Hi everyone, I started CP about 2 years ago, and while I enjoy CP, I just can't get the results. During summer 2025, I started grinding from 1200's to 1300 and then 1400s. I used TLE's CP31 sheet, did the typical cses problems and did a lot of the the recent problems. After that, I started grinding all recent Div 2. from A to C. However, during this phase, I could not take many contests because I had a job during that time.
Then life got pretty busy and while I tried, there were phases where I would have a whole week without practicing. Now I left that job so I could take more contests. Since the last 2 months, I was suggested to start solving in the 1600 rating (as they told me I had enough problems solved in 1200-1400 range). I have tried to take every contest in the past 2 months and every time I get poor results. I would skip over words, sometimes won't even get simple observations right, it's these stupid mistakes every single time.
I am starting to question my cognitive ability and maybe competitive programming is too far fetched for me. If you guys could go through my profile and figure out where do I fall short, that would be great? :)
Edit: Thank you everyone for your support. Based on all the valuable advice I will take these steps: 1. Use virtual contests more 2. Prioritize my energy (especially on the day of the contest) for cp (This includes taking more rest and mental peace) 3. Use a journal to test myself, track my learnings and spaced repetition
hi, what would you say is the biggest barrier from you getting C on the typical contest? i see that you upsolve it after the contest, how difficult is that and what do you think are those barriers preventing you from seeing the solution
Hi! Thank you for reviewing my profile. So during a contest, it feels like my thinking process throughout the contest is slower than normal practice.
If I did the same practice in morning, I would get it done within half the time probably. The contests (in my time zone) happen in the evening.
I would skip over words, and while I would make observations that could have actually led me to the solution, I tend to forget them or feed thought to them.
My schedule looks like this:
When there's no contest: CP is the first thing in the morning. 3-4 hours
When there's a contest: 6-8h work till afternoon, 1h of football, then the contest in the evening
So to sum it up, I just feel way slower in contests.
i guess if speed is the issue, then i can think of a few things
if you do a bit of timed practice (i like to do 1 problem with 40 minutes time, using this bot), then it can help focus on speed. also imo, i think there are diminishing returns for how long you spend practicing a problem, if you spent 40 minutes and don't have any ideas then i would just read editorial, if you are making progress then feel free to continue). 40 minutes is honestly not that much time, especially for problems at the edge of your skill level, so it kind of focuses on converting problems into ACs in contests with a bit of additional pressure
for sure the other activities you are doing that day can hurt your performance in contest, maybe think about skipping football that day so that you can focus on the contest? there's not much you can do about skipping work i guess. its up to you but a bit of additional rest can help make your thinking much clearer
Yeah it makes sense. Thank you for your kind advice. I will definitely focus on virtual contests and use this bot. Plus the day with contest will have less workload, and will distribute that work over the weekends instead.
fishorz
Its a matter of if you like CP for not.
If you don't enjoy this, then you should find your true interest.
Also, you can try out some other types of "interesting tasks" (if you enjoy CP). that includes but not limited to: interactive problems, communication problems, partial scoring problems, and subtask based problems that appear in OIs.
I love CP. I enjoy solving problems but I am not getting results. Thanks for suggesting them! I will definitely focus on OI style problems as well. But getting a good rating in cf was a personal goal of mine which I really want to achieve one way or another
my friend MohammadDallash faces the same thing....
back at it huh
dude whattt
DO one thing for at least one month :
Try solving a problem of 1300(start from here again), but if you cannot solve it then don't take any help, just think of all the solutions you can think of, if you still cannot solve it on that day, then bookmark that problem(after thinking for that on the whole day, as amount is not your issue) and solve a new one on the next day(if you can solve the problem then choose the next problem).
Now what you have to do with the bookmarked problem is that you visit then after 2 to 3 days again, if you still cannot solve it then do it again after 7 days, if still not able to solve it then see a hint for that.
This may sound hardcore but it for sure trains your cognitive abilty, you yourself will see the effects.
Thank you! I will definitely do more of that now. I was previously following Shayan's advice to solve as many and just look at solutions right away.
i guess it depends on the person but i'm not really a big fan of this strategy. first of all, it's a bit demotivating if you are spending a long time on a problem without making any progress on it. it might be specific to me as well, but keeping a problem open for a long time is an easy way for me to think that i am being productive while actually just wasting a lot of time. if it works for you then that's great, but i see this advice thrown around a lot (partially because its also a lot more satisfying to get a problem after spending a long time on it rather than admitting defeat), but its not for everyone
Only if you are actually trying, if you are not even getting the idea then its better to attempt is again later than seeing the editorial right away. But in case you are actually seeing the editorial then I would say to spend as much time as possible to understand the concept (not the solution).
This might be better in most case ig, this is what I try to do.
I will try to rely more on my intuition. Sometimes when exploring an approach, you feel like this tunnel is going nowhere and you are way off the track, then I would just be ready to give up and look up the solution.
i think codeforces problemset isn't best to grind , consider doing some math to some level t
Project Euler >>
I think that you should take the $$$AGCT$$$ on cognitivemetrics.com and use code $$$FREJARD$$$ at the end to avoid paying any money. The people who revived that test, they aren't the most honest people, but they do know their stuff, so it is an accurate measure of your $$$IQ$$$. Then you won't need to question your cognitive ability anymore.
Hey, how much should be for advancing in cp, like i did that and scored around 115, but mainly because i wasn't able to complete it, just filled randomly for around 20 — 30, so how high it should be
This test is one of those tests where the time pressure is a huge part of the score, kind of like the wonderlic (idk if you were ever sent one of those for job applications and whatnot). Also I think it's one of the tests where guessing, on average, shouldn't affect your score since guessing is $$$-\frac{1}{3}$$$ and a correct answer is $$$1$$$ so out of $$$4$$$ items u will average $$$0$$$ points if you guessed. I'm not entirely sure about this one, but I think that this is more or less how it works.
As for rating, there's only been $$$1$$$ study on cf rating vs $$$IQ$$$ and it used a bad $$$IQ$$$ test but it found that a $$$115$$$ $$$IQ$$$ corresponds to a rating of just around $$$1600$$$, but this comes with some caveats:
So, I would say that you could improve to the $$$1400-1600$$$ range if you made it a goal to practice every day and gave it some time.
Also thanks for taking the test, hardly anyone takes the tests I recommend, even tho it is an objectively positive experience to take such a high-quality test.
I just tried it and is 139 actually that good
Not to rain on your parade or anything, but I'm not sure how good that test is, since there seems to be very little information on it, and idk about the algorithm they use to score it / give you the next item. In general, it's best to stay away from community-made or heavily-modified tests on that site. So the best tests there are the:
Basically, any once-real test that was not modified too much is probably your best bet. Like for example the $$$AGCT$$$ is basically the same as it was when it was used in $$$WWII$$$, and the $$$GRE$$$ is just a combination of old $$$GRE$$$ forms when it was still a good measure of general intelligence. And the $$$AGCT$$$'s norms that they use are pretty close to the $$$1940$$$ ones (it appears that the $$$AGCT$$$'s items aren't susceptible to the Flynn effect, which is a sign that they are good items), and the $$$GRE$$$ norms are also very accurate.
It is unclear how good a lot of the other tests are, like almost all of them, since they use their community to make the norms for the tests instead of the general population, which isn't a problem for the above tests (except for the $$$1926$$$ $$$SAT$$$ I think) since they have norms derived from general-population-esque samples. I'm not sure why they don't have a Wonderlic form available since they have many real forms and it would be the best way to find out your $$$IQ$$$ in a very short amount of time.
If you have true love with CP, I won't say quitting is a good idea. If you are having problems with the problems, maybe you could practice easier problems to get in touch with mistakes. Also, don't be so stressful about it, just think CP as a joyful hobby :D
I hope my advice would be helpful to you in some way. I wish you will have better results in the future :>
Right now the main problem is talking to people, they ask you what do you do? I tell them I do cp, and then they ask my rating which is quite embarrassing for the time I have put to it. That is the only reason rating is a concern rn lol
I don't think that rating is everything. It doesn't reflect all your skills. In my opinion, hardworking is the most important skill that you must have acquired to achieve success for every job. Keep on the good work and I bet that you will have better results in the near future :D
Long answer
I know I may not have the rating/results necessary to say anything, but here goes...
But I would suggest maintaining a separate journal for all the problems solved and apart from upsolving(non — negotiable) . Do re — solving/re visiting the hard problems from past(spaced repetition) and see if you can derive the core idea yourself.
On thinking, keep 50 minutes at least for anything above ur rating, higher rated problem then more time think through each and every damn thing, work patterns in paper. All in all do whatever you mean as "serious effort/serious analysis", if not then leave/abandon the approach try new approach for the same(I know its hard to get creative here, but I feel creativity is one of a huge separator).
And also , no AI btw, that thing kills progress, ask friends, in blogs, view solves of higher rated people, don't ever AI for solution, but use it to analyse your weakness and areas of improvement. Turn off tags when solving higher rated probs. Refer the competitive programmer's handbook.
I am pretty sure Codeforces awards that level of effort over time, and also don't neglect the mental health, sleep, eat well, take breaks when necessary, and reward yourself if you solve hard problems.
Hope you liked it. Higher rated people, suggest any improvement,
Thanks,
Edit: Virtual Contests are goldmine No advanced stuff early on Math topics like graphs, combinatorics, number theory, probability are needed
And more importantly : like cr7 said, hard work beats talent when talent doesn't work hard, enjoy and trust the process, wishing you the best
Thank you. Will do! I actually set up a tracker in Notion which used a script to automatically put my cf submissions in it, but at some point I stopped using it. Will get back to that
Glad to hear, all the best
You seem to be training rather consistently, and have a good count of solved problems. If you keep it up, I'm sure you'll start to see some results; don't lose hope!
As someone who earlier this year was in a somewhat similar position with being able to solve problems fluently in practice but struggling in contests, I do have some training advice. Mainly, I suggest grinding a bunch of virtuals and upsolving from them after (you could do CF Div3-4's and AtCoder Beginner Contests). This really helps you get used to taking contests and improves your speed a lot. Also, while it is exciting to solve hard problems, in my opinion there's not much of a reason to grind problems 600 points above your rating at the moment — these are not the in-contest misses which are holding your rating back. Virtuals allow you to target those weaknesses much more directly.
That makes sense. I have realized many of the 1600s I am doing rn are actually Div 2. Ds, but my main problem is with div 2 Cs. Will prioritize vc more
bottom 10 questions oat but ill answer it anyways
if you dont like cp dont do it
if you like it do it
rating is irrelevant in whether you should do it or not
Auto comment: topic has been updated by ahmadexe (previous revision, new revision, compare).
FWIW, if you go "bruh I should've seen that", instead of "wtf how is anyone supposed to see this" that's a good thing because you will not miss that trick/strategy in the future.
Also yes the journal idea is very good, I like to try to write down 1 thing you can learn from each problem whenever I can, and you'll quickly see yourself writing down a lot of the same things.