Page 1 of 1

On 0.000 running time.

Posted: Fri Feb 13, 2015 12:04 pm
by zholnin
Hi,

I noticed that in every single problem top of the statistics board is occupied by bunch of people with 0.000 (or close to 0) running time. I mean every problem, even I have know clue on even if it is ever possible to achieve 0.000 time. Just as an example, you can look at the following problems:

http://uva.onlinejudge.org/index.php?op ... ategory=12
http://uva.onlinejudge.org/index.php?op ... &category=

I have figured out that there can be three options:

- I am hopeless coder if I can't even figure out how you can get close to such timing for those problems
- Some people cheated (e.g. by finding problems testcases in the internet and writing code which just spits out answers for prepared questions, thus doing just one look-up per question)
- There were / there are some issues with the way solutions are timed by judge

Can you give me some hint on what is correct answer to above?

Stanislav

Re: On 0.000 running time.

Posted: Fri Feb 13, 2015 10:30 pm
by brianfry713
I wouldn't worry about it too much.
Even an empty code like this might be judged WA with greater than 0 runtime:

Code: Select all

int main() {return 0;}
The same code submitted twice in a row will usually give a slightly different runtime.
Most of the 0 times were from years ago on the old server.

Re: On 0.000 running time.

Posted: Sat Feb 14, 2015 11:56 am
by zholnin
Thanks, brianfry713.

I've seen lots of questions on "0.000 vs 0.008 debate" - my question is different. My claim is that the tasks I listed could not potentially be solved neither with 0.000 nor with 0.040, and even 0.100 is very doubtful. While there are submissions which claim to do just that.

So this is not about precision of measurement by the judge, it's generally about submissions which seem to be by magnitudes faster then any fastest algorithm I can think of.

Re: On 0.000 running time.

Posted: Tue Feb 17, 2015 11:22 pm
by brianfry713
Yes there are some problems where the judge's I/O has been published and you can get AC by just printing the output.

Re: On 0.000 running time.

Posted: Tue Oct 20, 2015 3:44 am
by maiducnhu
You can use look up tables, memoization, and mathematically reduce the complexity of the problem to achieve zero time. It's a lot harder than it sounds in some case and that's why not everyone gets zero time.