Artificial Fish Intelligence and selectivity??

RWFF game announcements and discussion (off-topic posts will be deleted)

Moderators: Scottish Warrior, Yellowstone7, richwest, Pitti

Artificial Fish Intelligence and selectivity??

Postby Yellowstone7 » Fri Feb 13, 2009 1:10 am

Hi Everyone:

I have been enjoying the RWFF program for several months now and have fished all of the standard outings, plus some excellent and exciting ones from other RWFF members. Namely Geraldro, Oppimax, Richwest, Cutch, Fishing Musician and many others. I have a question re: fish intelligence and selectivity. I noticed while fishing some of the outings fish were not very selective, while others outings required very specific and shape and correct colored flies to catch fish. Since I am no electronics genius I am curious as to how the programer inputs the "feeding requirement" for these outings? Perhaps savvy people like Paul and Richard and other long time members can shed some light on this.

Case in point is the Upper and Lower Sacramento Rivers. Both are great outings and were beautifully done by two different people. I would consider the Upper Sacramento one of the most beautiful and challenging outings in the RWFF program. This outing requires very specific flies in the right size and color to catch fish. The Lower Sacramento is a much more "forgiving" river and requires less precision and the fish are much more willing and cooperative. I would compare these two rivers to the Yellowstone River in Montana, a great fishery that can produce 30 plus good rainbows and browns and cutthroats in one good day of fishing, to the Armstrong Spring Creek near Livingston, Montana, which is considered to be the Graduate School of Flyfishing by many avid flyfisherman. The Armstrong is gin clear, slow moving Spring Creek that requires stealth, exact imitation of hatches, and a perfect drag free presentation to achieve success. If one can catch two or three good fish at the Armstrong it would be considered a great day. It is amazing that this "artificial fish inelligence" can now be programmed and incorporated into the virtual fishing world. So..... how do they do it? Can anyone shed some light on this?

While I have been tying some specific flies for different outings and have had some success catching some large fish, I am just wondering how specific one has to be in size, color, shape, etc. to match the insects and minnows shown in the "food" table in the outings. Upper Sacramento is still the hardest one to catch a good fish and perhaps this is what makes this game so much fun. Would love to hear your thoughts on this.....how some virtual fish are smarter than we are! :D :D

Regards,

Stanley
Stanley (Yellowstone7)

Image
User avatar
Yellowstone7
 
Posts: 3083
Joined: Fri Sep 26, 2008 12:07 am
Location: Hong Kong (formerly Idaho & Montana)

Re: Artificial Fish Intelligence and selectivity??

Postby richwest » Fri Feb 13, 2009 10:46 am

Stanley,

Right now I'm at my job, I get back on this when I get home tonight.

Regards Richard
Image
User avatar
richwest
 
Posts: 803
Joined: Mon Nov 05, 2007 7:06 pm
Location: Woudrichem, The Netherlands

Re: Artificial Fish Intelligence and selectivity??

Postby paul » Fri Feb 13, 2009 11:12 pm

I can chime in on this one too. In each outing (or, if the outing developer wants to get more specific, each site) there are configuration files that determine a number of things about the fish. These include how active they are (likely to feed in general), whether they're cruising or holding in place, how easily spooked they are, and how selective they are. For each food item present the outing developer can also specify how interested the fish are in that particular food. The more selective the fish, the more precisely a fly must match the size and color of the naturals. If you get the match absolutely perfect, the fish will be just as interested in your fly as they are in the naturals. But if they're only moderately interested in a particular type of food, and/or if they're only moderately active, even a perfectly matched fly may still require patience and skill to catch fish. That's why it's a good idea to try several flies, matching several of the food types you find, until you have a good match for the food they seem to be most keyed in on.

Tight lines,

Paul
User avatar
paul
Site Admin
 
Posts: 936
Joined: Thu Nov 17, 2005 4:28 pm
Location: La Crosse, WI USA

Re: Artificial Fish Intelligence and selectivity??

Postby Yellowstone7 » Sat Feb 14, 2009 2:01 pm

Hi Paul:

Thanks for your very clear and precise answers to my question. So each outing, even sites, can be programmed in a very specific way as to how active and anxious the fish will feed on the recommended "food" items. This explains why I could catch a fish with a blue dun while the fish were feeding on orange duns in one site, and in other sites even a slight variation of the fly's size or color can make a difference between catching a fish, or having the fly just floating by untouched. This feature surely sets the RWFF apart, and ahead of other fishing simulators! Very interesting indeed.

This information will help when I try to get my son to produce a couple of nice outings from Idaho and Montana this summer. Thanks again for the great info. :D :D

Regards,

Stanley
Stanley (Yellowstone7)

Image
User avatar
Yellowstone7
 
Posts: 3083
Joined: Fri Sep 26, 2008 12:07 am
Location: Hong Kong (formerly Idaho & Montana)

Re: Artificial Fish Intelligence and selectivity??

Postby Yellowstone7 » Sat Feb 14, 2009 9:54 pm

Hi Paul:

I just found one of the unscrambled files in a recent new outing. Now I can see how the fish activity and various other factors are configured. Glad to understand this mysterious part about the rwff program. Great job! :D

Regards,

Stanley
Stanley (Yellowstone7)

Image
User avatar
Yellowstone7
 
Posts: 3083
Joined: Fri Sep 26, 2008 12:07 am
Location: Hong Kong (formerly Idaho & Montana)


Return to Real World Fly Fishing

Who is online

Users browsing this forum: Google [Bot] and 1 guest