V27's other half: slower respawns, slower chests, and the reward that went up
The last post checked a player's four gem numbers and found the
layer they came from: 13 tables with an _AB_ suffix that Version
27.0.0 added on top of the game's reward data. That raised the obvious follow-up.
If four payouts moved, what else did? Ten of the 13 tables reach things you can
feel directly, and the answers run in both directions.
The layer, in one paragraph
Version 27.0.0 ships 13 *_AB_ tables. Each row carries a
ref_id pointing back at a base table's id, an
ab_test_group, and only the columns it overrides. The client's remote
config selects group 1 (INGAME_RV_GROUP = 1), so the number you see in
game is the base row with the override merged over it. The base tables are mostly
untouched between builds, which is why anything reading only base data — including
this site until last week — still prints the pre-27 economy. The
gem-cut post walks through the
mechanism and the four gem payouts; this one is the rest of it.
Half one: the farm got slower
96 of the 214 monster spawn groups had their
RespawnCoolTime rewritten, and every single one got longer —
from ×1.1 to ×13.8, median
×5.0. The base table set each group's timer off its chapter, anywhere
from a couple of minutes to two hours. The override moves most groups onto a small
set of fixed cooldowns: 23 groups at 15 minutes, 31 at 2 hours,
15 at 6 and 6 at 9.
| Spawn group | base | now | longer |
|---|---|---|---|
CH2_Spawn_Spider_Boss | 26 min | 6 h | ×13.8 |
CH2_Spawn_Dragon_Elite | 39 min | 9 h | ×13.7 |
CH8_Spawn_UndeadElfQueen_Boss | 35 min | 6 h | ×10.3 |
CH15_Spawn_MagmaGolem_Boss | 36 min | 6 h | ×10.0 |
CH9_Spawn_Ragnaros_Boss | 59 min | 9 h | ×9.2 |
CH16_Spawn_SkullDragon_Boss | 1 h | 9 h | ×9.0 |
The biggest jump is CH2_Spawn_Spider_Boss —
26 min to 6 h, ×13.8. Look
at what those groups are: everything on 2 hours and up is an elite or a boss, while
the trash mobs land on the shortest timers. The rare spawns that carry the best
drops are the ones held back hardest.
Half two: chests take longer too
World chests were rewritten on two columns at once. Their reveal time went up on all nine rows — wooden ×5, rare ×6, legendary ×8:
| Chest | base reveal | now | longer |
|---|---|---|---|
| Wooden Chest | 2 min | 10 min | ×5 |
| Rare Chest | 5 min | 30 min | ×6 |
| Legendary Chest | 15 min | 2 h | ×8 |
In exchange, the reward bundle each chest rolls got bigger. All 9 chest rows point at a new reward group, and every one of those live groups carries reward rows its base group didn't. The catch is that it isn't pure addition: the second chest set (ids 4, 5, 6) also had its bluestone tier rows rescaled down to make room for the orb rows it gained, so "bigger" means different, not uniformly more.
Separately, ChestSpawnerData moves 12 stages' chest spawners
one set later — every chest id in the spawn sequence shifts by +3, so a stage that
spawned the first-tier chests now spawns the second-tier ones. If you know a stage's
chest rotation by heart, it changed.
The part that went up
Here's the exception, and it's a real one. 34 of the 71 7-day event missions point at a new reward group, and 33 of them pay more gems:
| base gems | missions | now |
|---|---|---|
| 30 | 14 | 100 |
| 50 | 4 | 100 |
| 60 | 15 | 100 |
| 150 | 1 | 100 |
Thirty, fifty or sixty gems become a flat 100. One mission went the other way: 150 → 100, on day 7. Across the changed missions that's 1,670 → 3,400 gems, and across all 71 missions, whatever their reward type, 3,170 → 4,900 — up 55%.
So is the farm better or worse?
Both, in different places. The 87 rewritten monster variants drop 893 → 5,335 gems in total, and the 24 variants that dropped gems before have a median ×4 per-kill payout now. Against that, the median spawn group waits ×5.0 as long. Those aren't the same rows, so it's not a strict per-hour ratio — but the two point in opposite directions and land at similar magnitudes.
And it isn't uniformly up: 2 bosses dropped gems in the base table and drop none now — Dark Lord and Hobchief. Both are listed on the Rewards page, in the same table as every other rewritten drop.
Add the sink from the gem post — 12 Key House slot unlocks repriced from orbs to gems, 5,570 gems for the twelve — and the shape of v27 is consistent: the repeatable income shrank, the time between payouts grew, per-kill drops rose to compensate, and one event got richer.
Two smaller movements
The rewarded-video roll had 1,672 of its rows rewritten. The gem slot went 35 → 10, and not every other slot went down: 623 of the changed rows paid nothing in the base table and now pay 3–33, against 1,048 that came down. The full per-type table is on the Rewards page.
What we still can't tell from the files: the override layer is shipped with one group, so there's nothing else to compare against, and a server can flip individual accounts without changing the APK. Your own numbers are the real check. If your respawns still look like the base table, tell us — that would mean group 1 isn't what your client is being served.
Where to check all of this
The Reward Payouts page now carries every table in this post — nine sections, every row side by side as base vs live, with a filter box. The Bestiary shows the live drop amounts on every monster card, and the World Map shows the chest spawners in place. If a number in this post disagrees with a page, the page is the source — every figure here is computed from the same tables it prints.
Sources: EnemySpawnGroupData,
ChestData, ChestSpawnerData,
SevenDayEventMissionData, RewardGroupData,
EnemyData, ChestSlotUnlockCostData,
IngameRVData and StackableItemData, plus their
_AB_ overrides, from the 27.0.0 extraction. Reward bundles
resolve through RewardGroupData by Group; base and
override rows are joined exactly as the site does it (ref_id →
id, group 1). XP Hero is developed by Supercent — XPHero Fans is
unofficial and not affiliated with or endorsed by them.