[donw]

We also need to track exactly how loud those corrections are published.

Calling out "hey, we totally got this one wrong, and here's why" even "louder" than original prediction is great. This is how you earn trust.

Quietly publishing a correction on the metaphorical back page? Nope. Not even close.

Silently changing the article after the fact? Oh hell no. Go directly to credibility jail, do not pass Go, do not collect $200.