What Cowardly actually does
On paper, Cowardly reads like the worst trait in the game, and the market prices it that way. The data tells a more interesting story. Two coefficients: courage −0.25 and threat −0.22. The courage cut is the second-deepest the wiki has datamined (only Brittle’s −0.26 goes lower), and it means exactly what you’d fear. When a delve turns bad, this hero’s combat AI scores “hold ground and keep fighting” far lower than a neutral hero’s does. They falter, hang back, and break early.
The threat side is the part everyone misses. −0.22 is by far the largest threat reduction in the game; the only other trait that lowers threat at all is Methodical, at a token −0.05. Enemy attention is a resource the dungeon spends every round, and a Cowardly hero attracts dramatically less of it. Put the numbers next to Unbreakable Will (+0.25 courage, +0.10 threat) and Cowardly is nearly its mirror image, except the threat swing runs much larger in Cowardly’s favour than in Unbreakable Will’s.
Within the fear family of Brittle, Shaken, Haunted and Superstitious, Cowardly is the only member that gets paid anything for its flaw. Brittle takes a slightly deeper courage cut and receives nothing in return, which is why the two deserve very different treatment despite near-identical courage numbers.
Reading it in the delve report
Watch how the same hero reads in the early rounds versus the late ones. Early, while the delve is going to plan, the courage penalty is dormant. Courage prices bad moments, and there haven’t been any yet. Meanwhile the threat reduction works every round: enemy pressure routes towards the other three heroes, and the report shows your Cowardly hero taking conspicuously few hits. Managers sometimes read that as the hero being carried. It isn’t. It’s the trait doing half its job.
The bill arrives when things sour. Wounds stack, a condition grinds the party down, and courage starts deciding who stays in the fight. A Cowardly hero is among the first to stop committing, and in a four-hero party that is a quarter of your action economy gone at the precise moment you can least spare it. Attrition conditions are the worst case: flooded halls and stacking curses push every fight towards the long, deteriorating shape in which courage gets consulted over and over.
One consolation the report will also show you: the hero who draws no attention and retreats early is usually the one who walks out intact. Under permadeath, that is worth something real.
Party fit and synergies
Sheltered backline roles are the natural home. A damage dealer or healer stationed behind a solid front line already avoids attention through positioning; −0.22 threat reinforces the arrangement, and the courage cut is rarely tested because the bad moments hit the front rank first. On an anchor the trait is disqualifying twice over: a tank who sheds enemy attention is failing the job description, and one who breaks early leaves a hole where the formation used to be.
Remember that the attention Cowardly sheds doesn’t vanish. It lands on the other three. Build accordingly. Partner them with heroes who want that attention and can carry it, the Brave, Dauntless and Unbreakable Will end of the courage ladder. A high-protectiveness teammate such as Loyal or Field Medic also helps, since a faltering hero benefits enormously from a party-mate whose AI prioritises covering others. Run the four through the synergy planner as usual; nothing about Cowardly disables named pairings.
The one hard rule: never field two fear traits together. One hero leaning out of a crisis is survivable. Two leaning out together is how routs start.
Market value, oaths and living with the flaw
Market prices punish the name harder than the numbers justify. Cowardly is a flaw you inherit (on an academy graduate, a bloodline heir, a bargain transfer) rather than one anybody seeks, and most managers dump these heroes reflexively. That reflex is your opportunity. A Cowardly hero with elite attributes, bought at the fear discount and stationed behind a stable front line, is some of the best value per gold on the transfer market. The recruitment guide covers the general method of pricing a flaw by what it costs your actual delve plan rather than by how it sounds.
When to pass: if your guild doesn’t yet have a settled, courageous front line, don’t buy the problem. The discount isn’t worth a hero you can’t shelter. Think twice before a promotion, too. Higher tiers mean longer, nastier dungeons, exactly the environment where the courage cut gets tested.
Living with one you already own is mostly a matter of honesty about role. Never make them the anchor, never build a plan that depends on them holding a collapse together, and their career can be long, because heroes who draw little attention and retreat early die less, which matters if a bloodline runs through them. Behaviour coefficients don’t fade with age, so training will not fix this; the arrangement is permanent, and the roster should be planned around it rather than in spite of it.
Version history
Both coefficients have been stable across every build this wiki tracks, v1.9.6 through v2.1.8. The courage-for-threat trade you’re evaluating today is the same one you’ll be living with next season, which makes discount-shopping for Cowardly heroes a strategy that survives patches.