Hello my friends, I was thinking of what things could improve our contest. Please tell me your criticisms or your suggestions in the comments for our future contests.
Hello my friends, I was thinking of what things could improve our contest. Please tell me your criticisms or your suggestions in the comments for our future contests.
| # | User | Rating |
|---|---|---|
| 1 | Benq | 3792 |
| 2 | VivaciousAubergine | 3647 |
| 3 | Kevin114514 | 3611 |
| 4 | jiangly | 3583 |
| 5 | strapple | 3515 |
| 6 | tourist | 3470 |
| 7 | Radewoosh | 3415 |
| 8 | Um_nik | 3376 |
| 9 | maroonrk | 3361 |
| 10 | XVIII | 3345 |
| # | User | Contrib. |
|---|---|---|
| 1 | Qingyu | 162 |
| 2 | adamant | 148 |
| 3 | Um_nik | 146 |
| 4 | Dominater069 | 143 |
| 5 | errorgorn | 141 |
| 6 | cry | 138 |
| 7 | Proof_by_QED | 136 |
| 8 | YuukiS | 135 |
| 9 | chromate00 | 134 |
| 10 | soullless | 133 |
I feel like the Div 2 was quite speedforces, and that the problems were very classic. It was a well contest overall, I just wished that there were more non-classic/non-trivial problems. Thanks!
what's your main acc
why problems nowadays are topic-less?
C and D wayyyy too easy for div2, A and B are fine, D-E gap is huge, I had no idea what to do on E but I headsolved D in like 5 minutes.
Also it was evil of you not to guard against the obvious D fakesolve in samples (the fakesolve being, just sort even indices then sort odd indices). A simple test like
61 2 3 6 5 4would have prevented a lot of unnecessary WA2
I mean if somebody thought that D was that easy, then it's really their fault for WA.
I actually liked that they didn't put this test on the examples, as this spoils the most important observation needed to solve the problem. This isn't just about covering a silly mistake, but is about an actual core logic of the solution, which makes it reasonable that people who missed that observation get WA on pretest 2.
I'm rather sad that I wasted so much time on 1C (2E) because I missed the only $$$n=1$$$ case that makes my first solution fail but the examples didn't have it. It's my fault not to test it myself, but it would've been a super good day for me only if $$$n \ge 2$$$ or the example had provided that case.
oops i wrote
if(n==1) print(1), rip 1 submission and 10 minutes of contest timeDid the same mistake, but lost 17 min due to it.
Btw the condition was redundant but I had it in my code cz initially I thought the problem was actually asking for minimum value, which is I think even <D2A.
Well after I thought of the fakesolve I just thought of the test case 1 4 2 3 within like a minute so I think its more on the participants to notice rather than the setters to provide
I thunk any problem where strong samples give away the solution of the problem, and the problem is more than 1200 rated, is a bad problem.
I don't think so, usually this is true if the strong sample means giving edge cases, but in this case proving that the odd/even solution works or doesn't work is a core part of the problem
D is 1700 according to clist, E is 2200. That is not a large difference.
As the other comment mentioned, I very much like the fact that such samples were not given. Learn to think without relying on samples.
Anyway I didn't want to think during the contest, and I just found all the permutations you can get starting from $$$[1, 2, \ldots, 7]$$$ with brute force. That's almost equivalent to getting all the possible samples with $$$n = 7$$$, and only required a few minutes.
A pure brute force is annoying to write, but in this case it was enough to do random moves $$$10^6$$$ times to get all the reachable permutations.
If the test case was included, D would be even easier
tbh, I feel like it is the first round that would be 11 on a scale of 1 to 10
Make the round 2h45min, and rainboy can AK.
BTW, congrats to rainboy!
Hi, here is my feedback on all the div1 problems.
1A — felt boring. The solution is obvious pretty much immediately and coding it is not fun (to me) even if its fast.
1B — great problem! Definitely my favourite in the round. Really liked how neatly the solution worked.
1C — good problem, no other comments.
1D — my solution does a fairly bashy technique, which is why I didnt enjoy it much. But i agree there are some better solutions as described by others. I think its ok.
1E — way too much implementation. MindSolving it only took about 15 — 20mins, and I liked this part mostly; but then implementation is so long which i definitely did not like
1F — (i got spoiled on this) The idea is fine, but again the implementation is very long. I am surprised 2 participants could finish A — F in just 2 hrs.
Overall, the round was fine. I think having more interesting problems and less implementation would help, especially with respect to 1A, E, F
Especially in problem E? It seems that most passed codes have a complexity of $$$O(n\log^3 n)$$$(some are even $$$O(n\log^4 n)$$$) and their max running times are all 3000+ms. Constant factors are so significant that my 3log code got TLE while my 2log centroid decomposition part has a little big constant factor. I optimized it by pre-initiation and passed at a speed of 4.6s. SAD:(
Not a criticism or suggestion, but I thought 1C was quite cool. Surprised that it had so many solves, although I guess this is partially due to my solution doing something silly/redundant. Thanks for the contest!
too many data structure problems....
Firstly, it was a great contest and it was balanced (one more contest like that and I'll be Expert :) ) I really enjoyed all of the problems I saw (but I was afraid from D) Also, the sample testcases for D was really funny, like hey guys let's solve a problem solved by only 400 people (at this time), by sorting odd indexes and even indexes, let's goo, haha I thought of it as a hint that this is the reason many people got WAs (or a hint to let us know that this will not always happen) and I tried to hack it and I did ^_^
And here are some notes:
A was quite hard, like normally I solve Div2A in ~5 mins or less (my friends did so), but today I took time to solve it more than the time I took for Div2B and Div2C (coz I know the idea of Div2C, saw something similar before, not saying A is harder than C, but for me it was, haha)
Also, I noticed that all of the contest, I didn't find any problem must be solved by some algorithms or data structure, like BS, greedy or DP, this is very strange like I tried to say I am wrong maybe it needs some DP and then I say, Oooh where there is no DP or BS ever, just some intuition and ideas then super ez implementation (I only saw Div2A-Div2D which is the ones are being seen by the majority of the community nearly) Maybe some where, but my sols where different than expected, but like some Adhoc, then sorting and Adhoc then spiral then inversions (I figured it by intuition and saying why not trying it, maybe)
I called Spiral EZ implementation coz it is known, but if someone saw it for his first time it is a nightmare (I had it in my template so I solved it in no time)
Where is your Tutorial???
I'd like to see the proof of D2C I'm sure most of the people guessed it
A with only 250 score makes me choke hard this contests. I think it deserves 500 points at least. (IMO, 750)
I need calculate all the way the answer to the case n % p ^ 0, there's always an answer: Last array with len < p, have sums of (n//q+1)*q-m.
After 2 WA by missing cases... it took me a while to work them all out for the solid solution. While all the points is already gone.
the round would have been good if i had gotten 1 more rating point.
Div2 F>>>>E?
I solved F used about 40 minutes but E about 80 minutes.
Maybe because Chinese are good at data structures but bad at greedy.
1C is too hard while 1D is too easy
I don't think everyone who solves C can prove their solutions
So it should be considered harder than it shows on the board
there are provable greedy solutions, like mine.
Honestly, the scoring distribution wasn't that good. The gap between C and D was way too large, I think D should be around 1250-1500. A and B scoring was pretty nice for a Div 2 considering it was mostly math.
Also, I think you should create more problems that require implicit observations because on problem C, I found the implementation of creating a spiral from the center out in the first 10 min of reading the problem.
I only did 1A, 1B and 1C, but I really liked the contest.
1A was a bit on the easier side but I didn't prove my solution. Just went "I want to put the low numbers close to the middle, so probably a spiral, right?"
1B was amazing.
1C was good.
what's that building behind u
The great pyramid, in Egypt. I visited after participating in the ICPC world finals.
1A was good, I proved my solution instead of guessing it
1B was very good
I didn't solve 1C but the solution looks cool
Time limits for some problems were a bit of a issue.(Also the solutions were running 1.5x slower in the contest which was kind of unpredictable.) I got -104 but appreciate the problems, nice round.
Also do you guys knew deninc dividng(https://mirror.codeforces.com/contest/1693/problem/D) is similar to 1D? I got the wrong solution for D and actually it passed the above problem, which is hilarious.