Story Time

It's been about a year since I started CP and reached Expert pretty fast. However, for some reason or the other, I couldn’t reach CM. I got pretty close multiple times but lost rating as soon as I was one contest away. It felt like I was collecting every 18xx rating lmao.

At least in the beginning, the reason I wasn’t CM was, well, I genuinely wasn’t at that level yet. However, towards the end of the year, I felt like I was of that strength, but for some reason couldn’t convert that into the rating jump I needed. I think some of you would be able to relate to this — you feel stronger than your current rating, but can’t reach there, and maybe there’s something holding you back psychologically.

Beating Psychological Pressure During a Contest

I tried to do something different in the last 2 contests I wrote to deal with pressure. Firstly, I stopped checking my standings — both global and friends’ standings. I used to check these very regularly in previous contests. My brain would go like this:

Ughhh can’t get any idea about D, let’s see the solve count. Oh wow, it’s pretty high, it must be something obvious I’m missing. Let’s see which of my friends have solved it. Oh, one guy solved it, interesting. Let me continue thinking… (15 minutes later) Ok I give up, but the solve count is even more now. Wait, this friend also solved it even though she’s rated below me, umm what am I missing :( Ok let’s check using Carrot how much rating I’m losing as of now… crap I really need to solve D to save my rating

… Why, just why? Clearly this is not useful.

Firstly, in order to solve a hard problem, you need calm thinking without distractions. I’m not saying you shouldn’t take breaks — sometimes taking a walk for 5 minutes can actually help refresh your mind. But seeing who all solved the question was just creating more pressure for me.

Secondly, with the state of AI and LLMs today, the solve count might be very inflated. In the near future (if not already), solve counts during contests might just indicate how easy the problem is for AI, and not its true difficulty for humans.

I went one step further and literally covered the solve count with my right hand during the contest whenever I returned to the problems page, just so I wouldn’t get spooked by it (yes, I know I could’ve just used a mirror website, but I’m not as comfortable with the interface lol).

“But aren’t standings and solve counts useful to see how well you’re doing, or whether a later problem is actually easier?” Sure, C vs B and E vs D difficulty swaps are common enough. But why do these exist in the first place? Difficulty is relative.

In fact, for Good Bye 2025, I got the idea for E instantly, but it took me a long time to get the idea for D, even though E was harder for most people. I think it’s better to just read the next problem if you’re stuck for a while, and developing this habit could be useful in general — especially in contests like ICPC, where problems aren’t arranged by difficulty.

Recognizing difficulty is a skill in itself, and I would argue it’s best to look at problems without any bias. Maybe you get the idea for a “harder” problem quickly, and you wouldn’t have tried it with your best effort if you’d seen the solve count.

If you aren’t psyched by how others are doing, and are actually motivated by it during a contest, then yeah, this advice probably isn’t relevant for you. But I think a lot of people would relate to this sentiment, regardless of rating.

More Story Time

I had a goal of reaching CM by the end of the year. However, I was pretty far away before Good Bye 2025 — not even in the 1800s — so I pretty much gave up on it after getting close so many times.

But after 2 great contests without the pressure of reaching CM, I thought I could actually get there. My rating was 1891 after the last contest of the year, so I thought, “oh well, didn’t actually get there, but I found a way to remove pressure — nice.”

But after the rollback today, I reached a rating of exactly 1900 :D (Did I hit my goal or not is up to technicality, but I count it — it feels so clutch)

It also feels like a belated birthday present hehe, since mine was yesterday.

Fun fact: my ICPC team name is literally “cantreachCM”, since all 3 of us had a peak rating of around 1850 while registering. We had our regionals on 3rd Jan, so we were joking that it’s actually bad luck to reach CM before that... and I thought I didn’t lol (or did I technically? idk).

Some Final Pep Talk

I actually think I’m still not at the level to be a stable CM, but that’s okay — I’m going to keep practicing and getting better anyway. All of us improve at different rates, whether due to different exposure to DSA, coding, math olympiads, etc. It’s best not to compare and instead focus on yourself.

Learn from the community and help each other improve, yes — but understand that there’s no deadline to getting good, and it’s okay if your rating doesn’t improve, or even goes down.

Let’s not lie — it’s especially hard to increase rating now with the number of cheaters. Even without cheating, rating might not reflect your true skill. It’s just a metric of how well you do in rated contests compared to other humans (and a few robots).

With enough time and effort, you will get better at CP, regardless of whether your rating reflects that. Trust the process, and have fun solving problems and learning new things. That’s enough :)