An important milestone has been reached with this project. It has a presentable name: StarCraft Human ‘N’ AI League, or SCHNAIL. I’d like to explain the goals, and execution of this ladder a little bit more.

First of all, we have a prototype of a logo!

The goal is to facilitate Human vs. Bot play. Conveniently, and with features common in modern gaming. This entails having a separate, self-contained launcher, where all you need to do is select a map, a bot, press play, and you are instantly thrown into a custom game vs. that bot.

The goals are multiple. First, many people only know the built-in AI, and gets quite surprised when they find out that the custom AIs are quite strong nowadays. If you watched any of Jealous’ man vs. machine streams, you know what I’m talking about. He is a 1900 MMR player, and sometimes got beaten by bots. So I’d like to provide a good challenge to people, who don’t necessarily want to play vs. other humans. The bots provide a consistent, stable challenge. Also, they tend to write “omfg retrad noob uninstall” somewhat rarely.

Besides casual gamers, I plan to introduce this to pros, or pros-in-training. A bot can do a lot of things, and it can be even used to test out specific setups, micros, etc. For example, I heard of a pro player trying to beat Locutus at goon micro. Which was a very good, but ultimately futile exercise. There was an attempt at this already in the past – CoachAI believe it was called. I aim to make the process considerably easier. And if there is a need, and enough exposure, this will be a stronger pull for bot authors.

I believe if there is a community around that, that incentivizes bot authors as well. There is a more immediate feedback, and it’s easier to find out bugs/stupid behaviors than it is now. Also, in the future, I’d like to actually pay bot authors somehow, even if just a token amount, or in some related swag. They are the true stars of the show, believe me.

I get inquiries every now and then, when it will be done, etc. The answer is always the same: It will be done, when it’s finished. We are making steady, albeit slow progress, and that has been the case since we started. Notice that I use a plural “we” here, as there is an actual team now.

I will post some answers to frequently asked questions here as well.

What version of BWAPI will you support?

4.4 will be a minimum when we go live. Until then, a grace period will be in effect, so bots that use older BWAPI versions can participate. For testing, we don’t actually need 4.4, but after that, I’d like to ensure bot quality. Eventually, BWAPI5 is the goal, of course.

What languages are you supporting?

C++ and Java currently, with plans for Python. I plan to accomodate as many as feasible.

Will you support SC:Remastered?

Currently we don’t have the means, but if it becomes possible, definitely. The SCHNAIL version currently is 1.16. However, we are also working on a custom hotkey plugin, so getting back to it will be easier.

Will it be public?

Not initially. During alpha, a select few will be able to test it, but since then it’s not even feature-complete, I don’t want to release that into the wild. After that, a closed and an open beta will follow. Also, the look and feel is not finalized yet. I have a pretty good general idea how it will look, but it’s still a lot of UI (and art) work. I will, of course publish updates as soon as it changes.

Will it be an actual ladder with leaderboards and stuff?

Not in a sense that you will be ranked – since this is, at the heart of it, a PvE training ground. I do plan achievements, and other social features, but as an opt-in. It is my intention that no one needs to know about your gaming, if you don’t want them to.

Do I have to pay for it?

As a player: No. As a bot author: Absolutely not. Funding will be secured in other ways. Supporting, of course, will be available to anyone.

What are the requirements of participating?

As a player, nothing, as a bot author, your bot should not crash, basically. During beta, the requirements will be very lax, then there will be a gradually increasing quality control, and a bot upload auto-test pipeline.

What maps will be playable?

At the start, the current SSCAIT map pool. The bots are not required to be able to play every map – during upload, you can specify which maps you want to enable your bot on. There is also a map upload functionality, and there is no technical barrier for “Use Map Settings” type maps – and I do plan enabling those for training scenarios.

How can I help? I have ideas!

While suggestions are welcome, implementation is the bigger issue. Currently, making a website is big step we didn’t really begin yet, and also, the client and the coordinator server is written in Java – so if you really want to help, that kind of development effort is welcome. I’d like to stress that I’m not begging or asking anyone to help out – I’m glad for any help, but I’m not relying on it to finish this project.

These are all the questions that came up so far. If you have others, please don’t hesitate to ask!

As a closing thought, here is a picture of the current state. It is kinda-sorta working, but a lot of stuff is hardcoded, and well, work in progress.