Javin: An update and possibly a name change

This project has been on the back burner for several months, but is picking up now that three MComp post-grads have been brought in to help 🙂 but the previous experience of these masters students suggests C++ is the most likely future platform. So, my own progress with Java (looking at an extension to VASSAL v3) is probably moot, but the C++ based 4th version of VASSAL (when it eventually arrives) might be better suited to this new approach. However, this might require a name change for the project; the J in Javin stood for Java after all.

Opinion on the UK X-Wing FaceBook Group about resurrecting this idea has been varied: One commenter, Allan, asked why we’d want to computerise a social, physical game when plenty of Star Wars video games existed already. That wasn’t an unreasonable question, but adding meaningful “single player” to any board/card/miniature game is a time honoured activity. Other responses showed support for the idea of being able to play/practice solo. Also, most Star Wars video games have been FPSs or Ship Sims, 3rd person combat MMOs and RPGs, or RTS games, and nothing like the X-Wing we know and love. That, in itself, is interesting, and leads me to think that a video game version of X-Wing turn-based dogfight style play is an underdeveloped concept; A.I. would make a big difference to the success of such a title, and creating a testbed for researching X-Wing bots would be a useful deliverable.

So, the aim of the project is now a C++ simulation (based upon the VASSAL implementation of X-Wing) that allows you to “play” X-Wing against a variety of bot opponents. Like VASSAL, it couldn’t have any copyright material, and wouldn’t be able to automate everything, such as special pilot abilities, etc, but might provide a simple opponent that goes a little beyond the web applet or the look up table for a basic A.I. that is available on BGG. However, augmenting the physical game, such as using VASSAL to reproduce “blow-by-blow” graphics to represent real games – here is one excellent example posted on FFG – should be considered. So, the ultimate aim of the simulator is threefold:
1) provide enough of a simulation of X-Wing to be at the level of VASSAL, but with rudimentary A.I. that goes beyond that provided by the current web applet. It would also be a testbed for further A.I. research.
2) Use AR to map a real tabletop game into the computer, by way of special tags on the base of each ship. This would initially just allow two human players to have their physical game tracked by the computer.
3) The two above combined for rudimentary single play, or the possibility of a computer aided/augmented game. This would use a projector underneath a translucent play surface to provide a projected HUD-style overlay for effects – e.g. damage, range, etc – similar to what is possible with VASSAL.

The obstacles are mostly the complexity of choice: Do I take a Focus or an Evade, or Barrel Roll? Knowing the opponent has a “Fat Han”, how would a bot guess where that YT1300 is going in order to block him with a TIE? This is why Jay Little was sceptical, when I talked to him about it a couple of years ago, and it has been slow progress working alone, or recruiting suitable students. However, I’d like the UK Nationals, one day, to have the top table be this interactive X-Wing thing, if only to have the game live-captured by the computer. Entering an artificial player in the tournament IS a pipe dream – Jay is right on that, for now at least! – because of its complexity. However, having the chance to play an Automated TIE Drone Swarm is a bit like Kasparov vs Deep Blue; likely for the droids to lose, but interesting to see how well they do. This is even canon; droids were fought against pre-New Hope and in the EU (The N64’s TIE/d anyone?), even though in Star Wars they weren’t as good as human pilots according to Wookipedia. The current students will start by investigating the first stage of simulating the same basic game play that VASSAL provides. Other features, such as the AR functionality will come later.

GLOSSARY
AR – Augmented Reality, which in this case uses a web cam to identify tags in order to ‘digitise’ the location of ships. Other applications involve overlaying data onto the real world; an example would be Google Glass(tm).
FFG – Fantasy Flight Games, producer of the Star Wars X-Wing Miniatures Game.
FPS – First Person Shooter, like Star Wars:Dark Forces.
MMO – Massively Multiplayer Online, such as Star Wars: The Old Republic.
RPGS – Role Playing Game, like Star Wars: Knights of the Old Republic.
RTS – Real-Time Strategy, such as Star Wars: Galactic Battlegrounds.

2 thoughts on “Javin: An update and possibly a name change”

  1. I have been thinking about something entirely different than AI. E-Sports is becoming HUGE. It seems that X-Wing is going the route of E-Sports; however, the coverage of the 2015 World Championships was NOT a very good experience for spectators.

    They had a fish eye camera rig suspended above the play table which distorted the play field. They cut other video feeds which was fine but in general, spectators have a hard time ‘seeing’ what’s going on in the play field.

    That is why I was thinking that having a system that tracks the miniatures in some way (I was thinking RFID to get the type of miniature and its location in 2D space). However, it seems you already have worked out this crucial bit of tech.

    Anyways, have you made any progress with the 2d location capture of the miniatures? Are you using a camera and doing image processing or something else?

    Once the 2d location data is captured, there are all sorts of interesting things one can do for visualisation especially for an E-Sport audience. I would be keen to know where you are with this.

    1. Hi Michael. Sorry for delay in posting 🙁 as I’ve not had too much time for blogging in the last couple of years. The ship tracking has taken a back seat of late, but we did some early tests using AR tags on ship bases, with a frosted Perspex board and IR lighting and an underneath web cam. This worked well enough, but ships did slip a bit; making the AR tags rubbery helped. The aim of the frosting was to allow projection of effects from underneath, which would enable SciFi style underplays to show ship stats, etc, and AI movement. We need to build a full size table, but funds are short 🙁

Leave a Reply