On Tue, Mar 10, 2020 at 11:19 AM Bartholomew Furrow <[email protected]>
wrote:

> I think the major difference between the previous rule and the new rule is
>> In the previous rule, the hidden test for a submission will be run only
>> when that submission is the LAST submission and it passed all the visible
>> test.
>> In the new rule, the hidden test for a submission will be run only when
>> that submission is the LAST submission that passed all visible test.
>>
>
> Thanks for refreshing my memory, and correcting my examples. I believe
> you're absolutely right.
>

If we are being pedantic, using always the last submission was 2018's rule.
2019's rule used the last submission that passed the samples (samples
weren't used in 2018, so that rule wouldn't even have made sense then)
since the #1 complaint was accidental submissions ruining your score, and
accidental submissions would usually fail the samples. 2020's rule is even
more forgiving, letting you try the hidden on the last submission that
passed all the visibles (i.e., the last one that even has a chance). This
addresses the bit of frustration left-over from the 2019 version AND also
handles interactive problems that didn't benefit from the samples addition.

If anyone is interested, the short version of the rationale is: we wanted
to keep a single try at hiddens because it is part of the hidden identity.
However, if you submit something that passes all visibles and then you
submit more but never pass all visibles, you are just encouraged to
resubmit your last submission that passes all visible cases near the end of
the contest. We wanted to save you from having to do that non-interesting
step and us from having to judge it. The additional impact of improving
your penalty compared to an actual resubmission is just a side effect to
keep the rule as simple as possible.

Best,
Pablo

-- 
You received this message because you are subscribed to the Google Groups 
"Google Code Jam" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to [email protected].
To view this discussion on the web visit 
https://groups.google.com/d/msgid/google-code/CANa5jcCR%2Bi%3DpXqz7RFYV4VZWkM_a7SfNV%3DRKSUBO7YeDY-8kOg%40mail.gmail.com.

Reply via email to