ChessCraft in 2023

Published on 2024-01-01 by Stuart Spence

I've often liked year-end reviews like from quanta magazine. So here's another short one for ChessCraft like last year.

660,000 Installs

It has now been 6 years and 2 days since my first git version of ChessCraft. Remarkably, the rate of installs per month is still increasing. We're now at 660,000 installs which is nearly double the total at the end of 2022. This is on track to hit one million installs sometime in late 2024. Wow!

There are now nearly 175,000 custom pieces and boards designed by the ChessCraft community.

Jobs

One milestone this year is the creation of a jobs page for ChessCraft. I haven't advertised this much. I spoke with one great applicant for community manager but they were overqualified - it's more of a university student tier position. Other candidates seemed like nice people but I'm hoping to see evidence of professional communication skills. No hires yet. So I'm still cranking through piles of Discord chats, emails, Google Play reviews, and Steam reviews all by myself.

Development

In 2023 I released 1.14 (popular designs, acrobats, and online backups) and 1.15 (folders! and eight new piece rules). I also replaced my MongoDB database with PostgreSQL which turned out to be a great decision.

Finally, I'm mostly done a major new feature to be released sometime in 2024. This is by far the most popular feature request. It involves S3 buckets and presigned URLs. Can you guess what it is?

LLMs

I first wrote about ChatGPT in the 2022 year review. In 2023 it has been an incredible tool for me. I've also been experimenting with open models for my new role at Environment Canada which is all about AI.

I was thinking of running an LLM (like ChatGPT) on all ChessCraft boards and pieces. This could maybe:

  • write a description for designs without one
  • categorize them
  • detect inappropriate content

This is not super useful to players, but lots of fun for me. Stay tuned!

chess.com

chess.com is one of the largest chess websites with 10 million daily users.

Much to my surprise, a few months ago their staff contacted me. I even exchanged some emails with the CEO. I admit this was flattering. The possibilities were a bit vague but included partnership, licensing, or acquisition. As some of you know, I have no serious business plan and just love building the game and community so this really caught me off guard. I lost some sleep!

Their team looked at two aspects of ChessCraft:

  1. number of users
  2. strength of AI

They concluded neither were strong enough to interest them.

My appraisal of the situation: the ChessCraft AI is good enough to beat 99% of casual players, but anyone working at chess.com is better at chess than that. If you play ChessCraft in classic chess it's easy not to be impressed. Imagine riding a bicycle on a highway. You don't see all the forests and alleys the bicycle can go to, you only see how slow it is compared to the hyper-specialized cars on the highway.

The chess.com developers also have a financial incentive to downplay how much clever work goes into features like weird board shapes and "blacksmith" pieces. In my opinion, no textbook chess engine can be patched to handle most ChessCraft features.

As for my player numbers, the growth is accelerating so I'm not sure what they were looking for.

Anyway, our communications died out within a couple days. I could list a dozen reasons why the variants section of chess.com is a graveyard compared to their 10 million daily users, but they didn't give me much of a chance to talk about this! The historic low interest rates disappeared around this time so maybe the prospect of throwing around cheap money became unnappealing.

Vision

This led me to think about my high level goals for this project.

  1. Build the best chess variant community and database.
  2. I don't plan to sell ChessCraft except for a life changing amount of money. And I'm not sure it's worth that much yet.
  3. Still no plans for any ads.
  4. In 2024, a focus on big features to unlock more creativity.
  5. I'm always considering porting from Unity to Godot. This would mean better performance, development velocity, and long term release stability.
  6. Promote, discover, and analyze the best user content.

Goodbye, 2023

I hope everyone had an okay year, despite the challenges that came with it. I look forward to 2024 and hope you do too. Thanks for your support and being such a great chess variant community! I wish you all the best in 2024.

white griffin sprite