Difference between revisions of "ScalarBot"
Jump to navigation
Jump to search
m (Scalar now means fast, precise, adaptive and accurate.) |
m (What's next) |
||
| Line 11: | Line 11: | ||
| − | : | + | : Fast, Precise, Adaptive and Accurate |
| − | + | : — [[Xor]] ''ScalarBot'' | |
| − | |||
| − | |||
| − | |||
| − | : — | ||
| Line 41: | Line 37: | ||
== Additional Information == | == Additional Information == | ||
; Where did you get the name? | ; Where did you get the name? | ||
| − | : ''Scalar replacement, when I was reading [[wikipedia:Escape analysis|Escape analysis]]'' | + | : ''Scalar replacement, when I was reading [[wikipedia:Escape analysis|Escape analysis]]. '' |
: ''Scalar used to mean fast, strict and precise. But being strict and precise doesn't necessarily make it accurate. '' | : ''Scalar used to mean fast, strict and precise. But being strict and precise doesn't necessarily make it accurate. '' | ||
| Line 47: | Line 43: | ||
; Can I use your code? | ; Can I use your code? | ||
| − | : ''Not yet. '' | + | : ''Not yet ;). But [[SimpleBot|what it's mostly based on]] has already been made open source. '' |
; What's next for your robot? | ; What's next for your robot? | ||
| − | : ''Make | + | : ''Make it fast, make it precise, make it adaptive and make it accurate —'' |
| − | + | :* ''Optimize precise prediction related code harder'' | |
| − | + | :* ''Take bot width into account in precise mea calculation (for accurate gun score)'' | |
| − | + | :* ''Tune movement against different kinds of targeting strategy (for better hit rate classifier)'' | |
| + | :* ''And then choose countermeasure accurately'' | ||
; What other robot(s) is it based on? | ; What other robot(s) is it based on? | ||
Revision as of 05:32, 10 October 2017
- ScalarBot Sub-pages:
- Version History - Understanding ScalarBot
Background Information
- Fast, Precise, Adaptive and Accurate
- — Xor ScalarBot
- How competitive of it?
- It ranks #3 in RoboRumble since Oct. 3rd, 2017
- What's special about it?
- Nothing. Really? Since you asked —
- By far the top bot without an Anti-Surfer Gun
- By far the top bot without a Flattener
- By far the top bot without using Gun Heat Waves
Strategy
- How does it move?
- kNN Wave Surfing
- How does it fire?
- GuessFactor with Conditional Density Estimation via kNN
- What does it save between rounds and matches?
- Everything between rounds, nothing between matches
Additional Information
- Where did you get the name?
- Scalar replacement, when I was reading Escape analysis.
- Scalar used to mean fast, strict and precise. But being strict and precise doesn't necessarily make it accurate.
- Thus it now means — fast, precise, adaptive and accurate.
- Can I use your code?
- Not yet ;). But what it's mostly based on has already been made open source.
- What's next for your robot?
- Make it fast, make it precise, make it adaptive and make it accurate —
- Optimize precise prediction related code harder
- Take bot width into account in precise mea calculation (for accurate gun score)
- Tune movement against different kinds of targeting strategy (for better hit rate classifier)
- And then choose countermeasure accurately
- What other robot(s) is it based on?
- It inherits ShieldBot's structure while using the gun and movement (actually, only the trees) from SimpleBot. It, however, evolved a lot since then.
- Although no code directly, its evolvement is inspired by a lot of bots, including DrussGT, Diamond, Tomcat, Nene and more on this wiki.
- Credits give to
- Skilgannon, for his DrussGT gun, which influenced me a lot, and for his wise advice to me.
- Voidious, for his Diamond gun, which is extremely strong, pushing me to work on movement more and more. Some gun attributes are also inspired by Diamond.
- Jdev, for his Tomcat movement, which is pretty unique, and its data analysis inspired me a lot as well.
- Chase, for his Nene, which told me that a flattener or an Anti-Surfer Gun is not essential. Its movement inspired me a lot as well.
- Jamougha, for his RaikoMicro, strong GuessFactor Gun with strong Random Movement, which is my favorite bot to test with ;).
