Meta

Reading a new patch before the guides catch up

·2 min read
Contents
  1. Quick answer
  2. Why this keeps happening
  3. What to do, in order
  4. How to measure it
  5. Mistakes that undo the work
  6. Frequently asked
  7. Read next
  8. Games worth trying

The two days after a patch are the only time when reading numbers beats copying builds. Four steps give a working picture before consensus forms.

Quick answer

  • Read only the numeric changes to characters or items you already understand.
  • Mark any change that crosses a threshold you rely on.
  • Play two casual games with the changed option before forming an opinion.

Why this keeps happening

Patch notes list changed numbers, and most changes do not matter. The ones that matter move a breakpoint: a damage value that now kills one hit earlier, or a cooldown that now fits twice in a fight.

Guides arrive after consensus, which means they describe what everyone already knows. The early edge comes from testing interactions nobody has written about yet.

What to do, in order

  1. Read only the numeric changes to characters or items you already understand.
  2. Mark any change that crosses a threshold you rely on, and ignore the rest.
  3. Play two casual games with the changed option before forming an opinion.
  4. Test one interaction you suspect has changed, rather than the whole kit.
  5. Write a prediction and check it a week later.
  6. Avoid tier lists for three days; they describe the previous patch.

How to measure it

Predictions that held after a week is the honest measure. Two out of five is normal, and the ratio only improves if predictions are written down.

Mistakes that undo the work

  • Assuming a nerf removes a pick from the game.
  • Copying a build without checking whether the numbers still support it.
  • Reading every line with equal attention.

Frequently asked

How long until the meta settles?
One to two weeks. Early statistics are distorted by experimentation.
Should I change my main?
Only if the change crosses a threshold you depended on.
Are early win rates useful?
As a signal, not as evidence. Sample sizes are small and behaviour is abnormal.

Read next

Games worth trying

Recommended games