Wrong leaderboard ordering and `N/a` scores discussion
Nobody has claimed this yet.
- Dominant language
- Python
- Stars
- 176
- Forks
- 74
- Avg merge
- 5d 3h
- Merged PRs (30d)
- 21
Description
1. Wrong leaderboard ordering
Sometimes the ranking on the leaderboard does not follow the order from the primary column. It happens several times and we can hardly reproduce it. Once a competition starts to have a scrambled ordered, it stays this way, but the problem does not appear again when re-uploading the same bundle and submissions.
Issues:
- #1284
- N/a scores still ranked first: #2321
- #1171 <-- P1
- Another example below in comments
2. N/a scores
We have several issues that report N/a scores on the leaderboard. N/a is a replacement of score shown for a task not found. The problem is that the scores and tasks are separate and are sent to the frontend where they are matched and then scores are shown for a task or several tasks. This opens room for inconsistencies.
I propose to make the leaderboard on the backend and just send it to be displayed without any extraction or computation. This way we will never have n/a scores.
-
a) N/a when submission rule is "Force Last"
-
#1055
-
#1275
-
b) N/a for multi-task when submission rule is "Force Last"
-
#2211
-
#1174
-
#1570
-
c) N/a score after auto-migration (#2239)
-
#1472
-
#1174
3. When multiple tasks, how should we order the entries?
Right now it seems that the ordering is based on a sum of scores across tasks.
4. Request: separate leaderboards for tasks
- #1392
Contributor guide
First steps
- Read the whole issue, then the project's contributing guide.
- Comment on the issue to say you are picking it up — it saves two people doing the same work.
- Fork the repository and make your change on a branch.
- Open a pull request that references the issue number.
Research direction
No files, tests, or entry points are named. Start by tracing the leaderboard ordering and score/task matching described across the backend and frontend, then review the linked issues for existing decisions. Done would require an agreed scope that resolves the ordering and N/a-score behavior without leaving the separate-task ordering questions open.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- backend, frontend
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Stale
- Clarity
- Needs clarification
- Newbie friendliness
- 25/100