mirror of https://github.com/dirtbags/moth.git
74 lines
2.6 KiB
Markdown
74 lines
2.6 KiB
Markdown
|
Scoring
|
||
|
|
=======
|
||
|
|
|
||
|
|
MOTH does not carry any notion of who is winning: we consider this a user
|
||
|
|
interface issue. The server merely provides a timestamped log of point awards.
|
||
|
|
|
||
|
|
The bundled scoreboard provides one way to interpret the scores: this is the
|
||
|
|
main algorithm we use at Cyber Fire events. We use other views of the scoreboard
|
||
|
|
in other contexts, though! Here are some ideas:
|
||
|
|
|
||
|
|
|
||
|
|
Percentage of Each Category
|
||
|
|
---------------------
|
||
|
|
|
||
|
|
This is implemented in the scoreboard distributed with MOTH, and is how our
|
||
|
|
primary score calculation at Cyber Fire.
|
||
|
|
|
||
|
|
For each category:
|
||
|
|
|
||
|
|
* Divide the team's score in this category by the highest score in this category
|
||
|
|
* Add that to the team's overall score
|
||
|
|
|
||
|
|
This means the highest theoretical score in any event is the number of open
|
||
|
|
categories.
|
||
|
|
|
||
|
|
This algorithm means that point values only matter relative to other point
|
||
|
|
values within that category. A category with 5 total points is worth the same as
|
||
|
|
a category with 5000 total points, and a 2 point puzzle in the first category is
|
||
|
|
worth as much as a 2000 point puzzle in the second.
|
||
|
|
|
||
|
|
One interesting effect here is that a team solving a previously-unsolved puzzle
|
||
|
|
will reduce everybody else's ranking in that category, because it increases the
|
||
|
|
divisor for calculating that category's score.
|
||
|
|
|
||
|
|
Cyber Fire used to not display overall score: we would only show each team's
|
||
|
|
relative ranking per category. We may go back to this at some point!
|
||
|
|
|
||
|
|
|
||
|
|
Category Completion
|
||
|
|
----------------
|
||
|
|
|
||
|
|
Cyber Fire also has a scoreboard called the "class" scoreboard, which lists each
|
||
|
|
team, and which puzzles they have completed. This provides instructors with a
|
||
|
|
graphical overview of how people are progressing through content. We can provide
|
||
|
|
assistance to the general group when we see that a large number of teams are
|
||
|
|
stuck on a particular puzzle, and we can provide individual assistance if we see
|
||
|
|
that someone isn't keeping up with the class.
|
||
|
|
|
||
|
|
|
||
|
|
Monarch Of The Hill
|
||
|
|
----------------
|
||
|
|
|
||
|
|
You could also implement a "winner takes all" approach: any team with the
|
||
|
|
maximum number of points in a category gets 1 point, and all other teams get 0.
|
||
|
|
|
||
|
|
|
||
|
|
Time Bonuses
|
||
|
|
-----------
|
||
|
|
|
||
|
|
If you wanted to provide extra points to whichever team solves a puzzle first,
|
||
|
|
this is possible with the log. You could either boost a puzzle's point value or
|
||
|
|
decay it; either by timestamp, or by how many teams had solved it prior.
|
||
|
|
|
||
|
|
|
||
|
|
Bonkers Scoring
|
||
|
|
-------------
|
||
|
|
|
||
|
|
Other zany options exist:
|
||
|
|
|
||
|
|
* The first team to solve a puzzle with point value divisible by 7 gets double
|
||
|
|
points.
|
||
|
|
* [Tokens](tokens.md) with negative point values could be introduced, allowing
|
||
|
|
teams to manipulate other teams' scores, if they know the team ID.
|