I have an idea for a tarot card mechanic based on a discord bot I know. This bot looks at all the chat messages in a server and tries to create sentences by putting words together. My tarot concept would work similarly but with cards instead of words. When you use this tarot, it would convert any 5 cards in your hand into the poker hand you play most often. I thought about making it give random cards instead, which might work better with certain jokers that add cards to your deck. But I want to stick with the original bot concept of analyzing patterns and creating something from them. What do you think about this idea?
honestly this sounds pretty cool but what happens if someone hasnt played enough hands yet? like early game when theres no real pattern to work with. maybe it could default to straights or something until it learns your style? also wondering if this would make the game too predictable once it locks onto your favorite hand type.
The pattern analysis angle is fascinating and captures that machine learning feel well. My main thought is whether five cards might be too restrictive for the conversion - some poker hands need specific card relationships that might not always be achievable from random draws. Perhaps the tarot could have a fallback system where it attempts your most frequent hand first, but if the five cards can’t reasonably form that pattern, it moves to your second most common choice. This would prevent situations where you get completely unusable conversions while still maintaining the predictive element. Another consideration is how this interacts with deck manipulation strategies - players might try to game the system by deliberately playing certain hands early to influence the tarot’s future behavior.
This mechanic has interesting strategic depth but might face some implementation challenges. The pattern recognition aspect is clever, though tracking poker hand frequency could get complicated if you’re switching between different strategies mid-game. One concern is that players who experiment with varied hands early on might get stuck with suboptimal conversions later when the tarot analyzes their overall play history. You might want to consider having it analyze only recent hands rather than all-time patterns, maybe the last 10-15 poker hands played. This would make the tarot more responsive to current strategy shifts while still maintaining that adaptive AI-inspired element you’re going for. The mechanic could also become more powerful as games progress and more data gets collected.