OODA Coefficients Reference

Complete catalog of every tunable coefficient in the OODA Loop System (Feature 032). All coefficients live in the OODADefines class within src/babylon/config/defines.py unless otherwise noted.

Access pattern:

from babylon.config.defines import GameDefines

defines = GameDefines()
ooda = defines.ooda  # OODADefines instance

# Example: get base observe time
ooda.base_observe_time  # 1.0

Cycle Time Weights 

Used in babylon.ooda.cycle_time.compute_cycle_time() to compute the four-phase additive OODA cycle time.

Formula:

\[T_{cycle} = T_{observe} + T_{orient} + T_{decide} + T_{act}\]

Where:

\[\begin{split}T_{observe} &= \texttt{base\_observe\_time} + \texttt{sensor\_latency} \times \texttt{latency\_weight} \\ T_{orient} &= \max\bigl(\texttt{base\_orient\_time} \times (1 - \texttt{coherence} \times \texttt{coherence\_weight}),\ \texttt{orient\_time\_floor}\bigr) \\ T_{decide} &= \texttt{decision\_base} \times (1 + \texttt{bureaucratic\_depth} \times \texttt{depth\_weight}) \\ T_{act} &= \texttt{base\_act\_time}\end{split}\]

Parameter	Default	Constraints	Description
`base_observe_time`	1.0	`gt=0`	Base Observe phase duration (ticks).
`latency_weight`	0.5	`ge=0`	Weight of `sensor_latency` on Observe phase.
`base_orient_time`	2.0	`gt=0`	Base Orient phase duration.
`coherence_weight`	0.6	`ge=0, le=1.0`	Weight of `ideological_coherence` on Orient phase.
`base_act_time`	1.0	`gt=0`	Base Act phase duration (fixed, not modified by profile).
`coord_weight`	0.3	`ge=0`	Weight of coordination on Act phase (reserved for future use).
`depth_weight`	0.4	`ge=0`	Weight of `bureaucratic_depth` on Decide phase.
`orient_time_floor`	0.1	`ge=0.0`	Minimum Orient phase duration. Prevents coherence from zeroing out Orient.

Code location: src/babylon/config/defines.py:2141-2177, used in src/babylon/ooda/cycle_time.py:17-39.

Decision Mode Base Times 

The decision_base term in the Decide phase formula. Looked up by DecisionMode enum value in babylon.ooda.cycle_time.compute_cycle_time().

Parameter	Default	Constraints	Description
`decision_mode_base_autocratic`	1.0	`gt=0`	AUTOCRATIC: single decision-maker, fastest.
`decision_mode_base_delegate`	2.0	`gt=0`	DELEGATE: leader delegates to trusted subordinates.
`decision_mode_base_democratic`	3.0	`gt=0`	DEMOCRATIC: majority vote among members.
`decision_mode_base_consensus`	5.0	`gt=0`	CONSENSUS: full agreement required, slowest.

Ordering invariant: AUTOCRATIC (1.0) < DELEGATE (2.0) < DEMOCRATIC (3.0) < CONSENSUS (5.0).

Code location: src/babylon/config/defines.py:2179-2199, used in src/babylon/ooda/cycle_time.py:42-58.

Initiative Scoring Weights 

Five-component weighted sum in babylon.ooda.initiative.compute_initiative_score().

Formula:

\[I = w_{speed} \times \frac{1}{T_{cycle}} + w_{inst} \times B_{inst} + w_{ci} \times C + w_{embed} \times E + w_{mom} \times M\]

Parameter	Default	Constraints	Description
`initiative_weight_speed`	2.0	`ge=0`	Weight for speed component (inverse cycle time).
`initiative_weight_institutional`	1.0	`ge=0`	Weight for institutional bonus (state advantage).
`initiative_weight_counterintel`	1.5	`ge=0`	Weight for counter-intelligence capability.
`initiative_weight_embeddedness`	1.0	`ge=0`	Weight for community embeddedness.
`initiative_weight_momentum`	0.5	`ge=0`	Weight for recent success momentum.

Code location: src/babylon/config/defines.py:2201-2226, used in src/babylon/ooda/initiative.py:22-67.

Institutional Bonus by Jurisdiction 

The \(B_{inst}\) term in the initiative formula. Looked up by JurisdictionLevel enum value.

Parameter	Default	Constraints	Description
`institutional_bonus_federal`	5.0	`ge=0`	NATIONAL (federal) jurisdiction. FBI, DEA, ATF.
`institutional_bonus_state`	3.0	`ge=0`	STATE jurisdiction. State police, National Guard.
`institutional_bonus_local`	1.5	`ge=0`	COUNTY or MUNICIPAL jurisdiction. Local PD, sheriff.
`institutional_bonus_nonstate`	0.0	`ge=0`	Non-state organizations (factions, civil society, business).

Code location: src/babylon/config/defines.py:2228-2248, used in src/babylon/ooda/initiative.py:148-167.

Momentum 

Momentum rewards consecutive successful actions and decays exponentially per tick.

Formula:

\[\begin{split}M_{t+1} = M_t \times \texttt{momentum\_decay} + \begin{cases} \texttt{momentum\_success\_bonus} & \text{if action succeeded} \\ 0 & \text{otherwise} \end{cases}\end{split}\]

Parameter	Default	Constraints	Description
`momentum_decay`	0.8	`ge=0, lt=1.0`	Exponential decay factor per tick. Higher = slower decay.
`momentum_success_bonus`	0.2	`ge=0`	Bonus added per successful action.

Code location: src/babylon/config/defines.py:2250-2261, used in src/babylon/ooda/initiative.py:127-145.

Action Cost Modifiers 

Community-modified action costs in babylon.ooda.action_costs.compute_action_cost().

Formula:

\[\begin{split}\texttt{modifier} = \begin{cases} \max\bigl(\texttt{min\_cost\_modifier},\ 1.0 - \texttt{overlap} \times \texttt{embeddedness\_discount}\bigr) & \text{if embedded (overlap > 0)} \\ \texttt{contradiction\_cost\_multiplier} & \text{if across contradiction axis} \\ \texttt{outsider\_cost\_multiplier} & \text{otherwise (outsider)} \end{cases}\end{split}\]

\[\texttt{effective\_cost} = \max\bigl(1,\ \lceil \texttt{base\_cost} \times \texttt{modifier} \rceil\bigr)\]

Parameter	Default	Constraints	Description
`embeddedness_discount`	0.5	`ge=0, le=1.0`	Cost discount factor for embedded orgs. Higher = steeper discount.
`contradiction_cost_multiplier`	2.5	`gt=1.0`	Cost multiplier when org and target are on opposite sides of a contradiction axis.
`outsider_cost_multiplier`	1.5	`gt=1.0`	Cost multiplier for orgs with no membership in the target community.
`min_cost_modifier`	0.5	`gt=0, le=1.0`	Floor cost modifier. Even fully embedded orgs pay at least 50% of base cost.

Code location: src/babylon/config/defines.py:2263-2285, used in src/babylon/ooda/action_costs.py:34-82.

Consciousness Effect Limits 

Per-tick clamping in babylon.ooda.action_effects.compute_consciousness_delta().

Parameter	Default	Constraints	Description
`max_ci_delta_per_tick`	0.05	`gt=0, le=1.0`	Maximum absolute CI delta per action per tick. Prevents runaway consciousness shifts.

Code location: src/babylon/config/defines.py:2287-2293, used in src/babylon/ooda/action_effects.py:85-88.

Action Base Consciousness Multipliers 

The action_base term in the consciousness delta formula. Zero means the action has no consciousness effect (e.g., AGITATE affects contestation instead).

Formula:

\[\Delta_{CI} = \text{clamp}\bigl(\texttt{tendency\_modifier} \times \texttt{cadre} \times \texttt{cohesion} \times \texttt{credibility} \times \texttt{action\_base},\ \pm\texttt{max\_ci\_delta}\bigr)\]

Parameter	Default	Constraints	Description
`action_base_educate`	1.2	`ge=0`	EDUCATE: direct political education. Highest CI multiplier.
`action_base_agitate`	0.0	`ge=0`	AGITATE: increases contestation, not CI directly. Zero base.
`action_base_provide_service`	0.6	`ge=0`	PROVIDE_SERVICE: serve-the-people programs. Moderate CI effect.
`action_base_recruit`	0.3	`ge=0`	RECRUIT: bringing in new members. Low CI effect.
`action_base_organize`	0.5	`ge=0`	ORGANIZE: building community structures. Moderate CI effect.
`action_base_propagandize`	0.8	`ge=0`	PROPAGANDIZE: media and messaging. High CI effect.
`action_base_repress`	0.8	`ge=0`	REPRESS: state violence. Backfire CI effect (always +CI on target, REVOLUTIONARY tendency).
`action_base_surveil`	0.2	`ge=0`	SURVEIL: state surveillance. Smaller backfire CI effect than REPRESS.
`action_base_assimilate`	1.0	`ge=0`	ASSIMILATE: institutional co-optation. Negative CI effect (pushes LIBERAL tendency).

Special case – PROVIDE_SERVICE tendency split (hardcoded in _get_effective_action_base):

REVOLUTIONARY org: uses action_base_provide_service (0.6)
LIBERAL org: uses action_base_provide_service * 0.3 (0.18)
Other tendency: zero effect

Code location: src/babylon/config/defines.py:2295-2340, used in src/babylon/ooda/action_effects.py:231-246.

Autonomy Tradeoff 

The autonomy-breadth tradeoff in babylon.ooda.constraints.apply_autonomy_modifier().

Formula:

\[\begin{split}\texttt{effectiveness} = \begin{cases} 1.0 & \text{if } n_{targets} \leq 1 \\ \max\bigl(0.1,\ 1.0 - \texttt{autonomy} \times \texttt{scale} \times \frac{n_{targets} - 1}{n_{targets}}\bigr) & \text{otherwise} \end{cases}\end{split}\]

Parameter	Default	Constraints	Description
`autonomy_effectiveness_scale`	0.5	`ge=0, le=1.0`	Scale factor for the autonomy-effectiveness tradeoff. Higher = steeper penalty for spreading actions.

Code location: src/babylon/config/defines.py:2342-2348, used in src/babylon/ooda/constraints.py:92-116.

Agitation-Contestation Coupling 

AGITATE actions increase ideological_contestation on the target community. When contestation exceeds a threshold, EDUCATE actions receive a bonus multiplier.

Parameter	Default	Constraints	Description
`agitation_contestation_delta`	0.1	`ge=0, le=1.0`	Contestation increase per AGITATE action.
`agitation_educate_bonus`	1.5	`ge=1.0`	EDUCATE CI delta multiplier when community contestation exceeds threshold.
`contestation_threshold`	0.3	`ge=0, le=1.0`	Contestation level at which EDUCATE bonus activates.

Mechanic: AGITATE first to raise contestation above 0.3, then EDUCATE with 1.5x multiplier. This models how agitation creates the conditions for effective political education.

Code location: src/babylon/config/defines.py:2350-2367, used in src/babylon/ooda/action_effects.py:78-82 (EDUCATE bonus) and src/babylon/ooda/action_effects.py:150-163 (AGITATE resolver).

Lifecycle Modifiers 

Elder legitimacy in babylon.ooda.lifecycle_capacity.elder_legitimacy_bonus().

Parameter	Default	Constraints	Description
`elder_legitimacy_multiplier`	1.3	`ge=1.0`	CI delta multiplier when organization has elder (D’-phase) members. Elders lend legitimacy.

Code location: src/babylon/config/defines.py:2369-2374, used in src/babylon/ooda/lifecycle_capacity.py:44-70.

Counter-Intelligence 

Counter-intelligence score progression.

Parameter	Default	Constraints	Description
`counter_intel_increment`	0.1	`ge=0, le=1.0`	Counter-intel score increment per successful COUNTER_INTEL action.

Code location: src/babylon/config/defines.py:2376-2382.

Base Action Point Costs 

Default AP cost for each of the 21 action types. All are integers with ge=1. Modified at runtime by babylon.ooda.action_costs.compute_action_cost().

Parameter	Default	Description
`base_cost_educate`	1	Political education. Cheap, high CI impact.
`base_cost_agitate`	1	Raise contestation. Cheap, no direct CI effect.
`base_cost_fundraise`	1	Generate resources.
`base_cost_employ`	1	Business auto-metabolism.
`base_cost_surveil`	1	State surveillance. Cheap, low backfire.
`base_cost_map_network`	1	Intelligence mapping.
`base_cost_propose_alliance`	1	Diplomatic overture.
`base_cost_denounce`	1	Public condemnation.
`base_cost_recruit`	2	Grow membership.
`base_cost_organize`	2	Build community structures.
`base_cost_propagandize`	2	Media and messaging.
`base_cost_provide_service`	2	Serve-the-people programs.
`base_cost_repress`	2	State violence. Moderate cost, high backfire.
`base_cost_protest`	2	Public demonstration.
`base_cost_counter_intel`	2	Counter-intelligence operations.
`base_cost_attack_infrastructure`	2	Sabotage community infrastructure.
`base_cost_assimilate`	2	Institutional co-optation.
`base_cost_strike`	3	Labor withdrawal. Expensive, powerful.
`base_cost_expropriate`	3	Seize property/resources. Expensive.
`base_cost_infiltrate`	3	Intelligence infiltration. Expensive.
`base_cost_build_infrastructure`	3	Build community infrastructure. Expensive, lasting effect.

Code location: src/babylon/config/defines.py:2384-2411.

Layer 3 Propagation Coefficients 

Used by babylon.ooda.layer3.process_layer3() to propagate action consequences to community state.

Parameter	Default	Constraints	Description
`repress_heat_delta`	0.15	`ge=0.0, le=1.0`	Heat increase per REPRESS action. Heavy state attention.
`surveil_heat_delta`	0.05	`ge=0.0, le=1.0`	Heat increase per SURVEIL action. Light state attention.
`build_infrastructure_delta`	0.1	`ge=0.0, le=1.0`	Infrastructure increase per BUILD_INFRASTRUCTURE action.
`attack_infrastructure_delta`	0.1	`ge=0.0, le=1.0`	Infrastructure decrease per ATTACK_INFRASTRUCTURE action.

Code location: src/babylon/config/defines.py:2413-2428, used in src/babylon/ooda/layer3.py:97-222.

Module-Level Constants 

Constants defined outside OODADefines, typically structural data or loop safety limits.

Structural Constants 

Constant	Location	Description
`_CONTRADICTION_PAIRS`	`action_costs.py:25-31`	5 tuples of `(CommunityType, CommunityType)` defining contradiction axes: SETTLER/NEW_AFRIKAN, SETTLER/FIRST_NATIONS, SETTLER/CHICANO, PATRIARCHAL/WOMEN, PATRIARCHAL/TRANS.
`_UNIVERSAL_ACTIONS`	`action_eligibility.py`	`frozenset` of 12 ActionTypes available to all org types: RECRUIT, ORGANIZE, EDUCATE, AGITATE, PROPAGANDIZE, FUNDRAISE, PROTEST, COUNTER_INTEL, MAP_NETWORK, PROPOSE_ALLIANCE, DENOUNCE, BUILD_INFRASTRUCTURE.
`_NPC_PRIORITIES`	`npc_stub.py:20-47`	Priority queues by org type. STATE_APPARATUS: SURVEIL first. POLITICAL_FACTION: EDUCATE first. CIVIL_SOCIETY: PROVIDE_SERVICE first. BUSINESS: EMPLOY first.

Loop Safety Limits 

All OODA modules enforce static upper bounds on iterations per the project’s loop safety rule. These are not tunable game parameters.

Constant	Value	Location	Scope
`_MAX_ACTIONS_PER_ORG`	20	`constraints.py:15`	Maximum actions processed per organization per tick.
`max_orgs`	1000	`layer0.py:39`	Maximum Business orgs processed in Layer 0.
`max_results`	1000	`layer3.py` (5 functions)	Maximum action results iterated per sub-processor.
`max_edges`	1000	`action_costs.py:102,141`	Maximum edges iterated for membership overlap.
`max_nodes`	1000	`action_effects.py:281`	Maximum nodes iterated for community membership.
`max_actions`	20	`npc_stub.py:78`	Maximum NPC actions selected per org.

Hardcoded Multipliers 

Two values are hardcoded in function bodies rather than in OODADefines:

Value	Location	Purpose
0.3	`action_effects.py:243`	LIBERAL tendency multiplier for PROVIDE_SERVICE (`action_base * 0.3`).
0.1	`constraints.py:116`	Autonomy effectiveness floor. Prevents complete ineffectiveness.
0.01	`action_effects.py:68`	Minimum overlap for credibility (`max(overlap, 0.01)`).

Coefficient Derivation Chain 

Every tunable coefficient in OODADefines traces back to one of five empirical irreducibles through a three-level hierarchy. Field descriptions use category tags to indicate provenance:

[A] Direct substitution from an existing source primitive
[B] Computed from source primitives via an explicit formula
[C] Grounded in empirical data (COIN studies, census, organizational theory)
[D] Theoretically justified but not computationally derivable
[E] Engineering constant (structural value, keep as literal)

Level 0: Empirical Irreducibles 

ID	Value	Symbol	Source
IR-1	7 weeks	\(\tau_{1/2}\)	Political half-life. FM 3-24 counterinsurgency doctrine: insurgent activity recovers to 50% within 6–8 weeks post-operation. Produces \(\lambda = \ln 2 / 7 \approx 0.1\).
IR-2	0.8	\(\alpha\)	Imperial extraction capacity. Amin/Emmanuel unequal exchange: ~80% of periphery surplus transferred to core.
IR-3	0.3	\(p_c\)	Network percolation threshold. For social networks with \(\langle k \rangle \approx 3\text{--}4\), site percolation \(p_c \approx 1 / \langle k \rangle\).
IR-4	1.5×	rent premium	Gentrification rent premium. Census/HUD data, UCLA Urban Displacement Project: median rent increase of 40–60% during rapid displacement.
IR-5	0.2	\(\Delta S\)	George Floyd solidarity shift. Pew Research 2020: 20 percentage point shift in white BLM support.

Level 1: Source Primitives 

Source primitives are GameDefines constants that OODA coefficients reference. Each derives from a Level 0 irreducible.

Primitive	Default	Derivation	From
`consciousness.decay_lambda`	0.1	\(\ln 2 / 7\)	IR-1
`consciousness.agitation_decay_rate`	0.1	= `decay_lambda`	IR-1
`consciousness.routing_scale`	0.1	= `decay_lambda`	IR-1
`consciousness.sensitivity`	0.5	\(\lambda / (1 - \alpha)\)	IR-1 + IR-2
`economy.extraction_efficiency`	0.8	Amin/Emmanuel	IR-2
`solidarity.activation_threshold`	0.3	\(p_c \approx 1/\langle k \rangle\)	IR-3
`solidarity.scaling_factor`	0.5	= `consciousness.sensitivity`	IR-1 + IR-2
`territory.heat_decay_rate`	0.1	= `decay_lambda`	IR-1
`territory.heat_spillover_rate`	0.05	= `heat_decay_rate / 2`	IR-1
`territory.high_profile_heat_gain`	0.15	= `rent_spike × heat_decay`	IR-1 + IR-4
`territory.rent_spike_multiplier`	1.5	Census/HUD	IR-4
`territory.eviction_heat_threshold`	0.8	= \(\alpha\)	IR-2
`struggle.solidarity_gain_per_uprising`	0.2	Pew Research	IR-5

Level 2: OODA Coefficients 

Coefficient	Default	Cat	Formula / Source
`repress_heat_delta`	0.15	A	= `territory.high_profile_heat_gain`
`surveil_heat_delta`	0.05	A	= `territory.heat_spillover_rate`
`momentum_success_bonus`	0.2	A	= `struggle.solidarity_gain_per_uprising`
`agitation_contestation_delta`	0.1	A	= `consciousness.agitation_decay_rate`
`momentum_decay`	0.8	B	\(1 - 2\lambda\)
`max_ci_delta_per_tick`	0.05	B	\(\lambda / 2\)
`action_base_educate`	1.2	B	\(1 + 2\lambda\)
`action_base_propagandize`	0.8	B	\(1 - 2\lambda\)
`action_base_repress`	0.8	B	\(\alpha\)
`action_base_surveil`	0.2	B	\(1 - \alpha\)
`action_base_provide_service`	0.6	B	\(k + \text{routing\_scale}\)
`action_base_organize`	0.5	B	\(k\) (sensitivity)
`action_base_recruit`	0.3	B	\(p_c\) (activation threshold)
`contestation_threshold`	0.3	B	\(p_c\) (activation threshold)
`agitation_educate_bonus`	1.5	B	= `territory.rent_spike_multiplier`
`embeddedness_discount`	0.5	B	= `solidarity.scaling_factor`
`elder_legitimacy_multiplier`	1.3	C	\(1 + \text{ideology\_institutional\_weight}\)
`counter_intel_increment`	0.1	C	\(\lambda\) (Sparrow network entropy)
`outsider_cost_multiplier`	1.5	C	= `territory.rent_spike_multiplier` (Prebisch-Singer)
`contradiction_cost_multiplier`	2.5	C	\(\approx \sqrt{4.2}\) (incarceration disparity)
`institutional_bonus_federal`	5.0	C	FM 3-24 force density ratio
`institutional_bonus_state`	3.0	C	RAND state police force ratio
`institutional_bonus_local`	1.5	C	Galula administrative presence premium

Runtime Validation 

Call OODADefines.validate_derivations(game_defines) to detect coefficient drift when source primitives are changed via YAML overrides or scenario configs:

defines = GameDefines.load_from_yaml("custom_scenario.yaml")
drifts = defines.ooda.validate_derivations(defines)
if drifts:
    for msg in drifts:
        print(f"WARNING: {msg}")

Code location: src/babylon/config/defines.py, validated in tests/unit/ooda/test_coefficient_derivations.py.