View Single Post
Re: mp_schwalbe (BETA) Waypoints
Old
  (#16)
the bindlestiff
Member
 
the bindlestiff's Avatar
 
Status: Offline
Posts: 280
Join Date: Nov 2005
Location: Washington state
Default Re: mp_schwalbe (BETA) Waypoints - 17-03-2006

I've been focused on completing my new computer for the past two weeks, so it's taken some time to get back to waypointing. The new FX-60/Geforce 7800GT is now online!

Here's some hard data about the route difficulties that I've observed. I've seen it in all the maps where I've used routes, but the easiest one to test is my original effort, mp_zion. There is a route #0 at the Allied spawn point that directs the engineers to take either of two paths to the dynamite objective. One is outside the chapel door (alt_roam #9) and the other is inside the door (alt_roam #8. The inside route is the most dangerous, so I'd llike the bots to mostly use the outside route that goes over the boxes and down to the outdoor patio. I have the pathactions set to "8 9 9 9" which should result in the engineer bots going inside 25% of the time and to the outside 75%.

I test this by kicking all the bots, then loading a special engineering squad of 10 Allied engineers. Restart the map and watch where they go. Repeat 10 times, which gives 100 bot trips through the route random-number generator, which should be enough to get a reasonable statistical count.

1. Results for the "8 9 9 9" scenario:
bots going outside - #9 = 32
bots going inside - #8 = 68

Not even close to 75/25 is it?

2. Changed the pathactions to "9 9 9 9 9 9" which should result in 100% of the bots going via #9.
Results:
bots going outside - #9 = 23
bots going inside - #8 = 77

This should have been 100/0.

3. Changed to "8 8 8 8 8 8"
Results:
bots going outsideto - #9 = 38
bots going inside - #8 = 62

This should have been close to 0/100.

4. Changed to "8 9"
Results:
bots going outside - #9 = 47
bots going inside - #8 = 53

This is the only case that looks right. The others make no sense at all. Either I'm doing something wrong (not at all unusual) or there is a bug in the code.

I hope someone else could repeat these measurements to verifiy my findings or not. I've attached my engineering squad to save you a little time (just remove the .txt extension)

edit: noticed that I had the wrong bot file attached -- fixed!
Attached Files
File Type: txt USengSquad.cfg.TXT (173 Bytes, 347 views)

Last edited by the bindlestiff; 17-03-2006 at 06:52..
  
Reply With Quote