This repository contains a toolkit for simulating board games with Google spreadsheets. The target group is board game developers who also knows some coding.

Sometimes images are better than words. Below is an example of when the Bot Playtesting Toolkit plays No Thanks.

Note that the time stamps in the log messages above show elapsed time in milliseconds.

Statistics for 10,000 games of No Thanks, played according to a given strategy.

You're building a board game. At some point you get interested in getting quantitative data: How often does X occur? Will strategy A generally win over strategy B? Will changing a card or a rule in a particular way shorten the game, and if so, by how much?

These questions could be answered by playtesters, but it would take a lot of sessions to get reliable data. And the results will often lead to tweaking numbers in the game, to get things balanced in the way you want. Which will require even more playtesting. Playtesting that could be better spent on getting feedback on how the game feels, what parts are more or less engaging, what makes the players laugh or grit their teeth, and things like that.

The idea behind the Bot Playtesting Toolkit is that human time is precious, while computer clock cycles are cheap. Use human playtesters for the things computers can't do.

Check out the documentation.

BPT is under development. To be sure to be notified about new stuff, either subscribe to this repository here on GitHub or follow my blog.

The Bot Playtesting Toolkit is an open source project. This means that the code is free to use, study, share and improve (and that the code must be passed on under the same conditions). It also means that you are very welcome to join in developing the toolkit. There is an issue queue for the Bot Playtesting Toolkit. You can contribute to the project by posting ideas and suggestions there, or discuss ideas that have already been posted. The issue queue is also used for reporting and tracking bugs – feel free to suggest code improvements too.