search

nus-cs2113-AY2324S2 / pe-dev-response

0 stars 0 forks source link

Inaccurate problems per minute #2432

Open nus-se-script opened 6 months ago

nus-se-script commented 6 months ago

I answered 0 questions before exiting, but the program still marks down my problems per minute as 300. It could be good to only give a problems per minute metric if any problems were solved. Screen Shot 2024-04-19 at 4.16.42 PM.png Screen Shot 2024-04-19 at 4.17.18 PM.png

[original: nus-cs2113-AY2324S2/pe-interim#2344] [original labels: severity.Low type.FunctionalityBug]

Geinzit commented 6 months ago

Team's Response

The program specifically states that all tests that are not finished will be marked as wrong, which means the test will still be viewed as finished and your time spent will be calculated as 'from the time you start until the time you hit exit', and your problem count will be counted as the number of the problem set. So there is no inaccuracy to the speed calculaton.

But it does make sense that users may feel weird that a problem set they didn't do or finish have an unreasonably high speed. So a better solution may be to set the 'problem count' mentioned above as the number of problems the user solved before they hit quit, and then calculate the respective accuracy as well as speed based on that.

Thanks for bringing this up!

Duplicate status (if any):

--