This is an Advent of Code leaderboard. It does not compare the times at which participants have submitted but instead it compares the speed of the solutions. This way you can see how long it took other people to solve a problem and the goal is to solve all problems within 1 second.
Since Advent of Code is a global competition which starts at the same time for everyone it is difficult for some contestants to start working on the problems as soon as they are released. This gives them a major disadvantage in the usual leaderboards. Additionally, this year there have been many instances of LLMs getting on the global leaderboard making it less interesting to compete for speed.
Instead this website strives to motivate participants to make their programs as efficiently as possible. In the past this has lead to interesting optimization efforts including rewriting parts of solutions in assembly. We also believe that this is a really good way to get familiar with different profiling and optimization methods for the programming language of your choice.
You will notice, that some people will have golden or green usernames on the global leaderboard. These colors symbolize different achievements. If you have solved all problems published so far, then you will have a green name tag. If you also managed to do this with a runtime of less than 1s your name will appear golden.
A common goal for participants of Advent of Code is to optimize your solutions to run as fast as possible. To quantify this challenge participants want their solutions to take less than 1s combined over all days and parts. This is by no means an easy challenge, especially the later days often require a lot of time and effort to solve them efficiently. To keep track of your and your friends progress, this website provides the runtimes entered so far.
For user authentication this website uses GitHub OAuth. It would have been too much effort to create a dedicated user management system for this little platform and almost every participant already has a GitHub account.
The runtimes will be measured locally by you. There is no reference system so feel free to run it on the fastest computer you have access to (maybe even a super computer cluster). We are aware that this might lead to some unfairness but in our experience algorithmic and language specific optimizations are way more important than hardware differences. Also note, that this leaderboard is not intended to be ultimately competitive. This is unfortunately not possible without building a measuring and judging system which benchmarks the code.
Of course, a major disadvantage is that people can cheat. Currently, we do not have the time and capabilities to build and host judging servers which execute your code and measure the runtime. Furthermore, it would restrict the flexibility in the choice of language that people currently have (like even Minecraft or Excel).
Additionally, it worked really well in the past as Advent of Code contestants have been trustworthy and there is no gain in entering incorrect values. Nonetheless, if something seems out of the ordinary we’ll investigate and check if the runtime is correct.
This is not the first year this leaderboard is up. It has already been up in 2022 where many students from ETH Zürich enjoyed it. It has been down last year due to other circumstances and has now been completely rewritten and renewed for 2024.
Most of the work for this website has been me, TecTrixer, with the support of my beloved friend, ChatGPT. That is also why the website uses “we” instead of “I” in most descriptions. Regarding myself: I am a computer science student at ETH Zürich who loves participating in Advent of Code and engaging with the community.