{c}
{tag=MIT}
{tag=Gridworld}
{title2=2003-}

* https://github.com/battlecode

TODO quick summary of game rules? Perhaps: https://battlecode.org/assets/files/battlecode-guide-xsquare.pdf

Some mechanics:
* inter agent communication
* compute power is limited by limiting <Java> bytecode count execution per bot per cycle

\Video[https://www.youtube.com/watch?v=oa4CAizd1Nk]
{title=Battlecode Final Tournament 2023}

\Video[https://www.youtube.com/watch?v=BLExWo9Empk]
{title=Introduction to Battlecode by MIT OpenCourseWare (2014)}


Battlecode

{c}
{tag=Apache License}
{tag=Gridworld}
{title2=2020}

* https://github.com/deepmind/lab2d
* https://deepai.org/publication/deepmind-lab2d

<Gridworld> version of <DeepMind Lab>.

<Open sourced> in 2020: https://analyticsindiamag.com/deepmind-just-gave-away-this-ai-environment-simulator-for-free/

A tiny paper: https://arxiv.org/pdf/2011.07027.pdf

Very similar to <gvgai>, <Julian Togelius> actually called them out on that: <DeepMind Lab2D vs gvgai>.

TODO get running, publish demo videos on YouTube.

\Image[https://web.archive.org/web/20221218211007im_/https://github.com/deepmind/lab2d/raw/main/docs/screenshot.png]
{source=https://github.com/deepmind/lab2d}


DeepMind Lab2D

{c}
{tag=Gridworld}

* https://github.com/google-deepmind/pushworld 2023 Too combinatorial, gripping makes it so much easier to move stuff around in the real world. But cool nonetheless.


Gridworld AI game

{c}
{tag=Gridworld}
{title2=2014-2020}

http://www.gvgai.net (dead as of 2023)

The project kind of died circa 2020 it seems, a shame. Likely they funding ran out. The domain is dead as of 2023, last archive from 2022: https://web.archive.org/web/20220331022932/http://gvgai.net/[]. Marks as funded by <DeepMind>. Researchers really should use university/GitHub domain names!

Similar goals to <Ciro's 2D reinforcement learning games>, but they were focusing mostly on discrete games.

They have some source at: https://github.com/GAIGResearch/GVGAI TODO review

A published book at: https://gaigresearch.github.io/gvgaibook/

From <QMUL Game AI Research Group>:
* Simon M. Lucas: https://gaigresearch.github.io/members/Simon-Lucas[], <principal investigator>
* Diego Perez Liebana https://www.linkedin.com/in/diegoperezliebana/
* Raluca D. Gaina: https://www.linkedin.com/in/raluca-gaina-347518114/ from Queen Mary
From other universities:
* <Julian Togelius>
TODO check:
* Ahmed Khalifa
* Jialin Liu


gvgai

A discrete 2D game on a rectangular grid: https://towardsdatascience.com/reinforcement-learning-implement-grid-world-from-scratch-c5963765ebff

This is analogous to many traditional <board games> such as <chess>, the concept is very natural and maps well into computer.

The downsides of gridworld games are:
* it is hard to model speed in discrete worlds. When you 10x faster, when do you collide with something else that is also crossing your path?
* they tend to not use vector representations of objects. So to have an object be 10x longer than another one, the naive implementation has to add 10 smaller objects. This becomes untenable as the number of objects increases.


Ciro Santilli @cirosantilli 37

 Tagged: Gridworld