### geniucos's blog

By geniucos, 4 months ago, ,

Hi there!

As you all are aware, IOI 2018 is now over, and I felt the need to write a few lines about it, about my experience of it, and about my conclusions. I don't normally do that, but this was my last IOI and likely the end of my intense CP career.

So, the problems were absolutely great! Big up to the scientific committee. However, I'm sort of addicted to the codeblocks debugger which ruined the whole experience for me. As a matter of fact, the only reason for which codeblocks remained my primary editor was because its debugger was awesome and the only alternative of a better editor that I saw (better on the long run, including future real-life programming experience) would be vim which I knew is not even close to as friendly when it comes to user friendliness of debugging. I've also had competitions on a regular basis for 7 years and never had a time interval where I could say: well let's try something new cause we don't have anything useful to do (as I would always tell myself that it's better to train on thinking than on getting to know anything else — even the basis of relatively useful stuff). It was also the fact that I was very familiar with codeblocks after using it for soo long. Many would argue that it's my fault that I was ignorant, I even start to feel a bit of regret, as I could've started using another editor or try to learn to use the gdb. I just never thought it would be useful for my olympiad experience and that I will have my whole life to learn stuff, and better use this freedom time to think free, to not acquire any information and just to invent stuff by myself and just think. I'm in 2 minds about it: if I saw this coming, I would've definitely learnt how to use some other editor and/or how to use the gdb, but not being able to predict the future, it still seems that for my purpose, the strategy of just doing CP in the free time (instead of learning something new) was optimal.

Anyway, as what happened to me (and probably to several other participants) sucked at a whole new level, I genuinely want to advise you not to count on that debugger (there are alternatives, like not getting used to use a debugger all the time, or using gdb or who knows what other built-in debugger). I don't want anybody else to go through this. In Romania, in the county olympiad (the phase preceding the national olympiad), in my county at least, codeblocks is the only available editor (and windows the only available OS) so it's pretty common to use codeblocks here. As a matter of fact all the romanian team used it during IOI this year (and we were unequally affected). So my advise goes like: don't count on codeblocks' debugger (preferably, don't count on codeblocks at all, as it would randomly shut down and freeze whenever it wants when unstable), if you are ever going to work on a different environment (for example, if it works fine on your PC and you don't compete in onsite contests, it's probably fine).

Another thing that I've learnt the hard way in all my 3 IOIs is that after so much work and so much desire to achieve a certain goal, in the end, except for a momentary happiness or sadness, it's the experience that you're left with and that you should make the most from. I've had so much frustration and obsession behind that I couldn't enjoy the IOI in the beginning. Take advantage of the social opportunity, make from it something more than a random online competition with higher stake. I was pretty disappointed with this year's organization from many reasons. In many ways it was bad, but I believe the worst part on the long run is that we didn't have enough many opportunities to socialize, to meet new people. However, with enough effort you could still do that. Also, please don't let IOI destroy your self confidence or let it become a goal (happened to me after last IOI and it took months to get it back). It's good to want to do well, but frustration is not healthy and also leads to worse results. After 2 years, this year I learnt how to control my nerves and emotions, and if it weren't for the codeblocks, I would've achieved my goal (as a matter of fact, even under these circumstances, I was exactly 1 character away from gold). And although this is the ending, and is by no standards, a good one, I'll take the best out of it and remember the nice parts of IOI the feeling you have when being part of such a select event.

I'm considering making a blog about my whole CP path, going back to 8 years ago, if there are enough many people willing to know about it, but I hope this one blog to help people set their priorities straight. I used to think that people graduating and telling me stuff like "in 3 years from now, you'll just forget about olympiads and not care at all anymore" or "it's important that you got there" were just bullshitters, people who didn't try hard enough because they didn't care enough. I came to understand part of what they said. It's not like a successful or failed IOI won't stay in your memories, but the huge difference is that right after it, or during the years when you're still eligible for it, you may think about it regularly whereas after that you would just remember when you especially try to do so or someone else brings up the subject. And yes, making it there is a great achievement. Some of us had higher expectations than mere participation or a certain medal and most of those expectations were justified through hard work and previous results. Still, if you had any reasons to believe that you deserve more, then those reasons themselves are enough for you to be content with your knowledge.

Last but not least, don't forget that IOI is nothing more than 2 days of contest, 6 problems, with (usually) slightly higher quality than in the usual contest. So why consider those 2 days more important than all the other contests that you've competed in (say for example the selection contests, where you definitely did well if you got to IOI)? 6 problems can't and will never tell apart the brightest minds in an effective way. IOI probably will never decide to host 3 or 4 days of contest and till then there will always be a decent luck-dependent part of it, but you can minimize that part though training.

This is not a complaining/whining blog and I genuinely want people who have gotten to be frustrated to understand something that I couldn't for too long. I though I've put my life into this thing, and sort of did by training for so long, but I've done that out of passion and IOI only helped me to keep being motivated, but probably I would've done the same amount of work anyway, and it was a lovely ride and IOI is definitely not something worth being seen as a final purpose (not even when considering only CP, let alone life)

•
• +359
•

 » 4 months ago, # |   +133 probably I would've done the same amount of work anywaythis was my last IOI and likely the end of my intense CP career. Uh-huh.This IS a whining blog. But it isn't a bad thing.
 » 4 months ago, # |   +139 but this was my last IOI and likely the end of my intense CP career. Yeah, my friend and teammate was repeating the same after his last IOI and after our icpcwf18. Do you know why his nickname is I_hate_ACM?
•  » » 4 months ago, # ^ |   +21 I would update the blog with my experience with this year's problems as well and when doing that I'll also add an important mention: that the sentence talking about the end of my CP career is not related to my result. It's something I kind of can't control that is dependent on my ACM teammates and on how much free time I'll have. Also, I have to prioritize some things at least in the close future as a new phase of my life begins. There's a "likely" there that doesn't completely exclude the idea of this career continuing to exist, but given the circumstances, it's less probable and definitely not a sure thing. Soo, why is his name I hate ACM? I could guess because he hates ACM but why?:))
•  » » » 4 months ago, # ^ |   +30 Competitive programming isn't only IOI then ACM, though. There are all sorts of other competitions with onsite finals that are hard to get to, but definitely not impossible.
•  » » 4 months ago, # ^ |   +115 Because he doesn't know it's called ICPC, and not ACM? :)
 » 4 months ago, # | ← Rev. 2 →   +90 I was using CodeBlocks when I was starting with programming. I dropped it almost a year before qualifying to IOI for the first time, learned to debug by hard work with command line and never looked back. (Sure, when working with a several thousand-line codebase I don't know perfectly, I'll probably use a debugger, but I never needed it in contests.) It was for the reasons you list, too. I used to think that people graduating and telling me stuff like "in 3 years from now, you'll just forget about olympiads and not care at all anymore" or "it's important that you got there" where just bullshitters, people who didn't try hard enough because they didn't care enough. I came to understand part of what they said. I never have. All sorts of people have always been telling me what my life will be; I always ignored it and my life never went the way they predicted. It's been so long and I still enjoy programming competitions, not any less than before (although I also have other stuff to do, like shitposting). The way I see it, people do a lot of things they don't really care about, so when it stops being convenient, they stop doing them. I just put much less effort into these things and instead focus on what I do care about. Failure in the latter only motivates me.Also, what about your experience with individual problems?
•  » » 4 months ago, # ^ |   0 I totally agree with your philosophy. However, I was exclusively talking about Olympiad. I agree to the fact that a particular event is easily forgettable no matter how important it is, if you no longer have something to do with that. I'm not saying that I won't care about CP anymore though. Like those people were former olympians who were telling me how their feelings regarding the IOI changed over time, and that really makes sense. I want to keep competing in contests from time to time to the extent of my spare time, but the IOI was the only event with an onsite final and with several other contests behind that I have had the chance to compete to. There's ACM as well, but I won't have enough much time for that, and now that I'm getting to a place with more smart people, I would like to also get a taste of life. So far, I just worked all day long because I literally had nothing else to do, and it's yet to be seen how my priorities would change when I would be faced with several options. So what I'll do in ACM is subject to how my life will change in university.And are you talking about the experience I've had with this year's IOI problems? I'm not sure what you meant.
•  » » » 4 months ago, # ^ |   +9 Yeah, experience with this year's IOI problems. That's what I thought this blog would be about based on just the title.
•  » » » » 4 months ago, # ^ |   +10 Ohh shit you're right, I'll actually just update it rather than answer in this comment.
 » 4 months ago, # |   0 What about plain old debug output? Did you even consider it? Just a small tip — learn how to debug using printfs, this is probably the most useful skill from CP that you will need in the "future real-life programming" career.Also, how come that you've previously been to 2 IOIs and didn't have any problems with debugger and this time you had them? Are you saying that computers were bad and couldn't handle CB debugger or Codeblocks got worse or you lost all your debugging skills?Actually, this blog is 99% whining and 1% talking about the real problem. How about you telling us what exactly the problem was? The only thing you said is "don't count on codeblocks at all, as it would randomly shut down and freeze whenever it wants when unstable" — which is not enough to understand the situation.
•  » » 4 months ago, # ^ |   +77 On some Ubuntu versions (like those in this year's IOI) a simple sudo apt-get install codeblocks would install an incredibly buggy version of Codeblocks (random crashes, broken debugger etc. — also quite old, if I remember correctly). The fix is to run sudo add-apt-repository ppa:damien-moore/codeblocks-stable sudo apt-get update sudo apt-get install codeblocks instead, which installs a stable version. There may be other fixes which don't require using a different repo, but I never looked into it.It happens that I was also there for Costin's other 2 IOIs, and all issues akin to this were quickly fixed after the practice session, but surprisingly not this time though. Costin is an amazing coder, and I would only blame the organizers for failing to fix this after it was signaled by multiple contestants during the practice session.Many will say "just use gedit and make aliases for the compilation and run commands, having a terminal in split-screen configuration with the editor", but I will argue that not knowing your environment inside-out can be a 50% slowdown, especially if you are dependent on a GUI for gdb, such as the one Codeblocks provides.
•  » » » 4 months ago, # ^ |   +45 You are poset.
•  » » » 4 months ago, # ^ |   +23 It's clear that you will never understand the struggle of someone who wishes to become better since you're stuck in your small ignorant world. It made me laugh, thank you :) You can't compare 2 people's speeds when one of them is used to the environment and the other is not. I agree with you. Your explanation makes sense and now we know the reasons of your poor performance during day 1. What about day 2? Didn't you have enough time to figure out how to install proper version of Codeblocks, didn't you have time to talk to technical committee of the olympiad or your teammates who apparently have recipes for your problem? I'm talking about Andrei1998's comment now. Last but not least. Dude, it's not only you who works hard towards the goal, it's not only you who has ups and downs — if you don't change your attitude you will have a hard time getting along with other people.
•  » » » » 4 months ago, # ^ | ← Rev. 3 →   +26 Clearly he tried to fix the problem, but he also said technical committee didn't do anything.Andrei1998 wrote I would only blame the organizers for failing to fix this after it was signaled by multiple contestants during the practice session.
•  » » » » 4 months ago, # ^ |   +76 Failure is not frustrating. Failure caused by inevitable and unfair issue is frustrating. This time, it was latter case. Do you really think contestant can install other stable version of Codeblocks? Everyone has a right to talk about unfairness of the contest. I'm not sure it's whining or not, but I think it's perfectly fine to do so.
•  » » » » » 4 months ago, # ^ | ← Rev. 2 →   -13 I doubt this is an issue that can be prevented. Organizers knew it only after the practice session of the contest, which is just several hours away from contest..Codeblocks is just one of the editor they provide, and they used apt-get to install it. It's ok if the issue was discovered before practice session (by image from IOI), but it was not, and I don't think we have a right to expect more from them.
•  » » » » » » 4 months ago, # ^ |   +9 I don't want to say it's organizers' fault. Maybe there was nothing they could do about it (although, it didn't load during practice session at all because of some configuration issue, and I think it was their duty to make a second round of practice for those affected who basically didn't get the right to practice session). Still, being biased, I believe it can be looked upon as unfairness. Unfairness that is not due a particular person, but actual unfairness. Why? Because simply people who weren't coding in a certain way (say without any nice-looking debugger) had an unnatural advantage. Again, nobody's fault, but still something sad, annoying and that can affect relevance of the competition. However, I still prefer to think about the blog as being mainly "how to get over failure" than about whining. Like I had to describe the sad situation in order to emphasize my conclusions.
•  » » » » » » » 4 months ago, # ^ | ← Rev. 2 →   +16 Ok, now I got what you tried to meant by "unfair". And I was shocked to see Codeblocks is unreliable. Our students also told me that Codeblocks crashed very frequently (But I remember that was not as frequent as you). I'll try to ban them in future IOI training camps :(This is a painful and surprising lesson which shows some IDEs are unsafe to rely on. I'm particularly sad because this lesson doesn't help you at all. I hope good karmas could always follow you in the future!
•  » » » 4 months ago, # ^ |   +113 I'm in a chat where one conversation in a week is a normal thing, we don't discuss much. But the last thing started with 'Opened CF first time in a couple of days and there is Romanian whining. At least not that epic as Chinese whining a year ago.'That's how much whining there is in your post. And it becomes worse and worse with every comment by you. Just stop.I'm answering on this comment because it's ridiculously bad. You know who writes useful posts? Radewoosh, not you. You want to whine? That's OK, I understand that you are broken, maybe there will be guys who can support you. But don't cover your whining with that 'I'm doing the right thing, you all just don't understand me'.
•  » » » » 4 months ago, # ^ |   +26 I find it funny how the author thinks that the blog post is supposed to be "motivational/educative" when it is full of negative.Everyone has had one bad contest but that's not how you should handle it.Please don't respond to this :)
•  » » 4 months ago, # ^ |   +36 Also, how come that you've previously been to 2 IOIs and didn't have any problems with debugger and this time you had them? On the previous IOIs they used Ubuntu 16.04 and Code::Blocks from PPA (see Andrei1998's comment). So, everything worked OK. Now, they installed a buggy version from the Ubuntu repos.The issue with Code::Blocks was discovered in the practice session. They also announced it and said it'll be fixed. But they couldn't fix it on the contests. (I didn't use Code::Blocks, but they were announcing this info)
•  » » 4 months ago, # ^ |   +34 Your username perfectly matches this scene xD
 » 4 months ago, # |   +63 I'm considering making a blog about my whole CP path, going back to 8 years ago, if there are enough many people willing to know about it Yes please
 » 4 months ago, # |   +53 No point for you to completely leave CP. You should be able to reach ICPC finals (100%) or even win medals, even if you won't train intensively but get some high orange / low red teammates (there must be some in Oxford).
•  » » 4 months ago, # ^ |   +18 I will try my best to compete in ACM, it's my teammates' limited desire and our limited time that are likely to make it not go so well, but we'll try that. The likely stands for the training part. Like I'll try to compete in random stuff from time to time (ACM being one of them), but there would be a huge difference from how it was so far (when I would train almost all my free time for those events). I'm not sure how likely it is to get to the finals, but there's no purpose in not trying. Thanks for the support:)
•  » » » 4 months ago, # ^ |   +17 There's no way you won't get to the finals. NWERC is a crazy weak region currently.
•  » » » » 4 months ago, # ^ |   +56 Don't underestimate the difficulty of advancing from regions with low quota. In certain years, a single mistake on a single task can be the difference between advancing or not; whether at other regions with larger quota you can have a brain freeze for a couple of hours and still advance if you're really good (assuming you're also not competing with other teams from your university).
•  » » 4 months ago, # ^ |   +42 Afaik many strong people go to Oxford/Cambridge but all of them instantly lose any will to compete in any contest as soon as they get there. Moreover there is completely no support from the university side in competing there — "You wanna go to NWERC? Ok, you're allowed to, but that's everything from us". That results in these universities being rather nonexistent in ACM despite having many skilled people.
•  » » » 4 months ago, # ^ |   +28 And then there is that you don't have much free time anymore. If your teammates go to different colleges, it's even harder to find a common time. The closer it gets to the exams (which in the case of CS at Cambridge, apart from a project in the final year, is 3-4 exams in consecutive days for the whole year), the less people care about ICPC (and quite rightfully so).
•  » » » 4 months ago, # ^ |   +8 "You wanna go to NWERC? Ok, you're allowed to, but that's everything from us" Isn’t it true for virtually all university? At least I can confirm it for SNU / KAIST. I don’t know how universities should support them, too.
•  » » » » 4 months ago, # ^ |   +20 Hm, at least on my university we have professors, coaches and even nontechnical people who help in organizing trips for many teams, booking hotels, airplanes/trains etc. and we don't need to pay for anything. I was convinced it is like that in all (or at least vast majority) universities seriously competing in ACMs. And regular training on weekly basis, 2 one week camps, championship of our university etc. Afaik on Oxford/Cambridge you have nothing or maybe one contest qualifying for NWERC. Do you have nothing of things I mentioned on your university as well?
•  » » » » » 4 months ago, # ^ |   +5 In SNU, the money for trip and hotels are supported from school, but we should organize trip, book hotels/buses for ourselves. We don't have regular training, camps from university. We coordinate a contest for our university once per year, but they are not supported by school.
•  » » » » » 4 months ago, # ^ |   0 people who help in organizing trips For the case of KAIST, we have to fully organize our trip. However, our department usually pays out for the airplane if we ask. Still, our professor was kind enough to inform, like mailing "you should do this in baylor, until due date". I guess for SNU this is in reverse, you have to inform the professor. And regular training on weekly basis, 2 one week camps, championship of our university etc. Regular training.. For KAIST 18WF team, we did most of trainings on winter vacation. It's hard to afford them in semester, but in March / April we were kinda crazy and somehow did 1 ~ 3 trainings a week. For KAIST 17WF team, they did about 4~6 trainings before WF in total.No KAIST student can afford 2 week vacancy for ICPC. I don't understand how people attend Barcelona / MIPT Pre-finals etc on their school period. For me it simply seems impossible. Please let me know if you have any secret tips... But we hold annual contest. SNU does it for fun + small money, KAIST does it to appeal for student council. Problems were trash, but we improved a lot recently. (Next wednesday we have our annual fall contest. We'll again try to make it into gym :D)Everything I wrote here is 100% organized by ourselves. We contact for sponsors, we contact for classroom reservation, we advertise, etc.. Afaik on Oxford/Cambridge you have nothing or maybe one contest qualifying for NWERC. We have national qualifier contest to replace it, so we have no such contest, and it's ok.
•  » » » » » » 4 months ago, # ^ |   +54 No KAIST student can afford 2 week vacancy for ICPC. I don't understand how people attend Barcelona / MIPT Pre-finals etc on their school period. For me it simply seems impossible. Please let me know if you have any secret tips... Looks like education is taken very seriously in Korea. In Russia, especially if it's not Moscow / St. Petersburg, students can afford not to visit university at all during the semester and then still pass exams.
•  » » » » » » 4 months ago, # ^ |   +20 I want to second dalex opinion. Do they require physical presence on classes? If yes, then it's stupid way of forcing students to attend courses which are stupid because otherwise they wouldn't need to enforce presence. If no, then what's the big problem? On our university many students don't attend classes at all and pass exams with good grades because pressure is put on learning by ourselves, at home. There are many professors that are very good teachers and are passionate about what they teach, but reality is that on many compulsory courses teaching is pretty poor and teachers are there because they need to, not because they want to. However final requirements are still pretty high.
•  » » » » » » » 4 months ago, # ^ |   +10 Not all, but about half of professors DO require physical presence in class. I also think it's stupid, but it affects to my grade, so...
•  » » » » » » » 4 months ago, # ^ |   +1 I'm surprised that physical presence is not an obvious thing. About 70% of our classes require that. I really don't know. It may be stupid, but I think it's up to them :p
 » 4 months ago, # |   +44 I'm considering making a blog about my whole CP path, going back to 8 years ago, if there are enough many people willing to know about it, but I hope this one blog to help people set their priorities straight. Include me in the list :D , I want to hear about your 8-year journey :)
 » 4 months ago, # |   0 How did you get into oxford? Is it due to your informatics results or also something else? Also did you get there with full scholarship?
•  » » 4 months ago, # ^ |   +31 There's no shortcut for it. I've just applied as any other student. There's an exam called MAT and you apply with one recommendation and one personal statement. Then, if you're selected, you go to an interview and if you pass that as well you're admitted with some conditions (like a certain level obtained in an English exam, a certain score in the national baccalaureate). There's no merit scholarship given by the university/EU, so I didn't get any. I'm making a loan for the tuition fee (there's a governmental loan available to EU students, which has some special characteristics) and will get a social scholarship (based on parents' income) from the university (but also only available to EU students). The results might have helped but I did pretty well both in the exam and in the interview so probably, provided my knowledge, the contests didn't make any difference
•  » » » 4 months ago, # ^ |   +10 Oh ok. Thanks for answering!
•  » » » 4 months ago, # ^ |   0 Do you need to have good marks in a high school to be applied to the Oxford?As far as I know some American universities require good marks in a high school.
•  » » » » 4 months ago, # ^ |   +13 It matters much less than it does for US universities. Probably it can be regarded as a good or bad thing, but as long as you do well in the MAT exam (say get over 80), and don't ruin your interview, you should be fine. Being orange, for example, should partly guarantee acceptance unless you completely suck at math. The MAT exam is mostly about thinking than about mathematical knowledge so it should be fine. I genuinely believe their selection system is among the best that I have ever gotten to see. You need a strong recommendation probably if your high school record sucks, though. I'm not sure as I don't know many people with low marks as most of them were just precautious, but they do care more about what you know. The referee should make them aware of the reasons your record is not as good as others' and as long as you do well in those phases you should definitely get in.
 » 4 months ago, # |   +8 Conclusion: it's good to use official environments on VirtualBox and write some contests/practice session in it to be aware about such bugs and maybe get used to another IDE.
•  » » 4 months ago, # ^ |   +10 It's actually funny that using the VirtualBox wouldn't reveal the bug. The organizers did that. They've even tried codeblocks on a random number of laptops and it appeared to work. I think the conclusion is just not to use CB. They've even decided upon a new rule that would warn the contestants months before that CB is unstable (like via the site, maybe emails), as a result of what happened
•  » » » 4 months ago, # ^ | ← Rev. 2 →   +30 I was using VirtualBox and codeblocks would crash randomly on there as well. I thought something is wrong with my computer so I thought nothing of it.
 » 4 months ago, # |   -135 pussy
•  » » 4 months ago, # ^ |   +142 Wow this shit has +15. Best community ever.
•  » » » 4 months ago, # ^ |   -18 Would have upvoted, but did not want to ruin it.
•  » » » » 4 months ago, # ^ |   +54 By discrete analogue of Darboux theorem, every comment/blog that has >=69 upvotes at some point had exactly 69 of them. That's very common event, I don't like getting excited about such thing :P.
 » 4 months ago, # |   +14 Wait didn't you have notepad.exe in domestic Olypiad Windows? rng_58 was coding with that (and somehow wordpad.exe as well) years ago.
•  » » 4 months ago, # ^ |   +3 We had notepad, yes. I've omitted editors that are not made for coding indeed. I think there's no way to delete notepad actually:)) Sorry for not making the information complete, I just thought it's obvious that we had some windows-standard editors (not sure about word for example, didn't check it, but probably since there were computers used for teaching stuff like how to use microsoft office, it was there as well)
 » 4 months ago, # | ← Rev. 4 →   +27 WARNING: Whining from a greedy random guy who wanted gold and got nothing.I am just sad that I spent majority of last seven years practicing informatics and I failed unimaginably, where I should have done my best (two times). I could blame statistics this year for making me believe sixth task was easy one and then spending all 5 hours on it. I think on IOI, strategy is way more important compared to any other olympiad I’ve been on. On second day if I just spent my time in optimal way gold would actually be possible for me (100 51 36 seems super ez after contest). But thinking my solution was right for third task and then spending all time blaming bugs in code is what got me to no medal zone. I think ordering tasks by difficulty would be a better solution than statistics if you want to lessen importance of strategy on IOI.(like IMO where I had no trouble achieving more than I deserved). Many from IOI will agree that statistics on second day(at least during first two hours) suggested that third task was the easiest one and likely many had it accepted. (I wrote pesistent li chao tree during contest thinking it was right solution, plus I thought I had all the ideas and also thought cartesian tree was excluded in syllabus). This is still my own failure and I just wanted to share my thoughts. I agree that all this time spent on training was totally not a waste and got me to meet lots of amazing people.
•  » » 4 months ago, # ^ | ← Rev. 2 →   +10 One last opinion about IDE-s. I was using codeblocks myself and it didnt really cause me much trouble. I mostly use vim at home, but uncofigured vim is totally unusable for me. I think one great way to make environment more comfortable would be to give contestants option to provide their own configuration files before contests. In my opinion it is totally feasible to check every configuration file, just as they check keyboards, mouses and dictionarieS.
•  » » 4 months ago, # ^ |   0 Sorting tasks by difficulty is harder than it looks. In hindsight it's easy to say "A lot of people solved problem X and not a lot solved Y, it's obvious which one is easier", but before knowing the results it isn't that easy. Firstly because it's subjective, since the IOI has a greater emphasis on partial scores (aka, lots of subtasks) than the IMO, a problem might be easy to score 50 points, but super hard to score 100, whereas the 100 point solution to another problem is doable but the 50 point solution is also not that easy.But I think the main reason this wasn't done was because the ISC ran some experiments on the past IOIs, where they tried to rank problems by difficulty before the actual contest and the results weren't very good (I'm not 100% sure how "not very good" they were, they just said the results weren't very good).To go around this, the ISC decided to introduce something that could have a similar effect and after much discussion they decided on the statistics gimmick. There was a lot of discussion (by the IOI GA) on whether this should be used this year, but it ended up being used. I guess contestants didn't have enough time to get used to it or think about how it works, but the statistics can be misleading since they don't really tell you the easiest problem but rather the problem with the most points at the moment (which might be because it's very easy to score 20 and everyone has done so already).Hopefully next year people will be more used to the system and what you described won't happen again (or at least will be rarer).
•  » » » 4 months ago, # ^ | ← Rev. 3 →   0 Just to share more details about my experience:As I remember, at about ~30-40 minutes into contest suddenly percentages turned from x% y% 0% to x% y% 70%. I didn't really have much time to actually analyze subtasks of all tasks and guess why this happened. So my intuition told me that it meant that somebody accepted it. This was probably the biggest mislead. Maybe disabling statistics during first hour, or giving it a 15 minute refresh rate could stop contestants like me from relying on statistics too much.Anyways I think stats is a good thing and I've heard good examples about it from other contestants. Spending whole contest on one task is totally not justifiably by this(from me especially as I did same thing last year).
•  » » » » 4 months ago, # ^ |   0 I see, that is unfortunate.The IOI GA discussed solutions like the ones you propose. Specially the disabling it during the first hour, but there were some technical concerns for doing so since such a thing hadn't been tested in CMS.Hopefully they'll fix and test that, but after so much discussion the contestants will be aware of these issues and will know better. To be honest, personally, I was afraid that since we haven't tried it before in an actual contest and contestants weren't used to it, some unpredicted and unpleasant situations (like the ones you described) could happen, but any change to the contest has unforeseen repercussions. So my strategy would have been to just ignore the statistics, but I know it's easy to talk from the outside.
•  » » » » » 4 months ago, # ^ |   0 What about having 2 types of stats, the current one and the percentage of full (or almost full) scores for each task ?
•  » » » » » » 4 months ago, # ^ |   0 That gives you too much information. The goal with this statistic is to avoid exactly that. If you give the percentage of full (or even almost full) scores, the best strategy for top contestants will be to hold off when they have 100 point solutions until the last hour, because they don't want to give off which problems are easier for 100 points. This was one of the key points discussed in the previous codeforces thread (http://codeforces.com/blog/entry/61049) and one of the things the ISC wanted to avoid.
•  » » » » » » » 4 months ago, # ^ | ← Rev. 3 →   0 The point is that the number of people who can solve the intermediate problem isn't so big that it would influence the stats, but it may. For me, fully blind is the best solution then. The grading system worked well in day1 because the difficulty of the subtasks was like combo->werewolves->seats and the difficulty of the full solutions was also combo->werewolves->seats, but when day2 came and meetings was the hardest problem with the easiest subtasks, the flaws of the grading system became clear and affected many contestants. The grading system has a similar problem like holding off your 100-pointers: many people may send easy naiives which mislead other people, intentionally or not.Btw, I told my leader the scenario you said will happen before day1, but it didn't. Combo was like 80% in the first 2 hours.
•  » » 4 months ago, # ^ | ← Rev. 2 →   +8 100 51 36 seems super ez after contest Lol more like 100 90 60. Highways wasn't hard to figure out given the constraints — finding 1 edge on a shortest path is basic binsearch and the idea with BFS trees comes quite naturally afterwards, since we're dealing with shortest paths. There are some 60-pt solutions for meetings that also aren't all that hard.There's even an online solution of meetings that's too slow for 750000, but maybe not for 100000: split the array into segments with size and for each segment and each right end of a query, compute the range of left ends such that the optimal center mountain lies in this segment. For each query, this can be used to determine the right segment for the center mountain, and with some more precomputed stuff (costs for queries covering exactly that segment for all center mountains, left stack, right stack for that segment), it's easy to find the cost for each possible center mountain. Something like time, memory.I didn't have statistics when I tried solving the problems, but figured out meetings was the hardest. I was alternating between thinking about meetings and highway, with a little bit of doll; I figured out a 90-pt solution to highway in about an hour and didn't have anything like a basis of a working solution for meetings (so it has to be pretty hard) in the same time. Doll seemed like an all-or-nothing ad-hoc problem, which would probably be the easiest — in the end, everything was easy after I found a solution by paper and pen method for M=1, N=3, so I was right about that. I'm still thinking about a full solution of meetings...
•  » » » 4 months ago, # ^ | ← Rev. 4 →   +22 100 51 36 super ez100 90 60 ez100 100 100 feasible
 » 4 months ago, # |   +84 It was a bad IOI because I wasn't there. IOI 2019 will be perfect <3
•  » » 4 months ago, # ^ |   -41 It probably will be the worst IOI in the history
 » 4 months ago, # |   -33 I didn't talk that much about my experience in this year's IOI, so I felt like I should say what's on my mind somewhere :).SECOND CONTEST WAS PURE SHIT.BTW I had lots of good times during this journey, thanks to my friends, teachers and online judges :). I hope I won't have to do CP anymore and can find something more interesting to do in my life. Good luck to everyone ;).
•  » » 4 months ago, # ^ |   +23
•  » » 4 months ago, # ^ |   -27 I just checked Codeforces again and saw your downvotes, fuck you all ;) XD.
 » 4 months ago, # |   +92 Let me also put a few words here.To topic starter: how come you used Codeblocks for so many years and didn't know it is super unstable, especially on Linux, especially in debugger part? I've heard it many times, and on many contests it was installed without working debugger.To future IOI (and any other important competitions) contestants: do not assume that everything will run smooth (I don't talk about your solving skills, it's another topic). Every single thing may go wrong in the environment and ruin the contest for you. Hashes may accidentally collide, Codeblocks' debugger may crash, g++ may have bugs. The contest floor may go to blackout, the judges may hang, the translation may contain mistakes, the contest may be postponed or extended for hours, police may enter the arena and ask everyone to exit immediately. Of course no one will blame you in case any of these events affects your performance, because these events are kind of not normal. However, there are many cases when people overcome these difficulties thanks to their good knowledge of related topics, their ability to adapt to changing environment, their ability to think out of the box. So make sure, if your goals are high, to train so that the negative effects of possible troubles are as small as possible.
•  » » 4 months ago, # ^ |   +5 "how come you used Codeblocks for so many years and didn't know it is super unstable, especially on Linux, especially in debugger part? I've heard it many times, and on many contests it was installed without working debugger."Well simply because I haven't heard this thing many times (no time actually). Any time that someone would say something about it, they would fix it. You can look up Andrei1998's comment to see how the thing is usually fixed. I've taken maaaaany contests so far, including 2 other IOIs and several other regionals (like BOI, CEOI) and never encountered anything close to this...Actually the IOI organizers decided to publicly announce on the websites of the future editions that CB is unstable. But so far, in both editions that I've attended, nobody had any problem with it. And because I wasn't aware of anything like that, I didn't see any valid reason to switch to something else. Seeing this on IOI's webpage is a wholly different thing that I wish I had. And I genuinely think that if it was really that clear that CB is unstable, I would've gotten to know that in this whole time. Probably in most cases, when there is desire to make it work, solutions can be found. Apparently they couldn't find any solutions on these particular laptops though. I'm definitely not the only one using it, and to be frank I think there are at least 50 people (which is a huge number) using it, and that was an extra reason to trust that it would work.I'm fully aware that there were things I could've done to stop that from happening but I just didn't imagine the scenario at all. As you said, police may come in and evacuate everybody. By the same judgement ninjas could fall out of nowhere and you'd have to fight them. Those that know how to fight would have an advantage, wouldn't they? It's not like you can foresee anything. Unfortunately I still think that I used the optimal strategy, given my information (that didn't include anything related to CB being unstable)
•  » » » 4 months ago, # ^ |   0 I once had two files opened with codeblocks and it compiled tbe first file every time! Simple reopen of files did help, but it took me 10-20 minutes of debugging, because changing the code did nothing :D Btw i was using windows.
•  » » 4 months ago, # ^ |   +65 It's not something I'd expect at my 3rd IOI in a row though. Of course, you're right that relying on your computer is a bad idea exactly for situations like this. Every single thing may go wrong in the environment and ruin the contest for you. I'm sure as CF coordinator, you have plenty of experience with that ( ͡° ͜ʖ ͡°).
 » 4 months ago, # |   +24 I was pretty disappointed with this year's organization from many reasons. In many ways it was bad, but I believe the worst part on the long run is that we didn't have enough many opportunities to socialize, to meet new people. Can you please elaborate in what ways have you felt that this year's organization (specifically, not including scientific part) was off and what you feel could have been done to improve it?
•  » » » 4 months ago, # ^ |   +13 No shower in hotel room? LUL. Is it a world championship in 2018?
•  » » » » 4 months ago, # ^ | ← Rev. 3 →   +41 I can explain. Like typical Japanese hotels or dorms, for example in NITS (main accommodation in IOI 2018), there are "big bathroom" which has about 15-20 showers and one big bathtub which the size is about 7-10 meters by 3-4 meters. I am pretty sure that it is not crowded and sometimes almost empty. All Japan Team members said like that. We were comfortable using them. But... one very very big reasons why this didn't work properly is, you know, only Japanese residents know this system. That's why many people used showers there. One interesting fact is that, a JOI committee member which also make part of IOI said to Japan Team before IOI like "your team should use big bathroom because foreigners don't know the system, and your team should not decrease the comfortability of showers." So, why did they choose the accommodation called NITS and KEK? Some people may think like "they can use university dorms like University of Tsukuba, which is pretty big and famous and supporter". The reason why they couldn't choose that is, quoting JOI committee member's words like "In many universities in USA or Europe or some other countries, when one semester ends, they will leave dorms until next semester starts. But there are no such culture in Japan." But in my opinion, the fact that there were no showers could be avoidable by some complex ways. One option is like "in order to use university dorms, temporarily (for 7-9 days) move all university stndents in the dorm which IOI contestants use, to another place like NITS or KEK, and then IOI contestants can use it!". Actually there are a convenient dorm for international contestants in University of Tsukuba (which is a co-organizer of IOI!), which is called "Global Village" (you can see infos in Japanese ), and it can hold 500 people (private room, 5 people per "unit"). Of course, it includes a shower per a unit, so they can use conveniently. There are big rooms there called "Community Plaza" and they can internationally interact. If there are no fatal flaw in this way, the IOI accommodation may be more improved. EDIT: I think many people thinks like "why IOI 2018 was not held at Tokyo or Osaka, and was held at Tsukuba?". I actually asked and IOI organizers said like "because, Tsukuba city became the co-organizer."
•  » » » 4 months ago, # ^ |   +41 After last year's IOI when Iran, an economically destroyed country (especially when compared to Japan) hosted us as if we were kings, this year's accommodation seemed horrible. I have attended several IOI's and noticed a surprising pattern: the richer the host country is, the worse is the accommodation quality. For example, in Canada and Australia, even team leaders had bad accommodation with shared toilets and bathrooms.On the other hand, I have also been a contest organizer (not IOI but a smaller contest) and my responsibility was to book the accommodation. There are many nice luxurious hotels in my city. However, I booked the rooms in a modest hotel. Why was I so evil that I didn't want to provide better accommodation? The reason was that our contest budget was very limited. The wealth of a country and the money that can be used to organize a contest are very different things.
•  » » » » 4 months ago, # ^ |   +14 They've proved through the fact that teamleaders were accommodated in a hotel, in single rooms, that they had enough money. Double rooms are not significantly more expensive than single ones, and staying 2 in a room at a 4-star hotel is definitely better in everyone's opinion. So why not accommodate the students (who are the more important part of the event) in the hotel and the team leaders in the dorms? The costs should stay relatively the same. I agree that there is, theoretically, a chance that a wealthy country doesn't invest into such an event, but as I proved above, they definitely had enough money at least to make that solution work. IOI should set its priorities straight: the students before the team leaders who are there only FOR the students
•  » » » » » 4 months ago, # ^ |   +3 So why not accommodate the students (who are the more important part of the event) in the hotel and the team leaders in the dorms? Because the team leaders would complain so much. I see your point but unfortunately this will never happen.
•  » » » » » » 4 months ago, # ^ |   +18 I think that with a bit of organization, students would be able to complain at least 2x more:)) but yep, unfortunately, as I said, they seem to have fixed this as a rule: care more for team leaders (whatever the reasons are). I think that even several team leaders would agree that this is a non-sense
•  » » » » » » » 4 months ago, # ^ |   +13 Personally, I think this is relevant issue, however looked at a wrong angle.It's always nice when the accommodation quality is amazing, but IOI is not a vacation, and it shouldn't be expected.The other angle to look at it would be location. As a contestant, I personally would find it annoying to travel more than half an hour to the contest, and this issue would be something I'd definitely flagged. However, there are also logistical issues -- many leaders were up until 3am translating tasks, while some left a lot earlier, so I can understand why the allocations were made as they were. Furthermore, considering the amount of GA meetings, leaders would always get preferred to the location closer to the main venue.Also, as a contestant I would've preferred this allocation as well -- as someone who sometimes had issues falling asleep before competition days, I'd heavily preferred a single room to a hotel's double. And comparing to my past seven IOIs, I've found student's accommodation to be good.Finally, don't forget, while most of the students are having off-days at school or in-between university; many of the leaders take out a week's long vacation (which is often 25% of yearly allowance) to go to IOI with the students. Some of them take their partners and children with them as guests to not let it go to waste. So I would find the way of thinking "who cares about leaders, let's only think about contestants" to be disrespectful.So, while I agree that IOI should be organised for the students, comparing accommodation quality is not relevant to that goal in my opinion.
•  » » » » » » » » » 4 months ago, # ^ |   +31 IOI is more of a vacation than of anything else. I have always looked at it this way. Fair enough, I guess there are different mindsets here. I personally never viewed IOI as vacation because for me it was all about competition. Sure, no one stops you from socializing and enjoying the event, but it wasn’t the reason I was doing 6-day a week training during summer while others were having school vacations. it's not a vacation, why keep them in luxury Wasn’t saying that this should be used as excuse, only that it shouldn’t be expected. There are so many other important expenses (e.g. volunteer accommodation) that “let’s book an expensive luxury hotel” should be done in my opinion only if everything else is sorted and you still have money. Which looks like has happened in few years, and it’s good to be aiming there, but it’s extremely unrealistic to expect that level of funding every year for a high school competition (you get way more interested sponsors from an industry in an university competition). IMO always gets better accommodation Not in my experience. It's not a business trip (and as proof, no country buys tickets to the business class). Business trips don’t work like that.
•  » » » 4 months ago, # ^ |   +140 I was force to learn how to iron a shirt Oh my god! That sounds terrible!
•  » » » 4 months ago, # ^ |   +15 Thank you for your feedback! Please make sure to fill the survey (which should be provided to you by your delegation leaders) about this IOI.I will make sure that these comments are not lost, however; and if you or anyone else have any feedback or comments about what should be considered for the future IOIs in organizational matters, feel free to contact me personally, and I will aim to share the summaries of that feedback with my colleagues at IC.
•  » » » » 4 months ago, # ^ |   +13 Actually I was waiting for the email with the survey and didn't receive it (as opposed to last year). Should we ask our teamleaders for it? Why weren't the surveys sent to all the competitors directly? I find that more efficient
•  » » » » » 4 months ago, # ^ |   0 Yes, the survey was distributed to GA. Don't know why.I've sent you a PM with the link.
•  » » » 4 months ago, # ^ |   +29 HSC from off-site here.First of all, I also agree with that the host organizer is shit. I am very disappointed that they didn't try hard to gather volunteers just a month ago knowing nothing about how hard it is to organize the 300-contestants contest and suddenly sought help to alumni to be that part, without any salary (while alumni were supposed to work). At last they failed to fill the quota of team guide (honestly I doubt there are many people who can speak English and who want to use that skills for free for a week, because it's a rare talent here). Some technical staffs were working hundreds of hours in August to fill the absence of staffs. I wouldn't get surprised they didn't think anything about food and served with the first glance.However, the most difficult part is, they were always in need of money. I have heard many times that this contest couldn't use a lot of fund, such as IOI 2014, 2016, 2017 (yeah IOI 2017 is the most luxurious one), mainly because the government's support is very little, and they have to find sponsors to get funded. They also tried crowd-funding as well. They have collected more than $10k through that. (but remember: what can they do with that little money? It's just$10 per one participant (not only contestant).) Quality of food correlates very much with how much they paid, at least while it is in cheap range ($1~$10 for one food).Speaking of the shower thing, I completely agree with square1001. It is standard that there are only 3 showers for a bathroom with the capacity of 20. You could've wait inside the bath tub.I am pretty sure that the poor excursion destinations are also related to lack of money. Long ride costs money. They didn't afford to. (Some might think trip to Tokyo wouldn't cost much, but is it really possible to gather 900 members to return to Tsukuba at certain time? It would be a big news if there were some contestants lost somewhere in Tokyo. I and a guest from Japan were almost lost in Venezia, in IOI 2012.)I have never seen an IOI in which leaders or organizers were using worse accommodations than contestants. (IOI 2017? Actually leaders' hotel was even better than that good contestants' palace!) If you say so, never say something like "this IOI organizer assigns contestants to good hotel while they send leaders to a hut" when you become a leader or something. Tsukuba is known for not having enough place to stay properly. When I went there for national biology Olympiad, the hostel didn't have a diner and I was served garbage bento no one likes for 4 days. When I went there for ICPC regional contest, we had to stay days in a hostel much worse than a jail. The IOI2018 volunteers were using that jail hostel. You should be at least happy not to have to stay in those sheds.Shouldn't the IOI 2018 be held in Japan? I agree. I have hoped thousands of times that this troublesome event goes elsewhere. Contestants see only the light side of IOI, but being an HSC allows you to see the dark side of IOI...
•  » » » » 4 months ago, # ^ |   +8 What is HSC ?
•  » » » » » 4 months ago, # ^ |   +21 Host scientific Committee I think.
•  » » » 4 months ago, # ^ |   +28 when Iran, an economically destroyed country (especially when compared to Japan) hosted us as if we were kings Which says a lot about the supposedly rich and supposedly poor countries.
 » 4 months ago, # |   +362 Why this post is posted on Codeforces main page? I couldn't find the reason that this post should be notified to everyone.
•  » » 4 months ago, # ^ |   0 I think the post was pinned primarily based on the title.
 » 4 months ago, # |   -44 But you were literally whining on your ass just right after day 1 contest, weren't you?
•  » » 4 months ago, # ^ |   +12 Dude that's rude.
 » 4 months ago, # |   +15 Thank you for sharing your thoughts with community, we are all humans here, we understand what state you're in. Talk about it, let this life phase pass. Hope this helps! This post will help future onsite finalists not to repeat your mistakes.
 » 4 months ago, # |   0 IOI is poset.IMO > IOI <3