Hello everyone! I have encountered a problem — I can't solve problems anymore that I had no problem with earlier.
I began CP in 2023, and was solving grey to cyan problems. In early 2024 I began to solve many blues and 1900s, after that I was just solving contests, and everything went very well, I really enjoyed it.
In late 2024 I decided to solve more hard (for me at the time) problems, and solved many 2000s. And that's when the problem appeared first: for some reason I lost the ability to solve 1200-1500s. I couldn’t understand what was going on, so I decided to solve a whole bunch of 1200-1400s. It didn’t help AT ALL. I began to avoid live contests for the first time.
Later I tried solving harder problems again, up to 2400-2500, and it went fine again. I solve them a bit slowly, but on my own, without or almost without editorials. And I really liked solving them.
But I still can’t solve 1200s… I do solve them, but not during live contests and much more slowly than I used to, a year and a half ago. When I see solutions to the problems I can’t solve, I realize that I’m overcomplicating everything. How can I get rid of it?
Should I give up on 2500s and solve only 1200s until I stop overcomplicating? (That may ruin the fun of CP for me though.) Should I solve 2500s, enjoy it and stop caring? (Kinda hard to not be demotivated by my low ratings, so it’s also ruining the fun. It's no fun to feel stupid.)
The more I can’t solve easy problems, the more I question my abilities, and it just poisons the whole experience..
Has anyone else encountered this problem before? And how did you solve it? Any help will be highly appreciated!
A Chinese proverb says,'The greatest truths are simplest'.You can solve problems using the easiest method.
Maybe you think that solving problems is the most important thing,but the process is more important.Enjoy it. And believe in yourself.If you don't ,you will become more and more insecure.
I hope my poor English won't cause any problems,I want to say many things,but my poor English let it go.
Thank you for the reply!
Perhaps I didn't express myself correctly. I know the solutions must be simple, but for some reason I cannot see them anymore. I'm sure I'm not the only one who has encountered this, there must be somebody else who has experienced the same, so I wonder how they got out of it.
idk if it's motivating or not, but you have by far the biggest rating difference I've ever seen on What If I Never Brick.
As for advice I'm not really sure what to do. What helps me when I realize I'm heavily overcomplicating something is to step back for a few mins and restart from scratch once I'm back. Sorry if this isn't the most helpful but I really think you have a "mental game" issue by overcomplicating things rather than a problem solving issue.
Also I'd recommend spending more time with what's fun for you. Yes, high rating is nice, but I wouldn't force myself to hours of something that's boring just to see a number go up.
Thank you for the reply! Never heard about this thing (What If I Never Brick), but thanks a lot for trying to motivate, it's really kind of you! <3
"What helps me when I realize I'm heavily overcomplicating something is to step back for a few mins and restart from scratch once I'm back."
Actually, that's what I did today! I had no idea how to solve a B, so I solved C, D and E, and then got back to B and solved it in a couple of minutes. I will now try to do this every time! It should help, simply to get distracted for a while. How did I not even think of this? (I did the opposite: tried to fit different approaches to the solution, but somehow it just didn't work so far, I was only getting more and more stuck.) Anyways, thank you again!
"I'd recommend spending more time with what's fun for you. Yes, high rating is nice, but I wouldn't force myself to hours of something that's boring just to see a number go up"
Earlier this year I tried that and only solved what I liked (and just avoided easier problems, and to do that I avoided live contests too), but with this strategy my problem got only worse. I'm not ready to write live contests in Rainboy style lol, so if I don't find a way to deal with easier problems, I'd have to give up live contests for good... that definitely wouldn't be fun.
I am recently also struggling about this issue, and my symptom is something like I struggle the same amount no matter if it is a hard problem or an easy one, like say my chances of solving a 2100 is P% it will also likely be P% for me to successfully do one in the range 1000-1300 without stumbling on some mistakes or get some ideas so intuitive that it overcomplicates the question and doubles(maybe more considering debugging) my workload. (If you check my contest history it normally takes up to a full hour to just do A B C for Div2 and solve D in less than 30 minutes).
I was originally going to practice more 1000-1300 problems but seems like you tried and it did not turn out to be helpful, would you like to describe how you practiced them? I would like to discuss and hopefully we can all find a solution to solve our issues!
Hi!
So sorry for not replying earlier, for some reason I didn't get any notifications and didn't see your comment before now.
I have the exact same issue, spending a large amount of time on easy and hard problems now. For some reason I overcomplicate the solution for the easy ones.
"I was originally going to practice more 1000-1300 problems but seems like you tried and it did not turn out to be helpful, would you like to describe how you practiced them?" Nothing special, just tried to solve a bunch of the easier ones and hope it would get better, though that didn't really help.
Right now I decided to take breaks more often. I have a feeling that might actually help a bit.
It seem you've overcome this though, mind sharing? :D
You seem to have an issue with greedy algorithms.
If you take more than 15 minutes on those, your practice is probably not good (you solved them already!):
I didn't only pick greedy, but problems you failed hard on. It would be interesting if you can reply to this comment with the time you took to solve them again.
You know the answer already. But in case you are not convinced, https://mirror.codeforces.com/blog/entry/116371:
Hi!
I resolved the problems in the list you sent, and here's the timing. (I had a few breaks in between them)
This is the first time I've solved problems this way with a timer. Perhaps I should do this more, with increasing difficulty, and train more on the level where I start solving them slower? Seems like that could be a good approach...?
These times sound pretty good. At least you don't seem to have a memory issue.
Maybe practice more greedy/constructive algorithms from the problemset, around 1100-1500, with a hard time limit of 10 mins? And in contest, force yourself to try a greedy that seems like it could work. These problems aren't all interesting anyway since they rely a lot on guessing, and the proof is often way more complicated (e.g. 1100 required to solve, 1800 required to prove). Also, skipping is always an option, if you are strong and fast enough for the rest.
To train this rest, you can practice at 1500-1800 rating. That could be the other half of your training. But 2500 is clearly overshot.
Last but not least, it's often advised to not spend too long on problems. The consensus among red coders seems to be about 15-30 mins. E.g.: https://drive.google.com/file/d/1J2x8pIYQ3MXANgvzOgBciWd3d79j_Exa/view. Which is why contests are an issue; it allows to practice under pressure but spending 1h30 on a problem we should tackle in 10 minutes is not a great usage of our time.
Sounds good, thanks! I'll try it out.
I see it like this.
In easier, wrt to your current level, problems you often need to guess rather than solve them -- that's the fast system at work (by Kahneman).
Even more so, in general the optimal strategy to have higher EV of delta is guessing rather than solving. For example, look at 2138B - Antiamuny Wants to Learn Swap, which most higher rated participants guessed (fourtunately wrongly, thus constituting a good example for arguments). That shows that they usually guess problems during contests.
Harder problems require you to think much more and experience more cognitive strain, that's the slow system -- harder problems are relative to your current level, and they are hard because you don't have sufficiently many relevant neuron connections to enable good guessing yet. Here, a word talent is lurking around, since you can't just have whatever connections you want to have, but I think the current hindrance is not talent.
So, I think your current problem is that you overfitted to solving problems rather than guessing them during your training process. I would suggest solving (just kidding, you should guess them) more easier problems and time them to force usage of the fast system more efficiently. Keep in mind, that you were solving a very large amount of problems and depending on how flexible your brain is, you might need quite a few easier problems to fix the overfitting.
Hi!
And how should I train 'guessing' problems? Try to solve a bunch of easier problems as fast as possible, with a timer?
Pretty much, yeah.
I think for you the easy range can be 1200-1400 of today's difficulty (the older the problem is, the more inflated it's difficulty, so, say, for 4-5 years old problems the range can be 1400-1600)
Today's difficulty is off because of AI cheaters. 2+ years ago prevails.
Thanks, I'll try this!
Nice topic. Overkilling isn’t bad — it often leads to simpler ideas later. If it feels too complex, there’s probably a cleaner way. Sometimes you can simplify a lot just by thinking a bit more.
I think that you should definitely not forget to take care of yourself, like exercise, eat well, like a lot of protein and stuff, sleep well (a very important one) and maybe like just sit still and try to not think about anything for a few minutes a day. Whenever I do the last one, it's just so boring that I can immediately focus on other things afterwards. Also maybe make sure you're not deficient in any micronutrients. Cuz I see that when you reached expert in $$$2024$$$, you were improving at a linear rate, so I can't help but feel like something else has changed. That's just my opinion, though.
Hi!
Yeah, right after reaching expert in 2024 I changed schools, and ended up with a lot of mental pressure. I know for a fact that it affected me negatively.
"sit still and try to not think about anything"
I'm pretty sure I needed this the last 1.5 years. Thanks a lot, I'm trying to relax more now. It's really important, everything you said about self-care, and food, and rest.
encountered the same problem, in olympiad finals i couldnt solve easy constructive problem, i tried to bruteforce it and got 74 ...
i once solved a div2c two pointers problem with lazy segtree and binary search lol
Respect! :D