Advanced score modding

Case number:845813-2013170
Topic:Game: Tools
Opened by:nspc
Opened on:Saturday, May 21, 2022 - 16:39
Last modified:Wednesday, May 25, 2022 - 20:02

In binder puzzles, recipes like "wiggle and mutate" or "band", use score to compare solutions it found.

Base score and objectives have a lot of influence in what a recipe found.
Sometime, depending of the current step of the puzzle round, we want to maximise a specific score or objective.

For exemple, we can try to maximise CS and have no BUNS at begin, and optimise shape complementarity only when main BUNS are resolved.

The idea is to have an interactive and easy control to influence what the recipe should found.
By "easy", I mean, not by modify script, but with a simple UI that every player can use.

This control can be a "score modding", I mean GetScore() called by the recipe is a false score that the user can specifiy :
-Add a score multiplier in BUNS bonus, CS, or other objectives (all objectives if possible).
-Add score multiplier to some sub scores like "Packing".
-"Voids" malus importance

This new UI can be a new separated windows with sliders.

Off course recipes will give a score that is not the real max at end, but it is what we want.

Hand folding stay an important step to find main AA. But recipes can give a real optimisation.

For now, recipes often destroy the current good hand folding by adding BUNS for exemple.
The new BUNS score (with less malus) can be usefull to find good CS, but it always add BUNS, and we realy need to control it.

Scientists give a score system, and change it sometimes, to help players find specific kind of solutions.

But it can be very interesting to let players change some score importances in different steps of the round,
and to experiment new methods to resolve each step.

It's an idea that I had mentioned in last office hour, but I wanted to give more precise details.

(Sat, 05/21/2022 - 16:39  |  1 comment)

Joined: 09/24/2012
Groups: Go Science

Some recipes try to make something like that, but you have to know the options. It's indeed a very good idea to have it as a generic tool.


Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, Boehringer Ingelheim, RosettaCommons