🎉 Celebrating 25 Years of GameDev.net! 🎉
Not many can claim 25 years on the Internet! Join us in celebrating this milestone. Learn more about our history, and thank you for being a part of our community!
My bot ! [Was : All bots so far...]
The problem with the woodbot version you''ve seen is that it react to fast if it loose sight of the enemy... the new vesrion will handle that a bit different...
ok for the prediction, isn''t the formula for the anglular velocity (change in angle / change in time). So would I have to implement a timer myself or is there a better way of doing it?
might be wrong bout the formula, long time since I did physics.
thanks
Author
We''re talking degrees/update here... You don''t have to implement timers : rememeber the angle from the previous update, get the angle from the current update, and your instant angular velocity for the previous update is (current-previous) degrees per update.
ToohrVyk
ToohrVyk
You can use the idea suggested above, or you can use this method for genuinely real-time stuff. It's all the same. For a timer, I just wrote a function
that returns the number of milliseconds since the match started (effectively). Additionally, I have a static int that grabs this value each frame and uses it for calculations, like how long the match has been going, prediction, etc. Works surprisingly well for being so simple, even if it is cumbersome to use milliseconds (of course you could always convert it to decimal seconds if you so chose).
Later,
ZE.
[edited by - zealouselixir on August 5, 2003 5:52:28 PM]
DWORD GetNormalizedTicks(){ static DWORD FirstCall = GetTickCount(); return GetTickCount()-FirstCall;} that returns the number of milliseconds since the match started (effectively). Additionally, I have a static int that grabs this value each frame and uses it for calculations, like how long the match has been going, prediction, etc. Works surprisingly well for being so simple, even if it is cumbersome to use milliseconds (of course you could always convert it to decimal seconds if you so chose).
Later,
ZE.
[edited by - zealouselixir on August 5, 2003 5:52:28 PM]
thanks, the prediction seems to work now. Didn''t improve the kill ratio much though. I''ll see what I can come up with to be ready to counter 2.0
I guess hiding isn''t the best tactic to use
A new version (0.2) of woodbot is avaiable:
ftp://ftp.woodoomagic.de/bots/woodbot.zip
it still has a lot of bugs ;-)
ftp://ftp.woodoomagic.de/bots/woodbot.zip
it still has a lot of bugs ;-)
for woodbot, yeah it still has bugs. Especially if you attack it from behind
. It''ll just start backing up, rotating and fireing.My bot also seems to be broken, I disabled the hiding stuff and the kill ratio jumped up to about 90% on 1.8
I''m gonna write a new hiding system, got a good idea for it. Watch out 2.0
@ ToohrVyk, hey man just wanted to thank you for posting a decent bot for me to test against, and warn you that in getting about a 9 out of 10 kill ratio.
@ Woodoo, your prediction is a bit stronger but a couple of sneak and peek tactics really confuse your bot. 7/10 kill ratio.
@ HappyDude, 4 wins /4 losses/2 draws. after 10 iterations, not a great sample beacuse we never spotted each other on the 2 draws. Your aim is a tiny bit better but i got a full grenade kill on 1 of my wins.
[edited by - Dreddnafious Maelstrom on August 5, 2003 7:35:59 PM]
@ Woodoo, your prediction is a bit stronger but a couple of sneak and peek tactics really confuse your bot. 7/10 kill ratio.
@ HappyDude, 4 wins /4 losses/2 draws. after 10 iterations, not a great sample beacuse we never spotted each other on the 2 draws. Your aim is a tiny bit better but i got a full grenade kill on 1 of my wins.
[edited by - Dreddnafious Maelstrom on August 5, 2003 7:35:59 PM]
Here's a link to my HappyBot. He wins against Woodbot about 95% of the time, against Devious 1.8 around 50%, although it's hard to judge that accurately, since Devious will win a few in a row, then mine will win a few, and I don't have the patience to do enough trials to do a proper sampling .
The major problem with HappyBot is that he has absolutely no prediction system implemented yet... he makes up for this by moving really close to the opponent bot, so that he doesn't have to aim so well. I've also got an internal map and object list system, both of which are currently disabled until wall detection works properly... once those work, I can start doing a few more complex things, such as not getting stuck on rocks...
Note for the woodbot person, your version 0.2 still seems to be named 0.11 in game...
[edited by - HappyDude on August 5, 2003 7:10:20 PM]
.The major problem with HappyBot is that he has absolutely no prediction system implemented yet... he makes up for this by moving really close to the opponent bot, so that he doesn't have to aim so well. I've also got an internal map and object list system, both of which are currently disabled until wall detection works properly... once those work, I can start doing a few more complex things, such as not getting stuck on rocks...
Note for the woodbot person, your version 0.2 still seems to be named 0.11 in game...
[edited by - HappyDude on August 5, 2003 7:10:20 PM]
This topic is closed to new replies.
Advertisement
Popular Topics
Advertisement
