plasmid

Sounds interesting. It's been unpredictable when and how much time I get away from work, but I'll keep an eye out for it if we reach critical mass while I'm free-ish.

It's been a long time since I last posted a riddle, and since most of the people have changed I should probably explain my peculiarly. I write riddles in such a way that they sound kind of like they're describing one thing, but are actually describing something else completely unrelated, and title them "I'm not a..." whatever the distraction thing is. In this case, the answer has absolutely nothing to do with anything even close to a chess piece.

As we face off, the white and black I stand beside my twin See sunbeams fall through patterned cracks That lie above this den Of diamond, sir, my heart's composed It makes me who I am Across that heart I keep exposed The mark of Superman If in the end our paths should meet Then head o'er heels let's go Should safeguards fail, there's no retreat As I bring forth the snow

When it comes to knocking evolution, I feel I have to step in. I get upset when people dis my profession. Particularly if they're drawing scientific conclusions from some fables about a spirit who came to earth and healed the lepers and the insane by driving out the demons that caused their illness. Fables which, as EDM even pointed out, don't explicitly exclude a figurative interpretation in which God set in motion the process of evolution as the means through which he created the various forms of life... it would be perfectly understandable to simplify that part of the story in order to convey the main point since people 2000 years ago wouldn't have been able to grasp such concepts. On a molecular level, my colleagues and I use evolution as a scientific tool. If you want to find out how HIV will evolve resistance to antiretrovirals, culture it in the presence of subtherapeutic concentrations and see how it evolves. On a larger scale, I think I put it fairly succinctly in a past post: Maybe you'll concede that viruses and bacteria can evolve resistance to antibiotics, but assert that humans could not have evolved from apes? The 1% difference between the human and chimpanzee genome (implying a roughly 0.5% difference between modern humans and chimps when compared to their last common ancestor) doesn't honestly seem that daunting, particularly if a majority of that is due to expansion or contraction of gene duplications. Expansion of gene duplications doesn't even take evolutionary time periods to happen. Examples are fragile X syndrome and Huntington's disease, both caused by expansion of repeated DNA elements on the X chromosome that can show expansion of repeat lengths and worsening of disease expression that's noticeable even from one generation to the next as the repeats are expanded in carrier women. If you still think that God created man, which form of man did he create? Homo habilis, homo ergaster, homo erectus, homo neanderthalensis, or homo sapiens? And why did evolution stop at that point and require divine intervention to take over? If the issue is not with evolution but with abiogenesis, then that's fine. Theories abound, but there has been no demonstration yet that self-replicating complex particles along the lines of the "RNA world" hypothesis can be generated from early-Earth conditions. To say that it would be impossible seems a bit premature though. Finally, about the second law of thermodynamics, it does not say that more complex things cannot arise from simpler things. It simply says that the entropy of a closed system will increase. You can be inside a box with a bunch of clay can sculpt it into an aardvark if you like, the entropy of the clay will have decreased, but as long as enough energy was converted to heat in the process, the net entropy of the box would have increased to satisfy the second law. And when it comes to the beginning of the universe, it might sound very strange to hear this coming from a scientist, but I don't think the second law of thermodynamics necessarily applies. To rehash an old post again (yes, we seem to revisit these topics a lot):

Had it not been for the closeness of EDM's guess, I would have said "camera", which seems like it could fit the clues if it has a flash and is on a tripod to make lines 9-10 work. But working off of her guess,

Had I only seen the opening riddle, I would have guessed the following which seems to be able to fit the clues rather well. But I can't make it fit with the added hints on page three, so no rhyming for this answer.

Hm, the very things you saw that make you have faith in God are the things that make me want to learn more about biology. All that stuff, arising out of a genome encoding only about thirty thousand genes (and of course a bunch of non-protein-coding regulatory elements that shouldn't be short changed). It's remarkable how well that little program works. But to me a genome seems less like evidence of a vastly more complex creator and more like an example of how intricate and beautiful emergent properties can be.

In my particular case I sort of opened the door myself, a little inadvertently. I was born as a Jehovah's Witness, and when I was about five I was taught about dinosaurs, and I refused to believe that dinosaurs ever really existed because wiping out an entire ecosystem of species would be very un-God-like. My parents showed me Nova episodes of paleontologists actually digging up dinosaur bones to try to convince me that they were real, but I saw them digging white bones out of white rocks and I was convinced that they were just carving out whatever they wanted in order to tell a neat story. Had I not had a dad who was scientifically minded and sort of told me that it's ok to trust science rather than religion, or even worse if I had one that told me that I'd better stay in the family's faith or else, then I might have stayed religious. Or at least not have mentally converted until much later. Like you, I also get the feeling that people are, to a large degree, a product of their environment. Sure, anyone can potentially leave their religion, but in some cases people have facts presented to them in such a one-sided manner, and are stuck in a social situation where they can't easily leave, that they can hardly be blamed for remaining religious. But there are other reasons why I'm not sure if I would qualify as a true atheist, depending on who defines "true". While I don't believe in any god, as you can tell from the Phronism thread I also don't dismiss religion as being unequivocally useless and destructive like many hardcore atheists do. I somewhat sympathize with the quote attributed to Einstein: "Science without religion is lame, religion without science is blind", although I find the religions currently in existence to be far too burdened with drivel and a little short on really inspirational stuff. Maybe that means I don't get to carry the badge of being a "true" atheist, but so be it.

Wow, I wasn't expecting the T2 defeator to do THAT well against a bunch of algorithms that it wasn't specifically trained to beat! It looks like only Jarze's was the same, and dawh's was similar but did have changes. Maybe learing to beat a lot of specific approaches ends up producing an strategy that's good in a generalized situation, the win/loss margin makes it seem like it was more than just luck.

Hm, if the navigators have the entire map of the maze they need to navigate, then it would be fairly straightforward to find the shortest possible route through the maze. Start with a 50x50 array of the maze. The walls are all "X" and the passageways are all blank. Place the number zero on the starting square. Then, place the number one on any open square adjacent to the square with zero. Then place the number two on any open square adjacent to a square with a one (unless the square already has a number in it). Keep repeating until you've filled the maze. Then every square will be filled with a number indicating the minimum number of moves it takes to reach it. Now, start at the exit with number "N", backtrack to any adjacent square with "N-1", then to any adjacent square with "N-2", etc until you reach the starting square. It might be more tricky if the navigator only gets a map of, say, the 7x7 grid of squares closest to its current position. It might also be more tricky if the starting and ending points don't need to be at the edge of the maze, and can be somewhere in the middle (as if you're being dropped into the maze and need to find a target inside it). That would prevent you from being sure that you will eventually reach the exit by sticking to a wall.

Looks like we had kind of the same philosophies in mind, but different approaches to it. For the other algorithms, do you still have the "English" descriptions of how they're supposed to work? Also, how about seeing how the T2 defeater does against all of these algorithms? I wonder if it really does give any indication of how "novel" the strategies are.

Some of us other programmers might be able to give it a shot if you PM us a description. Unfortunately I'm working overnight Monday, but I know I'm not the only other programmer on this topic.

One pair I can think of that fits most of the main clues, like one being bright and powerful, one of them being necessary for life while the other makes it easier, and not working well together

I tend to agree with octopuppy's analysis of this puzzle, but must admit that symmetry argument is not to be taken lightly. I think that it can be addressed as follows, though. The symmetry argument goes along the lines of: either both masters will cooperate or both masters will defect (for after all, they are both masters of logic), and of those two options the mutual cooperation option would produce the best result, so they must choose that one. I counter with the following argument. Suppose you are a master of logic. You're such a perfect master, in fact, that you know with certainty what your opponent will play, because you're both equally perfect masters of logic after all. (If this were not the case, then the whole symmetry argument would break down.) Now there are two possibilities: either you know for a fact that your opponent will choose to cooperate, or you know for a fact that your opponent will choose to defect. Consider the possibility that you know that your opponent will cooperate. Then what would your best move be? It would be to defect and gain two points rather than cooperate and gain one. But then the outcome of the game would be that your opponent chose to cooperate and you chose to defect, which is impossible for two masters of logic. From this, you can conclude that your initial guess (that your opponent would choose to cooperate) would lead to an impossible state and must be incorrect. The only remaining possibility, then, is that you know that your opponent will choose to defect. Then what would your best move be? It would be to defect and lose nothing rather than cooperate and lose beer tokens. You reach the conclusion that the only possible outcome is that you both choose to defect. So basically, my argument is that it is impossible for a master of logic to choose to cooperate (if there is only one round in the game), and therefore it is impossible for both masters to cooperate. The shortcoming of the original symmetry argument is that it fails to realize that mutual cooperation is an impossible outcome before choosing the optimal strategy of the remaining outcomes that appeared to be possible. There is much more to be said about the case of having multiple rounds. But I'll have to wait for another day before getting to that. As an unabashed side advertisement, anyone interested in this thread might also be interested in a thread going on in the Games section: "Rock Paper Scissors Algorithm Contest", especially if you have a knack for programming. In fact, the idea of having a similar sort of algorithm contest in the future based on the prisoner's dilemma or some other sort of game theory topic had crossed our minds...

Um, this problem gives you two objectives. One is to maximize your own profits, the other is to win more than your opponent. Does one of these goals take precedence over the other? Or is there any unambiguous way of telling which of two outcomes is better -- for example, is it better to win $4.50 and do no better than your opponent, or is it better to win $3.10 but beat your opponent by $2?

I already tested it out against the old T2 algorithms to make sure it works. As you would expect, it beat every one of them, with a total of 2159 wins, 24 ties, and 17 losses. This algo could expand on a question that came up earlier of how a program geared specifically to defeat one algorithm might do in a general matchup against other algorithms. The twist is that this algorithm is geared to beat a smattering of different algorithm types. It might be neat to see how it does against the new T3 algorithms as a side exercise, and see if it really does end up showing how "novel" a new algorithm is in any sort of intuitively meaningful way, but I don't think I'd include it as one of the participants in the big T3 tourney for scoring purposes. Especially if we're going with the idea of having one algorithm per person... cuz I really like my current plasmid! algorithm.

All that talk of "algorithm defeaters" earlier gave me an idea. This might be a sort of way to judge how "novel" a new algorithm is compared to the other ones. Behold, the T2 defeater! BTW, it would DEFINITELY be cheating to use this trick of invoking the rpsgo() function within an actual T3 entry, so don't even think about it! This is purely to gauge how original a new algorithm is. if (algoNum==36) // T2 defeater { int alg, round, alg_score, max_score = -1, defeater_play = ROCK; int[] T2_index = new int[] { 2, 4, 14, 15, 10, 11, 19, 20, 12, 18, 30, 13, 29, 17, 21, 28, 24, 25, 5, 22, 23, 31 }; // For each algorithm from T2 .... for (alg=0; alg<22; alg++) { // For each round that you have data for, see what // the algorithm in question would have played for that round // Tally how many times you would guess correctly if you assume // that you're up against the algorithm in question alg_score = 0; for (round=0; round<i; round++) { if (b[round] == rpsgo(T2_index[alg], round, b, a)) alg_score++; } // If the algorithm is the one that best predicts what your // opponent actually played, then play whatever would beat // that algorithm's next move if (alg_score > max_score) { defeater_play = (rpsgo(T2_index[alg], i, b, a) + 1) % 3; max_score = alg_score; } } return defeater_play; }

I think I read a Mathematical Recreations article on this a while back that got pretty sophisticated. Sure, if you're dealing with one other person (or set of people) in one instance and will never deal with them again, then it makes perfect sense to defect and be a traitor because it would give you the better outcome regardless of what the other person does. A more interesting question is, what happens if you will deal with the same person again? Then it would be better if you can cooperate with the other person repeatedly to get a better outcome for all than if you were both constantly defecting. But you also have something to gain if you can cheat every once in a while without blowing the overall cooperation. From there it delved into a game theory analysis which actually looks sort of like the rock-paper-scissors contest going on in the Games forum right now, except of course that it's a prisoner's dilemma contest between different algorithms instead. The exact values of how long each prisoner stays in prison for each possible outcomes are variables that can affect which strategies are optimal. I don't remember what the end result of the paper was anymore though. Maybe if any of us programmers get a significant amount of free time again we could try it out here after rock-paper-scissors. Unfortunately I'm once again kinda pressed for the forseeable future.

lol, can't tell you how many times that's happed to me before. That's why half of my program is comments, even with only 13 lines of commands (although I admit it still seems a little OCD-ish).

Perhaps we should not yet focus on the name of the patient, but first on the names of the other players. And I must admit that plainglazed's guess helped me with this one.

There are potentially limitless ways of scoring how well algorithms do. Margin of victory being one of them, in which case we add up the total scores from each of the individual matchups. Caring not about margin of victory but about how many different algorithms it can beat is another, in which case we care about number of wins. Grading an algorithm on how well it "learns" by considering only the last 50 rounds of each match might be another. Typically, the time-honored tradition of sports is to just count the number of matches won without considering the margins, or else you delve into the oblivion of the BCS monstrosity. If we do want to capture margin of victory in some way that doesn't completely "over"-reward an algorithm for having a blowout, then as long as we have a computer keeping score, we might as well avoid arbitrary step cutoffs between "big" and "small" wins by taking log (base 10) of the number of matches separating the winner from the loser. Get 2 points for a perfect thrashing, 1 point for a win by 10, -1 for a loss by 10, etc. Or make it 2+log base 10 so the scale goes from 0 to 4 instead of -2 to +2. But I think I'd be slightly more in favor of just counting 1 point for a win and 0.5 for a tie as we've been doing.