THE PRICE
OF POWER.

Decision

The $60 weekly catalogue can create meaningful paid advantages, but it is not ready to be treated as a balanced launch economy. The current source allows rapid economic growth financed by three Mints and credit-based Finish Now, while unlimited enhancement repeats can turn cheap Standard equipment into dominant combat gear. Weekly store limits alone do not control those effects.

In the primary prospective fresh-account scenario, all three cohorts eventually reach HQ50 and Tier 10 infantry. Their dates, equipment and combat performance remain substantially different. Dedicated free players can contribute useful damage and win selected coordinated fights in these tests; they are not duel-equal to the high spender. The proposed free reward program is essential to that conclusion and is not installed today.

This review recommends testing the four packs with explicit reward replacement accounting, complete conversion quotas and a separate decision on repeat scaling and credit acceleration. Additional standalone spending rapidly loses marginal value under the draft caps. Do not expand the catalogue merely to dispose of surplus gems.

Scope and authority

Analysis date: October 18, 2026. Authority: the existing Ashline Next Chapter Unity source, instantiated through AuthoritativeSession with its current world design and handwritten policies. Historical JavaScript and documentation were checked against current C# quotes; the older handoff is research evidence only. No game balance, billing configuration, source assets, APK, device, live host or real save was changed.

Outputs are three linked studies: fresh-account economic progression with segregated equipment funds; weekly equipment purchase/upgrade/maintenance histories using those funds, followed by actual current-source encounters; and a separate mature HQ45 research comparison. Neither deterministic test cases nor expected loot values are population forecasts or measured player win rates.

Pre-existing run-model.mjs, results.json, ledgers.json and weekly-results.csv in this analysis folder were preserved and not used as authority. Use the progression-, equipment-, reward- and mature- v 2 files named in the reproduction section. The supplied historical handoff was located locally. Three indexed handoff files were absent; the corrected model/ files were available.

Pricing and proposed catalogue

Normal packs include 100 gems per US dollar plus their stated contents. Standalone purchases give 50 gems per dollar, without a quantity limit. Weekly pack purchase limits do not reset when buying standalone gems. Promotions must explicitly disclose any exception. All products below remain proposals.

Current implementation versus hypotheses

Root ECONOMY-RULES.md contains historical formulas and cannot describe this branch by itself. For example, current repair uses 10% of enhanced hardware maintenance value, current equipment has seven slots, premium ammunition costs 5,000 credits per shell, and training credits are divided across the academy batch capacity. The rules in the following table were read from the current policies and corroborated by extracted quotes or current-source tests.

Free rewards and grinding workload

The developed-account target is 2,000 Data Drives, 400 normal shards, 40 Rare/10 Epic/2 Relic Tank Fragments, 200 gems in vouchers and 24 speedup hours per week. Tank Fragments craft equipment; they do not change the vehicle rarity. The new-account primary case ramps this TOTAL target to 25% below HQ10, 50% at HQ10, 75% at HQ20 and 100% at HQ25. That ramp is a proposed design choice.

A target is filled only by the shortfall after counted existing rewards and included Supply Shop drives. Current deterministic rewards, random loot, purchases and proposed top-ups are separately recorded. Random rewards are probability-weighted expected income here; the work does not promise fractional drops or a guaranteed weekly loot outcome. The full target is never stacked on top of every old reward.

The fresh primary workload is two 30-minute visits on six days/week: 156 active hours over 26 weeks. It assumes three eligible daily objectives, a qualifying cycle per visit, modest PvE exposure, collections and local Archive dispatch/return opportunities. Construction/training/gather tasks must actually qualify; late daily build and 600-cycle-point requirements make this an optimistic activity interpretation. The forecast is conditional on the proposed milestone program being earnable at this workload. It is not a phone measurement.

Current rewards are far smaller in shards, fragments, gems and speedups. As a separate illustrative audit envelope, four daily claims/day, two cycles/day and 120 mob wins/week imply about 54 loot boxes, 24.27 gems, 0.405 normal shards and 4.99 hours of speedups. This larger hunting workload is not the primary fresh schedule. Six days of Supply Shop drives add 300 drives for 4,500 credits, which must be counted inside any target that includes them.

700 Archive drives need 35 unboosted tank-hours. That is 28 fresh tier 1 trips at 25 drives each, or roughly 11 garage 45 tier 1 trips at 65 each. Travel, contested stock, fuel, occupied tanks and orders to bank cargo matter. Getting the proposed fragments purely from mob drops would require approximately 400 Rare-tier, 160 Epic-tier and 48 Relic-tier kills/week in expectation. That is about 101 kills per active day. Deterministic milestones are necessary if the intended workload is 60–90 minutes/day.

Three-cycle bonuses require a third event window or a valid boundary/completion opportunity. Six-cycle rewards cannot be assumed from two ordinary visits. Bounties, Titans, city control and PvP wins are conditional upside, excluded from the dependable baseline. Bounty payouts are particularly large and have no obvious daily cap; the source audit flags repeat/collusive farming for focused verification rather than claiming a demonstrated exploit.

Fresh account progression over 26 weeks

Start: the current HQ1 account with 4,000 credits, 1,200 food, 1,600 scrap, 100 drives, 60 Tier 1 troops and zero gems. All cohorts use the same economic policy: develop Mints and production/research speed, then military capacity and Tier 10 infantry. They build and pay for additional producers and tanks rather than assuming them at registration. Production accrues at the actual current rate, retains earned stock when rates change and stops at each producer’s eight-hour capacity.

20% of gross credit income is segregated for equipment and 10% for a separate operating reserve. Those funds are not available for construction/research and are not counted again as new equipment income. Training, promotions, healing, fleet/queue unlocks and Finish Now are charged. Hero training, workers, city bonuses, alliance help and money-only Builder 4 are excluded. Equipment is funded for one primary tank across all seven slots; other owned fleet tanks are not assigned duplicate copies of that loadout.

Income and expenditure are different from progression

Large balances are a source behavior worth addressing. Economic specialization produces substantial credit/food/scrap surpluses while scarce materials and queue use decide later research and equipment choices. The model does not turn those balances into a generic power multiplier. The totals below are cumulative flows, and reserves remain owned assets rather than fictional expenses.

Equipment stats and actual combat performance

The linked equipment model receives only the segregated weekly equipment funds, shards, fragments and transferred gems from the fresh simulation. It makes week-by-week purchases and upgrades, reserves and pays maintenance, and does not reallocate the same 26-week credit balance repeatedly at each checkpoint. Its Standard-focused policy maximizes a declared attack/health/armor heuristic, not proven global combat optimality. It fills all seven slots but leaves cargo unenhanced. This policy leaves earned rarity fragments unused; the separate hybrid sensitivity crafts stronger support equipment from those fragments. The main figures are a reproducible spending choice, not an account power ceiling.

Current repeats can make the optimal-looking item a low-rarity one. Per-cap repeat gains are 3 percentage points of base hardware for Standard/Rare, 5 for Epic and 50 for Relic, at the full final cost; each Relic repeat also consumes one pristine Relic donor. Maintenance excludes historical enhancement investment. Standard armor therefore supports thousands of cheap repeats. More expensive rarity does not ensure a better duel outcome.

All combat rows below come from the current C# encounter engine using the simulated buildings, research, gear and actual healthy troop inventory. Baseline is Kestrel versus Kestrel, no premium ammunition, full condition and the engine’s ordinary simultaneous-attack rules. Mirrored roles and a small seeded critical-hit sensitivity are included. These are controlled outcomes, not estimated population win rates. An attacker victory flag with both sides destroyed is reported as mutual destruction in control cases.

The free account contributes 31.2% of actual damage in the mixed free+pack-buyer fight. Contribution remains credited after that free tank is defeated. Two free attackers fail against the big spender in this matchup; three succeed. These are multiple independent attacks: the current rule prohibiting Tank Group Attack remains intact.

The pack buyer defeats the free profile with 72.6% of its original hull remaining; the $200-extra profile wins with 87.2%. Actual shots apply armor through 100 / (100 + armor), vehicle and troop counters, and a current-hull damage multiplier. Consequently, a 57% attack advantage is not equivalent to a 57% battle advantage. The primary result remains the same with attacker/defender roles reversed; three seeded critical trials are a narrow sensitivity, not a win-rate estimate.

A separate controlled cannon-only paid loadout reaches about 71,363 total attack yet loses in six rounds to a balanced seven-slot loadout with about 26,796 attack. The latter retains 98.1% of entry hull. These mature controls are separate from the fresh-account table. They demonstrate why the older one-cannon resource comparison cannot establish combat power or an absolute ceiling.

Troops, repairs and repeated losses

The primary progression exposure is twelve modest 10%-hull-loss fights and one defeat per week, using at most 100 troops per exposure. This is a declared stress workload, not a claim that the player wins that many fights. It applies 40% casualty scaling, 30% direct death and hospital overflow, then pays healing/replacement/promotion costs. Full late-game deployment losses are measured separately below; do not mistake the small-detachment baseline for frequent full-army war.

The fresh equipment histories fund 56 condition points/week: 12 victories×3 plus 1 defeat×20, with return-home repair before the next weekly combat block. This matches the primary progression exposure. The separate mature equipment controls use 76 points. Maintenance is paid from the equipment envelope. The unused 10% operating reserve is separate; it is not counted as maintenance already spent.

Training T9 at academy 45 costs 6 credits per troop before discounts because the 1,440 credit batch cost is divided by 240 capacity. It still takes 216 seconds per troop before training speed. Healing costs 3 credits per troop. Recovering from repeated losses is therefore often a training-time/housing/hospital problem, not simply a credit ratio. Premium overpressure costs 5,000 credits per shell for 25% extra outgoing damage; it must compete with development and repairs and is disabled in the primary comparison.

Spending choices and marginal standalone value

Primary fresh spending buys permanent gem queue unlocks before transferring surplus gems to equipment. A gear-first alternative spends on gear earlier and forgoes those queues. Neither policy is claimed optimal. The fixed $60 catalogue need not be purchased by every real payer; the requested all-pack cohort is a deliberate comparison.

Draft complete weekly store stock costs 3,454 gems: 1,000 Relic+500 Epic+200 Rare+390 for 780 normal shards+500 drives+864 speedup time. Gear/fragments reserve 16 shard-equivalents, leaving 784 capacity; whole 20-shard offers allow 780. An all-pack buyer earning 200 gems/week receives 6,200, leaving 2,746 before unlocks or protection. Adding $200 standalone adds 10,000 more gems/week, not another weekly material allowance.

Some greedy-policy sensitivities are nonmonotonic: the lighter pack buyer or hour-item variant can reach T10 sooner while producing less income or developing differently. These results expose scheduling and spending-choice sensitivity; they do not establish that fewer visits or wasted time items improve optimal play. The current-reward-only free case ends at T9, making the new dependable reward program material to the headline result.

Higher-loss cases apply three times the baseline damage-exposure schedule. The current-store alternatives retain the proposed free rewards and packs; only the shop rules and stated spending priority change. They therefore do not describe an entirely unchanged live economy.

Existing worthwhile surplus uses are earlier permanent queues, equipment for other tanks subject to the same account stock, speedups in the current uncapped store, and desired shields. A shield costs 100 gems/hour or 20,000 credits/hour: gems can preserve 200 credits per gem when that protection would otherwise have been bought. This is substitution, not a direct cash-out or guaranteed new income. Shields restrict hostility; stockpiling unnecessary shields is not economic value. Cosmetics or new convenience offers would be new proposals and are not revenue or progression assumed here.

With the current uncapped Gem Shop instead of the proposed caps, the fresh $200-extra profile can expand carrier capacity and deploy all 3,455 trained troops. It defeats the primary free profile with 95.3% of entry hull remaining, versus 87.2% in the capped primary case. The variant also prioritizing unlimited gem time retains 97.0%. Those are paid-profile sensitivities against a fixed baseline opponent, not comparisons of two fully reoptimized economies.

Separate mature account research comparison

This study starts identical HQ45/Lab 45/all-three-academy 45 accounts with T9, no paid research and+75% starting research speed. It assumes 10 million/week already allocated to research AFTER unmodeled operating/equipment costs, plus explicit 12,000 food and 24,000 scrap/week for research. It does not add this income to the fresh study. Proposed free targets apply in full from week 1. The primary strategy purchases the draft shop stock, spends free gems on shards, and uses credit Finish Now; jobs are claimed only during six-day, two-visit activity.

The all-three Tier 10 prerequisite union costs 64,854 drives, 13,981,581 credits, 7,500 food and 18,750 scrap over 261 upgrade levels. First Tier 10 requires 31,282 drives. More drives do not independently establish an unlock date: time, queue dependencies and credits are all enforced.

Under the 10 m credit allocation, extra standalone gems do not advance the paid Tier 10 dates: drives and chosen prerequisite order bind, while credits already remove timer pressure. Removing credit Finish Now changes the result substantially. This is why the earlier mature model could not answer the fresh-account question or prove a fixed paid/free power ratio.

Permanent unlocks and alternate grinding routes

Current one-time unlocks: Builder 2 at HQ25 for 500,000 credits; Builder 3 at HQ10 for 25,000 gems; Research Queue 2 at HQ25 for 100,000 gems. Builder 4 requires a cash product, with no implemented price or available billing. Do not invent its cost or include it in the $60 weekly total. At 200 earned gems/week, saving every gem takes 125 weeks for Builder 3 and 500 weeks for Research 2. That is not a credible six-month free path.

The draft 12 weekly builder assignments after HQ10 and 16 laboratory assignments after HQ25 would grant the SAME permanent entitlement as buying it. Their counters must start after the prerequisite is reached, and purchases must not stack duplicate slots. These routes remain unapproved. The mature comparison tests the 16 week route with eligibility at day 0; fresh assignment dates can be derived from the actual HQ milestones but are not silently granted in the main simulation.

Recommendation: offer deterministic assignment routes with clear progress, and test whether 8–12 eligible weeks produces a meaningful paid time lead without leaving dedicated free players permanently capacity-limited. Keep Builder 4’s paid-only status explicit unless the owner separately changes it. Cash unlock price, refund/restore behavior and unlock timing are unresolved product decisions.

For the primary free path, HQ10 arrives on day 5.5 and HQ25 on day 21.5. Twelve complete eligible builder weeks therefore suggest approximately day 89.5, and sixteen laboratory weeks approximately day 133.5, before visit/reset alignment. These are derived eligibility estimates, not simulated assignment rewards. In the mature assignment sensitivity the second research queue arrives during week 16; the free first/second T10 dates remain 108.3/168.3 days because the chosen path is still drive-limited.

Routes that bypass simple purchase limits

Normal shards and Cracked Shards are substitutes for enhancements; a cap on only one leaves the other open. Direct equipment and tank/base fragments must share rarity-equivalent counters. Twenty matching fragments craft one item; alternate offer IDs must not create another allowance. Gem-bought equipment also reserves its recycling equivalent from shared material stock at purchase, using integer subunits. Recycle must not debit that reservation twice.

Credit purchases create another route: the cheapest Standard cargo costs 75,000 credits and ten recycle to one normal shard, or 750,000 credits/shard. Rare cargo is 800,000 credits/shard, Epic cargo 1 million/shard. This is expensive but real; credits saved through gem shields, gem speedups or paid resource packs can support it. Credit recycling is included as a separate equipment sensitivity, so the main loadout is not described as an absolute ceiling.

Supply Shop speedups and Finish Now turn ordinary credits into time at 1 credit/second. Capping the gem time offer does not cap that acceleration. Reclaim returns floor 75% invested shards as Cracked Shards while destroying the item; it is a lossy reallocation route, not infinite duplication. Repeats, shared troop capacity, alternative loadouts, paid task event points, bounties and other conditional rewards can change total account strength beyond this chosen primary-tank policy.

Any future gem-to-credit/resource conversion, promotional material offer, subscription or battle pass must enter the same accounting. Otherwise the claimed weekly limits cease to describe purchased progression.

Bounty valuation also deserves a focused audit. The baseline free armor would generate a raw half-liability bounty of 579,129 credits from 20 condition wear, while its repair costs 2,502. But personal loot caps, matchup reductions, same-victim cooldowns and per-trip limits sharply reduce payment: the actual pack-versus-free fight pays only 3,029 against combined repairs of 4,221, before troop replacement and fuel. The larger-spender fight pays 1,942 against 4,728 repairs. A profitable loop has not been demonstrated; using the raw bounty alone would be misleading.

The separate flat-income equipment control budgets 5 million credits/week solely for gear, with mature free rewards from week 1. Allowing the free profile to buy and recycle credit equipment changes equipment attack from 7,549 to 7,592, hull from 92,540 to 92,900 and armor from 4,031 to 4,047 under that policy. This small improvement is not an upper bound: the conversion is real, but its high credit cost limits this particular strategy. These figures must not be added to the fresh model’s income.

Recommended plan adjustments

1. Keep the confirmed 100/50 gems-per-dollar rules and the four-pack catalogue as the first controlled test configuration. Publish exact weekly limits, reset time, guaranteed contents and whole-item denominations. For these studies, 32 hours means 1,920 one-minute items granted together; the hour-item sensitivity shows why denomination affects short tasks. No live products should be inferred from this analysis.

2. Implement and phone-test a deterministic total free reward program before using it in marketing or balance claims. Replace the disabled expedition route with existing PvE/weekly milestones unless expeditions are separately approved. Credit only the shortfall against the declared floor; keep random and competitive upside visible. Test 60-minute and 30-minute schedules without overnight event windows.

3. Resolve credit-driven acceleration and production growth before increasing paid resource quantities. The current optimized fresh paths reach the HQ cap early and accumulate large balances. Decide whether that pace is intended. Candidate changes for a separate balance experiment include a daily allowance on credit Finish Now, escalating credit cost per task duration, or a slower late building curve. These would change approved rules and were not applied.

4. Review unlimited repeat effectiveness, especially Standard armor. A20-repeat control is included as a hypothesis, not a live cap. A softer alternative is diminishing additional stat gains beyond a threshold while retaining repeat ownership and all prior investment. Test full combat and maintenance, not rarity labels. If approved repeats must remain unchanged, accept that low-rarity optimization and earned credits can dominate the intended premium rarity ladder.

5. Enforce account-wide quotas across all equivalent commercial routes. Keep standalone gems unlimited but do not promise unlimited combat growth from them. Offer durable cosmetic/convenience value and existing useful protection/queue choices; defer new uncapped power sinks and overlapping paid products until surplus-gem behavior is measured.

6. Protect meaningful free participation through achievable recovery, group objectives and matchup design. A candidate acceptance target is that a dedicated free tank contributes at least 20% real damage in a mixed three-player PvE team and that coordinated free players can beat one comparable-age spender in tested encounters. These are proposed acceptance criteria, not proved universal thresholds. Do not promise equal duels.

The hypothetical 20-repeat control holds the same Standard-focused spending policy and reduces its equipment to 720 attack, 2,160 hull and 63 armor, leaving about 128.1 million gear credits unused. In its controlled fight, Relic gear defeats that capped Standard profile in 12 rounds and retains 90.7% entry hull. This is evidence that repeat policy changes the rarity ladder; it is not a balanced final cap because unused resources were not reoptimized into higher rarities.

Decisions and limits remaining

Owner decisions: approve or revise pack contents; total free floors and HQ ramp; one-minute pack denominations; weekly equivalent quotas and reset; assignment-route timing; paid Builder 4 price; desired early/mid/late paid time lead; treatment of repeat scaling and credit acceleration; permitted shield/other gem utility; and bounty/event positive feedback controls.

Evidence limits: no real player telemetry, measured phone task times, random-reward population distribution, full contested-world travel/access simulation, repeated siege campaign, city ownership economy, collusive bounty runtime proof, retention/revenue forecast or profitability model. Combat probes test specified states and choices; their outcome is not a general win probability. The primary troop-loss workload is deliberately modest; full deployments can lose hundreds of troops per defeat.

The proposed baseline is optimistic economic specialization. Other choices are included, but no global optimum is claimed. Gear acquisition/maintenance is funded and sequential; the progression study reserves its money and later feeds endpoint combat. Gear-dependent combat success does not feed back into every earlier PvE reward. Paid event compounding and PvP farming could widen the gap. Optional branches should be validated in a disposable game environment before selecting final product quantities.

Reproduction and verification

Primary files: progression-model.mjs and progression-results.json; equipment-combat-model.mjs and equipment-fresh-integration.mjs plus equipment-fresh-results.json/equipment-fresh-combat-results.json; mature-model-v2.mjs and mature-results-v2.json. Progression and mature weekly CSV files contain all 26 weeks for every scenario. Equipment weekly ledgers are in their JSON results. Report numbers are regenerated from these files.

The C# analysis harnesses compile current Runtime sources into analysis-owned output folders and run memory-only synthetic worlds. No Unity Editor, APK or live server is needed. The raw current-rule extractions are progression-rules.json and equipment-authority-catalog.json. See progression-independent-audit.json, mature-independent-audit.json, equipment audit records and verification-v2.json for completed checks and input hashes.

Current-source references: Runtime/GameplayDurationBalance.cs; SpeedupItemPolicy.cs; SettlementQueuePolicy.cs; ResourceProducerExpansion.cs; GemCommercePolicy.cs; PhoneGameplayPolicy.cs; EquipmentEnhancementRepetition.cs; EquipmentRepairBalance.cs; CombatConsumablesPolicy.cs; LootBoxPolicy.cs; BountyHunterPolicy.cs; AuthoritativeSession.cs. Runtime paths begin unity-migration/Packages/com.ashline.gameplay/. Supporting pure definitions: progression.mjs; production.mjs; settlement-options.mjs; public/research-rules.js; public/troop-order-rules.js; public/troop-housing-rules.js. Detailed file/line references and SHA256 values are in reward-audit.txt/.json and equipment/progression audits.

Reference plan: https://chatgpt.com/space/page_a25122c9889081918d3634204281f8ba . It was read without editing. The historical source snapshots were not copied into game source. Preservation monitoring covers 520 fingerprinted records, not every file in the repository. The final check found 518 unchanged and two concurrent documentation updates in AGENTS.md and NEXT-CHAPTER.md for completed city territory borders. Those entries were read and retained. All monitored economy code and settings remain unchanged; this analysis wrote only its analysis folder.