Each crew member i carries a morale mi ∈ [0,1] and every pair (i,j) carries a cohesion score Cij ∈ [0,1]. Every simulated day, random pairwise interaction events fire; each one succeeds or turns into friction with a probability set by the current mission stress S(t):
P(positive interaction) = 0.78 - 0.5·(S(t) - S0) + 0.2·(C_ij - 0.5)
success → C_ij += 0.03(1 - C_ij), m_i, m_j += small boost
conflict → C_ij -= 0.05·C_ij, m_i, m_j -= small penalty, conflict++
m_i(t+dt) += dt·[ 0.15·(mean_j C_ij − 0.5) − 0.2·(S(t) - S0) − 0.05·(m_i − 0.5) ]
Mission stress S(t) combines three real, documented drivers of long-duration crew friction:
S(t) = S0 + 0.35·crowding + 0.15·(delay/24)
+ 0.35·exp( -(t/T − 0.75)² / (2·0.05²) ) ← "third-quarter effect"
where S0 = 0.25 is the unavoidable floor stress of any long isolation mission
Fix applied in this 2D build: the original 3D engine used the raw stress S(t) directly in both the interaction-probability and morale-drift terms instead of the stress above baseline, (S(t) − S0). Since S0 = 0.25 never goes away (it's the mission's floor stress even at zero crowding and zero comm delay), every run — regardless of slider settings — drifted to near-total morale collapse (verified numerically: default sliders reached ~0.2% average morale by mission end, and even the best-case 0% crowding / 0 min delay settings only reached ~1%). Recentring both terms on S0 restores the intended behaviour: best-case missions stay healthy, and crowding / delay / the third-quarter dip become the actual differentiators instead of everything converging to zero.
- Crowding — habitat volume per person; denser stations raise baseline friction (ISS / submarine crew studies).
- Comm delay — one-way Earth-Mars light-time (0–24 min) reduces real-time psychological support from mission control, nudging conflict risk up.
- Third-quarter phenomenon — morale in Antarctic winter-over and Mars-500-style isolation missions dips sharpest not at the start, not at the end, but around 70–80% of the way through; modeled here as a Gaussian bump in S(t) centered at 0.75·T.
The 2D view shows the crew ring from a raking angle: each astronaut's vertical offset from the ring encodes (mi − 0.5) — confident, cohesive crews float above the ring; strained ones sag below it. Connecting lines fade from green to red as Cij falls. Drag to pan the view, scroll or pinch to zoom.