[Home]David Alves/JALT 1 Discussion

Robo Home | David Alves | Changes | Preferences | AllPages

What is JALT short for? If, indeed, it's short for anything. Do you know what's holding this experiment back? If you have been using FloodGrapher to tune it's movement I can tell you that it is dangerous to do that. I've created I think five (5!) different movements that all show up flat as pankakes in FloodGrapher but that all gets trashed by DT and many other bots. I think the problem lies with the Flood targeting system refusing to fire above guess factor 0.8 or so. I have tried to follow the source code of FloodMini and Fhqwhgads to find out why this is so, but I fail. It's all a bit too compressed code I for my thick head I guess. Anyways, use DuelistMini to gather the data and save it the same way as FloodMini does and you can graph it with FloodGrapher and I think you are set to create a killer movement. -- PEZ

JALT stands for "Just A Little Test", of course. :-P

I don't use FloodGrapher, I didn't even know there WAS a FloodGrapher. I made a graphing version of DuelistMicro a long time ago, but I don't really use it much... I seem to make better robots by randomly messing with things than I do by being systematic about it. JALT was testing a fundamentally different way of approaching movement. Apparently the idea works in theory but not in practice. --David Alves

Of course! I'm illegaly stupid at times.... Maybe if it was named JALTMF I would have understood. =) Can you share that movement theory? I need some fuel into my constant MovementLaboratory failures. -- PEZ

About systematic/nonsystematic approaches: Mr Systematic himself seems to be successful with his approach. As does Kawigi and iiley. JALT maybe could perform better with just a few grains of this to its development process. =) -- PEZ

TrackMovement. --David Alves

Robo Home | David Alves | Changes | Preferences | AllPages
Edit text of this page | View other revisions
Last edited March 7, 2004 11:30 EST by PEZ (diff)