Difference between revisions of "Gaff"
Darkcanuck (talk | contribs) m (link to targeting page) |
(Fix broken link and update) |
||
| (2 intermediate revisions by 2 users not shown) | |||
| Line 7: | Line 7: | ||
| released = June 2008 | | released = June 2008 | ||
| rating = 1689 (95th) | | rating = 1689 (95th) | ||
| − | | current_version = 1. | + | | current_version = 1.50 |
| license = closed | | license = closed | ||
| − | | download_link = http:// | + | | download_link = http://robocode-archive.strangeautomata.com/robots/darkcanuck.Gaff_1.50.jar |
| isOneOnOne = true | | isOneOnOne = true | ||
| isMelee = false | | isMelee = false | ||
| Line 23: | Line 23: | ||
I got tired of wrestling with learning bots and made Gaff using a simple movement that popped into my head. This bot was originally a testbed for a new targeting method I was working on. | I got tired of wrestling with learning bots and made Gaff using a simple movement that popped into my head. This bot was originally a testbed for a new targeting method I was working on. | ||
| − | But Gaff's claim to fame is achieving the highest score vs. [[Shadow]] in the [[ | + | But Gaff's claim to fame is achieving the highest score vs. [[Shadow]] in the [[Targeting Challenge 2K7/Fast Learning Results|fast learning Targeting Challenge 2K7]] by a wide margin. |
; How competitive is it? | ; How competitive is it? | ||
| − | Currently in the top- | + | Currently in the top-60 of the 1-on-1 [[RoboRumble]]. |
| − | |||
== Strategy == | == Strategy == | ||
| Line 51: | Line 50: | ||
Between rounds, saves all neural network weights, targeting stats and bullet dodging data. | Between rounds, saves all neural network weights, targeting stats and bullet dodging data. | ||
Nothing saved between matches. | Nothing saved between matches. | ||
| − | |||
== Additional Information == | == Additional Information == | ||
| Line 73: | Line 71: | ||
; What other robot(s) is it based on? | ; What other robot(s) is it based on? | ||
Most of the framework is shared with [[Leon]]. | Most of the framework is shared with [[Leon]]. | ||
| − | |||
== Credits == | == Credits == | ||
Latest revision as of 00:24, 13 August 2017
| Gaff | |
| Author(s) | Darkcanuck |
| Extends | AdvancedRobot |
| Targeting | Neural Targeting (GF) |
| Movement | Random Movement |
| Released | June 2008 |
| Current Rating | 1689 (95th) |
| Current Version | 1.50 |
| Code License | closed |
| Download | |
Sub-Pages: Version History | Challenges | Targeting
Background Information
- What's special about it?
I got tired of wrestling with learning bots and made Gaff using a simple movement that popped into my head. This bot was originally a testbed for a new targeting method I was working on.
But Gaff's claim to fame is achieving the highest score vs. Shadow in the fast learning Targeting Challenge 2K7 by a wide margin.
- How competitive is it?
Currently in the top-60 of the 1-on-1 RoboRumble.
Strategy
- How does it move?
Gaff was originally based on a cornering movement which tried to push enemies into the corners to narrow their escape angle and expose near-wall profile flaws. But since version 1.10, Gaff uses a multi-mode movement, consisting of:
- the basic cornering movement
- a simple orbital-style movement with bullet avoidance
- a random orbital-style movement which attempts to put Gaff at a random GF every time the opponent fires.
The cornering and simple orbital movement have a bit of randomness mixed in, plus the bullet prediction/dodging code from Leon. The random orbit makes no attempt to predict or dodge.
- How does it fire?
All the gory details are here: Gaff/Targeting
- How does the melee strategy differ from One-on-one strategy?
No difference, but the movement algorithm doesn't take more than one opponent into account so melee performance could be pretty bad.
- How does it select a target to attack/avoid in melee?
Selects the closest target, with some protection against target thrashing.
- What does it save between rounds and matches?
Between rounds, saves all neural network weights, targeting stats and bullet dodging data. Nothing saved between matches.
Additional Information
- Where did you get the name?
Blade Runner again.
- Can I use your code?
Not yet. But feel free to ask if you're curious and maybe I'll post some snippets
- What's next for your robot?
Improve the NN targetingDONE- Add a melee strategy, since I prefer that field over 1on1
- Improve targeting against random movers
- Make the multi-mode movement sophisticated enough to secure a top-50 position
- Does it have any White Whales?
- CassiusClay is proving frustratingly difficult to hit in the Targeting Challenge 2K7
- DogManSPE is Gaff's worst ProblemBot, followed by Infinity
- What other robot(s) is it based on?
Most of the framework is shared with Leon.
Credits
Targeting originally inspired by Engineer, which uses a SOM rather than a backprop network.
Inputs to the targeting network was originally based on Raiko's segmentation but heavily refined since then. The current versions use a data arrangement heavily inspired by DrussGT, which was graciously shared by Skilgannon on the TC2K7 chat page. The new data also uses a modified version of Simonton's radial-distance-to-wall algorithm.
Also thanks to Rednaxela for inspiring the training approach used in Gaff's anti-surfer gun.
