What Greedy actually does
Three coefficients, and every one of them is about self-interest: kill hunger +0.25, recklessness +0.1, protectiveness −0.15. The combat AI weights finishing blows noticeably higher, tolerates a little more danger to get them, and scores “cover a party-mate” options lower across the board.
On the kill-hunger ladder, Greedy sits one rung above Ambitious (+0.15) and one below Vain (+0.3), with Glory Hunter (+0.5) at the top. The interesting part is not the ladder itself but the riders. Ambitious is clean: kill hunger and nothing else. Greedy and Vain both pay for their extra hunger with a protectiveness cut, while Glory Hunter, despite being the hungriest of the four, keeps its protectiveness intact and pays in threat and recklessness instead. In this engine, “greed” is modelled as self-interest: the hero doesn’t just want the kill more, they care about the party less.
The −0.15 protectiveness is a modest cut in absolute terms (Lone Wolf is −0.5 for comparison), but it points in the opposite direction from everything a supporting role is built to do, and that matters more than the size of the number.
How it reads in a delve report
Watch two things when a Greedy hero comes back from a delve. First, the killing blows: they will cluster on this hero, the same way they do for Ambitious, only more so. Fights end sooner, and on floors where curses or flooding tax you every round, ending fights sooner is straightforwardly good. With only four heroes in the party, every enemy removed early is a compounding action-economy win.
Second, watch for what the hero didn’t do. Protectiveness governs covering, guarding and aiding, and a −0.15 there means that in rounds where the AI is weighing “step in for the wounded healer” against “finish the wounded enemy,” the enemy wins more often than it should. The failure mode isn’t dramatic; it’s a pattern of small absences. Your support takes a hit that a neutral hero would have absorbed, in a round where the Greedy hero was busy harvesting a kill that would have happened anyway. Under darkness or other visibility-punishing conditions, where positioning errors get amplified, the recklessness rider makes those absences a touch more expensive.
Neither effect will jump out of a single report. Read three or four in a row and the shape becomes obvious.
Where Greedy fits in a party
Damage dealers wear Greedy well. Their job description already says “finish things,” the protectiveness cut costs them least because nobody was relying on them for cover, and the party structure around them exists to absorb exactly this kind of tunnel vision. On a dedicated finisher, Greedy is close to Ambitious-but-more, and that’s a compliment.
Healers and supports should not carry it. A support with reduced protectiveness is fighting their own trait every round, and the kill hunger actively tempts them into target selection they have no business making. Anchors are a middling case: the line holds, but a tank who scores guarding lower is a slightly worse tank, full stop.
Trait pairings within the party do real work here. A Loyal or Team Player hero (+0.4 protectiveness each) standing next to your Greedy finisher covers the gap the finisher leaves. And if a hero rolls Greedy alongside Loyal, the protectiveness maths nets comfortably positive, and the combination plays like a hungrier, slightly rash team player rather than a mercenary. Two selfish-rider heroes in the same four is where compositions start leaking; run one, and check the synergy planner to make sure your pairings reward the focused damage you’re now committed to.
Recruiting, oaths and when to pass
On the market, treat Greedy as a conditional buy: worth a fair price on a finisher, worth a discount on anyone else, and a hard pass on support recruits regardless of their other numbers. Behaviour coefficients don’t fade with age the way attributes do, so a Greedy veteran keeps the sharpened target selection deep into their career; the trait holds value longer than the stat line around it.
There’s a roster-management wrinkle worth budgeting for. Kill hunger concentrates killing blows, killing blows build reputation, and reputation walks into the oath renegotiation with its hand out. A Greedy hero who has spent two seasons topping the kill count is exactly the profile that turns a routine renewal into a standoff, and the name of the trait should tell you how the standoff goes. Run the numbers in the oath negotiator before the term ends, and decide in advance whether you’re paying or selling. With a hero like this, drifting into the meeting unprepared is how you lose it.
Selling is genuinely fine, by the way. Kill-hungry finishers show well: their delve records are full of exactly the moments buyers look at.
Version history
Every coefficient on Greedy (the +0.25 kill hunger, the +0.1 recklessness, the −0.15 protectiveness) has been stable from v1.9.6 through the current v2.1.8. Whatever you build around this trait, patches have given no reason to expect the ground to move.
Renewal modifier
Greedy multiplies the renewal ask by 1.10 and subtracts 15 percentage points from haggle odds. Oathbound or Loyal takes precedence if present. This is separate from the ordinary base-wage modifier; the Oath calculator applies the renewal adjustment after the base wage you enter.