Talk:WaterPort

Reverse engineering WP populations

Joe let this little info slip: "When using an 80/80 Oak pole with FG, you have an 9.45% chance of catching the Acorn Slug"

Not sure if his info is correct, but if it is, it may be possible to reverse engineer/guess the rest of the populations (note, Acorn is not in top-4 for additional pole 50 bonuses).

	
Miny: 		Fairly Prolific (in WaterPort) - 9.45%
Miss Miny: 	Fairly Prolific (in WaterPort) - 9.45%
Catty:	 	Fairly Prolific (in WaterPort) - 9.45%
Cubey:	 	Fairly Prolific (in WaterPort) - 9.45%
Loafy:  	Fairly Prolific (in WaterPort) - 9.45%
Acorn Sea Slug: Fairly Prolific (in WaterPort) - 9.45%
total= 56.70%

Rest (43.30%) is divided for:

Hippie: 	Numerous (in WaterPort) - A%
Mime: 		Rare (in WaterPort) - B%
Blossom:	Rare (in WaterPort) - B%
Vampire: 	Very Rare (in WaterPort) - C%
Sea Tortoise:	Very Rare (in WaterPort) - C%
Twig:	 	Critically Endangered - D%
Woodpecker:     Near Extinction (in WaterPort) - 2xE%  (prefers FG)
Ninja:	 	Possibly Extinct (in WaterPort) - F%
Poison Oak: 	Possibly Extinct (in WaterPort) - F%

So, A + 2B + 2C + D + 2E + 2F = 43.30
And A>B>C>D>E>F

It is possible that there is a fixed or proportional increase between the populations, that stays constante. (ie. maybe A=2xB or A=B+1 .. etc)

According to Liquid Gold harness list, only Poison Oak qualifies for the top-4 bonus for poles greater than level 50 (not sure if this even applies anymore).. so we may be able to disregard this.

Leaving this now if someone else wants to take a stab :-)

--Jayberwock 19:01, 27 August 2009 (UTC)

The complete population list is (from FAQ):

Non-Existent (Rarest Population)
Possibly Extinct (2nd Rarest Population)
Near Extinction (3rd Rarest Population)
Critically Endangered (4th Rarest Population)
Endangered (5th Rarest Population)
Threatened (6th Rarest Population)
Extremely Rare (7th Rarest Population)
Very Rare (8th Rarest Population) 
Rare (9th Rarest Population)
Average (10th Rarest Population)
Numerous (11th Rarest Population)
Fairly Prolific (12th Rarest Population)
Overflowing (13th Rarest Population)

Notice the gaps in the waterport populations, if going with the "constant increase" between classes theory ...

--Jayberwock 19:11, 27 August 2009 (UTC)

If we assume a constant multiplier, then a quick spreadsheet test makes each population ~86.4% of the previous.. this is not a very round number, so the populations are probably not directly related. They may just be arbitrary constant numbers :-(

--Jayberwock 20:58, 27 August 2009 (UTC)

If anyone is interested in playing with the numbers I can attach the spreadsheet I used for testing. I guess the only interesting point is that no matter what kind of approximation is used, the populations are not as far apart as you would think -- for example using a preferred chum for a fish will make it more probable than another fish in a "higher" population class (Woodpecker should be more probable than Twig, if using FG).

I guess next step to reverse engineer would be to collect a large sample set (5000+ casts) of data using FG and Oak 80/80 in WP. When we already know the expected rate for fairly prolific fish, the other populations can be fairly accurately estimated. (this would also uncover any code bugs if the fairly prolific fish fail to get into the ~9.45% range).

--Jayberwock 16:01, 29 August 2009 (UTC)

Have been tied up with a few other things IRL lately, but have two points to contribute:

The first point is regarding rarity of fish IN THE SAME CATEGORY.

When Joe released 13 different fish populations, my first thought was that he would make the fishes in each respective category equally common. However, that was obviously not the case, since when first released, most of the level 73 and 80 were BOTH "Possibly Extinct", at least for the Woodpecker, Sniper and Merlin. (The Mechanic was released as "Near Extinction" I think). They are quite obviously not as rare as the level 80s though, so we have one additional, annoying thing to contend with: that some fishes in the same category in fact have differing rarity too :)

The descriptions have since been changed for the Woodpecker, Sniper (I think?), and the Sea Croc was released as a "Near Extinction" and later "Crit Endangered". The easy explanation would be that the descriptions were WRONG initially, and they were in fact coded as "Near Extinction". But the Merlin is still "Possibly Extinct" last I checked. My gut tells me that Joe made the level 73s slightly easier to catch though, due to the relentless complains.

My second point is collection of data. If we were to concentrate on getting 80/80 OB, Fish Guts, WP, we have a ready-made HUGE pool of data for us, in the people doing OB #4. I think it'll be very possible for us to call for them to paste their log somewhere, but is there an easy way of tabulating the data? :) I'm not sure how the Mousehunt people do it, but they direct people to paste data in their log summarizer, and are able to come up with approximate catch rates once they get a big enough sample size.

I suspect the catch rate percentages are an arbitrary number too :/ He probably just reduces the rates of the common fish whenever he adds a new fish. (Like when he adds 4 new fish totalling 5%, he'll just reduce the rates of the common ones 5%, divided over the number of common fish)

202.63.155.240 04:19, 31 August 2009 (UTC)Zach

Fish Populations: Clues from the "Harnessing Liquid Gold" page?

I was in WP helping a friend out, and stumbled upon the Liquid Gold page. You mention "Only Poison Oak qualified for top4 Rare fish", but with my OB equipped, this is what I see:

• 1. Wanderer, Possibly Extinct
• 2. Poison Oak, Possibly Extinct
• 3. Ninja, Possibly Extinct
• 4. Mechanic, Near Extinction
• 5. Woodpecker, Near Extinction
• 6. Twig, Critically Endangered
• 7. Houdini, Endangered

I suspect Joe coded OB and BS together somewhere, which is why both set of fishes are appearing here.

Looking at the number of Wanderers and Poison Oaks caught to date (1300+ to 1600+, ~25% more), does anyone think it's safe to assume that Wanderers are rarer than Poison Oaks? Or is that down to more people attempting OB4 than BS4?

Logically, I would think that if the fishes were equally rare, they would be listed according to alphabetical order? Which is why my guess is that Woodpeckers are now exactly as common as Mechanics, but is lower because of alphabetical order. (They were obviously rarer initially; there were consistently many more Mechanics caught than Woodpeckers).

Has anyone tried this in other places? I'll be trying later in BC, MR and SC and report back.

Of course, these are just small and rather inconclusive clues at best. The best way to test would still be to obtain a 10,000+ sample size of casts, possibly more.

--Zach C 10:36, 31 August 2009 (UTC)

I know this convo is years old, but I still would love to be able to reverse engineer some of this. The problems however are that there are so many possible ways this could have been coded and there are still so many things that are not really known (such as pole strength impacts catching percentage, although we currently know several other boost types guessing there are still several more we dont know much about and even ones we do know info about still could be coded in so many different ways). For example we know top 4 fish get boost with max pole, but is that calc to overall percent; calc to fish base percent prior to its overall percent?

At this stage of the game, getting huge numbers of sample data is less likely too. A test sample size of 10,000+ is usually considered a decent starting point when talking about low double digit percent values. But really the more data the better ofc, and 50,000+ is more realistic sample size. The lower the values requires even much bigger samples. (ie: I have seen games that had 50k - 100k data and thought they had nailed down 2% value, Dev's later released info that it was really 1%. Although doesnt sound like much made huge difference as that item once thought to be one best became mediocre at best.)

Anyway, if we could collect some decent sample sizes, we might be able to start eliminating some possible ways this is coded. But would that still be enough to help narrow it down I really dont know. Just couple of possible ways off top of my head. I would think Fish have a base value however even that could be done many ways:

1) Assigned based on Population density simple values like Jayberwock implied, or

2) Assigned with range values on Population density like Zach C suggested. Where for example Possibly Extinct could be anywhere from 1-3%; Near Extinct from 4-6%, etc, or

3) Each fish assigned its own value which is than multiplied by Population Density, which could explain why Wanderer (possibly extinct) appears more or less rare than similar Poison Oak (possibly extinct), etc. (Joe has mentioned several times fish being 1 in 300 chance or 1 in 15 chance which could be its base value OR could be its overall chance, we dont know, but if its the fishes base would easily explain odd values like 0.45)

4) It could be something as simple as 1/most_populated_density. Overflowing would be 1/1; Fairly Prolific is 1/2; Numerous is 1/3; etc; Possibly Extinct is 1/12; and Extinct 1/13 (although as Joe specifically list this in reverse by most rare, unlikely but still possible)

Sure are many others too not springing to mind atm. Kevin "Hawk" Fisher (talk) 08:26, May 20, 2017 (UTC)