Thread history

Fragment of a discussion from Talk:Movement
Viewing a history listing
Jump to navigation Jump to search
Time User Activity Comment
No results

Another trick I think is that to use very large bandwidth, since WscB & C will fire at near 1.0 when you have some speed, being as far as possible helps a lot.

Since ScalarBot is using something more like uniform shape function (and very low weighted 1 / (1 + sq(x)) to hint true surfing), and it's weighting secondary wave equally for same bullet power, without evaluating stationary danger, all making it visit as much gfs as possible, it's not strange it will perform very bad against simple guns (where being precise and strict hurts).

Xor (talk)‎15:29, 6 October 2017

Maybe trying Normal Distribution would help. Sometimes I see Rechner just going between 2 possible bullet locations and get hit there.

Dsekercioglu (talk)‎15:35, 6 October 2017

I am using normal distribution in probability view. But this is not helping me against simple guns.

iirc something like 1 / (1 + abs(x)) works better.

Xor (talk)‎17:23, 6 October 2017

What is a probability view?

Dsekercioglu (talk)‎19:54, 6 October 2017

meaning traditional wave surfing buffers

as my style is more like shrapnel surfing, which is called peak views.

Xor (talk)‎21:23, 6 October 2017
 
 
 
 
Retrieved from "http://robowiki.net/wiki/Thread:Talk:Movement/Doing_better_against_simple_targeters/reply_(9)#Doing_better_against_simple_targeters_5889"