Testing Gravity in the Solar System

I can think of a lot of stupider things that Europe could spend a billion euros on, so let's go for it! A billion euros divided by 300m Europeans? I'll give them €3...

I, too, find this notion odd that there's vast amounts of the universe out there that we have no visual or tangible evidence for - just some mathematical formulae that only work if it's assumed to be there. I'm very conscious of the old 'ant on a basketball' situation - where the ant thinks it's on a flat 2D surface and is astonished to walk in a straight line and find out it's back where it began - ie, there could well be a lot more than we are currently capable of sensing. And in a thousand years the stuff we've learned in the last 100 will look like playschool astronomy. But, but, but... It just feels uncomfortable.

I have to admit the lecture was very fast paced, I couldnt quite keep up with the acceleration but Ali's delivery was nevertheless captivating, I just wished I could get a grip on those graphs.  I did of course get the gist of his plight.

So what I understood is that at some point between two celestial body's gravitational influence on each other, the forces cancel each other out which I presume is the centre of mass of the two objects, correct?  So Ali's conjecture is that this point of the space curvature may offer clues or even answers to why Einstein's equations do not fit in certain elements of cosmic observation.

Have I got that right?  Of course subtle gravitational influences will also be in effect from other distant objects too, so that would have to be also taken into consideration in calculating the zero force point, I imagine?

I think its a brilliant, rational and very logical next step to finding the next clue to the secrets of the universe.  Its not as if he has some wacky theory of the what the secrets of the universe is (not that I frown upon that, I LOVE wacky theories!), all he is offering is where to look which is absolutely achievable in both technology and cost.

Funding is always a political decision.  Nothing wrong with politics.  We have to accept politics, better than be ruled religion, ay?What is important is that the ideas and endeavours are not forgotten, promotion of idea must always continue until it has political influence.  So Ali's scientific suggestion may not happen now (which we all are personally disappointed with, I am sure) but it WILL happen as long as the idea is always pushed.  Fingers crossed it hcan happen sooner than he thinks (he does seem quite dejected by it, doesnt he, poor Ali)  because I think this is an exciting one.

Just wished I could understand that first graph because he promised we will all understand it by the end!

Btw, Ali had his powerpoint on slideshow advance timings mode which I dont think he realised.  he needed it set on manual.

I really enjoyed the talk. A big well done to whoever booked him.  Tej my understanding of the first graph is that the closer to the saddle point you get the higher the sigma ie certainty is.

You just gave me an idea for a challenge: instead of stuffing three Pringles in my mouth like I normally do, try to eat around a Pringle while avoiding the saddle point.

Much like a Pringle, you crack me up Andy.

Brian, what is the sigma ie representing?   I think I missed another main point that Ali is striving for.  Currently we think of gravity as a constant, right?  The forces exerted being the variables depending on the masses of nearby objects.  So this proposed mission is to find out whether gravity really is a universal constant or perhaps it is a also a changing variable itself?   We are talking, of universal graviational constant, right, the big G?  If other forces are detected at these saddle points  (though i much prefer Andy's pringles points!), then that could lead to explaining the mysterious dark energy or dark matter?

Sigma is the level of certainty of the validity of an experiment result.  For example the discovery of the Higgs was not announced until it had been verified to sigma 5. This meant there was 1 in 35 million chance the results were wrong.  Ali's point is that if the results he predicts are proved then they could be at 60 sigma if the measurements where taken very close in.  He did also say that if he was proved wrong then he would not lose sleep over it and keep on researching.

Brian, I've read that the quality management technique known as 'Six sigma' translates to 99.99966% accuracy, or 3.4 errors per million, or one error per 294,000 - which is a long way off from Ali's five sigma of 1-in-35 million!

Any idea why there's that discrepancy? Is it because 'six sigma' is actually plus-or-minus 3 sigma?

Each "sigma" is one standard deviation (standard deviation is denoted using the Greek letter Sigma... hence the name).

For normally distributed data:

1 sigma = 68% of data
2 sigma = 95% of data
3 sigma = 99.7% of data
4 sigma = 99.99%
5 sigma = 99.9999%
6 sigma = 99.9999998% (or one in approx 500,000,000)
etc...

The QM technique for six sigma is different, I understand, as it incorporates a 1.5 sigma shift, so, technically, a 6 sigma shift in quality management is the same as a 4.5 sigma shift based on a normal distribution. See here for an explanation as to why this is done.

Thanks for that Mike. The 5 sigma seems to calculate out at one in a million - no?

As for the reason why 4.5 sigma became 6 sigma, the explanation is interesting but, I suspect, wrong. Here's what really happened...

Marketing guy: "We're ready to roll out the four point five sigma project across the company."
Tech guy: "Brilliant!"
Marketing guy: "And we've had a great idea!"
Tech guy: "Er, yeah? Really?"
Marketing guy: "Yep! We're gonna call it 'Six Sigma'! Great, huh?"
Tech guy: "But - but that's wrong. Statistically it's 4.5, not 6! Why change it?"
Marketing guy: "Six sigma is much more alliterative. Same vowel sound. Easier to remember, therefore easier to get buy-in as we roll it out. They'll love it!"
Tech guy: "You can't change the laws of probability and normal distribution! You have to call it 4.5 sigma!"
Marketing guy: "Who's Norman in Distribution and what's he got to do with it? Besides - we've already printed 10,000 brochures..."

The 5 sigma seems to calculate out at one in a million – no?

Near enough... I didn't include all of the decimal points... it's actually 99.9999426697% 😉 So, therefore, one in 1.744 million.

Here’s what really happened…

I suspect you're close to the truth there!

Thanks for the extra decimal places - good to know I'm using my spreadsheet properly. But that 5 sigma 1-in-a-million is kind far away from the 1-in-35 million that Ali referred to - or is there an apples-and-oranges component to this?

But that 5 sigma 1-in-a-million is kind far away from the 1-in-35 million that Ali referred to

Ahhh, I see what the problem is. Ali referred to 1 in 3.5 million, not 35 million.

So, it's still not 1 in 1.744 million... why not? Well, it's because you can quote the probability of a normal observation away from an expected value in either the direction of interest (one-sided) or either direction (two-sided).

This article gives a great explanation as to why quoting the one-sided interpretation is correct.

Thanks for that, Mike. Clearly Brian and I are using the same ear-trumpet. We both thought he said 35 million.

Right, then: take the rest of the day off from double maths and go and enjoy some lovely photographs...