HUMIES 2016 ENTRY ================= # 1. The complete title of one (or more) paper(s) published in the open literature describing the work that the author claims describes a human-competitive result; Towards automatic StarCraft strategy generation using genetic programming # 2. The name, complete physical mailing address, e-mail address, and phone number of EACH author of EACH paper(s); ## Pablo Garcia-Sanchez email: pablogarcia@ugr.es smail: ETSIIT, Dept. Arquitectura y Tecnología de Computadores, Universidad de Granada / C. Periodista Daniel Saucedo Aranda, s/n / 18071 Granada / SPAIN tel: +34-958241778 ## J.J. Merelo email: jmerelo@geneura.ugr.es smail: ETSIIT, Dept. Arquitectura y Tecnología de Computadores, Universidad de Granada / C. Periodista Daniel Saucedo Aranda, s/n / 18071 Granada / SPAIN tel: +34-958243162 ## Antonio Mora email: amorag@geneura.ugr.es smail: ETSIIT, Dept. LSI, Universidad de Granada / C. Periodista Daniel Saucedo Aranda, s/n / 18071 Granada / SPAIN tel: +34-958241778 ## Giovanni Squillero email: giovanni.squillero@polito.it smail: Politecnico di Torino - DAUIN / Corso Duca degli Abruzzi 24 / 10123 Torino / ITALY tel: +39-0118997091 ## Alberto Tonda email: alberto.tonda@grignon.inra.fr smail: UMR 782 GMPA, INRA / 1 av. Lucien Brétignières / 78850 Thiverval-Grignon / FRANCE tel: +33-130814596 # 3. The name of the corresponding author (i.e., the author to whom notices will be sent concerning the competition); J.J. Merelo # 4. The abstract of the paper(s); Among Real-Time Strategy games few titles have enjoyed the continued success of StarCraft. Many research lines aimed at developing Artificial Intelligences, or ``bots'', capable of challenging human players, use StarCraft as a platform. Several characteristics make this game particularly appealing for researchers, such as: asymmetric balanced factions, considerable complexity of the technology trees, large number of units with unique features, and potential for optimization both at the strategical and tactical level. In literature, various works exploit evolutionary computation to optimize particular aspects of the game, from squad formation to map exploration; but so far, no evolutionary approach has been applied to the development of a complete strategy from scratch. In this paper, we present the preliminary results of StarCraftGP, a framework able to evolve a complete strategy for StarCraft, from the building plan, to the composition of squads, up to the set of rules that define the bot's behavior during the game. The proposed approach generates strategies as C++ classes, that are then compiled and executed inside the OpprimoBot open-source framework. In a first set of runs, we demonstrate that StarCraftGP ultimately generates a competitive strategy for a Zerg bot, able to defeat several human-designed bots. # 5. A list containing one or more of the eight letters (A, B, C, D, E, F, G, or H) that correspond to the criteria (see above) that the author claims that the work satisfies; (G) The result solves a problem of indisputable difficulty in its field. (H) The result holds its own or wins a regulated competition involving human contestants (in the form of either live human players or human-written computer programs). # 6. A statement stating why the result satisfies the criteria that the contestant claims (see examples of statements of human-competitiveness as a guide to aid in constructing this part of the submission); (G) To the best of our knowledge, at the time of the publication there were no automatically-generated strategies for StarCraft, a Real-Time Strategy game so complex that is used as benchmark for human-designed AIs. In March 2016, several months after the publication, Google's DeepMind Senior Fellow Jeff Dean said that StarCraft is probably going to be among the next targets for the AI team that tackled Go using AlphaGo. (H) In the SSCAIT 2016 (Student StarCraft AI Tournament), the biggest event for human-devised StarCraft AIs counting 88 participants, a strategy generated by our approach was able to reach an overall 12% victory rate, ending up in the lower part of the global ranking, but still much better than the 10% of the human-devised bots in the competition. # 7. A full citation of the paper (that is, author names; publication date; name of journal, conference, technical report, thesis, book, or book chapter; name of editors, if applicable, of the journal or edited book; publisher name; publisher city; page numbers, if applicable); @inproceedings{garcia2015towards, title={Towards automatic StarCraft strategy generation using genetic programming}, author={Garcia-Sanchez, Pablo and Tonda, Alberto and Mora, Antonio M and Squillero, Giovanni and Merelo, JJ}, booktitle={Computational Intelligence and Games (CIG), 2015 IEEE Conference on}, pages={284--291}, year={2015}, organization={IEEE} } DOI: 10.1109/CIG.2015.7317940 See IEEE Xplore paper page: http://ieeexplore.ieee.org/xpl/articleDetails.jsp?arnumber=7317940 # 8. A statement either that "any prize money, if any, is to be divided equally among the co-authors" OR a specific percentage breakdown as to how the prize money, if any, is to be divided among the co-authors; Any prize money, if any, is to be divided equally among the co-authors (as it is probably more convenient, authors that will be participating to the event may receive the whole sum and later divide it among the co-authors) # 9. A statement stating why the authors expect that their entry would be the "best" In the paper, the proposed framework proved to be able to defeat several human-designed bots. Our approach managed to beat them using the most difficult race to control, Zergs, which has been selected by just a small amount of entries. Moreover, due to the high number of possible units, feasible actions and size of the map, the range of possible strategies to apply is huge. Thus, it is an easier tasks to select the most appropriate possible set of actions for a human than for an automatic algorithm, which composes them almost from scratch. Indeed, the obtained strategies are sometimes quite different from those that a human could design intuitively, which is one of the advantages of applying a Genetic Programming-based method. These facts give our approach more merit. In addition, a strategy generated using the proposed approach managed to defeat several human-devised opponents in a high-level tournament, the Student StarCraft AI Tournament 2016. This is a quite tough competition, with most state of the art bots participating in it. Thus, even if our entry ended up in the lower part of the rankings, beating several of these participants is a very difficult challenge. # 10. An indication of the general type of genetic or evolutionary computation used, such as GA (genetic algorithms), GP (genetic programming), ES (evolution strategies), EP (evolutionary programming), LCS (learning classifier systems), GE (grammatical evolution), GEP (gene expression programming), DE (differential evolution), etc. LGP - The evolutionary core uses Linear Genetic Programming to automatically create C++ code, which composes the core of an StarCraft bot.