Its good to bring light for this issue, but not much the average person can do about it unfortunately.

I mean if your goal was 1200, rating deflation puts your goal at <= 1400 which is still doable.

So much aura

How did you got to plastic, i didnt make it out of usaco paper

I think "10-15% of people in contests are cheaters" is probably lowballing it as its hard to tell if someone asked AI for the idea and implement it themselves. No way to catch such cheaters reliably. I agree with your main points.

Should have used a bunch of em dashes to trick us

A similar alternative solution is to see that winning moves will remove as many inversions as possible from the original string, which can be accomplished by removing all 1s or all 0s.

For example, consider removing all the 1s from the original string. If there exists $$$s[i]=0$$$ such that there are an odd number of preceeding ones, we can remove all ones from the string immediately by selecting this 0 and all other ones. Similar logic holds for 0, but using suffix instead of prefix.

If all prefixes/suffixes have even sum, Alice will be forced to make a move that causes either count of ones or count of zeroes to be odd, and Bob will use the same strategy to win.

void oneMoreProblem() {
    int n; cin >> n; string s; cin >> s;
    
    int onesseen = 0, zeroesseen = 0;
    for (int i=0; i<n; i++) {
        if (s[i] == '0' and (onesseen&1)) {cout << "Alice\n"; return;}
        if (s[i] == '1') onesseen++;
        
        if (s[n-i-1] == '1' and (zeroesseen&1)) {cout << "Alice\n"; return;}
        if (s[n-i-1] == '0') zeroesseen++;
    }
    
    cout << "Bob\n";
}

https://usaco.guide/gold/

Look at the graph and tree sections

Finally, if I somehow manage to escape that, I reach the "Super Ultimate Best Coder Ever My Parents Love Me Finally My Life is Perfect and Nothing Ever Goes Wrong Ever and I Have Never Questioned My Life Choices" tier....and maybe learn to implement a greedy algorithm or two.

Nice order :skull:

Large language models are the same shift, just faster and more visible. The people panicking about AI in Div. 2 rounds are, ninety percent of the time, panicking about something that was already possible with a well-organized template folder and a search engine. The difference isn't capability. It's that watching someone paste a prompt feels more like cheating than watching someone paste from their own notes, even when the effect on the scoreboard is identical.

This is just completely wrong. AI can think and create an entire solution. It is NOT like a well-organized template folder. Literally anyone, even people with no coding experience, can get CM+ now with GPT, the only difference is that more experienced people are less likely to get caught.

True, but it should be made clear that any cheating is bad. Partial cheating is still cheating. But good on OP for confessing.

Nice extension, when is DopamineDetoxForces coming out

Edit: The extension is pretty robust and covered a lot of ways to prevent you from seeing your rating. One way you can add is maybe removing the rating standings element from all https://codeforces.me/ratings/* URLS, as you can technically look up your rating from here. But it is sort of convoluted, so maybe its not worth the effort.

The old advice isn't completely outdated, just use the old "advice to hit expert" and hope that you become specialist lol

W reference

Everyone is equally cooked, AIs can solve problems from every division. (XVIII) Although if you make it to the top of the standings, you are maybe a little safer, as (I think) admins prioritize the top of the leaderboard first.

Nice blog but you should clean up the formatting a bit, its a little hard to follow

In a similar boat lol. Try to really piece apart what goes wrong during contests and or vcs. Do you usually get the idea but can't finish the impl? Do you struggle to come up with the idea at all? If its the former, you need to practice impl on mid level problems, and if its the latter then you have to spend time thinking about hard problems. If what you are doing isnt working, you have to change what you are doing, i think. Obviously not in the best position to give advice here but most of what i said is common sense.

Also keep in mind you have likely improved, but the cheating is bad, so gauge your improvement via past virtuals or what problems difficulty you have solved. Last summer i used to spend 2+ hours on 1400s but i can solve old ones in 20-45 mins. My rating is lower but it is what it is.

Great, another AI doomer post. Thanks DNR! I found about this tool recently that blocks posts with a specific keyword, I'm going to use that.

It actually seems pretty accurate for me, I tried it out.

Edit: I will add that the topic competency is inflated if a user has solved more problems, which is to be expected, but solving more problems will not always fix a plateau.

https://codeforces.me/blog/entry/154620

Since the rating deflation is so high, if you are able to maintain mid-expert, you might be CM skill. If you want to be CM, you probably need master skill nowadays tbh

That was cool to look at, definitely legit. I was wondering, do you use scratchpaper while solving or just do most work in ur head

Edit: just saw u said u used scratchpaper, nvm. If you do want to show ur work theres ways to cast an ipad to OBS

I think u misunderstand, the point is there is SOME move that will let alice win. If alice plays 14522 she will win, since bob has 00004 and cant make a valid move.

The first player only wins if they make an optimal choice. I think goal is to make the first player make it so that on the next move, the second player will lose.

Two alternative ways to do it, the first one I think is the editorial's intended

 

Happening to me rn as well while trying to sign up for 1106. I think if you sign up later it might work

Thats cool, would not have expected such an approach to work

I imagine that some people tried an increasing greedy instead of a decreasing greedy for C. This does work but the solution is less clean. The idea is still to binary search on a valid $$$m$$$. We check if the first $$$m-1$$$ elements can be filled. For a given $$$m$$$, we can do the following:

Create a $$$cnt$$$ array, and maintain an unresolved left bound $$$L$$$ and right bound to pick elements $$$R$$$. Iterate $$$L$$$ from $$$0$$$ to $$$m-1$$$. If $$$cnt[L] \geq 1$$$, then the element already contributes to the mex. Otherwise, we need to find a location $$$R$$$ to take the next element from. Increment $$$R$$$ until:

1. $$$R \geq 2L + 1$$$ (mod condition)
2. If $$$R \leq m-1$$$, we need at least one element in the current location: $$$cnt[R] \gt 1$$$.
3. If $$$R \gt m$$$, we can use any nonzero element.

Such a two pointer approach works in $$$O(n\text{ log }n)$$$.

Submission: 372875264

As not a tester, this will be a good contest

Running the Carrot rating extension recently seemed to start failing, this could be related

Edit: it looks like CLIST stopped working too

Im pretty sure its because there are untrusted participants above you for the true standings, but clicking the standings page gives you the standings for the trusted participants (it should say "[trusted participants only]")

Great job, congrats!!