I ended up being the only one with a functioning bot by the start of this week, another programmer had attempted his but had to scrap it, and the other had not made an attempt at one yet. On Monday, my bot was winning against the test bot we were given, 73% of the time, and when testing against a bot I got from the last trimester programmers, my bot succeeded 20% of the time. By Wednesday, these values were up to 80% and 30%, just by making my movement a lot better. I just need to make my movement a bit better and make my prediction a little more successful, and then hopefully I will become victorious, but the bot that I received from last trimester programmers, was actually the winning bot for the first phase (open world) so it is kind of THE bot to beat. Another programming teacher, Iain McManus, apparently has a bot that is unstoppable, so if I end up becoming victorious against the winning bot of last trimester, I may have to summon Iain’s bot into the arena.
As stated in Last Week’s Weekly Roundup, I compiled a list of questions over the weekend from problems I encountered or was just curious about. I got some good ideas for improvements to my bot, so they were very beneficial questions. We also worked on the bots a little more during class, for about half an hour.
Greg, our programming teacher, showed us some tools he recommends we play around with a bit. Of course I have forgotten the names of all of them, but there was:
- A NavMesh generator
- A tool that you could build levels really easily and fast with the ability to place infinite rooms in a certain space, through what seems like dimensional magic (but just uses perspective trickery)
- A genetic algorithm program, where you input an image and the program will create generations of randomly generated polygon shapes, sizes and colours, and will attempt to recreate the image inputted by layering the polygons it generates and manipulating them as the generation count increases.
- An AI flanking example program (not so much a tool)
I have always been interested in Tools Development and was very intrigued with the programs that Greg presented, and inspired by them.
We got down and dirty with learning about matrices. I remember matrices to be more difficult than how Greg taught us, I don’t know if it was just the way my teacher at school was teaching us, or if we were taught a more confined version by Greg, one that suited programmers, but it all made so much more sense, and I can see how so very helpful they are to programmers (especially for graphic programmers, but definitely for overall games development as well).
We were also taught how genetic algorithms worked. It was like a Biology revision at the beginning, as at first we were told all the biological terms of genetics and then they were translated to programming terms. It is very interesting how similar the structure of a programming language is to biological genetics.
I think I am going to start making these blogs on Fridays instead of Sundays as those extra few days do not help when trying to remember what was covered in the current week’s classes. Doing the Weekly Roundup on Fridays will also mean I don’t have to worry about them for the rest of the weekend, a very positive move as it does take me a while to write these blogs, as I am not a natural blogger.
This is Daniel Jochem, signing out.