Conquest Gold: When Finishing a Player Is Worth the Cost

Direct answer: when is conquest Gold worth finishing a player?

Finish a player for conquest Gold only when the expected payout plus the territory and border improvement exceeds the troops, time, and third-party exposure needed to secure the elimination. In v0.34.21, a conquered Bot or Nation pays all of its Gold, an active Human pays half, and a Human who never attacked pays nothing. Classify the target before pricing the finish.

That first classification changes the decision more than the number shown beside a wealthy target. A Nation with an estimated 10 million Gold can transfer the full 10 million. A Human with the same balance can transfer only 5 million after the game has recorded any nonzero outgoing attack. A Human who has never attacked transfers zero even if starting-Gold settings or passive income left a large balance. The game does not expose a guaranteed live balance or a clean public “eligible for payout” flag, so these amounts are planning bounds, not promises. A visible attack proves that the zero-payout gate no longer applies; silence does not prove that the target never attacked somewhere outside your view.

Use the reward as one input, not the objective by itself. The finishing attack still spends troops, changes your border, reveals your commitment to every neighbor, and can let another player take the last tile first. A full Nation payout is bad value when reaching the final pocket opens two healthy Human borders and empties the reserve. A half-Human payout can be excellent value when the elimination closes a front, captures useful land, and immediately funds the Defense Post, City, Port, or SAM needed to hold it. A zero-Gold Human can still be correct to finish for a victory threshold, a route, a dangerous structure, or removal of a front; it is simply wrong to justify that attack with loot.

Run the CLAIM check before committing: Classify the target type, estimate Liquid Gold conservatively, confirm Activity if the target is Human, price the Investment required to finish, and name the Move the payout funds afterward. If any item is unknown, use the lower payout bound. If the plan only works with a precise hidden balance, do not treat it as a plan. If it works at zero Gold because the land and border are already worth the cost, the reward is upside rather than a trap.

This guide owns the elimination-reward decision. Threat Assessment owns which target improves the whole position; Nations Pressure owns when a Nation is a safe or dangerous neighbor; Land Combat owns the cost of the active push. Conquest Gold begins after those pages identify a plausible finish and asks one narrower question: does the actual payout class change whether you complete it now, delay it, or let another front consume the target?

What v0.34.21 actually transfers on conquest

The payout rule has two layers. Target type sets the first ceiling: a defeated Bot or Nation can transfer its full balance, while a Human can transfer at most half, rounded down to a whole unit of Gold. Humans then face an activity gate: if the defeated player has sent zero attack amount, the transfer is skipped completely. This is why “Bot or Nation equals all, Human with an attack equals half, never-attacked Human equals zero” is more accurate than the shortcut that every elimination steals half.

The Human half is calculated from the balance at conquest, not from lifetime income, leaderboard income rate, structures, or the Gold the target had when the war began. If an eligible Human has 12,500,001 Gold, integer division yields 6,250,000 for the conqueror; the fractional remainder cannot exist in the integer currency. The conquest handler removes the defeated eligible player’s full 12,500,001 balance, so the other 6,250,001 does not move to another player. It disappears with the defeated economy. Against a Bot or Nation, all 12,500,001 transfers. Those distinctions matter when a finish is being compared with a specific next purchase.

The never-attacked Human path is intentionally different. When the sent-attack statistic is zero, the handler sets both the considered balance and captured amount to zero, displays a no-Gold conquest message, and does not remove the target’s stored balance in that transfer block. The defeated player is still conquered, so this does not create a usable bank for the conqueror or make the target economically alive. For player decisions, the only safe conclusion is zero reward. Do not invent a way to recover that untouched balance later, and do not assume that passive Gold, high starting Gold, or an impressive leaderboard history bypasses the gate.

“Has attacked” is also narrower than “looked active.” Building, upgrading, moving a camera, receiving income, joining an alliance, or being struck does not satisfy the activity gate. The relevant value is outgoing attack amount, and any nonzero recorded attack is enough to leave the zero-payout branch. A tiny probe can therefore make a Human eligible for the half-balance rule. You may not have observed that probe if it happened on a distant border. Treat confirmed visible aggression as proof of eligibility, but treat apparent inactivity as uncertainty unless the match state makes it genuinely knowable.

This guide applies to v0.34.21. In that version, the player-facing outcomes are full balance from a Bot, full balance from a Nation, half the balance from an eligible Human, and zero from a Human who never attacked. Treat those four branches as the stable boundary for this guide; if a later version changes the target classes, the activity gate, or the event message, recheck the payout before committing troops around an old expectation.

The CLAIM decision framework

Classify comes first because the type sets the payout ceiling. Mark the target as Bot, Nation, confirmed-active Human, or activity-unknown Human. Do not combine Bot and Nation behavior with Human behavior merely because all three can hold land and Gold. A Nation that has never attacked still belongs to the full-payout branch; the zero-attack gate is limited to Human players. A Human that sent one small attack belongs to the half branch even if they later went idle. Classification is the only fixed input you can settle before the balance and finish cost are known.

Liquid means the balance likely to remain at the instant of elimination, not an old leaderboard snapshot or a guess from structure count. Watch recent spending signals: new Cities, Port or Factory levels, SAM construction, Warships, and nuclear launches all reduce the cash that can become loot. Income can rebuild it while your attack travels. Use a range with a conservative floor. If the target appears to hold 8M to 12M, price a Nation at 8M and an eligible Human at 4M. If the decision fails at the floor, Gold is not a reliable reason to finish.

Activity applies only to Humans and asks whether you have evidence of a nonzero outgoing attack. A visible attack notification, a border push you observed, or an attack shown in a replay is evidence. Merely being online, owning buildings, defending automatically, or taking damage does not pass that gate. When evidence is absent, calculate two cases: half the conservative balance and zero. A robust finish remains correct in both cases. A fragile loot run that only works in the half case should wait for better evidence or be justified by territory instead.

Investment prices everything needed to own the last tile and survive the next response. Count the troop commitment, travel or transport time, the border created after conquest, and the opportunity given to a third player. Gold is not normally paid to launch a land attack, but the troops spent have economic value because they take time and cap headroom to rebuild. Include a Defense Post, replacement fleet, or SAM in the finish cost when that purchase is necessary to keep the prize. Do not count captured structures at full value before confirming that the conquest path actually preserves them.

Move names what happens within one decision cycle after the payout. “Become richer” is not a move. “Use up to 1.5M of a minimum 5M payout for the two posts that close the new neck, while keeping the rest liquid” is a move. So is “buy the next City level because current troops are already at cap” or “reserve the payout for one confirmed SAM line.” The move prevents a reward from sitting unused while the newly exposed border collapses.

Target class	Payout planning bound	Evidence required	Finish only if	Default when uncertain
Bot	100% of balance	Type is visible	Land plus full conservative balance beats the finish cost	Use a low balance estimate
Nation	100% of balance	Type is visible	The finish survives Nation and Human counterpressure	Keep a reserve for the new Human border
Human, confirmed attack	50% of balance, rounded down	Any reliable nonzero outgoing attack	Half the conservative balance plus position beats the finish cost	Price at half the low estimate
Human, activity unknown	0% to 50%	No reliable attack evidence	The finish is already correct at zero payout	Treat Gold as optional upside

CLAIM is a stop test, not a scoring game. Stop when classification is unclear and the Gold is essential, when the balance estimate depends on stale information, when the finish investment consumes the reserve, or when no immediate move protects the result. Continue when the conservative payout is enough, the land is useful at zero, and the next move is already budgeted. This keeps conquest Gold in its correct role: a financing event attached to a positional elimination, not a reason to create a bad elimination.

Scenario one: a 10M Nation with a contested finish

Assume a public FFA where a Nation has about 10M Gold, 6% of usable land, one exposed City, and two remaining pockets. You and another Human both touch those pockets. Your current reserve is 2.4M troops; the estimate is that taking both pockets costs 700k troops and leaves a short new border with the rival. The rival has 1.1M visible troops near the second pocket. The target is a Nation, so the source ceiling is 100%: if the 10M estimate remains true at elimination, the player who completes the conquest receives 10M.

Run CLAIM rather than rushing the wealth number. Classification is exact: Nation, full branch. Liquid Gold is uncertain, so use an 8M floor after allowing for one last structure or other spending. Activity is irrelevant because the no-attack gate applies only to Humans. Investment is at least 700k troops plus the force needed to prevent the rival from taking the final tile or attacking through the new border. Move is a specific plan: reserve 500k for a Defense Post sequence only if the final neck is covered, keep at least 1M Gold liquid for a response, and spend the remainder only after the border survives one cycle.

Under the stated assumptions, the finish is attractive only if the 700k commitment leaves a reserve that can answer the rival’s 1.1M. A simple rule is not “10M is worth 700k troops,” because troops and Gold are not interchangeable. The useful comparison is whether the remaining 1.7M troops can protect the new geometry long enough for the 8M conservative payout to fund the named move. If terrain, density, or another border means 1.7M cannot hold, the correct action is to let the rival bind the Nation first or take only the pocket that shortens your border.

Now change one assumption. The rival starts attacking the second pocket and is likely to land the finishing hit. The full reward goes to the conqueror, not proportionally to everyone who damaged the target. Your earlier work creates no guaranteed share. You can contest the final pocket only if the contest itself does not empty your core. This is the counterplay around conquest Gold: another player can make you pay the troop cost and then secure the elimination, or can let you receive the Gold and attack the border before it becomes defensive infrastructure.

The scenario’s completion test is observable. Proceed if the Nation remains near the conservative balance, the two pockets can be finished without falling below the response reserve, and the final route produces a coverable border. Delay if the rival can take the last tile first, the Nation spends down its balance, or the finish adds an exposed Human front. The full payout is real, but it is only valuable to the player who owns the elimination and survives the next move.

Scenario two: an active Human with 12,500,001 Gold

Assume a three-player FFA where a weakened Human has exactly 12,500,001 Gold at the decision point, has visibly sent several attacks, and holds a narrow 4% strip between you and the leader. You have 1.8M troops and estimate that finishing the strip costs 450k. The leader has 900k visible troops on the opposite side. Because the target is Human and the game has recorded a nonzero outgoing attack, the maximum immediate reward is half with integer rounding down: 6,250,000 Gold. The conquered player’s full balance is removed, so the remaining 6,250,001 does not become a second prize.

Classification and Activity are confirmed. Liquid is still time-sensitive: the exact figure is an assumption for the scenario, not a value normally guaranteed to you in live play. Investment includes 450k troops and the risk that the leader uses the removed strip as a direct route. Move must therefore protect the new line before it grows the economy. Suppose the plan assigns 250k to a necessary defensive purchase and keeps 1M Gold liquid; the remaining payout can later support a City, Port, Factory, or other proven constraint. The finish works only if the post-conquest troop reserve of 1.35M can answer the leader’s visible 900k without abandoning another border.

This case shows why a half payout can be stronger than a full payout elsewhere. The 4% strip closes a player, removes one front, and can deny the leader an easy route. Those positional gains may be worth the 450k even if the target spends all Gold before conquest. At zero reward the action can still be correct. The 6.25M then accelerates a move already justified by the map. That is safer than attacking a remote 20M Nation whose full reward opens three borders, because the payout multiplier does not measure the value of the resulting position.

Counterplay comes from timing. The leader may stop attacking and wait for you to pay the 450k, then hit your 1.35M reserve as soon as the strip changes color. The weakened target may spend the balance on a last SAM, Silo level, fleet, or City before being eliminated, reducing the captured amount. The leader may also secure the last tile and take the 6.25M instead. Your response is to stop reinforcing when the final tile is no longer yours to secure, preserve a force on the leader-facing side, and treat every observed target purchase as a downward revision of Liquid.

If the exact balance were 12,500,000 rather than 12,500,001, the payout would still be 6,250,000. The odd unit illustrates rounding but should never drive the decision. The real threshold is whether the low payout estimate plus the 4% positional gain beats the finish cost and leaves a survivable reserve. The scenario succeeds when the new border holds and the named post-payout move executes; it fails when the Gold arrives but the leader captures the new strip before that move matters.

Scenario three: an 18M Human who never attacked

Assume a hosted match with 5M starting Gold where a Human spawns, builds passively, never sends an attack, and later appears to hold about 18M. The player controls 3% of usable land on a side route. Reaching the final pocket costs an estimated 300k troops and opens a coast that another player can reach. It is tempting to call the target an 18M prize. Under the v0.34.21 conquest handler, that label is wrong: a Human with a zero sent-attack statistic triggers the skip branch, so the conqueror receives 0 Gold regardless of the stored balance.

CLAIM changes the call immediately. Classification is Human. Liquid may be 18M, but Liquid cannot pass through the activity gate. Activity is confirmed zero only for this scenario; in a real match you often lack that certainty, so zero remains the conservative bound. Investment is 300k troops plus defense of a newly open coast. Move cannot depend on loot because no payout exists. The only valid reasons to finish are the 3% land, removal of a front, access to a necessary route, capture or destruction of a dangerous position, or progress toward the live victory condition.

Suppose you are 2.5 points below a falling Overtime threshold and the 3% strip is reachable without exposing the core. The finish can be correct at zero Gold because the land itself ends the game. Suppose instead that you are far from the threshold, already have a safe route, and the coast creates contact with a stronger player. The same target is not worth finishing now for economic reasons. Leave the pocket as a buffer, spend the 300k troop opportunity elsewhere, and revisit only when the map value changes. The zero reward does not forbid conquest; it removes one false argument for it.

The target can counter your inference by making one tiny attack. Once the sent-attack statistic becomes nonzero, the target moves to the Human half branch. You may not see that attack if it occurred on a remote edge or before you began watching. That uncertainty is why an apparently idle Human must be priced as a range from zero to half, never confidently as zero unless the match evidence supports it. It is also why the guide does not recommend baiting or coercing a token attack as a general plan: spending troops or opening diplomacy merely to make hidden Gold eligible can cost more than the possible half payout and can wake a previously harmless buffer.

The code leaves the never-attacked player’s balance untouched in the skipped transfer block, but the conquered player is still defeated. Do not interpret the untouched value as recoverable treasure, a delayed team transfer, or a bank that the next conqueror can claim. For your decision it is zero. The no-Gold event message is confirmation after the fact, not a signal that helps price the attack beforehand. Post-match review can use it to correct your activity assumption; live play must use the conservative range.

The scenario succeeds when the zero-payout finish buys a named positional result that was worth 300k troops without any Gold. It fails when the player spends 300k and opens the coast because the assumed 18M reward was the only justification. This is the strongest test in the guide: if a Human finish is activity-unknown, erase the Gold from the plan and ask whether you still want the land. If the answer is no, wait.

Failure modes, opponent counterplay, and signals you cannot know

The first failure is pricing displayed strength as liquid Gold. A large territory, many structures, or a high income rate does not reveal the balance at conquest. A target can spend immediately before elimination, and structures are not currency. Correct it by using a range and by lowering the estimate after every observed build, ship, upgrade, or launch. Conquest Gold rewards the balance that remains, not the economy you imagine the target once had.

The second failure is using the wrong type multiplier. Bot and Nation pay full balance, while an eligible Human pays half. A Nation that behaved passively does not become a zero-payout Human, and a Human who fought aggressively does not pay full because they were difficult. Correct it by classifying type before estimating value. If the player type is visually ambiguous in a replay or localized interface, verify it before using the payout in a decision journal.

The third failure is assuming activity from presence. A Human can be connected, wealthy, fortified, and still have zero sent attacks. Another Human can look idle after one remote probe and already qualify for half. You cannot normally inspect the internal sent-attack statistic during a live match. Correct it with two-case planning. Confirmed visible aggression permits the half case; absent evidence requires the zero floor. Never make survival depend on a hidden flag.

The fourth failure is confusing contribution with ownership of the elimination. Damage does not reserve a share of the reward. The player who secures conquest receives the configured amount. An opponent can let you drain the target and then take the last territory, or can force you to overcommit to prevent that outcome. Correct it by stopping when your path to the final tile is no longer controllable. Sunk troop cost is not a reason to donate more troops to someone else’s payout.

The fifth failure is spending the reward twice. Players often count the expected Gold as both the reason the attack is affordable and the budget for several follow-up assets. Until conquest occurs, the balance is not yours. After it occurs, only one set of purchases fits inside it. Correct it by naming one immediate Move and a liquid remainder. If the new border requires 3M of defense and the conservative Human payout is 4M, only 1M remains for growth, not 4M.

Opponents have four practical counters. They can spend down before defeat, make a small attack to change Human eligibility, contest the finishing tile, or attack the conqueror’s new border during the payout-to-defense gap. None changes the payout formula; all change the inputs to CLAIM. Your counter is not a cleverer multiplier. It is shorter exposure: reduce the troop commit, secure the route to the final pocket, pre-position the reserve, and delay discretionary spending until the border survives.

Some signals remain unknowable in live play. Exact balance, remote attacks, planned spending, and the rival’s timing may be hidden or stale. State assumptions explicitly: “Nation, at least 8M, finish costs 700k, rival can contest in one route.” If the decision changes when one uncertain value moves slightly, it is not ready. A good conquest remains correct across a useful range and becomes better when the reward arrives.

Mode and map adjustments

In public FFA, third-party capture risk is usually the largest adjustment. Every player can benefit from the target disappearing, but only the conqueror receives the payout. Use Gold as a tie-breaker between finishes that already improve geometry. A remote full-payout Nation is rarely better than a nearby half-payout Human when the nearby finish closes a front and the remote finish creates two. Re-run Threat Assessment after every elimination because the target list, beneficiaries, and legal routes change at once.

In Team games, conquest Gold belongs to the individual conqueror’s balance, not automatically to a shared team treasury. The team can coordinate who secures the elimination and what that player buys afterward, but it should not assume every teammate receives a fraction. Give the finish to the player whose route and follow-up purchase best help the team, provided that routing the last tile does not delay the elimination or sacrifice the front. A teammate may be better positioned to take the Gold and immediately fund defense, while another preserves troops for the next line. Use Team Roles for that ownership handoff.

In singleplayer or high-Nation-density games, full Nation payouts are more common and can chain into the next purchase. That does not make every Nation a treasury chest. Hard and Impossible Nations choose valuable bordering targets they can safely beat, so a finish that exposes upgraded structures can turn the payout into a lure. Use Nations Pressure to decide whether the new adjacency is survivable. Bot payouts are also full, but a Bot pocket with little remaining Gold may be strategically valuable only for land and route.

On compact maps, the interval between conquest and counterattack is short. Value the immediate border and reserve more heavily than the nominal payout. Pre-position defense before the final tile and keep the post-payout Move small enough to execute without losing awareness. On large maps, a remote pocket may give more time to spend the reward, but the troop travel time and disconnected geometry can increase Investment. Long distance also gives the target more time to spend down, so lower the Liquid estimate as the campaign stretches.

On island and strait maps, route ownership can outweigh cash. A full Nation payout across water is poor when transports and escort consume the reserve and the landing cannot be held. A zero-payout Human island can still be worth taking when it controls the only port route or removes a launch point. Price fleet losses and home-coast exposure as part of Investment, then use Transport Landings for the actual crossing. Gold received after a failed bridgehead does not make the bridgehead survive.

Overtime and fixed victory thresholds change the value of land. When the target’s territory crosses the live threshold, the finish is a victory action first and a Gold action second. When the threshold remains distant, the payout needs a specific follow-up Move. Do not wait for a richer balance if another player can end the match, and do not rush a zero-payout pocket merely because it looks complete. Winning Overtime owns the live percentage calculation; this guide only tells you how reward class changes the cost of the finish.

Map modifiers can also distort estimates. Starting Gold can make a never-attacked Human look extremely valuable while the anti-idle gate still forces zero. Disabled structures reduce spending signals but do not alter the type multiplier. Team placement, Compact Map, water rules, and public rotation affect access and exposure, not the full/half/zero formula. Carry the formula across modes; recalculate route, time, border, and victory value for the board in front of you.

Pre-finish checklist and next reading

Before sending the finishing attack, say the five CLAIM answers in one sentence: target class, conservative balance, Human activity evidence, total finish investment, and immediate post-payout move. A usable answer sounds like: “Nation, at least 8M, activity irrelevant, 700k troops plus one coverable neck, then no more than 500k on defense while 1M stays liquid.” An unusable answer sounds like: “Rich player, easy kill, I will decide what to buy afterward.” The first can be disproved by new information; the second cannot protect you from anything.

Then run four stop signals. Stop if the target spends enough that the conservative payout no longer funds the Move. Stop if another player controls the final route. Stop if the required troop commitment crosses your response floor. Stop if the elimination creates a border you cannot cover. These stops are stronger than a fixed payout threshold because 5M can be decisive on one map and irrelevant on another. A reward is valuable only in the position where it lands.

After conquest, verify the event message and the actual balance before buying. A received-Gold message confirms the transfer; a no-Gold message confirms that the Human activity gate skipped it. Use the difference between expected and actual payout in Replay Review: was the type wrong, was the activity assumption wrong, did the target spend down, or did another player secure the conquest? Change one assumption in the next match rather than inventing a new formula.

Keep the v0.34.21 boundary simple in play: Bot and Nation are full-balance targets; a Human with a recorded outgoing attack is a half-balance target; a Human with no outgoing attack is a zero-balance target. The event message after conquest tells you whether Gold moved, but it cannot reveal the hidden balance or activity state before the finish. Price the action with the conservative branch, then use the observed result only to improve the next estimate.

You cannot normally see the exact enemy balance or the internal activity statistic live. That uncertainty is part of the decision, not a gap to fill with confidence. The completion definition is practical: choose a target whose elimination is worthwhile at the conservative payout, leave a reserve, secure the last tile, and spend only on the named Move. If the result is still good when the balance is lower or the Human pays zero, the finish is robust.

Continue according to the input that failed. Read Threat Assessment when the target or third-party beneficiary is unclear. Read Nations Pressure when a full-payout Nation opens a dangerous border. Read Land Combat when the finish cost is uncertain. Read Winning Overtime when the land itself may end the match. Conquest Gold is complete when the reward improves a position that was already defensible, not when a large number briefly appears after a costly elimination.