Keep the dependable return first
Queen Jindeok’s Harvest Song gives 20% extra resources from Alliance Resource Spots. Its 40% World Map return is a separate condition. Princess Lucy’s Inspiration supplies 15% gathering speed without that World Map restriction. This makes Jindeok main with Lucy assistant a sound starting pair; Lucy’s base contribution does not require specialty spending. These are the publisher’s listed skill values, not a promise that an already-developed march becomes 15% faster.
Changing their positions is not automatically neutral. The official assistant rules allow compatible Special Skills, Skill Books and Specialties on an assistant, but exclude that assistant’s equipment attributes, cultivated stats and red-star additions. Keep the developed main’s applicable investment in the comparison. Ascending a general to improve its assistant contribution buys no red-star benefit in that role.
Separate the base skill from the specialty
| Assistant | What the base skill does here | The investment decision |
|---|---|---|
| Princess Lucy | The generic 15% gathering-speed effect is the useful baseline. | Jungle separately lists up to 10% gathering speed. Price that next increment against the speed you already have; it is not another resource-return bonus. |
| Gaius Marius | The Marian Reforms’ 30% ore and 15% lumber/stone speed are explicitly for World Map spots. Do not credit them on alliance tiles. | Smelting separately lists 20% ore speed without that location restriction. An already-developed Gaius deserves an ore-specific comparison before replacement. This wording alone does not justify new Runestone spending on an unverified alliance result. |
| Constance I | Her 20% extra-resource base skill is World Map-only; the pairing matrix records a conflict with Jindeok’s base skill. | Overgrowth’s separate lumber-return effect uses “Outside City Resource Spots” wording. Its alliance applicability is not established. Do not buy her as another confirmed 20% alliance-return source. |
| Shimazu Yoshihiro | The 30% food-return clause requires World Map farms; his separate generic 10% speed remains relevant. | Bumper Harvest’s separate food-return specialty has the same unresolved alliance scope. A world-food yield total is not an alliance-food forecast. |
The named skills and specialties are listed in the official gathering-general guide. The specialist gathering notes also distinguish Shimazu’s retained alliance speed from his lost world-farm return.
A grey base skill does not automatically turn off a different Specialty. A firsthand Constance test found Overgrowth working behind Jindeok on a normal level-4 world lumber tile despite the base-skill conflict. That is useful evidence about separate activation; it does not prove alliance return. Keep those two conclusions separate.
Nor does reversing the pair solve everything. The specialist compatibility matrix records Jindeok’s base skill as disabled when she assists Constance. Do not reverse them assuming Jindeok’s alliance return will survive; that alliance-specific reversal is not established by a current test here.
Do not move the ore book for the wrong bonus
The Guide Wiki’s general pairing matrix marks Ore Gatherer as assistant-only when Gaius assists Jindeok. Its separate book guidance lists level 3 or 4 for Gaius. The matrix marks the other ordinary resource-speed books as main-or-assistant in its documented pairings.
That is compatibility guidance for preserving a working world-tile pairing, not a current alliance-tile test. It does not establish a new alliance-only placement requirement. Do not move books merely to restore Gaius’s World Map-only base skill for an alliance march.
Ordinary Gatherer books increase speed; the maintained book reference lists 45% at level 4. Do not confuse them with Exploitation books, which concern extra return.
A firsthand 2023 Gaius test found that level-3 Exploitation produced no alliance bonus, despite working on world tiles. That historical test is not proof of every current version, but it gives no basis for paying for an assumed alliance-yield increase. Preserve a working world setup rather than moving scarce books on that assumption.
The expensive question is the next specialty
Queen of Silla’s generic extra-resource label does not establish another 20% alliance return. One third-party Jindeok guide lists a fully developed total that still gives only 20% on alliance spots. The scope is unresolved; do not treat either a universal bonus or a universal exclusion as proved. Jindeok development reference
That matters because the fourth specialty requires the first three to be maximized. Do not fund the entire gate solely for an alliance-return increase that has not been established. Count only your actual remaining costs when comparing an already-developed general with a fresh one.
The practical order is short: preserve Jindeok’s active alliance return; use Lucy’s available base speed; keep an already-developed Gaius in the ore comparison; and leave world-only or unresolved yield effects out of the alliance forecast. For a continuous alliance stay, compare collection rate and total return over the same gathering interval. Unlike clearing a finite world tile, faster collection can add resources during that same stay without another dispatch.
The useful upgrade is the effect that works in this slot, on this tile, at a cost your existing build can justify. A famous gathering name has not completed that calculation for you.
— 
Kat