A sim gives you one big DPS number and, if you ran Best Gear or Droptimizer, a ranked list of upgrades. The list is usually more useful than the number. Most people trip on the same two things with their first sim: reading the DPS as a prediction, and treating tiny upgrades as real.
What the headline DPS number means
The big number is your average damage per second over thousands of simulated fights, played perfectly against a target that never moves. The sim doesn't move, dodge mechanics, lag or make mistakes, so your real DPS will be lower. That's expected. The number is for comparing versions of your character against each other, and it won't match your combat log.
Sims are much better at relative questions ("is item A better than item B for me?") than absolute ones ("how much will I do on this boss?").
When an upgrade delta is real and when it's noise
Every sim has some random variation from procs, crit streaks and trinket timing. The error margin on the result shows how big that wobble is. When you read the upgrade list:
- If a delta is bigger than the error margin, it's a real upgrade. Equip it or spend your roll on it.
- If it's inside the error margin, call it a tie and choose on something else: whether you'll replace that slot next week anyway, set bonus progress, or which one looks better.
- Trust the order of the list more than the exact numbers. The gap between rank 1 and rank 5 means something. The decimals on a single row don't.
The ability breakdown
Under the headline number you get each ability's share of your total damage. If an ability you barely press is near the top, something is off with the sim profile or your talent import. The abilities doing the most damage are also the ones where a missed cast costs you the most in game.
What sims leave out
SimulationCraft models your character's damage output on its own. It doesn't model movement and downtime, boss mechanics, the target swaps you'd make in a real pull, raid utility, defensives, or how well you play. A spec that sims 3% behind another can still be the better pick on a fight where its mobility gives it more uptime.
The same goes for the trinket leaderboards on this site. They rank standard profiles, so use them as a starting point and sim your own character before you commit. The tier list is built from real logged parses instead of sims.
Frequently asked questions
Why is my simmed DPS higher than what I do in raid?
The sim plays your character perfectly, with every cooldown on time, against a target that never moves or goes immune. Real fights have movement, mechanics, deaths and downtime. The simmed number is your best case, and your logs will come in under it.
Is a 1% DPS upgrade worth chasing?
Usually yes if it's free, like an item already in your bags or a guaranteed drop. It's harder to justify if it costs a bonus roll or days of farming that could go toward something bigger. Go by the order of the list more than the size of one delta.
What does the error or confidence interval on a sim mean?
A sim repeats the fight thousands of times with random variation (procs, crits) and reports the mean. The error figure describes how far that mean could move if you ran it again. Two items whose deltas overlap within the error are statistically tied.
Do more iterations make a sim more accurate?
More iterations shrink the random error of the mean, making rankings between close items more stable. They do not fix model accuracy: if SimulationCraft models a spec mechanic imperfectly, a million iterations converge on the same imperfect answer. Default iteration counts are enough for gear decisions.
Why did my DPS change when I re-simmed with nothing different?
Some of it is run-to-run randomness inside the error margin. The rest is SimulationCraft updates. The engine picks up hotfixes and spec fixes every week, so the same character can sim differently after a patch. Compare items within one report instead of headline numbers across weeks.
Paste your /simc string on the home page to get a result of your own to read.
Sim your character