Hello CF community I have been doing cp for 3 years,I solved more than 1341 problems ,but I'm still specialist,I'm really frustrated and I feel that I'm mentally retarted,many of my friends who started with me at the same time are masters and candidates now. Every time I fall down,I summon my strength and fight again,every 10 days I create a marathon on vjudge containings problems with specific difficulty(current one is 2200) and solve them. 
I solved many problems ,and tried many strategies:solving easy problems,hard problems,virtual participation and upsolving but all this hard work was in vain .I'm not smart like others but I'm proud of being a hard worker . So guys you have the statement,any accepted solution please! (sorry for the very poor english)
stop being obsessed with rating and start enjoying CP.
I agree with you!
how could you enjoy cp without getting many accepted during live contests??????
wait kid!
When you start enjoying CP, you will automatically stop worrying about the 'accepted' part.
I will never enjoy cp then.
it's hard to enjoy failure lol.
I'm not talking about "not being able to solve problems". When you enjoy problem solving, the fear of underperforming during contests goes away. I've felt it happen to me too.
Once my mentor said : More hard you try, More better you become. The rating will automatically increase with Tougher Questions.
Many people keep saying this message again and again to him, but this isn't helpful at all. Obviously his rating hasn't automatically increased, and it looks like he has tried and has enjoyed cp enough to keep trying, and just wants to know how to get better, not "how to enjoy cp more".
I would suggest he try looking at harder problems maybe using editorial more, >2200, as maybe having exposure to more clever ideas will help him, as he already has plenty of practice with easier problems, and take part in more vc's..
If you dont care about rating in Competitive Programming , Then what are you competing for ?
If you dont care about rating , then learn and try , but if you want to compete , you have to care about rating ....
Rating is a consequence, not the goal.
what is the goal then ?
For a lot of people ICPC WF/IOI? Some want something to add to the resume but I'd say that's not a good reason. Some just like problem solving + coding + competing. Rating is temporary, you shouldn't obsess about it. I'd say that if I learned something from dota it's that "just queue, don't care about rating, it comes and goes" lol.
I think I'll try to solve 2934 prolems like you,maybe this will help
What makes rating a worse goal than any of those reasons? Rating is temporary, but so is your time at WF/IOI. Plus rating is arguably a better measure of your skill because it is the outcome of many contests too.
A few bad contests in a row doesn't make you worse, in the same way as a few good contests in a row doesn't make you better. So it's better overall to not worry about it otherwise you'll be a pussy and create 12873618237821738 fake accounts.
The point the above guy was trying to make is ICPC / IOI are worse goals than an average rating, because a single contest day can have many unforeseen and random fluctuations.
I mean, I agree with you, obsessing over rating won't help, but then why also IOI / ICPC? By that logic, shouldn't we also not obsess about them too? Because they are also unpredictable
Personally I'd be "Some just like problem solving + coding + competing", the best thing is doing the things you like.
Yes, truly, what a great life it would be if everything we do, we did in a wholehearted way, and just for the pure joy of problem solving ! The best and purest motivation would be enjoying the process itself...
I disagree, ICPC is a much better goal than rating. It is a better goal because it is longer-term, in the sense that the only thing that will help you get a better performance in ICPC/IOI is becoming better at contests.
Whereas focusing on rating will lead to strategies such as skipping contests if they look too hard, which in the long-term make you a worse coder. So even if you don't succeed at your chosen goal, the long-term goal will usually have you getting out more skilled than the short-term goal.
True, but then we must not get very attached emotionally to the ICPC/IOI result, I guess, and not be very sad if things don't work out.
tfg As you mentioned Dota, I wanted to know, do you still play the game?
of course the goal is improving my skills and the rating is the representation of this improvement
He is asking how to improve. and your answer is stop being obsessed with improvement.
No, you should not quit, for you it is a richer experience. I mean, if you fell — there's no point in throwing, you've already been on top — then you'll be there again. This is the rule. Look at my rating, I have a close situation, but I know that sooner or later I will be in the last rating again. And to throw after so many years is nonsense. (sorry for bad english)
Hey that is totally me. Same story, same result. It might be the case that not everybody reaches red some day.
it's really tough when you work so hard and get so little
I think it's your response to problems you can't solve, like what do you do when you are unable to solve a problem ? Do you read editorials straight away?
Bro i think you should take a break of atleast 10 days. , practice hard and dont participate in any random contest , participate only when your mood is good and you have enough confidence .
I think he should take a break too, but don't practice too, Eddagdeg take a break of 30 days! I have not your experience but sometimes take a break is the best way...
Taking a break from contest participation is one of the worst things XD, I have experienced it. You cannot perform to your true potential if you give contests occasionally.
Your hard work will not go in vain if you've completely understood all the problems you solved.
i think consistency and way of practicing matters a lot and i also agree you should enjoy cp more
ps:- i lack consistency :(
Hi bro.
I know this feeling, because my rating also looks like zig-zag).
First of course, you should stop caring about rise/fell of your rating less than 200.The reason is there is two type of rating:
Your real rating rating that is steadily increasing
Codeforce ratinп is just relatively accurate reflection of your real rating.
It doesn't matter did your rating fell or rised it eventually get closer to your real rating.
Second, your rating is rising/falling probably, because you can't solve certain kind of problems, but very good at other kinds. Every time you meet such problem your rating is falling, but if contests doesn't have such problems your rating is rising.
That's why your rating can either grow by 100 or fall by 100, which is not very normal.
What should you do?
The same thing I did: You should figure out what kind of problems are your weak side and concentrate on them.
the problem is that I cannot leave problems,for example today I get stuck to B1 and B2 without even reading problem C but such thing I can't control
Why can't? After reading problem, check which problem have the most number of accepted solutions. B1 and B2 were very hard, compared to C and D, but if you looked at number of accepted solutions, you could see it. Also you should stop caring about your rating; you get nervous and make stupid mistakes, or miss obvious solutions. I mean man, you added prayer to your code. You are just taking it too close to heart.
P.S Your solution for A, was not accepted because alphabet in this problem have 20 letters not 26. I changed all 26 to 20 and it was accepted.
if u haven't any idea on task around 10-15 minutes u should read next (even u had stuck on A). it helps u sometimes, and improves ur upsolving after contest (bcs u already red problem and thought for a while)
thanks , bro this also help me.
try to chill and have a rest. stop solving so much problems, u need not to work hard. u get result when start getting a good emoties from ur life. try to do sport sometimes, play videogames, procastinate. im doing cp ~1,7 years and this is the best thing i got
the reason why I'm not taking a rest,it's because I'm not getting good result,but your rating graph shows the opposite,I'm really confused,what shall I do ,I'll try again
Okay, your solves are promising, better than me in higher difficulty levels. But I think you're doing something really wrong, you need to find it. And I can't help you with that. Something was wrong with my way doesn't mean you're doing the same mistake, but when I got stuck between 1600-1700 for about a year(alts...shhh!), that's when I figured out what I was doing wrong. For me it was, I read editorials too quickly, but that wasn't the worst part, there is a fine line between understanding a solution and fooling yourself that you indeed understand the solution, I fell towards the later, when I used to go back to some problems that I already solved, I failed at them, yet again, even after having solved it like 2-3 months ago. That made me realize I was freaking fooling myself, I still do, but I tried to keep it in check since then, result was crossing the 1800 barrier for the first time(Maybe it was inflation?).
I am still facing the same problems you do, still sometimes getting screwed up in ~1700 rated problems even after having solved 150+ of them. But maybe it's because I have got a sedentary lifestyle and mostly I am kind of sleepy during CF rounds(or maybe because I am too dumb??), I really think I am not giving my 100% in contests, but I guess I never did anyway so that makes it okay :`)
I feel a little embarrassed to come out here, but I also think, I'm moving towards a similar kind of thing. I'm new to competitive programming (started doing it seriously, ~3 months ago), and although I made good progress and reached 1621, relatively quickly, I have been quite stagnant in rating since then. I have been solving a good number of problems, but I'm unable to perform very well in contests because I tend to give up on slightly harder problems.
It often happens that I get stuck on a problem that needs deeper thinking or handling multiple corner cases, and I just give up, sometimes finding out post-contest, that I could have solved the next question(s) (which was maybe based on some standard Graph Algorithms, etc.)
I have a lot to learn, but I'd love to know how experienced coders tend to handle this kind of stress, and how I could possibly improve my ability to think with patience.
I'm at 2000 rating and I still feel like I know nothing.
I'm at 2700 and I feel like I know a bit but I'm too dumb.
what we know is a drop, what we don't is an ocean.
a man lives three lives: The first ends with the loss of naiveté, the second with the loss of innocence, and the third with the loss of life itself
The better you are at something, the better you know what you dont know of that subject
When I didn't start doing codeforces, I used to think I already know enough. Now I am noob of noobs
Having seen the number of questions solved by you and the time invested by you, I don't think that competitive programming is for you. You can only push yourself to a certain extent. Having said that if you still enjoy cp then I would suggest you practice giving some virtual contests to get a hand on strategies that may work for you during live contests. I would still suggest trying something else and not forcing yourself further because if you judge a fish by its ability to climb a tree, you know the rest.
yeah i think you are right cp is not for me maybe i should find a new field to compete in
I kind of have the same problem, it sucks especially when you think about the hours of practice wasted. This is my alt account where i practice [1700-2000]R problems , During practice i can solve most problems in that range but i just suck at contests.
"Whether you think you can, or you think you can't--you're right." --Henry Ford
How much time do you spend per problem? What is your success rate in R1800 problems if you choose ~15 problems randomly and give at most 60 minutes per problem? If your success rate is above 70% then the only reason I can think for you being stuck is competition anxiety. If your success rate is less than 70%, I would suggest you to try practising more in R1800 or easier problems instead of spending too much time on too hard problems.
Hey man!
I cannot tell you how much I can relate to your post. It felt like I had written it. I am also feeling quite low after the last 3 contests. Unable to get anything accepted kills your morale completely.
Unfortunately, I don't have a solution (as you can tell by the color of my handle), but I wanted to write this here to let you know that you are not alone. Don't give up hope. You have shown great perseverance over the years, so I hope you get the results you aim for one day.
Till that day, they only thing we can do is put our heads down and continue grinding.
Cheers!
PS: if you need something to cheer you up, watch some ozzy man reviews on youtube :-)
Check this person I don't know who he is. But I just follow him for motivation also It's interesting to watch him in every contest. He shows up in every contest but still cannot grow. I randomly found him in some contest long time ago. eugalt
If you look at his submissions You will understand why his rating fluctuates (he tries to solve problems with as few lines as possible)
1)
I don't think stagnation is a good reason to quit. Some good reasons are: you'd rather spend your time doing something else, you no longer enjoy competitive-programming, a different activity better aligns with your long term goals.
You suspect your stagnation is due to your innate stupidity. Well, let's say you quit competitive-programming. Then what? This fundamental problem, your stupidity, will be a handicap in any other intellectual pursuit. In other words, you will face the same problem if you replace competitive-programming with chess, poker, physics, mathematics, literature, or any other intellectual activity.
To be frank, I doubt you have reached your potential. Almost nobody does. You are already at a decent level and it is expected that improvement requires more work now than it did before.
2)
This is just my hypothesis, but I think being strong in mathematics is more important than people realize. If you look at the top competitors on this site and their background, many have very strong math backgrounds (e.g. competed in math Olympiad). Um_nik's advice, to just solve many problems, might be a sensible plan for him and others with similar background. But is just solving many problems a good strategy for you? In other words, if solving many problems is not producing the results you want, maybe try something different, like study compsci theory and mathematics?
Just my two cents.
just solve the problem, its that easy!
People have different aptitudes and it's very possible that cp isn't quite what clicks for you and that's ok!
If you enjoy solving cp problems, then definitely keep going. I believe the decision to quit has less to do with improvement and more to do with the joy cp in general brings to you.
I was in the same situation with math. I did competition math for years, practiced, but it was never really my thing. I loved problem solving though and I still do, so I tried out cp and it just felt 100 times more natural to me (and having that math experience helped with the switch). In less than a year doing cp, I easily surpassed what I was able to accomplish in multiple years doing math contests. All it took was getting out of my comfort zone.
Everyone has something out there that they can click with. If cp is frustrating for you at the moment, take a break and explore other topics that interest you. You may be surprised at how quickly you might improve at something else.
Remember, at the end of the day it's never worth doing something that frustrates us. Don't make the decision to give up based on your rating but rather base it on your current enjoyment of how you spend your time. That's my suggestion at least, take it for what it's worth.
May I offer a small suggestion?
I think one thing could be that maybe you are not struggling much with individual problems. I mean, without spending much time on actually patiently sitting and making some observations in a problem, what you are probably doing is that you are looking at editorials to fast.
Of course, all this is my guess, I could be wrong, but my logic is that if one is solving 1300+ problems without much improvement, a reason could be that the element of struggle is missing.
I would suggest, don't aim to maximise number of problems solved, instead, just spend more time on a lesser number of problems, but if you solve them by yourself, taking time, writing whatever observations you make on paper, and then slowly build towards the solution, then maybe it will help a lot.
You may try this : (just my suggestion, I think it might be helpful)
For a new problem, sit down and keep writing some stuff, small test cases etc and then slowly keep making as many observations as you can (like "a number n is always odd" or "if n is even, then the first player of the game can at least draw the game" or "number of possibilities of something is exponential, but the useful ones are actually much much less, almost linear, so we just need to check them" or "no matter what happens, after each operation regardless of anything, the graph has the same number of nodes with even degree" or things like that) and then as you write down observations, usually some combination of these might lead you to the correct answer approach.
Of course, if it is taking too long, you may check the editorial. But the main point is, please struggle, because in contests, usually it is the "struggling to find solution" part which is needed.
This is a very valid suggestion. But what's the upper bound? 30mins? 60mins?
I am able to make observations but many times unable to add them to derive the final answer. So after 30-40 minutes, I code up what I come up with (even though many times I am not confident of it) or I look up the solution.
Another reason of doing that is that I suck at implementation as well. So looking up a solution and then coding it myself is also a valuable learning for me.
Well, truly there is of course no such fixed upper bound, you can do whatever you want. Some people say leave the problem after 60 minutes or 30 minutes.
But,
I think, in this case, what is the best advice is what Errichto says : "try till you are out of ideas" , this sentence truly captures everything.
Sometimes, if you have a lot of ideas, then maybe try all of them and see if they work (this will take more time), and other times maybe you don't have many ideas, then try the one idea that seems obvious and then if you are really out of ideas, check the editorial maybe.
Also, sometimes it is said that instead of reading the entire editorial, it is better to read some small part of the editorial and then attempt the rest of the problem yourself.
Sometimes, people (like once Um_nik had said) have also said that if they can't solve a problem (tried many things but they didn't work), then they leave it for the time and then maybe a month / week later they realise that they can now solve it, or have got some idea for it. Basically they don't look at the editorial (or almost never), instead try to solve the problem themselves, and find that it is much more enjoyable and satisfying to solve the problem without any editorial and it develops a lot of skill too.
There are many different views on this, but the main idea is clear, that the skill they are trying to develop is the skill of making observations and struggling with the problem. And this is only developed by actually struggling. How long you wait before seeing solutions is not some quantity everyone agrees on and is well defined, but the basic idea is to spent some decent time.
I also almost never look at the editorial. I have problems enqueued since almost a year ago and I'm still trying to solve them from time to time. IMHO trying for 30 minutes is ridiculous.
Wow! Even though 30 mins seems less sometimes to me too, yet I never imagined that this number can be bumped to 1 year!
One question, how do you keep track of so many old problems? Does your browser always have a lot of tabs open?
When I have free time, I read a random 2700+ problem and try to solve it. Usually I can solve them in one day, sometimes within the time frame of a single contest, sometimes in a few days, but a very few of them have evaded me for months now. I remember revisiting a single 2700 problem at least 10 times this year, without luck.
but i have one doubt , what if the concept used in the problem which i was unable to solve came in some contest ? still i would not be able to solve that problem because i don't know it's way of solving , how will this method improve me then ? please tell me your way of dealing with this kind of situations . Thankyou
There's a difference between common knowledge (something that applies to many different problems) and original problem-solving. Of course you have to master these (just examples):
std::set, fenwick tree to become cyanBut not without practicing coding skills — this can be done by solving many problems. But to reach red and beyond, you need to be prepared for original problem-solving. Apparently, this means having more advanced mathematical training — mastering concepts from abstract algebra, number theory, combinatorics, but more importantly, building the intuition about when and how to apply these concepts to problem-solving.
This intuition building cannot be done by reading editorials. The editorial will merely give you the solution, not how the author came up with it. You have to do that yourself by trying many, many times.
For the most part, anyone who's motivated enough can reach at least high blue after learning about all the concepts found in The Book and solving a total of a few hundred problems across many different topics.
tl;dr For what your goals probably are, reading editorials is perfectly fine, as long as you don't rely too much on them. But not if you're going for the top 1%.
Well, you solved a lot of 1700-1900 problems and your rating is sometimes around 1700, it isn't that bad or surprising. And CF rewards heavily for being fast with easy problems, while it's not the case with e.g. IOI, ICPC and Codechef Long.
Hey Errichto!
What are you proposing then? Work on speed by solving plenty of easy-medium problems? OR solve more hard problems? Given the primary goal is to increase Codeforces rating.
PS: love your videos man!
Solve problems slightly above your level (difficult but not impossible). That's it, unless you really see that you're very slow with easy problems, then just practice fast implementation (and reading?), I guess.
But, I always solve A very fast and get stuck on B and C most of the times, What would you suggest? I think I am solving very easy problems from a long time.
practice solving problem B
Is this claim true that rating of problem you are able to solve implies your rating will be roughly equal to it?
Are the rating of a problem is given in that way?
Rating of a problem is based on rating of people who solved that problem in contest (with probability of 0.5). Suppose your rating is X, then if you choose 10 random problems of X rating, then you should be able to solve 5 of them.
PS. Correct me if I am wrong. This is what I understood from this blog
Finally, a red coder praising Codechef long on CF.
By the way, love your videos man!!
I think there's some crucial details missing about your training process. You say you are solving problems up to 2200 and indeed some of your submissions show that this is the case, but you don't typically solve anything higher than 1700-1800 in contest. What level of problems are you solving without looking at the editorial, and what percentage of your practice problems do you solve without using the editorial?
I'll start by saying that I think you should only do it for fun, and if you feel you're not at your best under the pressure of a live contest, and it makes it not fun, then either don't do it and just practice (for now), or practice under pressure. Personally, before I came to codeforces (I didn't know it existed), I only did long challenges on codechef (you got days to think) and I liked it alot. Now, if you care about rating anyways —
What does it mean that you've solved all these problems? Did you solve these all on your own? Did you practice doing some of them under pressure of time? Did you take the time to learn from your mistakes?
I haven't been here long, but from what I see it's mostly about speed up to 2000-2200. I see many users that very rarely solve a problem harder than 2000 in a contest but are consistent and have rating of 2100-2200 because of sheer speed. In the same way, I see users that don't solve problems harder than 1700-1800 but are purple because they solve them very fast. For rating, my advice would be to try to practice solving easier questions faster and more consistently. I, on the other hand, will continue trying hard problems because that's what I like! (screw ratings)
I have a dream, one day I will be able to answer such questions.
Learn some patience form ruban and PAG.
You should be focusing on grinding bar by bar. Try 50 problems in 1600 seriously then move to harder ones might help. I think you have enough dedication but lack of practicing strategy.
I am in a similar situation to be speaking. Been stuck at green for past 4~5 months. (I have almost around 500 problems solved varying from 1400-1700)
I found recently that I may have underlying issues with string type-implementation/constructive algo type problems so I currently plan on working on them If that is going to work out. Maybe you should see what causes you to occasionally go down steeply. Analyse if its due to some major weakness and perhaps work on that.
I didn't feel like putting up a blog since too many people have already done so if anybody can suggest me something they are most welcome.
It is unfortunate to see the disproportionate difference between your effort and improvement. I wonder if there's something lacking in your training. Or maybe you need a high IQ for CP? It would be very interesting to see how well you do on https://test.mensa.no/ compared to other people who have reached CM/Master in less time/effort.
it's less than the average 92
That seems to be on the lower end (most high rated programmers probably have an iq >120). Also for reference, most college graduates have an iq ~115 (source: https://psycnet.apa.org/record/1960-03446-001).
It seems that a high IQ involves pattern recognition, which could be very useful for competitive programming. This really sucks and I hope it isn't true, but your IQ could be your bottleneck.
IQ Tests Are not Reliable. And the metric of intelligence absolutely cannot be determined by looking at some random black and white pictures.
A lot of other factors comes into play when you are talking about IQ. Because if you actually practice a few of these IQ tests you'd score better in subsequent ones.
I believed and hoped for a very long time that what you said is true. However, I then saw the research around IQ tests and it seems pretty solid and unequivocal. IQ is one of the most tested and robust metrics psychologists have come up with so far. I know it's a hard pill to swallow, but it's the truth.
Here are some sources corroborating what I said (may add even more later): https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3950413/ https://www.newscientist.com/article/mg23931870-400-the-truth-about-intelligence-do-iq-tests-really-work/ https://www.gwern.net/docs/iq/2015-roth.pdf https://www.gwern.net/docs/iq/2007-strenze.pdf https://pubmed.ncbi.nlm.nih.gov/31378724/ https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4270739/pdf/mp2014105a.pdf https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1288305/
Correlation doesn't imply causation. You can get pretty good at this tests after doing a few of them. If my memory serves me correct, few years back, I took the very test that you linked(casually just for fun), and scored in the near gifted range. GM Hikaru, a chess grandmaster, who is a child prodigy, who is also among the super GMs, scored around 100 IQ points. So, definitely these tests are by no means be all and end all measure of intelligence.
As for the guy who wrote this blog, here is a solution for his problem.
Don't focus on how many problems you solved rather focus on how you solved. Now, go back to the problems you've solved and reconstruct the solutions entirely by yourself, step by step, using logic. You may use the Feynman technique for doing so. And for goodness sake, stop obsessing over internet points!
But correlation in large sample implies probability.
P.S. I am not defending IQ tests.
I'm not saying IQ isn't a real thing. It's just that measuring it accurately needs to take a lot of things in account which simply cannot be measured just by some random online quiz.
Dude I got 121, then why do I suck so bad?
Try solving more problems in the [1800, 2100] range.
I just got 107 on that test. I'm wondering what others get.
I've got 115, could not solve 7 problems on time :(
I got 118:)
I got 112 without using pen and paper :)
I've just taken this test, what a waste of time. Most puzzles are very repetitive, and the mistakes aren't shown at the end. I got 133 but this just measures how good you are with puzzles and patterns.
I am very skeptical about the whole IQ concept, but "how good you are with puzzles and patterns" definitely helps with CP performance, doesn't it? A lot of competitive programming is about noticing familiar patterns and solving random puzzles.
I took the test (got 133 also, lol), and a lot of questions are about guessing or noticing simple patterns like XOR, complements, and so on. I would expect performance on this test somewhat correlate with CP performance, especially if fixing all other possible "parameters" like algorithms knowledge, working style/productivity, and experience.
Sure, there is a huge correlation between good CP performance and being good at puzzles. We just shouldn't call "being good at puzzles" our IQ.
And as you noted earlier, there is a huge correlation between IQ and "being good at puzzles". This is also confirmed by research: http://jhpm.ir/browse.php?a_code=A-10-1061-2&sid=1&slc_lang=en.
If the above is true, why not just call it IQ?
P.S — The immediate reason could be 'correlation != causation'. However, the fact that IQ has such a great predictive power must mean that it measures the underlying true intelligence/puzzle solving quite accurately right?
There's also a positive correlation between intelligence and your salary. Or intelligence and the number of books you have read. Can I call that number my IQ?
Good point. I guess we can agree then that a low IQ is probably a bad sign for problem solving skills but still isn't a measure that is accurate enough to be taken too seriously.
A better measure could be actually trying to solve problems. But if you keep failing at that, is it safe to conclude you don't possess a natural talent for it?
This. Well put! There are a number of factors to how successful you are at anything, and IQ in competitive programming is just one of them. It'll also have a positive correlation with how much you've practiced, how much you know etc.
Also I feel like this test doesn't measure much other than spotting few classes of patterns, which though helpful, shouldn't really be called IQ. (I got 135 on this one). It's time bound, I feel like
People can get better at this by practice as well, since it's pattern recognition under bounded time.
A lot of people will perform much better with more time, once again related to 1. that if you start practicing, average time to solve will reduce
I got 125 :)
I got 102. I didn't use any pen/paper, not sure if that matters though. To be honest, I always feel I am no good with puzzles and stuff.
I got 131. But what does that mean? I am a noob :(
Lol half questions in the test are like "take xor/and or their combination of this and that"
Lul xor and pattern hunting
I got 133. I skipped 5 or 6 questions and couldn't finish them in time. I'm sure others were correct.
I got 133 (99%), I think I got around 4-5 questions wrong. Nothing super fancy.
Take rating as consequence. We don't programme for rating, do we?
Yes, but we do program for the joy of problem solving, and to get better at it. If one can't solve problems it isn't really problem solving anymore. Just problem reading.
Also, isn't a person's rating is highly correlated to their skill level? Anybody spending time in some hobby expects get more skilled as time goes on. I can imagine how frustrating it would be to plateau despite persevering.
Well, yes. Grading makes difference, difference makes depression...
A lot of comments, bro please write a summary in end of your blog, as it is a problem many of us face.
Summary, you have two options :
If you want to be better, solve more hard problems. Don't stop.
or
Say I am not good enough and leave like it is. There are many other cool things in life.
NEVER GIVEUP!
I sometime prefer to change platform, like switching to codechef, spoj and occasionally to hackerearth as well, solve on them for some days and then maybe change to some other, there is no permanent platform for me, but too many platforms are also not right as u might end up solving only easy questions on all of them, that will be a time waste, apart from that I have a barrier in mind like I don't practice question below a certain rating, even if the question is hard or tempting I will leave it (And just wont touch it). Try solving easy DP problems on hackerearth (available in their practice section), If u are able to solve them then ur basics are all right, then if u r competing in like every competition then try avoiding it, I usually practice for some days and then compete in a few competition when I feel I have something new or if just feeling boring. Also Try to compete on some other platform, like codechef lunch time or cook off, u might just be able to get good ranks their with ur kind of practice.
I would suggest doing more virtual contests, and mashups. You've done a lot of questions, but you don't seem to be able to solve them under a timed constraint. So doing harder questions will not help you. The problem is time.
Just my 2 cents. Multiply by 25 and you get a rapper.
Everyone grows at a different pace because of different grasping power.
exactly my story, but please dont quit, I am also fighting, so shoul you, change your goals, I have been trying to get 1900 rating but because of that I am performing bad due to pressure. focus on proving, solving the problems in contest, you have nothing to lose (i dont think you can go lower than specialist now or even 1600 rating for that matter). But please dont quit , keep fighting until you have given your best.
Recent CF contests are just memes. Don't feel bad about poor performance in them.
what do you mean?
First of all , no one is born with blah blah blah . The thing that I felt is that you have to be honest with yourself . I just want to say that if you are solving some problem of any difficulty , then don't just solve it for the sake that you have solved it . Solve it to improve your skill . It happens that people get obsessed with their submission list that what people will think if I do not solve tough problems or maybe people will be amazed if I solve hard problems . But the fact is you do not have to solve hard problems to get better , you have to solve problems which are let's say within +200 of your current rating. But solving means that do not look at editorial before giving it your 100%.
Also, I think you shouldn't compare yourself to your friends, if possible, ask them how they trained.
How can we see how many problems of which rating we had solved?
Codeforces Visualizer
You can use discord for obtaining any statistics related to your CF handle . Use this : https://discord.gg/2CJ6qvY
One person who can motivate you.
No one is talking about actual problem. Although i am not the best to answer this, but still you may read----->
1.You have played 130 contests, which is very big number but I have just played 30 contests and i still feel that i will never go to specialist back (confidence has huge role in your rating).So don't be under confident !2.Main reason of poor rank is wrong submissions which cause 10 minutes penalty(correct me if i am wrong), never ever submit solution directly even if it works for pretests given in problem, try to test the code with some random test cases of your own.Never hurry to submit the code, (at the worst scenario you will lose your 5 minutes) but you will be sure to get right answer.3.Fast solving of questions will improve your rank. (Actually most people confuse this with typing speed).Your major time to solve a problem goes in thinking and paperwork and very less times require to actually type (it takes me 30 minutes to think and 5 minutes to type for D problem).4.Never look at standings, but you can(actually you must) look at number of submissions for a particular problem, it tells you how hard is the question and then you can decide weather to spend time on the question or not.Hope this helps !!
I think the main time eaters differ a lot from person to person.
For me it is most of the time one of two things: Not understanding or misunderstanding the problem, and second being not able to implement the solution (in time, because of lengthy debugging).
Bro, stop thinking about your rating since rating doesn't matters much. What matters is your in depth knowledge of topics. If you will stop thinking about your rating and will try to improve more, definitely you will get results :)
Also you can try to give some Virtual contests as I also got improvement by giving them continuously. You can also try other platforms like leetcode, atcoder etc where I think nowadays problems are more conceptual based as compared to codeforces where nowadays only adhoc and contructive questions are coming which doesn't describe you at all.
Since you're a hardworker guy, you'll reach what you want... maybe next time is your golden one ... no one really knows about it but the thing is just to beilve that no effort will be underestimated :)
Forget about rating....you have 85+ contribution!!!
you took too much time to solve these problems i did almost more than half of these problems in almost 3 and a half months!! solve more as unmik always said !!
Read this blog by E869120.
The most overrated blog
Would you mind explaining why you think so?
I understand that his english isn't perfect and some of the information in it is outdated, but it's certainly a very good resource and is still very relevant right now.
Nothing wrong with that, but it is not as good as it is considered by CF users. It's not the Bible.
Your hard work is inspirational, at least you explored CP and now you can take a call on that, but whatever you decide weather to continue or not, One thing I am sure is that your efforts have not gone in vain, through this you created a nice work ethic, a habit of working hard, perseverance, and I am sure it must have required a lot of motivation and discipline to do this. So now these soft skills that you gained would be valuable and be useful in whatever you do in life. And you will have the satisfaction of giving your best.
the best answer ever
What is that tool you used to generate problem rating graph?
cfvisualizer
solve problems fast and by yourself!!!!dont use the editorial(idk if you do)
you did only around 300 problems without reading the editorial, is this how you became orange on codeforces?
yes. and im sure its so much more than 300 because mashup submissions are not public.
That's nice. I've just started CP and will try to rely as less as possible on editorial
You have done it wrong. You are so hasty, and toxic. And your mindset is really problematic.
Trying does not mean that it will guarantee your success. Solving many problems does not mean that it will make you better, unless you truly understand the concept.
3 years for a wrong path, what a waste of time.
Remember: only problems solved 100% by yourself, without reading editorials, should be counted. Is it still 1300+?
FACTS!
Then how do u learn anything...
I'll argue if you never need editorial you're doing too easy problems.
dalex isn't saying you need to never read the editorial. He seems to just be saying that the title of the blog "I solved 1341 problems up to 2200 difficulty and have no improvement" should be amended to only include problems that the author solved without the editorial. Doesn't make sense to complain about a lack of improvement if you read the editorial too much.
I will still disagree. I understand the point he's trying to make, but even if he read editorial on every single one, and almost certainly he didn't do that, he should have improved more if he's seeing and understanding problems that he's learning from. And putting in your solve count with and without editorial shows a more accurate depiction of how much time you have put into cp. If you were gonna take out any problems from his solve count, it should be those that are really easy like d2 A assuming he can do those quick without gaining much insight in contest.
Kognition I am also in somewhat similar position as op. During the starting I began improving quickly and managed to go to 1850 rating by two months which looked quite good for me. Then, even after practicing somewhat regularly, my rating is dropping and I am now 1680, nowhere near the current position. I don't know what's wrong with me given that I solved 270 problems. Are all problems solved using editorials discarded from counting? In some problems, I made some progress and then couldn't go more myself and saw the editorial. In some problems, I even had right idea but couldn't find proper way of implementation and in some problems, I made stupid implementation mistake, couldn't find it out and saw the editorial then understood the mistake. Would these cases be removed completely or counted partially?
Maybe this is why I am struck. And what would I do to improve myself?
Uh, I don't know that there's any meaningful formula for counting these things to that level of specificity. I just meant that the title as it stands is a little clickbait-y and doesn't really tell the full picture.
The reality for you is that you just haven't been competing for that long, so you can't even really say that you're "stuck" yet. It has not even been 3 months. I think I was low blue at 3 months in. Just keep practicing problems slightly above your level and keep competing. Seriously, that formula works for a long, long time. You can't expect to just go up every single contest. People have bad contests, hell even bad months. You can see that I dropped to mid blue shortly after my first time touching orange, does that mean I was actually that bad? No, I was just having a rough month of contests since I wasn't training much. Keep competing, if another 3 months go by without any real improvement, then maybe there's an issue.
Thanks a lot. I would sure try this formula.
I'd say using editorial is great! If you can solve a problem without editorial (in a reasonable amount of time), then you can probably solve it during a contest. So you don't really gain anything, except maybe now you can solve it faster.
On the other hand, if you can't solve it in a reasonable amount of time, you're probably spending too much time on it which also isn't a very good thing.
You did great in the last Div. 3, ranked 75. Congratulations on becoming blue again. Are you still thinking to quit?
there is no quitting after this hard work
I am just starting cp. You will be an inspiration for me. I will also work hard like you.
Congratulations Eddagdeg on Being Expert again
thank you bro
[del]
life is full of ups and downs bro, compare it with your climb to Expert, maybe you will feel better.
I Have been doing CP for over a year now. Am i killing it ? No , but i enjoy when i reached expert mark for the first time, the feelings i have at that moment cant be described. I do this for this little moments .Everyone is different . Dont compare yourself to others just compare to yourself. You will get a lot of these moments
number of wa is double than your ac, that means u keep submitting without finding corner cases or giving a second thought
.
[DELETED]
I gave up ^__^
Same here bro. I have solved around 600 problems but still I'm hanging in between 1100-1200 from last 2 months. I don't know in what I'm lagging. Take today's contest as a example in which i was just able to solve 1 while my peers solved around 4 and i was like wtf is my brain doing. So just take it easy bro eventually it'll all pay off and just be optimist and look forward. My warm wishes are with you.
I am just a low expert/high specialist, but looking at the problems you have solved, most of them are in your rating range. I think that you should practice 1400 problems. I was stuck in Pupil for a few months after which I started solving ~1600 problems while trying to not look at editorials. Soon after that, my rating went up to ~1700 within a few contests. And now I dont practice much, and I have fallen back to speacialist :( (although I am getting a +113 in today's div3)
I am inspired by you. I suggest to be very patient while giving contest. You will definitely gain high rating.
Did you solve without looking at tags and tutorial?