PineapplesOnPizza's blog

By PineapplesOnPizza, 6 weeks ago,

TL;DR I complain about the popular contest in my region last year, which will be held again this year, that had so many mistakes.

As Codegoda — a programming contest in my region — is about to begin this year, I recalled the bad memory I had with Codegoda 2021. So this post is gonna be me complaining and telling the story about one of the worst, if not the worst, contest I've ever participated.

First, you can look at the problem set and editorial here. I'll summarize the issue about some problems below. In this section I'll not focus about the problem idea quality.

Problem 2. Origami

In this problem, the statement didn't tell the exact definition of the word "fold", and told you to look at the example case, which I considered bad. The other issue was, they didn't even proof-read the example input. If you look at the input format, you will see that line 2 (The first line of the first test case) should be integer n and m, but the sample input 3 looked like this.

1
- -1 2
- 1 2
5 6


I know that the issue wasn't that big, because you can assumed what they meant from the output explanation, but this really showed how much attention they paid for the contest (which is none).

Problem 3. Ransomware

At the beginning of the contest, the problem's text wasn't loaded correctly. It took quite a time until it can be read.

Problem 4. Codegoda City

This was the problem I hate the most. If you look at the input format specifier, you can see that it stated that

Next N lines: city number, colon, then city numbers separated by commas.

Ex. 2:3, 4 means that City 2 is connected to City 3 and City 4.


The example was contrast to the statement. The statement specify that the city will be separated by commas, but in the example it was separated by comma and space. I didn't see it and implement the wrong parser. Although it may partially be my fault that my parser wasn't general enough, the specifier should still be written clearer.

But that wasn't the biggest issue in this problem. In this problem, the biggest issue was that the problem didn't specified the label of nodes(Like $1$ to $N$ or $0$ to $N - 1$). So, as I look at the sample output, I thought it's $1$ to $N$, but the test case in the grader was $0$ to $N - 1$. I only knew it when I changed and got AC.

Edit. My friends also point out that there was no constraint and the statement is ambiguous.

Spoiler

Problem 7. Agojiberries

This problem was the just a little modification of problem L in atcoder dp contest. Nothing more or less.

Problem 8. Agoji Print

For this problem, the problem setter decided not to write the number in latex, and it was more difficult to read than the other problem.

My own opinion

In term of problems idea quality, the contest doesn't stand out, there is only one interesting problem, and the rest is just standard. But I don't mind that much since it's more like hiring contest than competitive programming contest. The thing I felt unpleasant the most is that, there are problems that the statement is OK, but some shouldn't be included in any contest. That meant some problem setter knew how to write a good problem statement, but chose to ignore some unqualified problems. That make me feel like the company that held this contest just held it to PR their company. They didn't even wish to make the contest good. That make me feel really irritated.

Finally, I wish that my blog will reach them and this year contest will be better even though I won't take part since I'll travel that time and I still reserve the hotel through their website.

• +53

 » 6 weeks ago, # |   0 Sad to know