Here is a selection of some of my current and past projects.
Heretical Book Club
I read and study about a lot of different subjects, but when I sit down to write or produce some of my own thoughts, I usually go down a rabbit hole and find that a) My thoughts aren't original, b) someone has already countered them and now I'm questioning my own thoughts, then c) well this entire subject goes a lot deeper than I thought, and writing anything would just expose my ignorance.
However, I came up with an idea to get around that. I would read books and books from a layman's perspective, and then review them on video in a semi-scripted, mildly edited format and upload them to YouTube. If the channel grows, then I can get feedback and do follow-up with viewers, and it can be a collaborative learning experience.
And that's how I started my channel, the Heretical Book Club. Currently on slight hiatus until around mid-May, due to other things consuming my time, but I already have my next few books lined up. I don't expect this to be a big project, just a fun, niche learning experiment.
Fantasy Baseball Tools
I've been playing Fantasy Baseball for about 15 years and studying baseball longer still. With my experience, knowledge, and programming skill I've devised a number of unique strategies, tools, and perspectives that consistently let me finish in the money year in and year out. I'm currently working on refining and productizing some of these tools to be released within this baseball season.
YouTube Content Creation
I thought it would be a fun challenge to see how I could do at becoming a content creator in my free time. I could learn about marketing and video editing, as well as use my problem-solving skills to rise above hundreds of others in a competitive market.
Since I was tackling two new skills (video editing and marketing), I thought I'd make things easier on myself and choose a subject I was confident in. My goal was to become one of the top content creators for a new (at the time) strategy PC game called Gwent. I was able to become a top 300 player within a couple of months of picking up the game, and started making content soon after.
In just over half a year, I went from initial concept to having 300+ YouTube subscribers, 300+ Twitter followers, 700+ Twitch.tv followers, and over $400 in profit. I was produced videos that reached thousands of views, and created an interview podcast where I talked to top players from around the world. And I did it all in my spare time while working a full time job.
Unfortunately, the game I chose had a pretty short lifespan, and interest eventually dried up. I decided to abandon the project rather than try to transition it into something else, but I learned a lot in the process. In addition to video editing and marketing, I also gained live production, live broadcasting, and interviewing skills. All-in-all, it was a fun and rewarding project.
Daily Fantasy Baseball Projection System
This project combined three of my favorite things: programming, baseball, and making money.
I was a marginally profitable daily fantasy player in 2012 and 2013, but I wanted to use my programming skills to cut down my time spent researching and make more money overall. The results? Over the 4900 games I played in the 2014 season: $15,128.00 entered, $17,022.04 won, $1,894.04 profit, 12.52% ROI. You can see the game-by-game results of my season here. I’ve got to say, that’s better success than I thought I would achieve! Fanduel takes about a 10% cut of every entry fee, and despite this I was able to manage a profit of $12.52 of every $100 I put in. Even with this success, I know there are a number of improvements that could be made to my system. There are pros who make a living playing daily fantasy sports, and my program isn't quite at the level to hang with them as of yet. Some of the problems that I had to tackle when facing this problem include:
- Data normalization: Scraping a number of different sites for data and matching it all together.
- Data analysis: There is a lot of data available about every baseball player. Separating the important details from the unimportant, and weighting them correctly took a lot of research.
- Algorithm design: Creating an efficient algorithm was important. Players get scratched just before games start, lineups get announced late, and my program had to be able to create lineups on the fly. Going through every iteration possible would take more than a day of computing, but I was able to get my program to generate lineups in an average of 10-20 seconds.
I learned a lot from the experiment, and really enjoyed the challenges. Since daily fantasy is pitting one's skills against another player (or players), it's really a game about outwitting your opponents more than anything else. Currently I live in Japan, and none of the major daily fantasy sports sites will let me play, so this project is on hold for now. I had a lot of fun playing, adding new factors into my program, and watching the last few at bats of the day when the results could be the difference between $20 and $200 on a single contest.