I guess the theme for today is games.
How to lose your faith in humanity: read the comments on any YouTube video about a game.
Anyway, I had heard that Google had teamed up with Udacity to host a MOOC on Android development. Except the first time I checked it out, I couldn’t see past Udacity’s paid guided version to see the free-as-advertised version. There’s also a course on UX design, which may be useful when we design Beast Academy online (may be too late for AoPS version 3). There look like fun things to try out when I have time.
So thoughts naturally ran to what games I could make for a project. Finger Ball would be easy to implement: the rules are simple, and I’ve already worked out the AI (though adding in an adaptive AI using n-grams might be cool).
However, my quidditch design (the one where each house plays like its element: Gryffindor = Fire, powerful but only in bursts; Hufflepuff = Earth, slow and steady wins the race; Ravenclaw = Air, quick and mobile; Slytherin = Water, fluidity means tough to pin down but can still hit hard) is the design that’s worrying me gameplay being dragged down by too much extra junk, so maybe having a computer handle some of the junk will help.
Then I thought about my command & control game, but the current iteration uses cards to mimic command & control problems by preventing players from ordering any unit they want. It would be gimmicky to port that to the computer, so I’ll leave that as a boardgame.
Instead, what a computer can do that a boardgame can’t do easily is hidden information and uncertainty. For example, a wargame where the player’s orders aren’t put into effect immediately, the player is uncertain how long it will be before they take hold (or how the situation will change in the mean time), or even where exactly units are at the moment. Dean Essig has made a slew of games simulating similar command & control issues.
Meanwhile, Calculords has come out on Android. It’s probably the nerdiest game on my phone right now. I wrote a Python program to help me score Calculord bonuses while deploying the maximum number of cards, which I’m pretty sure was not the way the game was meant to be played. Luckily, my program only really works for up to 6 cards, 7 if I’m willing to wait minutes, 8 cards is impractical, so I have been doing some of my own thinking.
In non-game news, I fixed my disposal. Internet to the rescue!