Boosters and Buffs
Timed boosters and permanent buffs: what they multiply, what they cost, and the level at which a booster stops paying for itself.
Two systems that share one word
Two different things on this server get called a boost, and they behave differently enough that it is worth separating them before anything else.
A booster is a timed, levelled purchase. You buy one level of one booster, it runs on a wall clock, and when the clock runs out it stops. The catalogue holds 13 boosters and 11 of them are enabled. Boosters live under /boosts.
A buff is a permanent tiered purchase. You buy a tier and you keep it; nothing expires. Buffs live under /buffs (the alias /buff works too). The price of tier n is base cost x multiplier^(n-1), and both the shop card and the charge read that same formula, so the number on the card is the number you pay.
Both systems have a personal axis and a faction axis. Personal purchases leave your own wallet and follow you. Faction purchases leave the faction treasury and apply to the whole faction, so if you leave the faction you lose that half immediately while your personal rows carry on untouched.
The booster ladder shipped in update 1.5.0 and the plugin's own status line for it reads code-written, not server-verified. Nothing in the 2.0 line changed it: the catalogue still holds the same 13 rows with the same 11 enabled, the ten levels, the same nine effects and the same caps. Every figure below is the figure the code and the catalogue file contain. None of it is a report of measured behaviour on the live server.
The ladder, and where it stops paying
Every enabled booster has 10 levels. Level n grants bonus-step x n. On the income ladder the step is 50 percent, so level 1 is +50% and level 10 is +500%.
Prices are derived rather than hand-typed. Three settings produce the whole table: an assumed baseline income of 12,000 coins per hour, a cost ratio of 0.55 and a cost curve of 1.15.
break-even(n) = 12,000 x bonus(n) x hours, which is the most that level can possibly return.cost(n) = break-even(n) x 0.55 x 1.15^(n-1).
For the one-hour job income ladder that gives:
| Level | Bonus | Cost | Most it can return | Cost minus return |
|---|---|---|---|---|
| 1 | +50% | 3,300 | 6,000 | -2,700 |
| 2 | +100% | 7,590 | 12,000 | -4,410 |
| 3 | +150% | 13,093 | 18,000 | -4,907 |
| 4 | +200% | 20,076 | 24,000 | -3,924 |
| 5 | +250% | 28,859 | 30,000 | -1,141 |
| 6 | +300% | 39,825 | 36,000 | +3,825 |
| 7 | +350% | 53,432 | 42,000 | +11,432 |
| 8 | +400% | 70,225 | 48,000 | +22,225 |
| 9 | +450% | 90,853 | 54,000 | +36,853 |
| 10 | +500% | 116,090 | 60,000 | +56,090 |
The crossover sits between level 5 and level 6. Levels 1 to 5 can pay for themselves if you actually play the hour you bought. From level 6 upward the booster costs more than it can return in the same time, and level 10 costs 193 percent of its own break-even. That is the design, not an oversight: the top of the ladder is a prestige sink, a way to burn a large balance for a visibly large number, and it is not a route to earning faster.
Read the return column as a ceiling rather than a forecast. It assumes you earn the full assumed baseline for the entire duration, with no travel, no idling and no diminishing returns, which is not what an hour of play usually looks like.
The nine things a booster can change
A booster row picks exactly one effect from a closed set of nine. Each effect has one place in the code that reads it, so a booster with no reader cannot be sold by accident.
| Effect | What it changes | Notes and limits |
|---|---|---|
| Income | Multiplies coin payouts | Applied to the gross, before the tax is quoted, so your faction's cut rises with the boost |
| Job XP | Multiplies job XP | Capped at x6.0 across all active boosts |
| Cooldowns | Shortens job action cooldowns | Floored at 0.35 of the base, so at most 65 percent off. The action bar reports the real percentage |
| Side talents | Multiplies secondary-trait gain | Applies at the trait increment itself |
| Harvest yield | Raises the yield of the custom crop line | Custom crops only, not vanilla farming. It sits at the crop funnel's yield computation, which is the same funnel the Garden's species are harvested through |
| Find chance | Raises reward find rolls in rifts | Resolved per payee, so a group split is per person |
| Orders | Multiplies the coin ask on a writ | Applied before the writ fund reserves the money |
| NPC output | Raises your hired workers' daily output caps | Keyed on the NPC's owner |
| Fee discount | Cuts the market listing fee and forge repair fees | Capped at 25 percent |
The two rows that ship switched off
Two catalogue rows exist with their keys and their reasons written into the file, and both ship disabled rather than being sold in a broken state.
- Prospector would multiply ore yield, but ore has four separate drop authorities, and a single Tunnelling Pick swing dispatches four real block-break events. A per-event multiplier would therefore be a fourfold bonus on one swing, and the dupe audit would register the result as duplication.
- Treasure Sense was written against Ancient Ruins, and it shipped disabled because there was never a player-facing roll for it to hook into. Update 2.0.1 then deleted Ancient Ruins outright, along with
/ruin, the ruin loot table and the Ruin Key recipe, so the content the row was written against no longer exists. The row stays in the catalogue and stays switched off, and no booster on sale raises find chance anywhere except in rifts.
What the Garden did and did not add here
The Garden did not add a tenth effect and did not add a booster row. What it changed is what some of the rows already in the catalogue now reach.
- Its six new professions, Botanist, Florist, Beekeeper, Arborist, Barista and Chef, are ordinary jobs. The income, job XP, cooldown and side-talent boosters reach them exactly as they reach the other fifteen, because those effects are keyed on the job action and not on a list of job names.
- The harvest-yield effect sits in the crop funnel, and the Garden's species are harvested through that same funnel. The funnel now carries its own multiplicative terms for quality, season or greenhouse, and skill, recorded against the count the earlier steps produced and floored at it, so a term can raise a harvest but never cut it below what it would otherwise have been. A working greenhouse replaces the season term instead of stacking with it.
- Those terms ship pointed at the Garden species only. The retrofit switch is off, so hemp and cannabis keep the arithmetic they always had.
- The terms are gated on credit, so a hired NPC farmer's automated sweep is not multiplied. That is the same rule that already keeps the crop-yield buff and the Fiber Hands trait off an NPC harvest: the NPC is a throughput device, not a multiplier device.
The growing layer itself, the species, the soils, the greenhouses and the bees, has its own articles. This page only covers where a purchase touches it.
Which earnings the income multiplier reaches
The income multiplier lives at exactly two places in the payout code, and it is keyed on the reason a coin was paid rather than on the system that asked for it. That is why the same booster covers earnings that look unrelated to each other, and why an earning the catalogue does not name silently receives no boost at all.
| Earning | Boosted |
|---|---|
| Job actions | Yes, and that now includes the Garden's six new professions |
| Loot Hoarder coins on a mob kill | Yes |
| Cartographer chunk coins | Yes |
| Company labour while you are clocked in | Yes, if the operator leaves that switch on |
| Crop sneak-sales | Yes, and for a Garden good inside a daily budget |
| Admin shop sales | Yes |
| NPC worker output sold on your behalf | Yes |
| Company sales | Yes |
| Writ payouts | Yes, through the separate orders effect |
| Savings interest | Never |
| System grants and event prizes | Never |
| The post-vote coin drip | Never |
The three permanent exclusions are a deliberate safety line. A per-minute drip that pays you for being online has no player action standing behind any individual coin, so multiplying it would mint money for nothing. Interest and grants are the same argument.
The Garden's daily sell budget
Selling a Garden good by sneaking is the one genuinely unbounded way the Garden could mint coins, and update 2.0.1 bounded it. Every Garden good draws on one shared budget of 25,000 coins per player per day, reset on the UTC day, so you cannot raise your own ceiling by spreading a day's selling across several crops.
Three details matter if you are running an income booster alongside it.
- The fill is partial and counted in whole items. A stack that straddles the remaining headroom sells the items that fit and leaves the rest in your hand. It is a cap, not a wall, so a spent budget does not refuse the whole sale.
- The budget is asked before anything is taken or paid. That read books nothing, so a refused sale spends no budget.
- Selling into your own company does not touch it. If you are clocked in and the good is inside your company's charter, the consignment funnel takes the goods and the company pays, so the sale is your employer's transaction and is not charged to your personal daily limit. That ordering was wrong in the release that introduced the budget and was corrected: a spent budget used to block the company route entirely, and a partial fill used to destroy the rest of the stack.
For the first release the budget existed it bounded nothing at all, because nothing in the sale path called it. The class had a full set of green self-checks the whole time, every one of them asking a question about the budget's own arithmetic and none asking whether anything used it.
Boosted coins are still taxed
The multiplier is applied to the groß figure before the tax is quoted, which means your faction's cut scales with your boost by construction rather than by a second calculation. If your faction taxes crop sales per item, that tax rises in step with your income booster.
One thing that does not scale: the faction income share paid into the treasury on job actions is computed from the unboosted figure. That is the safer outcome and it is deliberate.
Writ coins take a different route. They never pass through the tax service, so a boosted writ payout is untaxed, and the writ fund has a hard ceiling of 2,000,000 coins above which nothing is minted at all.
Stacking, caps and what happens when things end
Two boosters carrying the same effect add together additively, and then one clamp is applied to the sum. The purchase card names the relevant cap before you click, so you can see whether a second purchase would actually do anything.
| Limit | Value | Meaning |
|---|---|---|
| Income cap | x6.0 total | +500% is the ceiling of the whole system, not of one booster |
| XP cap | x6.0 total | Same shape as income |
| Generic cap | x6.0 total | Applies to the remaining effect kinds |
| Cooldown floor | 0.35 | A cooldown can never drop below 35 percent of its base |
| Fee discount cap | 25 percent | The booster's share of any fee reduction |
| Banked time | 8 hours | The most time one booster can hold. The purchase refuses before charging past it |
Endings and edge cases
| Situation | What happens |
|---|---|
| The boost expires while you are offline | It is dropped when you next join. Expiry is an absolute time, so downtime is not refunded |
| You leave your faction | You lose the faction axis at once; your personal boosts are untouched |
| Your faction disbands | The faction boost goes with it |
| You are clocked into a company | The boost still applies. The coins reach the firm and you see roughly 10 percent back through the production bonus. The action bar shows the boosted figure credited to the firm |
| You are employed but off the clock | Off-clock income is discarded, so the boost is worth nothing in that window |
| You die | Nothing. A boost is time, not carried state |
| The boost expires mid-action | The payout reads the multiplier at payout time. One read, no caching |
| You buy while the anti-AFK gate is holding you | The purchase succeeds. That gate refuses earning, not buying |
| A booster's level range is lowered below what you hold | You keep your row and it is clamped when read |
| A booster is removed from the catalogue | Your row is kept rather than deleted, and contributes nothing while the definition is missing |
What happened to the old purchases
When the ladder replaced the old flat boosts, four things were migrated and each kept its original expiry, so the first restart after the update changed nobody's numbers. The old faction income and XP buffs became level 1 of the faction income and faction XP boosters. The Cooldown Pass became level 1 of the cooldown booster, and the Trait Boost became level 1 of the side-talent booster. The old shop buttons still work: the legacy key is now a permanent alias for level 1, so a purchase made through the old surface buys exactly what it always bought.
Buffs: the permanent tier model
/buffs opens a hub of categories, and a book on the hub opens My Active Buffs, which lists everything you personally hold. Faction buffs sit in the faction shop and are paid from the treasury by a ranking member; personal buffs come out of your own wallet.
Buff definitions are fixed when the plugin starts. The ones whose numbers come from configuration receive them at boot and the resolved table is logged, which is what keeps the shop lore and the enforcement in step. Every card carries a Current: n / max line so you can see your standing without opening a second screen.
A few examples of what the ladders actually pay, with their real per-tier values:
| Buff | Scope | Tiers | Per tier |
|---|---|---|---|
| Loot Hoarder | Personal | 5 | 10 / 30 / 60 / 100 / 150 coins on a qualifying mob kill, with an action bar showing the amount |
| Redstone Efficiency | Personal | 5 | 5 / 10 / 20 / 35 / 50 coins per minute near powered redstone |
| Swift Traveller | Personal | 5 | Speed I through Speed V, one distinct level per tier |
| Shardseeker | Personal, Exploration | 3, base 4,000 | +25 / 50 / 100 percent relative to your Rift Ore Fragment drop chance, capped at 100 percent absolute |
| Rift Ward | Personal | 3 | Reduces rift mob damage, and only while a rift is actually active |
| Territory Expander | Faction | Up to 40 purchases | +10 claim chunks each, priced 25,000 x 1.2 per purchase already made |
Territory Expander is the shape worth understanding, because it shows what the cost multiplier does over a long ladder. Purchase 1 costs 25,000. Purchase 10 costs a little over 129,000. The bonus per purchase never changes.
Two buffs stopped working in the 2.0 line
Both stopped because the feature underneath them was deleted, and the two were handled differently for a reason worth knowing.
Expedition Veteran only ever registered itself when the expedition subsystem was switched on, and that switch shipped off from the day it was written. The expedition package was deleted in update 2.0.1 and the buff went with it. Nobody on this server ever saw the card, so nothing was owed to anybody.
Ruin Seeker is the opposite case: real players bought it, at 1,200 coins for the first tier and double for each one after. What it bought was reach on the ruin locator, and update 2.0.1 deleted Ancient Ruins, their sites and the /ruin command together. The tier you hold is still registered and now grants nothing at all. It is not a weaker bonus than it was, it is no bonus, and the only reason the card is still visible at all is the refund described below.
- The definition stays registered. Every screen walks the registry, so removing the id would have made an existing purchase invisible rather than refundable. The card renders as Ruin Seeker (retired), says plainly that it does nothing, and both purchase gates refuse it so nobody buys another tier of a dead buff.
- The refund is the sum of every tier bought, not the top rung. A player sitting at tier 3 paid 1,200 plus 2,400 plus 4,800, so the refund returns 8,400. Paying only the 4,800 would look, in a log, exactly like a refund that worked.
- It pays by player id, so offline players are paid too. There is no join hook and nothing to replay; the level is removed in the same pass as the payment.
- It is a refund and not a payout, and the distinction is the money's. Buying a buff burns the coins rather than moving them to anybody, so giving them back is a re-mint of something destroyed: untaxed, and not booked as income. Paying it out instead would have taxed a player on their own refund.
The Garden added no buffs
Nothing new appeared under /buffs with the growing layer. The Garden's own effects come from things you grow and make, coffee and the garden potions, which you drink rather than buy as a permanent tier, and the Garden articles carry them.
Registering a buff only inserts a shop line; it wires no enforcement. That is why one legacy buff id stays deliberately unregistered: adding it would put a purchasable card in the faction shop that charges real treasury coin for a bonus no line of code reads. Ruin Seeker is the same rule read from the other end. Once an id has actually been sold, the answer is to retire it in place rather than unregister it, because an id nobody can see is an id nobody can be refunded for.
The village buffs and the NPC caps
Four faction buffs and one personal buff govern how many NPCs you can run and how much they produce. All five are permanent tiers, and their costs and magnitudes are resolved from configuration at boot rather than written into the shop text by hand.
| Buff | Scope | Tiers | Base cost | Per tier |
|---|---|---|---|---|
| Garrison | Faction, Combat | 3 | 30,000 | +2 guard cap slots and +4 percent guard damage |
| Workforce | Faction, Infrastructure | 3 | 25,000 | +2 to the shared faction worker cap |
| Overseer | Faction, Infrastructure | Unlimited by default | 25,000, then x1.6 per tier | +20 percent to worker output budgets and daily caps |
| Upkeep Subsidy | Faction, Economy | 3 | 20,000 | -10 percent NPC hourly upkeep |
| Personal NPC cap | Personal | 7 | 5,000 | +2 to your own NPC cap, taking it from 6 to 20 |
Overseer is the odd one
Overseer has no tier ceiling in the shipped configuration, and its output stays additive: tier N gives a multiplier of 1 + N x 0.20. Its cost is 25,000 x 1.6^(N-1), so the first three tiers are 25,000, 40,000 and 64,000, and it reaches unaffordable around tier 8. A purchase that would overflow the price arithmetic refuses with a message rather than charging a clamped figure, and the tier is clamped to a sane range when read, so a corrupted stored value cannot turn into an enormous multiplier.
The ceilings the buffs cannot cross
- 60 NPCs globally. A hard backstop for server performance, and no buff raises it.
- 20 personal NPCs. The personal cap buff maxes out exactly there.
- The guard cap and the worker cap are separate axes. Garrison raises guards only; Workforce raises workers only.
Where boosters meet NPCs
The NPC output booster raises daily output caps for workers you own, which stacks on top of the Overseer multiplier that raises the same caps for the whole faction. The Lumberjack's daily caps of 3,840 logs and 30,720 leaves are the numbers those multipliers act on, so it is worth knowing what your caps are before paying for more throughput.
Fees and ore drops: the two biggest multipliers
The market fee
Listing an item on the market board costs 5 percent by default. The buff ladder takes that down to as low as 1 percent. One method resolves the rate, and all three readers use it: the label on the create-listing screen, the actual charge, and company consignment. The number you are shown is the number you are charged.
Three separate discounts now compose multiplicatively on that rate: the buff ladder, the booster's fee discount (capped at 25 percent) and the Commerce skill (whose own factor is floored at a quarter of the base). Their product can approach a free listing. There is no hard floor on the composed rate, and that is on record as a known trade-off rather than something that has been quietly fixed.
The same booster discount reaches forge repair fees. The reforge path carries no fee discount at all, from neither the forge discount buff nor the skill, and that asymmetry is deliberate rather than an omission.
The two ore buffs, and what they now do
Ore Fortune and Double Drop were on sale for a long time while producing nothing on ore blocks. The ore drop authority clears the drop flag both buffs were gated on, so both were inert on every ore, by construction. Update 1.20.0 folds them into that authority instead of running a second drop path alongside it, which is the arrangement that produced an unstamped duplicate the first time it was tried.
There are two modes, and the ladder text a player is sold is generated from whichever one is active.
- Effective level (the default). Your buff level competes with your tool's own Fortune and the Precision Pick bonus. The highest one wins and feeds a single drop roll. It is a maximum, never a sum, so a Fortune III pick with Ore Fortune II gains nothing from the buff.
- Extra drops. The level adds that many extra copies of the resolved drop set, which is what the older ladder text promised.
The order is fixed: Fortune resolves first, then the doubling, so the two never multiply each other's variance. At level 5 of both buffs on a non-rare ore, the products are:
| Mode | How the two combine | Versus vanilla |
|---|---|---|
| Effective level (default) | 3.14x from Ore Fortune, times 1.50x from Double Drop | 4.71x |
| Extra drops | The flat extra copies, times the same 1.50x | 9.00x |
The 1.50x from Double Drop is an expected value: tier 5 is a 50 percent chance per break to duplicate the whole drop set, so half your breaks yield twice.
This is a material faucet and the figures are stated rather than buried. The faucet-to-sink ratio the economy carries is 3.18, which means considerably more value enters the economy than leaves it, and multipliers of this size are part of that picture. That number had a strange run: it was marked discredited for a release, then re-derived in the 2.0 line from the only ledger snapshot the repository holds, which gives 3.13 over fourteen days and 2.97 over seven. What had actually gone wrong was its provenance rather than its value, and those are different defects with different fixes. The snapshot it is read from is old, so a live read now answers that the ratio is unmeasured instead of printing a stale one.
Rare blocks are excluded. Buff bonuses are suppressed on the rare material set, so diamond, emerald and ancient debris are never multiplied by a purchased buff. Your tool's own Fortune and the Precision Pick are not suppressed there, because they are not purchasable buffs, and keeping that asymmetry intact was an explicit requirement of the fix.
Known limits and open questions
Several things about this system are unfinished or uncertain, and they are worth reading before you spend a large balance on a high level.
- The 12,000 coins per hour baseline is a guess about how much players actually earn, and every price in the booster catalogue derives from it. If it turns out to be too low, the whole ladder is underpriced; too high and nobody buys anything. There is a calibration tool that prints the measured figure so the assumption can be replaced with a real one.
- Whether the ladder is a net faucet or a net sink is not decided in advance. At low levels it adds more money than it removes; from level 6 it removes more than it can add. Which way the whole thing lands depends on which levels people actually buy, and the meters that measure it shipped alongside it.
- The vote shop's job XP boost is still unpersisted. A restart deletes it after the vote tokens have genuinely been spent. This is flagged in the plugin's own notes and has not been fixed.
- The Loot Hoarder tier lookup has an unclamped index in the mob-death handler, which is a latent crash rather than a live one. It is recorded and deliberately untouched by the release that found it.
- None of the booster ladder has been verified on the live server. The plugin's own status for that release reads code-written, not server-verified, and the sections above follow that wording rather than improving on it. The 2.0 line rewrote a great deal elsewhere and did not touch the ladder, so nothing about it has been verified since either.
Boosters and buffs are both readable from Bedrock. Every screen renders as a native form, and /boosts keeps a plain chat fallback that lists your active boosts with text countdowns, so nothing about either system depends on a Java-only interface.
One small thing moved in update 2.0.5. /boosts help now opens the same screen the bare command opens, for a player on a server where both the booster service and its screen are live. /boosts help text still prints the old lines, and so does the help verb from a console or on a server with the booster system switched off, which is the case the text version exists for: a help verb whose point is answering on a switched-off server must not be the thing that tries to open a screen against one. /buffs has no help verb at all, by choice. The hub is the help.