"Managing OCs" for Legitimate Business is not "fill the queue." It is maximizing faction return per unit of the binding resource. There are two binding resources, and they bind at different layers:
So LB runs two coupled optimizations. Rank the Tier-4 group by EV per scope. Rank low-tier crimes by EV per member-day — they keep the cheap bench busy and top up scope, but they are a "just in case" reserve, not the engine. The faction is not choosing "which crime"; within the T4 group it fills every spawned crime (fill discipline, §1.2).
rewards block on completed crimes (live pull) supplies verified cash payout and scope yield (5) per crime — see §2. The board has now confirmed the missing spend: spawn cost = 4 scope on every Tier-4 crime. With cost and yield both known, EV_per_scope resolves to a real number. The one input still outstanding for a fully dollar-net ranking is live consumable market costs (the Firewalk tax) — tracked in §7.
Allocate the roster across three functional layers. The Tier-4 group (L2) is the engine and the mandate. The lower tiers (L1) are a "just in case" bench reserve, not the funding source; they are excluded from the core EV calculation.
Fill discipline — the canonical rule (leadership-confirmed 2026-08-08). The four crimes the gameplan optimizes are Break the Bank (~$285M avg live, reusable-heavy, low consumable tax), Clinical Precision (one heavy Imitator carry at 43%, 4 slots, cheap on member-days), Blast from the Past (~$126M, Firewalk now cheap — §4), and Window of Opportunity (item payout, fixed floor — §2.2). All crimes get completed, priority to T4; no crime is ever allowed to expire. There is no "spawn priority" — you can only select the tier you spawn from, not the crime (§6). The ranking BtB > CP > BFtP > WoO is a member's pick order when choosing which spawned T4 to join. WoO is the lowest-priority T4 pick — but like everything else, it does not expire unfilled.
The cheapest reliable crimes the low-CPR bench can clear. Purpose (leadership-confirmed 2026-08-08): the roster now carries more low-CPR members than there are recon slots in the T4 group, so lower-tier crimes are spawned to keep them productive rather than idle. This is a reserve, not the engine — excluded from the core EV calculation. Best candidates: Cash Me If You Can (3-slot L2), Mob Mentality (4-slot L1), Counter Offer (L5), Sneaky Git Grab (L6). Standing rule: every member runs the HIGHEST-level crime they meet the CPR requirement for — the bench is where you land only when nothing higher qualifies.
Tier-5 crimes (Stacking the Deck L8, Ace in the Hole L9, Crane Reaction L10) when carries are idle and roster depth allows. These out-rank Tier-4 on difficulty but need verified roster depth before the model recommends spawning them. Not part of the standing T4 mandate.
This is the menu the gameplan reasons over — not just what LB happens to be spawning. Weights are MAJOR-share % from the tornprobability role-weights API — the standing source per leadership (2026-08-08): CrimeHub lags on updates (they map every checkpoint before publishing), so it is not used for weights for now. CH tier is the scope band; L is per-crime difficulty. Disposition: running what LB runs today · add model-recommended · deprioritize · refused · bench overflow.
| Crime | CH | L | Slots | Heaviest role → wt | Disposition |
|---|---|---|---|---|---|
| Mob Mentality | 1 | 1 | 4 | Looter1 34 / Looter2 26 | add — scope factory |
| Cash Me If You Can | 1 | 2 | 3 | Thief1 50 | add — scope factory |
| Market Forces | 2 | 3 | 5 | Enforcer 29 / Negotiator 27 | bench |
| Smoke And Wing Mirrors | 2 | 3 | 4 | Car Thief 51 | bench |
| Gaslight The Way | 2 | 3 | 6 | Imitator3 41 | bench |
| Snow Blind | 2 | 4 | 4 | Hustler 48 | bench |
| Stage Fright | 2 | 4 | 6 | Sniper 46 | bench |
| No Reserve | 3 | 5 | 3 | Techie 38 | bench |
| Leave No Trace | 3 | 5 | 3 | Imitator 37 / Negotiator 34 | bench |
| Counter Offer | 3 | 5 | 5 | Robber 36 | add — scope factory |
| Guardian Angels | 3 | 5 | 3 | Hustler 42 | bench |
| Honey Trap | 3 | 6 | 3 | Muscle2 42 | bench |
| Bidding War | 3 | 6 | 6 | Robber3 32 | bench |
| Sneaky Git Grab | 3 | 6 | 4 | Pickpocket 51 | add — scope factory |
| Blast From The Past | 4 | 7 | 6 | Muscle 34 / Engineer 24 | running — 3rd pick |
| Break The Bank | 4 | 8 | 6 | Muscle3 32 / Thief2 29 | running — keep |
| Clinical Precision | 4 | 8 | 4 | Imitator 43 | running promote — graded |
| Stacking The Deck | 5 | 8 | 4 | Imitator 48 | bench — Tier 5 |
| Manifest Cruelty | 5 | 8 | 4 | Reviver 46 | bench — Tier 5 |
| Ace In The Hole | 5 | 9 | 5 | Hacker 36 (variant) | bench — Tier 5 |
| Gone Fission | 5 | 9 | 5 | Hijacker 25 / Imitator 25 | bench — Tier 5 |
| Crane Reaction | 5 | 10 | 6 | Sniper 41 | bench — Tier 5 |
| Window of Opportunity | 4 | 7 | 5 | Looter2 26 / Muscle1 23 | running — last pick, never expire |
tier is the scope band; difficulty (the L-number) is the per-crime knob. Tier-5 crimes exist (Stacking the Deck, Ace in the Hole, Crane Reaction at L10) and out-rank Tier 4 on difficulty. A wider model must not treat Tier 4 as the top of the menu.
Live get_faction_crimes_v2 (planning + recruiting, faction 9305, 2026-05-29): 17 active Tier-4 crimes. Planning: BFtP 2, BtB 2, Clinical Precision 1, WoO 4. Recruiting: BFtP 3, BtB 3, Clinical Precision 1, WoO 1 (plus 1 Mob Mentality on the L1 bench). Payout/scope figures are live averages from completed crimes, paid at the standing 90% convention (confirmed in rewards.payout.percentage). All four T4 crimes are currently spawned and must be filled (fill discipline, §1.2).
| Role | Material | Tier | Min CPR | Weight |
|---|---|---|---|---|
| Muscle 3 | Zip Ties (consumed) | MAJOR | 70 | 32% |
| Thief 2 | Hand Drill (reusable) | MAJOR | 70 | 29% |
| Muscle 1 | Hand Drill (reusable) | CORE | 50 | 14% |
| Robber | Hand Drill (reusable) | CORE | 50 | 13% |
| Muscle 2 | Zip Ties (consumed) | CORE | 50 | 10% |
| Thief 1 | Hand Drill (reusable) | RECON | — | 3% |
Model verdict: keep. Highest net cash in the catalog, low consumable tax (only Zip Ties ×2 burn; 4 reusables). The L2 cash-carry anchor. Note: live success 86% (20/3) is below the 100% the March KB carried — sample now larger.
| Role | Material | Tier | Min CPR | Weight |
|---|---|---|---|---|
| Imitator | Syringe (consumed) | MAJOR | 75 | 43% |
| Cleaner | Blood Bag : Irradiated (consumed) | CORE | 50 | 21% |
| Cat Burglar | None | CORE | 60 | 19% |
| Assassin | Chloroform (consumed) | CORE | 50 | 16% |
main-TSO3AJSP.js (2026-05-28). Imitator at 43% is the heaviest single role in the entire gradable catalog — heavier than any role on BtB or BFtP. This is a one-carry crime: it lives or dies on the Imitator, and only 4 slots means it is cheap on member-days. The other three roles are CORE; no slot is below 5%, so there is no true RECON throwaway here.
House CPR floors confirmed by leadership (updated June 6, 2026): Imitator 75, Cleaner 50, Cat Burglar 60, Assassin 50. The 43%-weight Imitator carry is gated hardest, as the model recommended — the earlier under-gating flag is resolved.
| Role | Material | Tier | Min CPR | Weight |
|---|---|---|---|---|
| Muscle | Zip Ties (consumed) | MAJOR | 80 | 34% |
| Engineer | Core Drill (consumed) | MAJOR | 80 | 24% |
| Bomber | Shaped Charge (consumed) | CORE | 70 | 16% |
| Hacker | Firewalk Virus (consumed) | CORE | 40 | 12% |
| Picklock 1 | None | CORE | 60 | 11% |
| Picklock 2 | None | RECON | — | 3% |
| Role | Material | Tier | Min CPR | Weight |
|---|---|---|---|---|
| Looter 2 | — | MAJOR | 80 | 26% |
| Muscle 1 | — | MAJOR | 80 | 23% |
| Looter 1 | — | CORE | 60 | 19% |
| Muscle 2 | — | CORE | 55 | 17% |
| Engineer | — | CORE | 50 | 14% |
Item rewards observed: Priceless Painting, Shabti / Ganesha / Vairocana Buddha Sculptures, Naval Cutlass, Companion Script: Ubay, Medieval Helmet. Painting pawn floor ~$70M FD 16549320. The rewards.money = 0 on every WoO success is a valuation-system artifact megathread 27274035, not a worthless crime — a manual-payout headache, not a reason to skip.
Once the model selects a crime, grading is mechanical and uses tornprobability weights only (leadership call, 2026-08-08 — CrimeHub excluded until it catches up). MAJOR = top weights summing to ~50–60% (carry slots, gate hard). CORE = mid weights (lower floor). RECON = lowest weight(s), typically <5% (no floor — bench and clock-starters). CPR floors per MAJOR are a house-rule decision (BigTasty420); the model sets the role, leadership sets the number.
| Crime | MAJOR (gate hard) | CORE | RECON |
|---|---|---|---|
| Break the Bank L8 | Muscle3 32, Thief2 29 | Muscle1 14, Robber 13, Muscle2 10 | Thief1 3 |
| Blast from the Past L7 | Muscle 34, Engineer 24 | Bomber 16, Hacker 12, Picklock1 11 | Picklock2 3 |
| Clinical Precision L8 | Imitator 43 | Cleaner 22, Cat Burglar 19, Assassin 16 | none <5% |
| Counter Offer L5 | Robber 36, Engineer 28 | Picklock 17, Hacker 12 | Looter 7 |
| Sneaky Git Grab L6 | Pickpocket 51 | Imitator 18, Techie 17, Hacker 14 | none |
| Cash Me If You Can L2 | Thief1 50 | Lookout 28, Thief2 22 | none |
| Window of Opportunity L7 | Looter2 26, Muscle1 23 | Looter1 19, Muscle2 17, Engineer 14 | none <5% |
Within any selected crime, the backbone is cycle time, not per-slot CPR maximization. Every member's planning timer is 24 hours from join. Carries are the highest-impact players; their idle time is the most expensive resource in the faction.
Recon (lowest-weight slot). Members who don't clear CORE thresholds. Their job is to start the clock. Recon joins first, on the throwaway slot (Thief 1 for BtB, Picklock 2 for BFtP). This burns the first 24h of cycle time on the cheapest player. Recon does not block Core or Carry — it accelerates them. Note: 4-slot crimes like Clinical Precision have no true RECON slot (no role <5%) — staging there is Core-then-Carry only.
Core (mid-weight). Mid-weight slots filled by members who clear CORE thresholds. Joins second. Priority: lowest countdown timer first, then highest-weight slot they qualify for, then best CPR match. Core members never wait for a Carry — they backfill behind Recon.
Carry (MAJOR weight). Members who clear MAJOR thresholds — they own the outcome. Carries join last on the crime with the lowest remaining countdown timer. By the time their 24h timer completes, Recon has burned 48h, Core 24h+, and the crime is at execution within hours of the Carry coming off cooldown.
Net effect: a 6-day BFtP that would take 8+ days end-to-end if everyone joined in parallel finishes in roughly 6–7 days when staged. Across 10 cycles per Carry per quarter, that's a 15–20% throughput multiplier on the faction's most expensive slot. Staging math is cleanest on 6-slot crimes; 3- and 4-slot crimes (CP, Counter Offer, Cash Me If You Can) compress the ladder but the order holds. The decision engine encodes the per-tier action tree — see the cross-reference panel.
The real EV of a crime is payout − consumable burn, not gross payout. This is what flips BtB/CP above BFtP. Reusables loan out and return; consumables drain on every execution.
| Crime | Reusables | Consumables (per success) | Tax |
|---|---|---|---|
| Break the Bank | Hand Drill ×4 (reusable) | Zip Ties ×2 (consumed) | low |
| Clinical Precision | None | Syringe ×1, Chloroform ×1, Blood Bag : Irradiated ×1 (all consumed) | mid |
| Blast from the Past | None | Zip Ties ×1, Core Drill ×1, Shaped Charge ×1, Firewalk Virus ×1 (all consumed) | high |
| Window of Opportunity | Wire Cutters ×1, Ladder ×1, Angle Grinder ×1 (all reusable) | Razor Wire ×1, Floor Cleaner ×1 (consumed) | mid |
rewards.payout.percentage = 90 confirms the 90% split is active across the queue.
BFtP Muscle MAJOR (34% weight, 75 CPR). March 6 KB identified only ~3 members clearing the lowered 75 threshold (Dr-Frog 81 clearly; FuzzyNaga 75 and Furai 76 at the line). The faction cannot staff more than ~3 simultaneous BFtPs at full quality. With 5 BFtPs active right now, at least 2 are running sub-threshold Muscles or stuck recruiting. Single highest-impact roster bottleneck.
Clinical Precision Imitator MAJOR (43% — newly graded). The heaviest single role in the catalog now has a name and a weight. Live success is 57% (12/9), well below BtB (86%) and BFtP (84%) — consistent with the carry slot being under-gated. Setting a house floor on the Imitator (Blocked-On B3) is the single highest-leverage grading decision available.
BFtP Engineer MAJOR (24% weight, 80 CPR). Higher bar than Muscle, similarly thin pool. The decision to hold Engineer at 80 while dropping Muscle to 75 was deliberate, but its math depends on whether 80 produces a meaningfully better success rate than 75. Open question §6.
BtB MAJOR depth. Staffable (Gregorian 70, vinye1mustaine 70, Noblewolf 74, Paprika 72 all clear), but thin enough that 5+ simultaneous BtBs strain it. Current queue has 5 active — at the line.
CPR data coverage gap. The faction_cpr endpoint only returns TornStats-tracked members. Untracked members do not appear in the CPR roster at all — this blinds every tier assignment for them. TornStats / YATA onboarding is a tractable, high-ROI fix. Blocked-On B6.
Scope is the binding constraint, not Carries — but the Tier-4 group is self-funding. Megathread 27274035 — Tier-4+ spawns cost scope, and the self-funding economics (§1) make the real constraint keeping the T4 success rate high enough that the net never goes negative: every failure is a scope loss with no replacement, which is the forum basis for fill discipline (§1.2).
The CPR-vs-weight balance rule. FD 16553833, Wood (karma 4073) — A low CPR in a low-weight slot is harmless padding. A 60 CPR in a 40%-weight slot is a disaster. Optimize the assignment, not the average. Especially load-bearing now that CP's Imitator is graded at 43%.
Crime XP is the dominant CPR input. Megathread 26823005, BigManBilly — Successful Crimes 2.0 history is the largest known CPR driver. Battle/job/skill stats are secondary multipliers that taper above 1B in any one stat. Idle members raise OC eligibility by grinding crimes, not BSS.
The Firewalk Virus structural shortage — superseded. Megathread 26842655 / 26957181, Allenone — projected structurally elevated prices. Leadership confirmed 2026-08-08 that the market moved the other way: Firewalk Virus is ~$4M. The old BFtP cost-deprioritization built on this thread is retired (§4).
The Tier-4 spawn distribution problem. Megathread 26957181, Allenone — Pre-WoO the Tier-4 pool skewed 50% BFtP / 25% CP / 25% BtB; the WoO release pushes toward 25/25/25/25 steady state. You cannot freely pick your Tier-4 spawn — you get a distribution across all four (BFtP, CP, BtB, WoO). The distribution hands you the crime; fill discipline (§1.2) is what stops it lapsing. Spawn-priority ranking governs which crimes we chase, never which we skip.
Window of Opportunity is the easy L7 — and it pays (fixed, now). FD 16549320, Crystal_Castles + DarthRevan — Higher CPRs across the board and item payouts (Painting + Shabti / Cutlass + Companion Script). The Elpexo-era $0-money artifact is fixed: Torn now assigns a fixed payout when any reward item carries a $0 market price (leadership-confirmed 2026-08-08). WoO is graded (tornprobability weights, §2.2) and sits last in the T4 pick order.
Consistency beats heroics. FD 16553833, Wood — 80% success on a moderate crime over a year beats 50% on a harder one. The decision engine's "70% sweet spot" encodes this; it is why we don't always chase the biggest crime.
The structural model (§1) is complete. These are the missing inputs that convert it into a ranked, dollar-valued recommendation. Each is a hard blocker, not a nice-to-have. Three original blockers — live payout, scope yield, and scope spawn cost — are now resolved (the last by board confirmation of the scope economics in §1) and removed from the list.
net_payout — especially Firewalk Virus (BFtP), Syringe / Chloroform / Blood Bag (CP). This is what turns gross payout into the real ranking and quantifies the Firewalk tax. Source: Torn market API per item ID.
faction_cpr only returns TornStats-tracked members. Untracked members are invisible to grading regardless of model quality. Onboarding mandate would unlock every downstream analysis.
rewards pulls to refresh the ~$87M planning figure. Low priority.
rewards supplies payout and yield (BtB ~$285M / BFtP ~$126M / CP ~$91M net cash; scope yield 5 on all Tier-4); the board confirmed the spawn cost, fixing the self-funding economics canonically documented in §1. With both inputs fixed, EV_per_scope resolves. Wired into §1 and §2.2; removed from this panel.
Status after the 2026-08-08 leadership review (Fox12_): Q1 confirmed (T4-group framing is standing doctrine, with completion-over-expiry as the hard rule), Q4 superseded (Firewalk ~$4M — no cost deprioritization; BtB > CP > BFtP > WoO stands as member pick order), Q6 answered (WoO is lowest-priority T4, never expired; payout now fixed by Torn). Remaining open items below.