Optimizing Pokeshell morph chambers. What is the best layout to maximize morph egg chance?

[blakemw]

You may have noticed that critters have a bias towards resting at the ends of paths. This can be used to influence Pokeshell morph egg chance, which increases only when a Pokeshell is in the appropriate liquid of adequate depth. Because grooming stations don't work when submerged, we either need the grooming station built on dry tiles or some kind of complicated scheme to drain the chamber to permit grooming, but it'd be better to have the grooming station on dry tiles and manipulate the critter's pathing.

Here there are 4 chambers, each with 2 dry tiles and 2 ethanol tiles but with a different topology, and each with 4 Pokeshells. Starting with 0% Oakshell egg chance, I let the simulation run for 5 cycles, and noted the Oakshell egg chance after 5 cycles.

 

(B = Block, ~ = liquid)

1. ~BB~ : 44%, 44%, 44%, 44% 
2. B~~B: 0.02%, 0%, 0%, 0%
3. BB~~: 2%, 3%, 0.9%, 2%
4. ~~BB: 2%, 4%, 4%, 2%

The numbers were overall highly consistent between the 4 Pokeshells in each chamber, only the version with the liquid at the path ends produced meaningful morph egg chance while the version with dry tiles at each end produced no morph egg chance at all.

And the versions with dry tiles at one end produced a tiny amount of morph egg chance, possibly with left side liquid resulting in a slightly higher egg chance, but also quite possibly just statistical noise.

Is it possible to improve upon perfection? How about some other layouts, and repeating the best layout for control purposes:

 

1. Floating Island: 7%, 6%, 7%, 8%
2. Control: 43%, 43%, 44%, 44%
3. Deeper pit: 44%, 44%, 44%, 44%

We can see here that the layout without cleanly terminating paths ("circular paths") performed much less well than the control, though it performed a bit better than paths cleanly terminating on dry tiles. It seems the most important thing by far, is just that paths terminate in liquid, with it not mattering measurably how deeply in the liquid the path goes.

A final thing to test. What if we give even more terminating paths in liquid? Logically this shouldn't help much because if all paths terminate in liquid it wouldn't seem easy to improve upon perfection, but lets try some more complex layouts anyway.

1. Deep: 44%, 43%, 44%, 43%
2. Control: 44%, 44%, 44%, 44%
3. Double: 36%, 34%, 34%, 33%

I had some hopes for the "double" layout, but it performed significantly worse. Why? From further observations, of the "control" layout but with a wider island, it seems that Pokeshells have a minimum pathing distance of 5 tiles before they can stop unless stopped earlier by the path terminating. Hence if there are 5 tile long paths that end on the island Pokeshells can stop on the island.

 In this video it can be seen that only with the 4 tile wide island do Pokeshells ever stop on it. On the shorter islands they are only transiting across it.

Conclusion:

If you want Oakshell or Sanishell morphs, always have path ends terminate in liquid and the grooming station / fountain on a non-floating island rather than at the end of the chamber. 

It seems that there are two possible layouts which are perfect, and it is likely that all other layouts perform considerably worse unless you give pokeshells MUCH more liquid to hang out in.

The two perfect layouts for maximizing morph chance that involve no "hanging out" on dry tiles and probably can not be improved upon.

[LadenSwallow]

I think 1 thing I tried was to lock doors when only 1 pokeshell was out of the desired liquid to force evolution chance on the other pokeshell. Relying on grooming/feeding the other trapped pokeshell to trigger dupe access which would unlock the doors for the others, who would then go to feed and then get groomed also.

[Sasza22]

Interesting. I`s cool that it takes just one tile so a dupe can jump over it when entering the room. I wonder how would it work with bigger rooms with more liquid on the far side. I might try that later.

[asurendra]

Well, 44% is still quite low. Critter Flux-O-Matic offers 50% of needed morph. 

Or you can make complicated designs like " groom critter for 4-5 cycles than flood room with necessary liquid". That will force you to wait longer a bit but guarantee about 90% chance of necessary egg

[blakemw]

These numbers are for only 5 cycles starting at 0%. The Morph chance will get much higher left for longer. By the time a Pokeshell actually lays an egg the perfect designs will have reached nearly 100%.

Also if you are ranching morphs they need to be "maintained" with liquid.