Why am I still green after 7 months of starting CF?
I spent 2.5 hours / day on average for CF and solved 400 non *800 problems on various topics.
read & follow advice from "how to improve", "how to become expert"
"practice problems slightly higher than your rating", "learn binary search", "do virtual participation", etc
I did everything I could and still green.
Normally, if regular person spends this amount of time and effort on single topic, then it should be *1600 (elo rating, expert in cf) but that was my personal belief.
this is far far below average, I've never experienced this level of suffering in my life from middle school to college entrance or even gaming.
I'm getting serious and start to reconsider my life.
I think that you should do more harder problems, especially from recent contests.
isn't *1300-*1600 enough on my level?
I've heard that you should do problems which have 200-300 higher rating than your desired, so in your case 1600-1800 problems. Also do more recent ones, because at least for me current problems labeled as hard as those let's say from 5 years are harder.
I think that is normal if you didn't have math background. There's no need to be depressed over that you're doing great!
is 800/800 score on SAT math & AP calculus counted as math background?
math background = experience in MO (math olympiads)
if you never did MO then it's quite hard to improve at the low levels (pupil or less)
Disagree
I never did MO but I feel it's not very hard to improve at the low levels
I'm already cyan now
Learn new topics like dp trees and graphs improve your dsa and algorithms
what can I say then lol
sometimes even by doing everything correctly it takes more time for certain people to see the results of their hard work,So we gotta be patient and keep practising
Focus on your weaknesses. Upsolve problems and pay attention to why you couldn't solve them. Was it some observation you missed? Just a lack of understanding of the topic? For example, from a brief look at your recent contests you tend to struggle with dp and number theory. Do some topic-based practice on those tags until you feel comfortable. Then rinse and repeat.
maybe do ioi style problems?
IOI problems are rarely easier than 2200, there is no point solving them if he is a pupil.
don’t necessarily mean ioi problems. But problems from national olympiads that aren’t so mathy and more problem solving imo
Hi, this may not be really helpful,but personally,for me it was crucial to have two things:
Good friends with same goals (because you need someone you can rely on)
Worthy opponents (you only progress when you play with stronger opponents)
The first aspect doesn`t let you burn out,whether the second one prevents you from stopping
To be really honest with you (and I'll be honest because I want to help you), if someone showed me a picture of your problem rating graph (the one you get with CF Analytics, for example), I would guess that you are green.
And that's not a bad thing, I'm sure that the you now is a lot better than the you 7 months ago. For me, I remember that I had like >= 70 problems of each rating x, x+100, x+200, before I got to rating x. I'm not saying that you should do 70 problems of rating 1400, that's not it. Your rating is a reflection of a lot of things, such as speed, debugging ability, knowledge of some very specific ideas that are common on Codeforces. Sometimes you got better but it just doesn't reflect a lot on you rating at the moment. I believe that you can reach cyan if you keep going.
Just do you. Forget all those advices (including mine, I guess) and do what you love and want to do. Try different things, see what works for you and what doesn't. Forget all those metrics people love to brag about (including doing x problems of y rating) because humanity is really good at optimizing the metric, but that doesn't mean you're optimizing your training. Everytime I tried to set some goals and metrics I spent more time trying to optimize the metric than really doing what I wanted to do.
If you are on your own solving harder problems (1500+) at least sometimes (20%+), just not in time or for whatever in contest, i think you're on the right track and will see contest improvement in the next 1-2 months. Keep doing harder problems and also start doing some more vc's for div3/div4 and atcoder abc. However i'm assuming if you're stuck this is not what's happening, and in that case...
I would recommend doing 1-2 chapters of usaco training gate without looking at editorial ever or extremely minimally, then come back to work on cf.
As in my practice guide blog, I believe cf is best for training once you have some level of background, but what i did not lay out in much detail is that too low rated problems (< 1500) often don't teach much, and you need to build background on longer impl, standard ideas, and forcing yourself to think longer first. This is why i recommend training gate to complete beginners (or people stuck in low rating).
If you're still stuck (I wouldn't think so), I'd probably recommend studying for the AMC math test a bit...
And on cf it only looks like you have 5 months of practice. If you don't have any coding background maybe that's normal, but otherwise you do need to be doing something quite different than now. If you don't want to do things above, at least consider things like if you're thinking hard enough per problem before giving up, need to do more vc's, try harder problems where look at editorial more, etc.
only 24 *1600 problems and 180 *800 problems! Also, dont forget about upsolving