Difference between revisions of "File talk:RobocodeMaximumEscapeAngle.jpg"
(Note about formula) |
|||
| (One intermediate revision by one other user not shown) | |||
| Line 6: | Line 6: | ||
Er, my fault. I'll put up another picture explaining why not shortly.[[User:Ceasar|Ceasar]] 2:04 AM, 23 March 2011 (EST) | Er, my fault. I'll put up another picture explaining why not shortly.[[User:Ceasar|Ceasar]] 2:04 AM, 23 March 2011 (EST) | ||
| + | |||
| + | I'm rather sure this image isn't right. Recently I had been working on some code to accurately and very efficiently calculate the maximum escape area. And it's results look like the following:<br/> | ||
| + | http://rednaxela-robocode.dyndns.org/data/EscapeArea2.png <br/> | ||
| + | Note, this picture for a bulletpower 2 wave, and a starting velocity of 0. From what I'm seeing, the maximum escape angle is closer to the old formula on the page, in just about all cases, due to the effect of bullet flight time. If anything, moving away might sometimes have a very slightly higher maximum escape angle. --[[User:Rednaxela|Rednaxela]] 12:42, 23 March 2011 (UTC) | ||
| + | |||
| + | Which is why many mid levels bots use a MaxEscapeAngle * 1.1 hack. — <span style="font-family: monospace">[[User:Chase-san|Chase]]-[[User_talk:Chase-san|san]]</span> 15:34, 24 March 2011 (UTC) | ||
Latest revision as of 16:34, 24 March 2011
I'm not entirely sure this is right, but I see what you're saying and I'm not sure you're wrong either. Using a circle implies that the amount of time the enemy has to move in any direction is the same, but the truth is the bullet flight time will decrease if you move towards the bullet source instead of away from it, so you won't get as far. It's also worth noting that the ratio of the bullet-side of the triangle to the bot-movement-side of the triangle is a maximum of 19.7 : 8, while the ratio looks larger in this picture - though I think adjusting that would make the areas you're pointing to bigger, anyway. The most common ratio is about 14 : 8. Curious to hear Rednaxela weigh in on this, I know he's done a lot of MEA research... --Voidious 02:09, 23 March 2011 (UTC)
Yeah, I can't say for sure it's entirely right. I only know for sure that the standard MEA formula isn't entirely right either. :P But my guess is that this is the case, and that typically the difference is too small to notice. I've recently been experimenting with shooting into a segmentation strategy that minimizes the number of segments necessary (partially achieved by shooting into the middle of buckets) and my gun was having trouble hitting orbiting targets by a hair's length. I'd definitely like to hear more as to the exact situation as well. --Ceasar 11:40 PM, 22 March 2011 (EST)
There are some interesting discussion about this at oldwiki:MaxEscapeAngle, you might me interested. --Nat Pavasant 04:25, 23 March 2011 (UTC)
Er, my fault. I'll put up another picture explaining why not shortly.Ceasar 2:04 AM, 23 March 2011 (EST)
I'm rather sure this image isn't right. Recently I had been working on some code to accurately and very efficiently calculate the maximum escape area. And it's results look like the following:
http://rednaxela-robocode.dyndns.org/data/EscapeArea2.png
Note, this picture for a bulletpower 2 wave, and a starting velocity of 0. From what I'm seeing, the maximum escape angle is closer to the old formula on the page, in just about all cases, due to the effect of bullet flight time. If anything, moving away might sometimes have a very slightly higher maximum escape angle. --Rednaxela 12:42, 23 March 2011 (UTC)
Which is why many mid levels bots use a MaxEscapeAngle * 1.1 hack. — Chase-san 15:34, 24 March 2011 (UTC)
