So this thingie tells you which instances are tooting a hashtag the most. It’s part of my goal to make instance discovery more fun and worthwhile.
You can try it out here. My score calculation is pretty dumb at the moment —a simple count of toots that use the tag. Any suggestions how to weight the data so populous instances don’t hog the top spot all the time would be greatly appreciated. Have fun!
Fibonacci average! 
Divide each subsequent toot by the golden ratio in relation to the value of the prior toot.
E.G.
1 + 0.618 + 0.382 + …
That should flatten out your data a bit :3
edit
To clarify, this won’t displace the largest instances from the top, you’d need to add another datapoint like posting activity or weigh each toot against the total toots that instance has produced to do that, but it will make the data flatter so the gross disparency is lessened.
Thanks! I tried just the unweighted Fibonacci and it didn’t do much, but folding it into the toot activity of the instance worked pretty well!
score += 1/(statuses*fib) // for each toot found at the instance
Fib changes to next F in the sequence per count. You don’t know how happy this makes me. 