Green zombies bite humans who then become Zombies, but recover after the "infection duration." Alter duration of infection. Alter benefit, cost of wearing protective armor. Individuals may learn to wear armor. All learn by observing all their next-door neighbors' outcomes.
Zombies (green) and humans (peach) start randomly scattered. Zombies spread between unarmored neighbors. Recovery occurs but immunity doesn't last. All choose to wear armor or not. Armor is costly and uncomfortable but protects against zombie bites. Everyone observes neighbors and learns what is the highest payoff. All see neighbors' payoff (benefit of having healthy neighbors minus the cost of wearing protective armor). An unprotected person next to 1 healthy neighbor gets benefit (b). A person protecting pays the cost (c) but if they are next to 8 healthy people gets benefit 8b, so their payoff is 8b - c (you can alter b and c with sliders). Some learn to "free-ride." That's avoid the cost of protecting with armor (the armor's heavy and hard to make). Such free-riders see a benefit of letting others bear the brunt of wearing armor while avoiding the cost themselves. Try to find a benefit and cost that fosters learning of wearing armor. Or find levels that make an interesting story in the fight between zombies and humans. Zombies are green then, as they recover, turn towards yellow (unless yellow? button is off). People recover (peach) for a period then are susceptible again. All imitate the next-door neighbor with highest payoff. As an interesting result, eventually, humans cluster, and Zombies invade in clusters. It's an Evolutionary Learning model. It's based on infection model I published in J of Theor Biol. Here infection is becoming a zombie, masks are armor.
