I've found that writing todo lists enormously increases my productivity while programming. Programming is all about breaking larger problems into smaller problems over and over until the problems are small enough to solve easily. Lists provide a map of the problems that need solving and make it possible to work on a smaller subset of them without having to simultaneously worry about the global structure of the program.
I'd like to learn more about creating computer programs able to play board games; to that end, over the next day or two I'm going to be working on an AI for 8×8 Connect 4 and try to accomplish the following:
Add alpha-beta punning to the Minimax algorithm.
Create a Monte-Carlo Tree Search algorithm for Connect 4. I'm not currently very familiar with this technique, but based on what I've read so far, it looks like it is state of the art.
Have the programs play each other (with similar run times) and post the results.
Other assorted ideas I'm thinking of trying (updated occasionally):
export google location log as a KML, plot a connected scatter plot of lat dist v long dist
Bump chart with 5 people?
2016 in everyone's graphics
scrape everyone's graphics twitter accounts and find uniq url
look at election graphics
2016 chart diary
Take reddit notes and throw in one post
play by play
Most boring playoffs ever, how many minutes was each champion behind?
link to reddit comments and nlp. which player is the most exciting?
During a rare happy stretch of Philly's blowout loss to the Hawks on Friday, the Sixers' broadcast team pointed out that Philly was on "an 8-3 run." I know the Sixers aren't good enough to put together normal NBA runs, but an 8-3 scoring gap is not a run. It is a randomly occurring blip that happens every game.
Which players are the streakest/have the most improbable runs? www.espn.com/nba/story/_/page/presents-19573519/heating-fire-klay-thompson-truth-hot-hand-nba
Which players do the best/worst after hiting/missing a FT? Do some players always go 7/10 or do some go 5/10 and 10/10 more often than they would randomly.