My wife and I played about 20 rounds in a row last night. She wasn't a fan of being locked out of an answer if the other player says it first, because it gives a big first-mover advantage, but I think it's the right design choice given that all answers are shown on the same screen. Another nice choice you made that I might not have thought of is allowing any given word to be used only once per game, so you can't just answer "man" or whatever to every prompt. I was surprised at how many prompts repeated over the course of 20 matches. Given the size of the data set, I'd have expected far fewer repeats.
Thanks for the link to the data source. As a puzzle- and game-maker, I think I'll find a lot of interesting uses...
Really love the idea and loved playing solo. Someday, when friends can gather close enough again, I look forward to trying it at a party. How did you generate the data set? I understand it's gathered from a large corpus. Did you write code to generate the set, or did you acquire the data from elsewhere?
I see this was released in 2017. I hope you haven't set this aside. The game has a lot of potential for mobile, with either live synchronous or asynchronous play, and this would make a great game in the Jackbox style.
