Repository navigation
Add ORAS wild encounters for Routes 101–110 - #1732
Conversation
Grass, surfing and Old/Good/Super Rod encounters for Omega Ruby and Alpha Sapphire on Routes 101-110 (Route 104 on its north-oras and south-oras areas), generated from PKHeX's Gen 6 encounter tables. 26 encounter_slots rows (shared per-position slots, numbered from 1) and 424 encounters rows, appended. Replaces the Routes 101-109 part of PokeAPI#1542 by ConnorMolz, which these rows match once its review fixes are applied.
jemarq04
left a comment
There was a problem hiding this comment.
Thank you for your work with this! I've taken a look at the existing changes, and everything looks correct and clean. To respond to your questions:
- Route 107 underwater. PKHeX has a Dive table for Route 107 (Chinchou
50%, Clamperl 30%, Lanturn 15%, Relicanth 5%, Lv 25–35; the same in
both games, and the same as #1542). PokeAPI has underwater areas only
for Routes 124 and 126, from the Gen 3 games, and none for ORAS. Should we
add an underwater area to hoenn-route-107, like those two, or put Dive
on the surface area as #1542 did? Later chunks have the same question for
Routes 128, 129 and 130, and whether ORAS rows on 124 and 126 should reuse
the existing underwater areas. The 24 rows are ready either way.
This just looks like we're missing a location that was unique to ORAS. If you're willing, please add a new location area for the location hoenn-route-107 (along with its prose) and the missing encounter data. If you have any questions on these things, I'm happy to answer questions as well.
- Slot numbers from 1 or 0? This follows the existing ORAS rows and the
example in the #1542 review (from 1). X/Y numbers its slots from 0.
I'd say start from 1 as you have, that feels more consistent with other games we've done.
- Shared slots. This adds one set of slots reused by every area, as X/Y
does, rather than a set per table with summed rates (as ORAS's hordes and
#1542 do). OK?
Shared slots is preferred – I recently added information for SWSH but couldn't find distinct slots so I made one for each area/method, which became a bit large. If the slots can be shared, please add them as such.
- Route 104 rows go on the existing north-oras and south-oras areas,
which have different grass and match PokeAPI's hordes, not on the unnamed
Route 104 area that #1542 used.
Yes, we want them split like that.
- Northern Route 104 Zigzagoon horde: fix PokeAPI's level (2 → 3) here,
or separately?
Please do – this is a quick fix and related to ORAS encounters so it's not out of place. I'll check the changes when you apply the fix.
Separately, I think #1542 has gone stale. I'll close it shortly, but feel free to add in the town encounters as well if you're willing.
Once you've updated the PR with these comments, I'll check again! Thanks!
From the PokeAPI#1732 review: a new location area for underwater Route 107 (id 1667, with en/de/fr names like Routes 124 and 126) holding its 24 ORAS Dive (seaweed) rows on 12 new shared slots, and northern Route 104's Zigzagoon horde corrected from Lv 2 to Lv 3 (rows 70606, 70609).
|
Appreciate the review @jemarq04 ! New updates:
|
jemarq04
left a comment
There was a problem hiding this comment.
Looks great, thanks a bunch!
Change description
Part of #1456. Adds the ordinary wild encounters Omega Ruby and Alpha Sapphire
are missing on Routes 101–110, in both games:
walk)surf)seaweed)That is 38 new
encounter_slotsrows (ids 3811–3848) and 448 newencountersrows (ids 123041–123488), all appended. Two more changes came out of review:
hoenn-route-107/underwater, game index 0 like the other ORAS-onlyareas), with English, German and French names following Routes 124 and
126's underwater areas: "Road 107 (underwater)", "Route 107 (unterwasser)",
"Route 107 (sous l'eau)".
70606 and 70609, one per game), to match the game.
Other existing rows (hordes, gifts, statics) are untouched.
Not included:
the Add oras surfing and fishing encounters (Part 1) #1542 review DexNav has no known per-slot rate.
Slots
Each new slot is a table position shared by every area, as X/Y
(version group 15) and Emerald already do: one
encounter_slotsrow per(version group, method, slot number) holding that position's rate. Each
encountersrow is one position in one area's table, so a Pokémon in twopositions has two rows, and their chances add up in the API. Slot numbers
start at 1, like the existing ORAS horde rows (see question 2).
walk,seaweedsurfold-rod,good-rod,super-rodSource
The species and levels come from the wild encounter tables in
PKHeX by kwsch and contributors.
Those tables are dumped from the games. We used
PKHeX.Core/Resources/legality/wild/Gen6/encounter_or.pkl,encounter_as.pklandtext/locations/gen6/text_xy_00000_en.txtat commit6c9583883c7551ac74a4b74180cead249d4277b8. We read the files with our owndecoder, took only the facts in them, and used none of PKHeX's code. Thanks to
the PKHeX maintainers.
The rows are generated, not typed. The generator maps each PKHeX table to a location area. It fails on
any table with slots that it can't map, and produces the same rows on every run.
How the rates were derived and checked
PKHeX doesn't store rates; in Gen 6 a slot's position sets its chance. The
rates above are those positional rates:
Before this work it was also confirmed against Bulbapedia's tables for 11
ORAS grass, cave and Dive tables.
(see below). The surfing rates also match X/Y's in PokeAPI.
Checks run on this branch:
(species, summed rate and level range per game, area and method). These are
Add oras surfing and fishing encounters (Part 1) #1542's 66 tables on these routes; its Route 104 tables are on the unnamed
area, so they were compared with both halves. Without the fix, Route 103's
surfing would differ. That fix moves it to 65/30/5, which is what the slot
positions give.
101, 102, 103, 104 north and south, and 110, both games), PKHeX's horde
table has the same species and levels. The one exception was northern
Route 104's Zigzagoon horde (PokeAPI Lv 2, the game Lv 3), which this PR
now fixes.
(grass, including the OR Seedot / AS Lotad split, surfing and all
rods), 104 (both halves: grass, surfing and rods, plus hordes confirming
which PKHeX table is north) and 110 (grass, including OR Minun / AS Plusle,
and rods) all agree in species, levels and rates. I then checked Routes
101, 102, 104 north, 105, 108 and 110 again by hand, every method, side by
side with Bulbapedia. Everything agrees except Route 110 surfing (below).
make build-dbandmake testpass (69 tests). The new rows show up at/api/v2/location-area/1207/,/api/v2/location-area/402/and the new/api/v2/location-area/1667/.Known mismatch: Bulbapedia's Route 110 surfing
If you compare Route 110 with Bulbapedia, its surfing won't match. This PR
has Tentacool 65%, Wingull 30% and Pelipper 5%; Bulbapedia's Route 110 page
has Tentacool 60% and Wingull 35%. I kept the PKHeX rates, because:
surfing table on all three (Tentacool in slots 1 and 3, Wingull in slot 2,
Pelipper in slots 4 and 5, same levels). Bulbapedia's Route 103 and 105
pages give 65/30/5 for it. The same table can't have different rates on
different routes.
are 50, 30, 15, 4 and 1, and no combination of them adds up to 60.
Tentacool's two slots give 50 + 15 = 65.
So Bulbapedia's Route 110 page most likely has a typo. This matches #1542's
review, which corrected Route 103's surfing to 65/30/5.
Relation to #1542
This replaces the Routes 101–109 part of #1542 by @ConnorMolz, which added
these routes' surfing and fishing by hand and has had review changes outstanding
since June 2026. Thanks to ConnorMolz: #1542's tables were the check for these
rows, and its review set the rules followed here (slot numbers, no 60/35/5
guess for PokéNav).
#1542 is being closed as stale. Its town tables (Oldale, Petalburg, Rustboro,
Dewford, Slateport) will follow in a separate PR. Its PokéNav rows stay out,
since they'd need a new method.
Settled in review
north-orasandsouth-oras.AI coding assistance disclosure
I used Claude Code to help build out the changes here and do the grunt work of copying values from PKHeX to PokeAPI. Also built a small electron app to let me do side by side comparison with Bulbapedia to check work.
Contributor check list