: 34
Views: 11793

Re: 1478: "P-Values"

p<0.03 "highly significant"? Maybe in the imagination of some scientific disciplines with questionable analysis methods. I don't find the study now, but there was one that got more than 50% probability that a study with "typical" analysis methods can find some p<0.05-effect in t...
by mfb
Fri Feb 06, 2015 9:45 pm UTC
Forum: What If?
Topic: What-If 0128: "Zippo Phone"
Replies: 24
Views: 11763

Re: Disappointed

I was expecting a section about converting the mass of the smartphone/lighter into energy. Me too, and I was coming here to find such a comment. The downside: we'll need an "antiphone", and I heard those are really expensive. In the meantime, we can start with fusion. The metals are not s...
by mfb
Mon Jan 26, 2015 9:29 pm UTC
Forum: Individual XKCD Comic Threads
Topic: 1478: "P-Values"
Replies: 34
Views: 11793

Re: 1478: "P-Values"

p<0.03 "highly significant"? Maybe in the imagination of some scientific disciplines with questionable analysis methods. I don't find the study now, but there was one that got more than 50% probability that a study with "typical" analysis methods can find some p<0.05-effect in th...
by mfb
Tue Nov 25, 2014 11:25 am UTC
Forum: Mathematics
Topic: How many ways can you prove x + 1/x >= 2?
Replies: 23
Views: 16311

Re: How many ways can you prove x + 1/x >= 2?

Taylor series! As established, it is sufficient to consider the range 0<x<=1. Consider f(x)=x+1/x. Extended to the complex plane, the function is analytic everywhere apart from x=0. Therefore, the taylor series around x=1 has a radius of convergence of 1. f(x) = 1+(x-1) + 1-(x-1)+(x-1)^2/2-(x-1)^3/3...
by mfb
Sat May 24, 2014 10:48 pm UTC
Forum: Science
Topic: A new way to produce hydroelectricity
Replies: 38
Views: 9149

Re: A new way to produce hydroelectricity

Santa-Klaus, Do you understand that gravity is a conservative force? That means that you can convert gravitational potential energy into useful work but you can't get more energy out than the difference between the initial potential energy of the system and its final potential energy. And you'll ge...
by mfb
Sat May 17, 2014 10:33 pm UTC
Forum: Science
Topic: What if we buit a planet with a dyson sphere?
Replies: 12
Views: 3144

Re: What if we buit a planet with a dyson sphere?

Current commercial enterprise uses fusors to manufacture Molybdenum and Cobalt. But not to get those elements, they get mined. 99 Mo is produced (in small quantities) as it decays to 99 Tc m , which has numerous applications as radioactive tracer. 60 Co is a nice source of gamma radiation. There ar...
by mfb
Wed May 07, 2014 9:15 pm UTC
Forum: Science
Topic: Orbital coilgun accelerator
Replies: 38
Views: 5787

Re: Orbital coilgun accelerator

The momentum transfer enables the coilgun to act as the second stage; since it never leaves orbit, the weight cost associated with a second stage is zero even if the fuel is not. The rocket would need substantially more fuel to achieve the same velocity with an SSTO design, which is part of what I'...
by mfb
Wed May 07, 2014 8:48 pm UTC
Forum: Science
Topic: Green Lasers in a vacuum: Visible indirectly?
Replies: 13
Views: 4572

Re: Green Lasers in a vacuum: Visible indirectly?

oxoiron wrote:For the sake of irony, I hope it is monitored with lasers.
Yes.
(does not have to, but it can be)
by mfb
Mon May 05, 2014 7:48 pm UTC
Forum: Science
Topic: Orbital coilgun accelerator
Replies: 38
Views: 5787

Re: Orbital coilgun accelerator

Incidentally, the nature of the momentum exchange is such that even using the same bipropellant liquid rocket fuel on the coilgun station, you'd still only be using a fraction of the fuel. Around 70 tonnes (as opposed to the ~300 tonnes the rocket would need to use for itself). You would still have...
by mfb
Thu May 01, 2014 8:21 pm UTC
Forum: Mathematics
Topic: Distance to the moon
Replies: 5
Views: 2552

Re: Distance to the moon

I mean, dont you all want to know how far away the moon really is? 392934km . While it is certainly possible to do this today, it needs proper equipment as the moon is far away. In addition, it is a three-dimensional problem, so you cannot solve it graphically. And we have way better methods availa...

Go to advanced search