With a special dedication to all the people from Ukraine.
Hi, Codeforces!
I have the pleasure of inviting you to participate in Codeforces Round 880 (Div. 1) and Codeforces Round 880 (Div. 2), which will start on Jun/18/2023 17:35 (Moscow time). You will be given 6 problems and 2 hours to solve them in both divisions.
The problems were authored and prepared by Arti1990, Okrut, w0nsh, kempusss, Fly_37, kobor, maras, and me. As the tasks are based on problems from Polish Championship in Team Programming for High School, which takes place earlier on the same day, participants from the original competition are asked not to participate in this round nor share the problems.
I want to thank:
- 74TrAkToR for coordination, the round wouldn't be possible without him;
- Alexdat2000 for translating problem statements;
- jacynkaa, oogerbooger, Swistakk, natofp, GreatEagle, Grzmot, Maliniarz, viki88830, AlexLuchianov, Little09, Ecrade_, Kieray, pshat, piratZnachor, piegusek, E_huan, njupt_lyy, p_b_p_b, KeNaj712, Lavine, RUSH_D_CAT, and AquaMoon for testing and many, many advice how to make this round better;
- MikeMirzayanov for Codeforces and Polygon platform!
The scoring distribution:
Div. 1: $$$500$$$ — $$$1250$$$ — $$$1500$$$ — $$$1750$$$ — $$$2250$$$ — $$$3000$$$
Div. 2: $$$750$$$ — $$$1000$$$ — $$$1250$$$ — $$$2000$$$ — $$$2250$$$ — $$$2750$$$
Have fun and good luck!
Anadi Orz
This didn't age very well
For what ? Highly downvoted blog ??
Because of the toughness, D problem was way too tough.
Codeforces is a community for Competitive Programming not for politics, they said
i do not listen to a guy whose username is "anal"
Sir please support Botswana and Russia.
EDIT: Lord Sparky_Master_WCH1226 will punish all downvoters.
EDIT 2: As I foretold, Lord Sparky_Master_WCH1226 did punish you with the amazing problems in this round.
is it rated? don't downvote please, it's my birthday tomorrow
Advance Happy birthday:)
Everyone: 1 Upvote to early-morning-dreams = 1 birthday gift
xD nice trick
omg Anadi round
Dziękuję — дякую — thank you!
Give me Upvote ☺
Thank you for supporting the oppressed people, good luck to all
Score distribution is scary.
Now you know why. Terrible Div1.B.
As a cyan, hope to pass 1500
well, i did :)
I also hoped
Can somebody plz made my contribution positive I am only able to comment once every 10 min so I am not able to ask questions after end of contests.
you may use
talksand ask someone privately.KGB joined the chat
Where did the "With a special dedication to all the people from Ukraine." come from? Not that I dont feel empathy towards the people in Ukraine, but i always thought CP and codeforces was out of politics?
" codeforces was out of politics "
Lol
The authors probably thought expressing empathy towards the people is not the same as supporting politicians
one of the hardest contest in these days
More like stupid contest.
why i felt this contest too tough
C is really hard
I hate this contest
ya more of maths olympiad than a programming contest
mathforces
Fact: 74TrAkToR doesn't know how to coordinate a contest properly.
feel like Div 0.001
What kind of round is this? Very tough.
difficult to understand the statement of problem B in Div2
also even the author is bit more confuse than me I guess :(
Hard. I did A and B but didn't submit.. Im too scared xD. I will check after the contest if they are correct or not. In this case I don't get any rating update, right? I mean, I didn't submit anything during the contest.
yes in case if you didn't submit any problem solution whether it is accepted or Wrong answer verdict it is not mean that you participate in the contest
but if you have a single submission either its AC or WA its conclude as you participate in the contest :)
Still waiting for first D solve in div 2
Standings seems to be frozen XD
Speedforces !!
it's not an interesting competition at all
I apologize for such an expression, but I think the creators of this competition made it up through the ass.
whatever rating will be assigned to problem C, it wont be the actual rating, but way lower. Biggest part of people solving C are clearly secondary accounts from higher level coders.
stuck here between feeling bad about giving up on the contest and being demotivated by the miserable amount of people that were successful with the next problem
Div 1.0000000000000000000000000000000000000000001
Authors be like
ITS DIV 2 THEN
they rounded up just like the guys from problem B
This is my first official competition. I could only solve A , I'm broken .
Take it easy. Most of Div.2 rounds have much easier problem B(even C) than the B of this round.
Div.-1 round
this contest should be unrated lol,the problems are too bad.
look at the gap of the number of participants solved B and C
in div1 only 150 people solved problem B
And look at the gap after C (Div. 2).
It seems that I'll lose "Master" in less than a day.
Thanks for Div1/B.
welcome back to purple after an orange week
Oops! I didn't.
But still thanks for Div1/B.
this is exactly the definition of the word "Balance"
Tougher, compared to recent contests. Especially B.
Why am i getting a penalty for submission of Div2B when the problem statement was changed after my submission? I can't predict the future.
The problem statement has changed, but it actually doesn't change anything.
unbalancedforces makes meh rating accelerates towards 0 :<
Negative delta is in fact a buff for future deltas : )
Div 0 round
why do i always do like shit in div2 rounds that has a div1 also, that B was evil
C was extremely evil, much more than B
In my opinion, C is much more straightforward than B.
Idk, I understood the problem statement of 2 fairly quickly but had to read C like 5 times because it was so confusing and corner-casey
I think C is two sum problem in disguised.
Did you switch the contests? It felt like this was the one meant for Russian school students from my past experience of those rounds.
Lol,i skipped russian round to give this round as russian school round was always shit but i guess fate follows you everywhere.
Insanely unbalanced round (and I don't have the proof of D1B and D1C, though I got PT passed)
please never write anymore contests
Div. $$$0.5$$$ :/
Worst contest I have ever seen!..
What a huge difficulty difference between Div.1 A and Div.1 B
How to do B? It frustrated me to no end
Greedy. First check, for each scientist, fill it with the maximum possible ( the max is g/2 if g is odd and g/2 -1 if g is even ). And for the remainings, we only care about the remainings MOD g since we can add g coins to a scientist and the result remaining the same. And from that remanings, if it's bigger than or equal g/2 + 1, then we can add it to one of scientist but that exact scientist will not get the optimum. If it's lower than g/2, then it's definitely gonna give u a minus. Sorry for my bad english
hes div 1
why now we only care for remain MOD g not remian
let's consider this equation remaining = a*g + r, where r = remaining%g and a is a non positive integer. for a*g, we can give it to the other scientist without changing the answer. let's say the number of coins that scientist has at this time is x. we know that x%g is equal to g/2 or g/2 -1. then, adding a g is not gonna be problem since (x%g) == ((g + x)%g).
but my question is that why such distribution.. like say we 300 ... g be 100 k be 3 n be 3.. then do 49 49 49 ..remainig is 153 ...here why not distribute like 77 and 76 ..then 126 and 125 and 49 still answer is 100 save but prove such distributions are not better than distributing all to 1st one.Write proof if u have for this
well, i didn't say it's better or not. note that here we don't care about the distribution. we only care about the sum of its modulo. so, we can just greedily give each scientist a optimum value
who said urs give is optimal .passing test case don't mean .we should prove ours is best
a series of exchange arguments can work here
split that not in g,g by giving(u said u got mod g ) and try to save more .why not this. urs method is not correct until u prove
I was 9 points for purple, so close and so far :'(
lmao i should stop crying then
After easily proving that E must have a solution for sufficiently large k:
"now finding a solution shouldn't be that hard" — famous last words
I am wondering, how many problems did the testers solve? Seems to me that they should have noticed that the round is unbalanced.
I have never participated in a codeforces so bad...
D1B ?????? Is this round tested?
If it was it would've been canceled. I am kind of honoured of losing rating on these tasks...
Idk, I just absolutely hated something about B, maybe the complicated statement, or maybe its just the problem thats bad
I think its wording, I got what it means and samples in like 20 mins, so that quite demotivated for little while. I think with proper wording that would be normal Div2B (though hell if I know how to word this problem).
What was the approach to Problem 2C?
I can't seem to find the edge case in my submission, so I must have missed something.
I can't see your submission at the moment, but for me the reason I got WA on pretest 2 was forgetting to check that the lower bound on the second number had the right amount of digits.
Same lol, i was trying something with binary search, finding the minimum values of a and b which can give the correct answer. i had 4 5 binary searches XD
"Each input file has at most 5 test cases which do not satisfy A,B,C≤3." is the key point. For above test cases,brute-force with pow(10,A-1) to pow(10,A)-1 and calculate possible range of second number.O(5*10^6) For remaining test cases,t<=10^3.So,repeat the method above.O((10^3)*(10^3)) Total time complexity:O(6*(10^6))
Bad round.
I thought, that in $$$Div1B$$$ the answer is one of $$$[i-2, i+2]$$$, where $$$i$$$ is one of input numbers. This passed several simple tests against bruteforce solution. But it happened, that this is incorrect. I can check for some point the probability, but which points should I check?
I think you also need to check position 0 due to the tiebreaking (e.g. when $$$n < k$$$ you should output 0).
I had bug in implementation related to lower_bound/upper_bound (I always suffer, when use such functions).
But test $$$0$$$ was also required; when I remove it, I get $$$WA11$$$.
They don't ask for any best solution, they ask for the lowest one (you need to check in ascending order and check 0)
this contest has crushed all my hope and dream to get back in Pupil in today, very cruel. Let's believe in the CF Round 881 (Div. 3) will get me back to Pupil ✨
Congratulations you are officially pupil now
C was quite a troll prob. I did binary search. 210159473 Hopefully this passes :pray:
so bad contest, especially 2B :) 9min submit my first 2B solution, about 30min send 2B announcement
Took me almost an hour to realize they ask for the minimum ticket number in B. Too used to "If there are multiple answers, print any of them".
try to take a shower
Very bad experience in div1, for most of the users, your ranking will depend on how fast and accurate you solve div1A. And the rest of the time is garbage time.
For me, this wrong answer submission ruined my whole contest. 210117710
This contest is the shame not only for authors, but also testers.
WOW! We made the same WA. Forgot to check if the count can be negative
Cool round!
Good problems! Maybe just a bit too technical-cornercasy-longlongy.
The contest is bit abstract.Hope the more reasonable difficulty in 2D/1B.
The worst C I have ever solved so far.
very bad contest. it was unbalanced, but more importantly, statement of d2B was wrong!!! It said that limit is g / 2, without ceil parathesis, so i assumed it is floor. I got one wa for this. Just one minute later, you give announcement that it is in fact ceil. Then got accepted. I want delta compensation.
r >= g/2 has a very clear meaning mathematically. It is not the same as r >= floor(g/2).
If you've learned math from writing C++ code, then tough luck.
Eh, g/2 does not have to be an integer and they meant the actual value of g/2 (if it were to be rounded, then it would be either in upper or lower bound or specified that it's an integer)
I don't think so, if you have a integer $$$x$$$ and $$$g/2$$$ which could be non-integer, $$$floor(g/2)$$$ is not greater than or equal to $$$g/2$$$, so in order to compare $$$x$$$, you should use ceil.
I misunderstood "unidirectional edges" in Div1D. I did this problem on undirected graph for several minutes (sad.
The wrong equation ruined D2B. It should always be mentioned whether we have to round up or down during division.
You don't have to round up or down. g/2 is a real number
Are solutions with random in 1C really correct???(I don't think it applies to the birthday paradox) And if they are incorrect,will them FST?
UPD: I got AC. I made stupid mistake in my code. Welcome to hack.
Let $$$a_i=\text{XOR}_{j=0}^{i} g_j$$$, we want to find 2 pairs of numbers with same XOR.
If the 2 pairs share 1 common number, it is easy to find.
Then we have $$$2^{2k+1}-2^{k}$$$ pairs, and the XOR is less than $$$2^{2k}$$$, so we only have about a half pairs that can not find the other pair to satisfy the condition.
Randomly choose the first pair, and we have a possibility of around $$$0.5$$$ to find the second pair, just use std::umap.
Is this Div.2 Round is worst Div.2 of all time ?
What is pretest 17 of Div.1D?
I've been stucking on it from 00:50 to the end of the competition and lost a chance to gain rating :(
Setters should try to include more DSA questions, it's been 28992343 years since I saw a tree or graph problem in div2 C. Mathsforces as usual
no, it's only been 3 weeks. Round 875 had a div2c about trees.
all I'm saying is they're very rare
yeah I agree
Why is this approach incorrect for B:
I fix the first person who we have to beat, and binary search on the second one. For this the condition I use is no. of people b/w these two people should be less than K.
Suppose we get two people using this, L and R. The number of winning tickets is ceil of half the segment length b/w these two, except if one endpoint goes to 0 or M, which I handle with casework.
For the smallest winning target, I take ticket[L]+1 if L!=0 and R!=M, and those cases I handle with some dirty casework.
Naahh this round is full of text, problem statements are very very ... (x3.14) unnecessarily long!
Worst contest I have ever seen!
Never seen such trash round
ShouldBeUnratedForces
I don't know about the difficulty of the questions and the beauty of their solutions But I know that this was the worst contest that I participated in my life (I'm not saying this because I got negative :) )
Worst contest of my life :'(
Mike Mirzayanov forgot to inform that Div.0 appeared on CodeForces
I don't understand how people got the right answer so quickly on Div2/B. I feel like its very evil to speak in terms of integers the entire time, then make the assumptions that contestants would understand g/2 as float division when it is presented.
Because always there is using of floor/ceil brackets if there is integer division (or any other way to tell that it is integer division). Without brackets sure it is float division (and that was in all previous contests).
thanks, this is good to know. in the grand scheme of things it only took 2 WA to learn this lesson, which is fine with me.
that is a separate thing .did u get why distribute the remianing to one 's proof
MathForces
Ceiling division in problem B was not mentioned, so I assumed floor and solved it using that, giving me a wrong answer :/ They should modify testcases to accept both floor and ceil
lol, it was normal division like in math...
it was as trash as weapons US dedicate to ukrain!
can anyone convince Mike to make it unrated because of rounding-up mistake? :)
Reminder
This is the worst problem B I have seen, statement is unnecessarily complicated. The problem statement was changed midway through the round, did not notice it and kept getting WA. At least that case that differentiates the old and new problem statement should be in the samples.
I spent the ENTIRE time thinking it is FLOOR DIVISION instead of CEILING DIVISION since the initial statement was g / 2 and made it impossible to debug
And problem D ?? WTF ?
Why should there be such a silly constraint in Div2C/Div1A? Is this an April fool contest or something?
I got FST on Div.2 A. Can anyone tell me in which case my code missed?
submission: https://mirror.codeforces.com/contest/1836/submission/210114174
For the case:
Your code gives "Yes", it should be "No".
I can't understand statement in this contest...
Now I realise why there were no "As a tester" comments.
Div2B is probably gonna send me back to newbie level :(
Hey , just asking are you really from IIT Bombay ?
Yes
Div1B is really tough.... 1. The overall solution is easy to come up with, I think, and maybe that is why so many was willing to spend time on it without moving on 2. But there are several details need to be considered, in my case, the boundary, the odd position, the left middle point...
Feel I would have a breakdown if...
The way I thought about it helped reduce the "casework":
Say we're considering a position $$$x$$$ for our ticket and want to know how many spots have us as a winner as we move $$$x$$$ slightly (i.e. it stays in the same position if it were inserted into the sorted list of tickets). Consider the $$$k$$$th person to the left of $$$x$$$ (say they are at position $$$l$$$) and resp. for (right, $$$r$$$). As we shift $$$x$$$ slightly, it can be seen that the number of positions that have us as a winner depend only on the parities of $$$x-l$$$ and $$$r-x$$$. So to get an optimal answer it is more than enough to try near every point in the input (as we can walk towards one of the points in the input in steps of 2), plus near the endpoints. Indeed this gives us the leftmost optimal answer as well. So just write the code to calculate the number of points that have a certain location as a winner and try all the stuff that matters.
Here is my bad (but it hopefully highlights what I'm trying to say) code: 210150664.
My main difficulty was in considering what happens when we choose a point that's already in the input. I am very bad at accurately doing these sort of calculations.
Downvoting solely based on the difficulty of a problem is a laughable and shameful act.(I agree that it was hard to read.)
Codeforces round? more like Corner testcases handling/ implementation round
trash round
Div1C is solvable for only 2^K numbers, though I think the non-deterministic won't work in that case.
Please nullify the wrong submissions before the problem statement was changed for div2B. Many of us considered the division to be floor since nothing was mentioned.
What a trash round. Guessforces
pls, don't downvote this contest so much :( Let respect for the problem-setters for this contest !
no.
D2C / D1A is a terrible problem as well. It is so thoughtless and its greatest difficulty is primary school arithmetic. Absolutely no joy in trying to solve that one
for me, problems in div 2 (especially B) were hard to understand except A lol
It seems that C++ double does not store large numbers reliably...
I was using
long long goodCeil(long long n, long long k){ return (ceil((double)(n)/(double)(k))); }but it gives inaccurate results with big numbers
you should instead do the following
(n + k - 1) / kor(n - 1) / k + 1for ceilings.long long has 64 bits for storing digits while double has 64 bits for digits AND exponent (meaning less than 64 bits for digits). If you want to preserve the precision of long long you can consider using long double. And if you want to calculate ceil value you can also use ((n + k — 1) / k) for non negative integers.
only thing to note is that with
(n + k - 1) / 2you might overflown.problem statement of div2 B was very poor.
Take about 90mins to check the weird boundary conditions for WA on test 2 in Div2C and couldn't solve it. Then, going to challenge E and feeling more disappointed. I might as well go to sleep.
I think that 1 second time limit for div.2 C is kinda low
Wrong submission that I couldn't afford cause $$$\lceil \frac{g}{2} \rceil$$$ was not clarified from the beginning. C was hard but we understand we're here to solve hard problems. Can't criticize anything else beyond the range of A-C. Wouldn't down nor upvote this round, imo these mistakes don't deserve this massive $$$1200+$$$ downvotes.
problems were bullshit, the round was trash :)
Mass downvoting for any non-extreme issue is always terrible & childish. People need to learn to walk and cool off after a round before engaging in keyboard warfare. None of it were as horrid as your attacks make it out to be.
Finding the solution to B was nice. Coding and debugging it was torture. C has so many random solutions passed, has more accept count than B. But its editorial is very neat. Wish I could come up with that. Wonder how many actually did in contest.
Very bad calls on the author side. Hope we do see you all again without mistakes like these.
+1 to this, imo the only case when round is deserve to be downvoted is a conscious plagiarism of problem(s)
Let people react how they want to react Mr.Adult.
Learn to comment from your main ID with some guts if you want to contradict someone, Mr Child
Unclear statement for B, but that's not a big deal, round was still trash even if the statement of B didn't have any issue.
Today's first contest was great, tho. I made a huge mistake when I participated in both of them :D
Why is the round this heavily downvoted? I really don't get it, did something happened in Div2?
There were issues with Div. 2 B and C statements, but overall, people didn't like the problems.
I see, statement issue sucks. I'm not a fan of the problems either, but don't understand this much downvote at all
Overcomplicated Div2B statement And only 11 success Div2D submissions
And as far as I understand for big majority of Div1 it's became "who fastest to solve Div1A"
P.S. I don't get downvotes too, but I didn't like it too)
People are just going with the flow, some people started downvoting from all of their accounts because they didn't like the contest and rest (mainly newbies ig) started downvoting, crowd on CF is somewhat immature
From the statement of the problem Div.2 C / Div.1 A:
"Each input file has at most 5 test cases which do not satisfy A,B,C≤3"
Guys, there was no need to hide this sentence at the end of the input description, as this was the key idea to solving this problem quickly. You just made the unclear statement even harder to understand.
.
Can we go for -1000? We know to round percents up! Cheers from Wroclaw!
finally , you touched -1000 , dreams do come true just believe in yourself
Anadi being 62686th in the Top contributors' list. Currently, only 3 users have fewer contributions than him.
Based on my experience, a problem A with 750 points or a problem B with 1250 points are usually much harder than a normal 500 points A or a 1000 points B. So I simply skipped problem B today...
BTW, I like Div1 C / Div2 E. It is not a typical problem that you know you need randomization at the first glance. (Or I didn't practice enough to make randomization an obvious choice)
Shit Round
Both D1B&C are bullshit.
I'm shocked that no testers or even authors have made any clarifications so far.
I have to say, this round deserves to be heavily downvoted as a warning to future rounds.
I was invited to test this round only two days before it started. During my virtual participation, the problemset was like D2B,D2A,D1A,D1C,D1D,D1E,D1F (D1B was added one day before the contest).
A is too hard for the position. It may be better for B or C.
B & C, I have to say, are both bad problems and should be replaced. Also, it can be seen from the dashboard that they do not reach the expected difficulty.
D is okay for this position. I tried a randomized method but ran in 986ms (slowest testcase), so I wonder whether the intended solution was related to randomization. If it is, please extend the time limit for 1~2s; otherwise, please add some stronger testcases.
E is too similar to a classic problem which asks you to find the index of convergence and the period of a directed graph. Solutions to these two problems are almost the same, so I suggest, if possible, replacing it with a better one.
F is perhaps more suitable to substitute for E.
I wrote this shortly after my VP, but the discord group was created one day later. What's more, the round coordinator joined only six hours before the contest. It's obviously tough for authors to deal with the feedback in such a short time, not to mention replacing the problems. I didn't know why the preparation time had to be compressed like this.
After the new problem 'Lottery' was added, I tested and passed in 30 minutes. Fully aware that it was no longer possible to change the problems, I gave the following feedback.
The new D is a nice problem, but it requires too much detailed implementation. I recommend adding some stronger samples to help contestants debug.
But I got no response. Later I @ the round coordinator and repeated my feedback, adding:
Also, it seems that the statement format of some problems is not standard and may cause unnecessary reading trouble. Could you please fix it?
He only replied that they would not change the problem statement unless there was an error in it, still ignoring the issues in other problems. Having nothing to say, I went to prepare for the upcoming ARC. Shortly afterwards, the CF round began, predictably being a disaster.
I hope that in future rounds, testers' feedback should be taken more seriously, and the preparation time should be more abundant.
so why didn't you guys postpone the contest for a day or something for improving the problems. I think that will increase contestants' experience a lot.
Yeah that's what I'd expected, but unfortunately it was only two days before the contest when I tested this round and gave feedback. Postponing a round is certainly not allowed within less than a week.
Sorry, I haven't seen your feedback. The lesson is that the testers' feedback should be shared with CF coordinators and authors.
I shared my feedback in the discord group and repeated it to the coordinator, but sadly he seemed to attach no importance to it and didn't share with you authors.
Just to take some blame off the coordinator — we didn’t give him a lot of time for testing and that is on us.
Thank's for Div2D/Div1B. +252
d is very hard
Bing Chilling
?
A+
your jokes are really funny, you must have a lot of friends at school, right?
Ching Cheng Hanji
Bro I’m sorry if I offended you I thought it was funny for everyone and yes, I do have a lot of friends at school.
Is this the most disliked blog in history?
No. A few months ago there was a round with deliberately copied problems. The announcement of that round received 3000+ downvotes.
More people solved D1C/D2E than D1B/D2D, in both div 1and div 2.
Nice contribution Anadi
Nice contribution Anadi
Is it most down voted blog? :/
What if I tell you that those numbers of downvotes are just one third of the numbers of downvotes of this slightly old announcement.
cool
Problem D in div. 2 and Problem B in div. 1 were the worst
WTF with all the downvoting? I'm not saying the round was great, but it was certainly ok: nothing wrong, no copied problems (as far as I'm aware), no queue, TLs were super generous (tourist's E is $$$O(m^3)$$$, and my first submit in $$$O(m^3)$$$ doesn't pass only because of slow input (and $$$n \le 10^6$$$ didn't help, but it's ok); in F everything can be done in $$$O(n^3 / 64)$$$, and it is even written in editorial, but for some reason TL is 7 seconds, so $$$O(n^3)$$$ passes).
If you don't like one problem and you spend all the contest on this problem — it's on you. Why do you guys even ask for scoring distribution if you can't infer that if B=1250, C=1500 and D=1750, and you don't like B, IT MIGHT BE A GOOD IDEA TO LOOK AT C and D?
B is a fine problem and I don't think it's too hard. For some reason, I couldn't see that calculating a score for one position is simple in 5 minutes, but now I feel like it's me being stupid, not the problem being hard.
I personally think that D is the worst problem in this contest, just because it's kinda just a jumble of well-known ideas, but it's not a bad problem by any means, and it's fine to have one standard-ish problem in a round. And among BCD we have 3 problems with a wide range of topics with roughly the same score, so you have a choice. The fact that you are not using this choice is your fault.
Most people are not good enough to have this luxury of skipping the first problems. In div 2, with 8 min remaining, 1300 people solved C and only 4 solved D (div 1 B). 8 had solved E and none F. I noticed this trend right after solving C and went to problem E first, with no success, then D, with no success also. Basically only 0.08 % of participants in the round had enough skill to solve problems past D. It is not unusual for very few contestants in Div 2 solve E onwards, but it is for problem D. Because D was particularly hard, the contest was unbalanced, that is why so mony people are downvoting.
After reading the D2B and spending some time on it, I saw that the problem probably has a random and intuitive way (which I usually can't solve this kind of problem).
Then I switched to see the D2C, and after reading it, I closed it because it seems to have an ugly solution and nasty code and switched to D2D.
And I was stuck on D2D for a long time and it was at the end of the contest that I realized that many people failed to solve that :)
In the middle, I finally found the solution to the D2B.
PS: I know I shouldn't be so chilly and should spend more time on problems. But two other contests (CF, AC Regular [In which I made the same mistake and I was thinking about problem D almost the whole contest because that seems interesting]) were held before that and I was a little tired.
It seems that the contest made some people have a time travelling