01
Match the upgrade to its action
No universal best upgrade order is verified for Wanderburg in the supplied pre-launch evidence. The most useful documented distinction is between improvements to an automatic action, an ability and the interval between repeated activations. Historical cannon and fire builds illustrate different priorities, while current Early Access values remain 待确认.
The taha.log() guide for Demo 1.5, published August 8 and updated August 9, 2026, distinguishes Automatic, Ability and Cooldown. Its account also warns that rarity does not replace compatibility with the build. Read which part of your configuration an option affects before treating a rarer choice as the better one. This is a reading framework drawn from that Demo account, not a claim that every module shares the same upgrade structure.
Steam describes between-run unlocks as an expansion of the combinations available to the castle. That is a different decision layer from improving an action during the current run. Keep the question of what you can choose next time separate from what the present upgrade actually changes. Current costs, reroll conditions and complete upgrade pools are 待确认.
02
The historical cannon example
Failloop's March 6, 2026 video description and automatic transcript describe a cannon configuration involving a Pirate captain, Side Cannons and Magma Core. At 00:00:18–00:00:24 the recorded commentary says the captain increases cannon damage on both sides, then describes choosing Magma Core at 00:00:41–00:01:00. The creator discusses selecting automatic cannon damage again at 00:03:13 and 00:09:20. This is commentary from one pre-Early-Access video with an unidentified exact build, not current percentage values or a frame-by-frame verification.
The important tradeoff is that the creator describes an enemy benefit as well as a player benefit. Dropping the enemy side would turn a conditional cannon example into misleading free-damage advice. The Demo 1.5 guide also associates Magma Core with cannons, but the two sources cannot be aligned to the same patch. That supports discussing the pairing historically, not calculating a present-day damage multiplier or declaring it the best artifact choice.
03
The historical fire example
The automatic transcript of BaronVonGames' December 7, 2025 pre-release video names Norbert, Fire Wizard and Carpenters at the opening. In the recorded commentary, the creator prioritizes reducing the fire wizard's cooldown at 00:00:28, discusses improving the repair crew at 00:04:23 and compares fire-ability variants at 00:06:30–00:06:56. This described route combines repeated activation, recovery support and an ability choice rather than following the cannon video's automatic-damage emphasis. The exact test build and any equivalent Early Access timing are 待确认.
Read that sequence as a change in the question being answered by each upgrade. Cooldown concerns the repeat interval in the Demo guide's terminology, whereas the later repair choice concerns support for continuing the run. Neither the observed order nor the video's enthusiastic title proves that cooldown always beats damage. Old cooldown seconds, stacking formulas and minimum intervals have not been established for the planned Early Access version.
04
Compare choices without a false ranking
These examples suggest three useful comparisons: the affected action, the existing equipment match and any attached drawback. For the cannon case, retain the enemy-damage caveat; for the fire case, distinguish cooldown investment from the later recovery decision. Those questions preserve the reasons described in the commentary without inventing a global priority list. They are a way to interpret choices, not a tested promise that a particular selection will win the run.
This page uses the morning September 8, 2026 Beijing snapshot, before the planned 21:00 Early Access opening. Actual release and current upgrade balance were 待确认, and the December and March videos do not supply a synchronized comparison. The evidence therefore supports conditional examples, not a DPS table or a universal first-pick recommendation. Keep the version and the action affected beside any upgrade advice you take from these runs.