If someone believed a card to be a win condition, but that card was not listed as a win condition by RoyaleAPI, how could they prove it to be so?
I believe Clone and Mirror to be the cards that are most win condition-ey of all cards that are not commonly accepted by the community as a win condition.
If a deck was able to reach top ladder with Mirror and or Clone, and did not contain any cards that are accepted as win conditions nowadays, would that be enough to call one of them a win condition?
I think most people would agree that if a super intelligent AI learned to play Clash Royale, it would probably use a deck that has not been used by a human before (since it could utilize certain interactions and cards that cannot consistently replicated by a human). I’m not sure why I think this (it very possibly is my own biases coming into play), but I think it is very possible the deck it would play would look nothing like a normal deck, and may even not contain any card that we consider to be a win condition.
This brings me to my final question. If a player is able to reach #1 with a deck, and they are the only person who can use the deck with success, can that deck be considered viable?
I know this is just a random assortment of questions, but I was just curious what you have to say about this topic.
This is an interesting question. I have long had issues with the concept of “Win Condition” to be applied to Clash Royale. I think that @Gerry would have a lot to say about this topic.
Win Condition is traditionally defined as the card that win games, and for CR, many people define it as a card that can hit the tower. Personally, at least for me, I prefer the term Key Cards which is the cards that define a deck. Example: what differs a Lava Loon deck from a Lava Clone deck — obviously Clone is the key card. So your example about Mirror and Clone are very valid. But should we redefine these as Win Conditions? Probably not.
For a long time, we have left Pekka as part of the list of Win Conditions because what we intended to say is Key Cards. Interestingly, that’s the exact words used by some other languages (e.g. Chinese). They define Win Conditions as “Core Cards” of a deck. The concept of Win Conditions in fact does not exist in those other languages.
We have actually done some machine learning analysis to see if we can find these key cards algorithmically so that we ensure that there are no human preconceived notions of what make decks unique. This investigation has been ongoing for a while. I work on it with some data scientists when time allows. Results have not been overly conclusive and so we haven’t published anything about it.
Very interesting. I love the topic of some words not translating perfectly to different languages.
This also assumes that an intelligent AI would play as we would. It is entirely possible it would be able to exploit a certain glitch to never lose, use Arbitrary Code Execution, or any of the other possible ways to win a match in a non-traditional method. I guess there is no point even discussing these methods of winning though since it is not something we could analyze or predict.
You mentioned intelligent AI but our machine learning has nothing to do with game play — we were using ML to analyze decks. It’s also not really possible to analyze game play to group decks together because of how the data is structured atm (as well as availability of data).