Deductions and endgame#11
Open
felixbauckholt wants to merge 2 commits into
Open
Conversation
This just tweaks the "risky plays" part of `decide_wrapped()` to take much more risk once the deck ran out. Also, once the deck runs out, if there is exactly one card that might be playable, we can deduce that the card is playable and play it. I just hardcoded this (but we can take out that chuck of code and only incur a tiny penalty).
… other places This is currently somewhat hacky, but I don't see a way to make it less hacky.
Owner
|
these changes look good. the extra boolean is awkward but not too bad - why would card ID help? in general though, I'm starting to get wary of making the strategy too complicated for small gains. would be interested in brainstorming ways we might get more big gains before we local-optimize too much! (i don't think it's too much a risk with these changes though!) |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
6720f33 makes some mild improvements by implementing some straightforward endgame optimizations.
e76722e marks down weights of cards in other places once they've become determined in one place. This requires us to track whether we already did this for each card; this is somewhat ugly in the current architecture, sorry! Maybe if we do expose some sort of "card ID" for each card to the strategy, this could look nicer.
(Sorry for the lack of detail here, the commits are almost a week old so I forgot about them.)