tag:blogger.com,1999:blog-29632375.post4031276597754406777..comments2015-07-26T10:48:18.008-07:00Comments on vexorian's blog: SWERC 2011, D, F, G, H and almost EUnknownnoreply@blogger.comBlogger5125tag:blogger.com,1999:blog-29632375.post-37856287899605581592011-11-28T06:46:39.346-08:002011-11-28T06:46:39.346-08:001. Sure, there are some easy problems. And yes, th...1. Sure, there are some easy problems. And yes, there is a reason. They are needed to let weaker teams solve a couple of them, and to have few teams solving none. In other words, to differenciate among the weaker teams. The contest is not all about the top ontestants. On the other hand I do not really agree that there was any ``filler'', even if some are easy or known.<br /><br /> Also, do you know any programming contests that do not contain some problems like that? Topcoder for example is much worse in this respect, there are only three problems, and the 250 is *always* filler, unless it's the final of an onsite contest or something. (And then you need to be quite fast to solve the other two even if you know how). IOI used not to have any easy problems, but now they are including them (although in this particular case I think they went too far).<br /><br />2. You did say it was confusing (not that you were confused).<br /><br />3. You certainly wrote it: "It seems that in every ACM-related contest there is a problem that is about evaluating a parenthesis expression".<br /><br />4. As I said, it is not meant to be a corner case, so I don't care if it's worthless or not. But you did imply it was a corner case: "Knowing ACM problems, it is likely there were cases with .... I bet that's true". What's the point of that sentence then? And how would that have been different in a non-ACM contest?<br /><br />5. I know what you said, I was letting you know that solution was expected to be accepted. On the other hand, for a slightly harder problem it wouldn't have been such a bad idea to let Omega(n log n) solutions time out if it were possible to tell them apart, since there are several O(n) solutions.<br /><br />And I repeat, *none* of the problems in this set had a complicated implementation. One can write the simplest of parsers for F. It takes 50 lines to solve the whole problem with no prior knowledge. Even if the parser were more difficult I do not know what would be wrong with that, parsing is algorithmically interesting too.<br /><br />Finally, I do find the Google Code Jam format the most enjoyable of all as a contestant, but that's another matter. ACM comes a close second to me. I never tried CodeForces though.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-29632375.post-43083631842840748482011-11-28T06:17:24.585-08:002011-11-28T06:17:24.585-08:00Annoying is so subjective. I think something like ...Annoying is so subjective. I think something like Peer review and Guess the numbers are that because they are mostly implementation.<br /><br />I did enjoy the contest, and the problems were good practice for me because I have been slow in implementation lately and the problems that are not Guess the numbers and Peer review did actually need me to think things. If I sounded overly negative it is probably because I really think the ACM contest format has aged a bit too much and needs updating in comparison to things like CF or GCJ.vexorianhttps://www.blogger.com/profile/09588316922172217808noreply@blogger.comtag:blogger.com,1999:blog-29632375.post-76837638937134014042011-11-28T06:02:54.901-08:002011-11-28T06:02:54.901-08:001. Is there a particular reason to have 9 or 10 pr...1. Is there a particular reason to have 9 or 10 problems? When 4 are really filler? Could easily go with 7 problems and keep 2 straightforward ones. I think 2 straightforward problems are good enough to avoid ties with 0 (which is a good and fine objective). <br /><br />Of course there are hard problems in these regionals and never stated otherwise, but the point was that there are always so many small problems, and the purpose of it has always evaded me, to say the least.<br /><br />What I dislike about them is that they are a nice way to lose time before you get to try the more interesting problems (That the scoring system is designed to punish you for not solving the easiest problems first does not help).<br /><br />2. Indeed it was <i>clear</i> and I eventually made the right conclusion, mentioned it mostly because it was my own confusion.<br /><br />"Crystal clear" is not a compound adjective I would feel is fair to use.<br /><br />3. I never said every regional has one, but just that every single time I try a problem set in an uva contest there is one of these problems. Admitely this is the first ACM problem set in a while I try, but I can go through my contest history and show you that for some reason there is always a problem like this.<br /><br />And not just any kind of parsing problem, but a problem about evaluating arithmetic expressions with parenthesis.<br /><br />4. I didn't said it was tricky, or wrong to have that corner case. Though now that you make me think about it, as a corner case it is just quite worthless.<br /><br />5. I never said it was ACM's fault. I explicitly said it was an UVA thing, and I know that goes on UVA's shoulders.<br /><br />"It seems that as usual uva wants you to care about the constant factor."vexorianhttps://www.blogger.com/profile/09588316922172217808noreply@blogger.comtag:blogger.com,1999:blog-29632375.post-9749699014222153292011-11-28T05:49:58.001-08:002011-11-28T05:49:58.001-08:00By the way, I forgot to say. None of the problems ...By the way, I forgot to say. None of the problems in this contest had an annoying implementation. I wonder why you would say that.Anonymousnoreply@blogger.comtag:blogger.com,1999:blog-29632375.post-6585537740054181472011-11-28T05:33:47.364-08:002011-11-28T05:33:47.364-08:00Some comments:
1. If you are looking for an ACM re...Some comments:<br />1. If you are looking for an ACM regional without "3 or 4 straightforward problems", check out e.g. SWERC 2010. But with 9 or 10 problems for five hours, 3 or 4 of them should be relatively straightforward even if the teams are good, as the disastrous scoreboard of SWERC 2010 shows.<br /><br />Also if it's hard problems you're looking for, what about problems A, B and C this year?<br /><br />2. The statement for "Peer Review" was crystal clear. I agree it's easy to make a mistake here, but there is no ambiguity if you pay attention to the statement. It's not meant to be tricky.<br /><br />3. No, not every regional has a parsing problem. But parsing in problem H was particularly easy, as there is not even parenthesis removal.<br /><br />4. Of course there are cases with n = 3 and p, z, x (say). If they weren't the statement should say so. Again this is not tricky at all as you seem to be implying. And how does this have anything to do at all with it being an ACM Regional?<br /><br />5. The O(n log n) solution would have been accepted at the real contest. It's UVA that sets too tight time limits (in particular one of the references solutions for C timed out badly at the online contest). The times at SPOJ should be more lenient.Anonymousnoreply@blogger.com