Your protests would be mistaken, assuming enough time to cool down to equilibrium in spite of the lower heat transfer speed, and sufficient insulation to maintain equilibrium.
But in a short time or a thermally leaky cooler, more surface area would have a performance advantage at the expense of endurance.
Every bit of heat-absorption inside a picnic cooler is coming from the fact that the ice is melting (other than a relatively small contribution from the ice/melt water warming up). If the ice isn’t melting it isn’t cooling anything.
A huge chunk of ice is melting. Just more slowly, because of its smaller surface area compared to cubes or chunks. Hence the qualification about time it takes to get cold (attain equilibrium) or the ability to overcome thermal leakage (maintain equilibrium ) because of the lower thermal transfer rate.
But since you want the inside of the cooler to be refrigerator-cold rather than ice-cold, a large block of ice that cools less efficiently but for a longer period of time is actually what you need, right?
Not true…it can go from e. g. -18c to 0c, without melting… And based on weight you can calculate the btus that went into cooling
True, but the latent heat of phase changes is equivalent to the specific heat of a lot of temperature change.
“It struggles a little with complex positions, like when there are an even number of moves and it has to round down, but when run against itself it’s capable of finding some novelties. At one point I saw six knights on the board at once; Stockfish rarely exceeds four.”
I’ve actually wondered (before seeing this comic) if chess engines are too greedy, in the heuristic sense of always picking the strongest position. The traditional algorithms look ahead a certain number of moves and then picks the “strongest”, using some metric for that. If the engine can look ahead until the end the game, that is provably the best. But no current engine can look ahead that far in the middle of a match. So how much gamut of the game space is being completely missed?
I’m not sure how much that applies to neural game analysis, human or electronic.
There is a ton of pruning of the search tree. (Advanced forms of alpha-beta searching.) So an engine will look only slightly ahead in most of the tree and further down for just a few promising moves. For openings, it will have a large book of better options for standard openings. Grandmasters playing computers will often try to come up with a novel variation of an opening to get the computer out of its book and then the computer will have to spend more time and not have such an early clock advantage. That in turns means the depth of search will be smaller sooner.
I had a chess-playing friend who beat Bobby Fischer in his prime. He’d then add, “Of course he was playing 49 other people at the time but I was the only one who beat him!”
He put it down to all the others saying OMG, it’s Bobby Fischer! and playing a defensive game while he went on the offensive right away. By the time the others were defeated his position was so strong even Fischer couldn’t pull it back.
Speaking to this bit which struck me funny until I thought some more:
One of the oddities of chess is that pawn promotion doesn’t have to be to a queen. That’s sure the way to bet though.
This got me thinking. Among the chess world is there any serious analysis of the benefits of pawn → knight promotion?
By rule you can’t make another king. A rook or bishop is just a nerfed queen, so they’re pointless. But knights are … different with powers beyond those of a mere queen. More limitations than a queen too, but you see the point: Knights bring their special je ne sais quoi to the fight.
Underpromotion to a knight does show up occasionally, in chess puzzles. And even a rook or bishop can be the right move, in extremely niche situations (i.e., where promotion to a queen would result in stalemate).
A few years back, I asked if it’d ever actually happened in a real game that there were two bishops on the same color, and it turns out that it has, a few times, as either a joke, a mistake, or an insult.
Actually there are situations where underpromoting to a rook is better than promoting to a queen. For example, in this position with white to move, if white promotes to a queen, black is stalemated. Promoting to a rook would let white win.
Although, in that situation, white could also move the king up behind the pawn, and then promote the pawn to a queen on the next move. It takes some serious work to come up with a situation where underpromotion to a bishop or rook is the only winning move.
An example of that case arises in the Saavedra position. In this position, the winning move is to promote to a rook. If white promotes to a queen, black Rc4+, then white must capture Qxc4 to avoid losing his queen, but that is stalemate.
OK, I can clearly see that promoting to a queen is stalemate, and that even the rook still has the potential to mate via Rc8-a8, but couldn’t black prevent that by responding to the promotion with Rd4-a4? At which point both sides have a king and a rook, and nobody’s immediately threatened, and I’m not sure where the game goes from there.
…or, wait, now that I post that, I think I do see it. Kc2-b3 still leaves Black’s king just as cornered, and now Black either wastes his turn saving his rook followed by Rc8-c1!!, or Black lets the White king eat the rook and it’s an elementary checkmate (maybe slightly delayed, but still completely inevitable).
“If only my ancestors had been fortunate enough to marry into the branch of the bacteria family that could photosynthesize, like all my little green cousins here.”
“It was great until my thumb slipped and I accidentally launched my telescope into the air at Mach 8.”
Is “72” just an error for “78,” or is it a joke I’m not getting?
Maybe he searched record players on Amazon and went with the first (incorrect) result?