This… feels like one of those “99% of people can’t do this basic math trick” memes. I mean, yeah, the puzzles are actually cool, but it feels strange to see such posts on a news website.
I solved it, to my own surprise, by considering a more general problem. I wanted to get an interval for the solution, so I replaced 3 with y. Now for y=5, the solution is obvious. x(y) is a increasing function and its minimum is... oh!
Given the (required) note that the misleading set of "triangles" is not drawn to scale, there is a range of solutions between 1 and 5.56776, since the angle of BCD is not specified to be 180 degrees (just as angle BAC is not given)
You could instead fold up the lines connected to the D point, into a V shape with the D point at the bottom, such that the points B and C (the top tips of the V) are separated horizontally by length 2 (and vertically by 0), and use the length 5 and 3 sides to go 4 units straight left or right from the top center of the V, and place point A and connect back to the far side of the V. Now point A is 3.87298 units above and 4.0 units to the side of point D, a diagonal length of 5.56776.
You could continuously deform this shape back down to the straight line case.
Now that I think about it, in the process of that deformation, there might be a point where slightly widening the V would allow you to place the point slightly higher and get an even larger max distance.
This… feels like one of those “99% of people can’t do this basic math trick” memes. I mean, yeah, the puzzles are actually cool, but it feels strange to see such posts on a news website.
What news site? Hacker News or The Guardian?
I solved it, to my own surprise, by considering a more general problem. I wanted to get an interval for the solution, so I replaced 3 with y. Now for y=5, the solution is obvious. x(y) is a increasing function and its minimum is... oh!
The solution is available if you look but this one gave me a good chuckle to myself.
It's like that Amazon Hanging Cable Interview Question
I started with cosine law and then went "wait a minute" - but still jumped to the wrong answer as pointed out in the solutions.
I was very happy with myself for seeing the first trick... but fell for the second trick.
Their answer to the first question is actually wrong. The correct answer is 1 less than the given answer.
Given the (required) note that the misleading set of "triangles" is not drawn to scale, there is a range of solutions between 1 and 5.56776, since the angle of BCD is not specified to be 180 degrees (just as angle BAC is not given)
You could instead fold up the lines connected to the D point, into a V shape with the D point at the bottom, such that the points B and C (the top tips of the V) are separated horizontally by length 2 (and vertically by 0), and use the length 5 and 3 sides to go 4 units straight left or right from the top center of the V, and place point A and connect back to the far side of the V. Now point A is 3.87298 units above and 4.0 units to the side of point D, a diagonal length of 5.56776.
You could continuously deform this shape back down to the straight line case.
Now that I think about it, in the process of that deformation, there might be a point where slightly widening the V would allow you to place the point slightly higher and get an even larger max distance.
2 replies →
SPOILER AHEAD:
Why? You're still measuring the distance between the two points on the line. One is in the middle and one is not.
I thought that initially but their answer is correct.
Not sure why this is supposed to be hard. Apollonius's theorem works as it's intended to here.
AB^2+AC^2=2AD^2+2BD^2
https://www.wolframalpha.com/input?i=5%5E2%2B3%5E2%3D2x%5E2%...