DarkHallow

From Robowiki
Revision as of 11:24, 27 September 2017 by Xor (talk | contribs) (Migrate from old wiki)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to navigation Jump to search

"Use no way as a way, No limitation as a limitation."...Bruce Lee


Help a targeting dummy out Help a fellow geek out Things I have noticed What Lesson is Here /OldBulletinBoard

Bulletin Board

September 10, 2005: -- Small segment change. Bug fix in the gun. Still can not stand in with a surfer. Score:

March 18, 2005: Ver 0.90.7.3 -- Small segment change.
Score: 1989.88 details graph 2184

March 14, 2005: Ver 0.90.7.2 -- Small segment change.
Score: 1993 after 899

March 13, 2005: Ver 0.90.7.1 -- Changed from 41 visit bins to 29.
Score: 2005.96 after 728

March 13, 2005: Ver 0.90.7 -- Added the segments back. Quick stats segment is still unsegmented.
Score: 2000 after 850+

March 10, 2005: Ver 0.90.5 -- Totally un-segmented visit buffer.
Score:1986.98 after 1027

March 1, 2005: Ver 0.90.4 -- More untested changes.
Score: 1997.25 after 1163

February 27, 2005: Ver 0.90.4 -- More untested changes.
Score:1996.53

February 24, 2005: Ver 0.90.3 -- More untested changes.
Score:1982.68

February 24, 2005: Ver 0.90.1 -- I have the day off, I am bored so I looked into the code. Found a couple of bugs so I fixed them. Not tested too extensively so who knows how it will do.
Score:1985.48

November 18, 2004: Ver 0.86.2 -- Reduced suffering segments.
Score:#12 overall. Not sure of the score.

November 17, 2004: Ver 0.86.1 -- Changing the surfing again. After chatting with PEZ I have some ideas on whats wrong with my sufring. Unfortunately those changes are not in this release. Just need a baseline.
Score: 1987.97 after 947

November 15, 2004: Ver 0.86.1 -- Same as .86 with bot width calculations removed. Added a better power management section which I hope earns more survival points.
Score: 1979.24 after 900. Another dissappointing release.

November 14, 2004: Ver 0.86 -- Orbit closest Wave origin, or target if no waves. Small change to danger calculations. Added stop evaluation although I do not think it is working all that well.
Score: 1989.16 after 1056. This was very disappointing as I thought this version was better than previous versions. It seemed to test better. Ah well.

November 12, 2004: Ver 0.85.3 -- Weighting waves according to bullet power.
Score: 2010.25 after 909

November 11, 2004: Ver 0.85.2 -- Roll back parts of movement to an earlier version.
Score: 2004.56 after 959

November 11, 2004: Ver 0.85.3 -- Removed setMax's to see if it makes a difference.
Score: 2001.72 after 810

November 10, 2004: Ver 0.85.3 -- Gun tweak, Rolling averages.
Score: 1999.65 after 955

November 7, 2004: Ver 0.85.2 -- Gun tweak, power management tweak, restored movement.
Score: 1991.93 after 1139

November 3, 2004: Ver 0.85.1 -- Weighting waves by shot power, segment tweaks.
Score: 2002.67 after 1291

October 16, 2004: Ver 0.85 -- Roll back and a bug fix. Damn you David_Alves =^>.
Score: 2010.09 after 829.

October 16, 2004: Ver 0.84.3 -- Movement tweak.
Score: 1980.38 after 1463

October 11, 2004: Ver 0.84.2 -- Movement tweak.
Score: 2019.34 after 1385

October 11, 2004: Ver 0.84.1 -- Added an even more (hopefully) precise movement prediction engine.
Score: 2015.77 after 695.

October 11, 2004: Ver 0.84 -- Added (hopefully) a more precise movement prediction engine.
Score: 2021.08 after 1159

October 10, 2004: Ver 0.83.1 -- Movement tweak.
Score: 2020.02 after 576.

October 10, 2004: Ver 0.83 -- Significant bug fix. I had an error in my directional calculation that resulted in movement waves that were backwards (-1 was in the direction of movement). Rolled back to 0.82.2.2 and am now testing the fix.
Score: 2022.71 after 901

October 9, 2004: Ver 0.82.4 -- Bug Fix. Still with PEZ's suggested change.
Score: 1998.05 after 1111.

October 9, 2004: Ver 0.82.3 -- Bug Fix. Still with PEZ's suggested change.
Score: 2011.2 after 1162

October 9, 2004: Ver 0.82.3 -- Roll back .0.82.2.2. Adding a suggestion by PEZ.
Score: 2008.83 after 535.

October 9, 2004: Ver 0.82.2.3 -- Experimental release. Added more GF's for hopefully better performance at longer ranges. Probably the last of the experimental releases before a rewrite of my surfing with some suggestions by PEZ thrown in.
Score: 2020.64 after 756

October 8, 2004: Ver 0.82.2.2 -- Experimental release. Adding more effective (I hope) protection from diving in.
Score: 2024.94 after 1367

October 8, 2004: Ver 0.82.2.1 -- Experimemtal release. Removed distance segment to see what effect it has.
Score: 2021.39 after 1130

October 8, 2004: Ver 0.82.2 -- Bug fix. Now distance segment works. I may disable it again later.
Score: 2026.15 after 584. Best score yet. Temporarily over 2040.

September 27, 2004: Ber 0.82.1 -- refinement of the bugfix in the previous version and a gun tweak.
Score: 2020.3 after 631. Great crash affected scoring.

September 26, 2004: Ver 0.82 -- Now bug fixing the .59.3 line. First bug fix: (int)Math.round(latVShotTime/3) can be rounded to 3. This muddied my movement related latvel stats.
Score: 2018.75 after 854

September 26, 2004: Ver 0.81.5 -- Now to make sure everything is back to normal.
Score: 2019.86 after 341

September 26, 2004: Ver 0.81.4 -- Testing the last of the .7x line of "improvements". This is the one I really hope works. It replaced AndrewsCoolWay with Paul Evans' classic RollingAverages.
Score: 2001.57 after 697

September 25, 2004: Ver 0.81.3 -- Testing some other options from the failed .7x line.
Score: 1999.38 after 603

September 25, 2004: Ver 0.81.2 -- Bug fix
Score: 1989.46 after 144.

September 25, 2004: Ver 0.81.1 -- Bug fix. Not sure what it was though.
RECALLED I had to download the bot and run it locally to find the error. Sorry to flail like that.

September 25, 2004: Ver 0.81 -- Adding back in functionality removed when I rolled back. Looking to see if it makes a difference or not.
RECALLED! Something is not right here. When you see this:

Fighting battle 10 ... jekl.DarkHallow 0.81,nic.Nicator 2.4
RESULT = nic.Nicator 2.4 wins 6063 to 488

you know something is not right.

September 24, 2004: Ver 0.80 -- Rollback to .59.3 to prove that it will not be as good. PEZ was right again.
Score: 2022.34 after 707.


Currently on my todo list:

  • I need to solve the "diving in" problem at some point. I am not sure what the correct solution is but I know there is one.
  • Wall avoidance. Can you believe I still hit the walls? I can not either.
  • Data saving at some point.



What's special about it?

It is my first WaveSurfer.

Great, I want to try it. Where can I download it?

http://www.robocoderepository.com/BotDetail.jsp?id=2296

How competitive is it?

It is in the rumble now so time will tell.

How does it move?

WaveSurf

How does it fire?

Uses a guess factored targeting system copied and slightly modified from BlackPearl.

How does it dodge bullets?

It surfs.

How does the melee strategy differ from one-on-one strategy?

one-on-one only at this point.

How does it select a target to attack/avoid in melee?

No melee capability in there yet.

= What does it save between rounds and matches?

No data is saved.

Where did you get the name?

From a song by the Grateful Dead: I would rather be in some Dark Hallow, where the sun don't ever shine, than to be in some big city, in a small room with a girl on my mind. I know it has no bearing on RC but I love the Dead so.....

Can I use your code?

Code is in the jar. It is not very tight right yet but feel free. Use the RWPCL please.

What's next for your robot?

Not sure yet.

Does it have any WhiteWhales?

None right now. More to come later I am sure.

What other robot(s) is it based on?

It started as BlackPearl, which is based on Jekyl.


Comments, questions, feedback:

"Can you believe I still hit the walls?" Try this:

setMaxVelocity(150.0 / getTurnRemaining());

Combine that with a large enough wall margin and you should be fine. For example, use 150.0 / getTurnRemaining() and don't pick destination points that are within 25 pixels of the walls. Using a higher number than 150 will prevent you from slowing down so much when you turn, but will require a larger wall margin to ensure that you don't hit walls. --David Alves

But this is a risky business for a WaveSurfer unless you also build it into your movement predictor. CassiusClay too hit walls at times. I think it is my WallSmoothing algo that don't take into account that it takes quite a while to reverse direction. Jim, check if maybe DH has decided to reverse direction just before it hits the wall and see if you share the same glitch. -- PEZ

So build it into your movement predictor. ;-) --David Alves

Nah. It's complicated enough as it is. -- PEZ

Right now I am testing all of my "improvements" from the .70 line. It seems that for the most part they were anything but. Once I am done with that I will move on to new things. One of the first orders of business will be to protect myself from firing enemy waves when they collide woth the wall. I am not sure how to work that out though. -- jim


November 2 2004 (CET) - Welcome back man. I've missed ya! -- PEZ

Lets see how I feel in a week. I was pretty burned out. This version will finally beat DT over 500 rounds. It is the only test I ran too. -- jim

So much for me having the top US bot. :-( Oh well. Congrats! --David Alves

Thanks. It was really just old code. Nothing new. Some small bug fixes. Still plenty of room between my score and infinity for you to slip into ;-) -- jim

Cool man, welcome back. ;] -- iiley


Pulsar is waking us all up. Isn't he? =) You found 10 points there. Good work. -- PEZ

I would not read too much into it. I just had the day off work due to a snow storm in the region. Like I said last time, lets see how I feel in a week. -- jim

How do you weight your visit stats compared to hit stats? Do you roll the stats? How fast? I find it interesting that you gain lots of points by segmentation of the visit stats while I lose a few... -- PEZ


Hey PEZ. I segment my stats like this:

hits[accelIndex][outIndex][(int)Math.round(myLatVel)/3][distIndex] fastHitsBuf[0][0][0]

The distance index is 5 buckets of 180 pixels, the out index is 5 buckets of 45 pixels. I am currently using 29 GuessFactors. I got a 6 point boost going from 41 to 29. I roll the stats using AndrewsCoolWay. My gun stats and my movement stats are indentical in most ways. My gun stats have more segments (a closingIndex and a timeIndex) and are not rolled. I have tried not rolling my movement stats but after about 15 or so turns I seem to get stuck between some local maxima and oscilate in place. As you are well aware, thats not so good =^> -- jim

Cool. And how (if) do you weigh the visit stats? Mine has weight 45. Unsegmented hit stats has weight 50 and most segmented hit stats weighs 100. I roll my movement stats using Paul's RollingAverage. -- PEZ