What the letters mean
Tiers (S, A, B, C, D) are just buckets we draw across a continuous stat — average placement. On brains.gg an S-tier comp is one whose boards average roughly 4.1 or better; A-tier sits a little above that, and so on down. The cutoffs are consistent for every comp, which is what makes the ranking fair, but they're still lines drawn on a spectrum. A high B-tier comp and a low A-tier comp are nearly identical in strength.
The practical takeaway: don't obsess over the letter. Two comps in adjacent tiers are interchangeable in power. The tier is a starting filter, not a ranking you should follow blindly.
Average placement is the anchor
Every tier on this site is driven by average placement — the mean finish across all boards we classified into a comp. In an eight-player lobby the expected placement is 4.5, so anything below that is beating the table on average. A comp averaging 4.2 is genuinely strong; one averaging 4.7 is a coin flip. Because placement rewards consistency, it's a far more honest measure than win rate alone, which only counts first places.
Use pick rate to find value
Pick rate tells you how contested a comp is. A comp with a great placement and a low pick rate is often the best-kept secret on the patch: few players are fighting you for its units, so it's easy to hit. A comp with a mediocre placement but a huge pick rate is usually over-forced — everyone piles in, the units get contested, and the average player's result suffers. Reading placement and pick rate together is how you spot the lines worth forcing.
Always check the sample size
A tier built on a few hundred games is far more trustworthy than one built on twenty. We show the game count beside every stat and hold back rows below a minimum threshold, but early in a patch even published comps can be volatile. If a comp rocketed to S-tier overnight on a small sample, treat it with healthy skepticism until the numbers settle.
Put it together
Read a tier list like this: shortlist the top one or two tiers, cross off anything with a suspiciously small sample, then pick a flexible line whose pick rate isn't sky-high. For the full definition of each metric, see our methodology, and when you're ready, open the comps tier list and apply it.