Micky Holdorf Nov 13, 2018, 3:25 PM (5 days ago) to me Bot Name: Microwave Bot Race: Zerg Author Name(s): Micky Holdorf Affiliation(s): Independent Nationality(s): Denmark Occupation(s): System Developer, Web Developer (These will be listed on the competition website) Bot URL: SSCAIT https://sscaitournament.com/index.php?action=botDetails&bot=Microwave Personal URL: Affiliation URL: Questions about your bot (please answer as many as you can, especially Q 1-3) Q: What is the overall strategy/strategies of your bot? Why did you choose them? To play safe games, but also play cheese (5Pool) once in while. Q: Did you incorporate any of the following AI techniques in your bot? If you did, please be as specific as possible a) Search-Based AI (Path-Finding, A*, MiniMax, MCTS, etc) No b) Offline Machine Learning (Supervised or Unsupervised, but not RL) No c) Offline Reinforcement Learning No d) Online Learning of any kind (Including competition file IO for strategy selection) Yes, UCB1 e) Influence Maps No f) Custom Map Analysis BWTA map analyser for base positions and defense regions BWEM map analyser for chokes used for calculating positions of defense structures like sunkens, spores and simcity BWEM map analyser for chokes used for simple ground pathfinding g) Hard-coded or rule-based strategy / tactics Openings are hardcoded. When out of openings rule-based strategy is used. h) Analysis of bots from previous competitions / hard-coded specific bot counter strategies No, did not have time for that. I let my bot find what works best. i) Any techniques not mentioned here FAP, Fast APproximation combat simulator Q: How did you become interested in Starcraft AI? I stumbled upon the SSCAIT stream on twitch and watched it for months before I slowly started to play around with UAlbertaBot and Steamhammer code. Q: How long have you been working on your bot? 1,5 year Q: About how many lines of code is your bot? 20.000 Q: Why did you choose the race of your bot? Zerg is cool and has interesting mechanics. :D Q: Did you use any existing code as the basis for your bot? If so, why, and what did you change? I used Steamhammer 1.0 as the basis. It had the foundation and basic bot functionality in place. I changed strategy rules, unit selection, target selection, added calculation of sunken/spore/simcity positions, added more defensive reactions. I have added new functionality from Steamhammer which I felt was useful. Q: What do you feel are the strongest and weakest parts of your bot's overall performance? - There are a lot of weak parts… :) Unit squads are the weakest part. They are too rigid when units a spread across the map. They should be broken up into smaller groups that can do combat sim. - I am not sure Microwave has any parts that do not need to improved… Maybe the macro calculations… in many games Microwave can be behind and still win. Q: If you competed in previous tournaments, what did you change for this year's entry? - Adjusted openings - Changed when to drone - More sunken/spore reactions - Changed unit selection - Changed macro adjusting for number of enemy bases - Changed combat sim usage - Updated FAP - Changed strategy rules - Added ability to pull drones from main or natural - Added simcity buildings (hatch) in natural - Added sunkens/spores at expansions - Queue reordering so when waiting for items being able to build, certain other items can be made - Base recon - Attack weakest enemy base - Storm dodging - Scarab dodging - Added guardians - Added devourers - Ground pathfinding using chokes - Updated distance map calculations Q: Have you tested your bot against humans? If so, how did it go? No Q: Any fun or interesting stories about the development / testing of your bot? Q: Any other projects you're working on that you'd like to advertise? Optional Opinion Questions: Q: What is your opinion on the current state of StarCraft AI? How long do you think before computers can beat humans in a best-of-7 match? StarCraft AIs are too predictable. They should be able to change / adjust / react strategy more easily like humans. Q: What do you feel is the biggest hurdle (technological or otherwise) in improving your bot's AI? Not enough time and too many skills and features needs to be added. ;) Q: Which bots are the most interesting to you and why? Steamhammer. Its main race is zerg and the author is very active adding new features all the time. AIIDE Specific Question: Q: Do you feel that the current format of iterated round-robin win percentage is a good indicator of bot skill ranking? If not, how would you change it? Yes, in Bot vs Bot tournaments, I think round robins is the best indicator of which is the best bot.