The trait system is the part of Halls of Torment that new players say they do not understand, and the confusion is reasonable: it looks like you are choosing from a list, but you are choosing from a pool that the game filters by your level. Understanding that one detail changes how every level-up feels.
What to take — in priority order
| Priority | Take this | Why it is this high |
|---|---|---|
| 1 | **Anything that supports your starting ability** | Your starting ability is your main weapon for the whole run. A trait that makes it better is worth more than a bigger number attached to something you are not using |
| 2 | **Percentage multipliers over flat bonuses** | A percentage compounds with everything else you take. A flat bonus is worth the same at minute five as at minute thirty-five, which is to say almost nothing by the end |
| 3 | **Critical chance and critical damage — if your hero uses them** | Enormous for the Sorceress and the Beast Huntress, close to wasted on a summoner. This is the clearest example of why "best trait" depends on who you are playing |
| 4 | **Survivability, once it is what is killing you** | Health and recovery are not exciting, but a run that ends at minute twelve produces no gold. Take enough to reach the part of the run where your damage exists |
| 5 | **Movement speed** | Undervalued. It is how you collect more, avoid more, and stay in a position to keep attacking in a game where standing still is what kills you |
The rule underneath all five: ask whether the trait compounds. Anything that multiplies what you already have beats anything that adds to it, and that gap widens the longer the run goes.
Ranks and the pool — the part nobody explains
Two facts change how level-ups read once you know them.
Traits have ranks, written in Roman numerals. Rank II of a trait is simply a stronger version of rank I.
A rank only enters the pool when you reach the level that allows it. This is why the early game keeps offering you the same modest options — the better versions are not being withheld by bad luck, they do not exist in the pool yet.
The practical consequence is that holding out for a better offer does not work. The pool improves when your level improves, and your level improves by killing things. Take the best option in front of you and keep moving.
Traits last only until the end of the current run. Nothing you pick here is permanent — the progression that carries over lives in the Shrine of Blessings, your unlocked heroes, abilities and items.
The Cupbearer changes the maths
Once you unlock the Cupbearer, potions let you modify trait selection in three ways, and they are not equally useful.
| Potion effect | What it does | How good it is |
|---|---|---|
| **Remove a trait from the pool** | Deletes an option permanently for that run | **The best of the three.** It improves every future level-up, not just this one. Cut the traits your hero can never use and the pool gets denser with things you want |
| Reroll the offer | Redraws the current selection | Useful when a single level-up is critical, but it only fixes one screen and the underlying pool is unchanged |
| Take the same trait twice | Stacks one choice in a single level-up | Excellent on a run that has already found its multiplier, wasted early before you know what the build is |
The removal effect is worth understanding properly because it is the only one that compounds. A pool with five useless traits cut out of it produces better offers for the rest of the run, and that is a bigger effect than any single reroll.
The Shrine of Blessings matters more than this page
It is worth saying plainly: optimising level-up choices is less valuable than investing in the Shrine of Blessings, especially early.
Blessings are permanent, they apply to every hero, and they are fully refundable with no penalty. There is no way to waste gold there. They raise the floor of every run, which means every trait decision you make afterwards starts from a better position.
Because refunds are free, treat the Shrine as a build you reshuffle rather than a commitment. Switching from the Swordsman to the Sage? Pull your blessings out and put them into what the Sage needs. It costs nothing.
How this connects to your hero
Trait priorities are not universal, and the reason is that different heroes have different dead options in the pool:
- The Sorceress wants critical chance and critical damage above almost everything, because her kit is built on crits and Spark application
- The Shield Maiden turns Block Strength into damage, so defensive traits are never dead picks for her — she has the smallest pool of useless options in the game
- The Warlock and Beast Huntress lean on summons, which means many direct weapon traits do less for them than they appear to
- The Sage chooses its own main ability, so it can look at what the pool is offering and build toward it rather than around a fixed weapon
That last one is why the Sage tops the characters tier list — it is the only hero that can adapt to the pool instead of hoping the pool adapts to it.
Bringing it together
Support your starting ability first, take multipliers over flat numbers, and stop waiting for better offers that the pool cannot produce yet. When you have the Cupbearer, spend potions on removing bad traits rather than rerolling, because that is the effect that improves every remaining level-up.
And before any of that, put your gold in the Shrine. It is free to undo, it applies to everyone, and it raises the floor that all of these decisions are made from.
