4 ms·
I was able to get the first five exactly with a bit of luck and think with a bit of thinking it should be possible to always land near 1% or 2%. But is there a
by V__ 3y ago
I was able to get the first five exactly with a bit of luck and think with a bit of thinking it should be possible to always land near 1% or 2%. But is there a good way to cut the cup, bird or map one without calculating it? I got lucky and got the cup exactly but I don't think I could get close without just beeing lucky on the bird and map onea.
- mhandley 3y agoWith the cup, I moved the doughnut up from below, and just eyeballed when the area above the midline of the cup equalled the area in the hole. Turns out I can do this fairly accurately.
- latexr 3y agoThe cup is symmetrical and so is the donut shading, so the trick is to flip it on its head: instead of cutting the cup with the donut, cut the donut with the cup by ensuring either the top or bottom half of the donut overlays the cup. With that line of thinking I get a perfect split every time. No insights to share on the other ones, though.
- V__ 3y agoOh that is clever, I eyballed it from the left side, but your method makes logical sense. Just tried it and easily got it exactly.
- gowld 3y agoThat only works if the cup and donut have the same area.
- latexr 3y agoI believe you, I don’t know. Show me different areas for both and maybe instinct will suggest something different. As it is, and because this was a game, I went for what seemed like a logic-y puzzle-y solution.
- sinkwool 3y agoBut they don't have the same area. The donut seems to be at least 1.5x larger than the cup (check out the screenshots posted by latexr in another comment. The area covered by the donut is half that of the cup but it's clearly less than half that of the donut, it looks more like a third or even a quarter of the donut)
- sinkwool 3y agoCan you explain more? (I might be stupid and don't see some obvious proof of why your approach is correct), but intuitively it doesn't seem that the fact that the cup is symmetrical is enough. Say if the handle of the cup was really really big, while still preserving symmetry, then its outstanding area would be huge compared to what the doughnut could cover.
- latexr 3y ago> Can you explain more? https://i.imgur.com/yE43nun.png https://i.imgur.com/yE43nun.png https://i.imgur.com/NTfbt5U.png https://i.imgur.com/NTfbt5U.png Note how the middle of the donut aligns with the bottom or top of the cup. > don't see some obvious proof of why your approach is correct I’m not a mathematician; maybe I got lucky with my approach. Crucially this is a brain teaser so it plays by the rules of a game. I formulated an hypothesis by starting from the assumption that they give you a satisfactory fighting chance.
- sinkwool 3y agoThanks for the reply and for the screenshots. before writing my initial reply, I went back and tried to get a 50/50 split based on your comment and came up with the same solution. But I didn't understand why it worked. I now believe it's either pure luck, or there's something about the sizes that result in a 50/50 split. I think it's easy to see that if the central hole of the doughnut were either smaller or larger, or if the handle of the cup was smaller or larger, these splits would not work. > I’m not a mathematician; maybe I got lucky with my approach. Crucially this is a brain teaser so it plays by the rules of a game. I formulated an hypothesis by starting from the assumption that they give you a satisfactory fighting chance. Sure, that's fair, it's just that your original explanation for how the solution is to swap the problem from splitting the cup to splitting the doughnut was so intriguing and puzzling that got me curious.
- iamwil 3y agoI didn't put much effort into it, and I got pretty close. However, I noticed that I always favored more on the right hand side.