CRATEDROP

HowtheHOTFeedDecidesWhatRanks

3 min read·10 April 2026

CrateDrop has two feed sorts: NEW and HOT. NEW is chronological, no ranking involved. HOT is a scored ranking that factors in upvotes, comments, and how old the post is. This is how it works.

The formula

Every post in the HOT feed gets a score calculated as: (upvotes times 2 plus comments) divided by (hours since posting plus 1) to the power of 1.8. Posts need at least one upvote to appear in HOT at all. If a post has zero upvotes, it stays in NEW but does not enter the ranked feed.

Why upvotes count double

Comments and upvotes both signal engagement, but they signal different things. A comment means someone had a reaction strong enough to type something. An upvote means someone thought the post was worth surfacing to others. Both are useful signals, but the upvote is a deliberate signal about quality rather than just activity. Counting upvotes at double weight reflects that.

Why the exponent is 1.8

The exponent on the age term controls how fast posts decay out of the top positions. An exponent of 1.0 would make decay linear. At 1.8, a post that is 10 hours old has already lost significantly more position than a post that is 1 hour old, even if both have the same number of upvotes. This means recent posts with fresh engagement can overtake older posts with more total upvotes.

The effect is that the HOT feed rewards posts that are gaining traction now, not posts that accumulated upvotes over several days. A post from yesterday with 20 upvotes will rank below a post from two hours ago with 6 upvotes and active comments. If you want to appear in HOT, timing and early engagement both matter.

What upvoting actually does

Upvoting a post is not a like. It is a ranking signal. When you upvote something, you are increasing its score in the HOT calculation and helping it surface to other people in the feed. A post with zero upvotes is invisible in HOT regardless of how many comments it has. One upvote is the minimum to enter the ranked feed. From there, additional upvotes combined with recency determine where it sits.

Random records from 16 million in the Discogs database — heard instantly on YouTube.

Start Digging →