Below the cut is a very geeky process-oriented post about THE PEOPLE’S POP POLL, which starts tomorrow on Twitter (@tomewing).

The main thing, though, is that over 100 people have nominated the songs which will fight it out in this poll, and donated over £2100 for Refuge (the fundraiser is still open!). They are amazing and make this whole thing something I’m very proud of.

You can listen to 186 of the songs here (the others are not on Spotify)


The big problem with setting up this poll is the diversity – not just in genre (that’s a good thing) but in fame: you don’t want a situation where every result is driven by how well-known a track is.

With the Number 1s and 2s polls this wasn’t so much of a problem. There was a demographic crunch in the 1s poll which meant the non-70s/80s stuff got ditched pretty fast, so for the 2s one I arranged it chronologically. This also had the happy effect of making the first round matches a lot tougher, which helps keep the poll interesting early on.

So I needed a way of bracketing the People’s Pop Poll which wouldn’t lead to many first-round mismatches and great songs going out without a struggle. Also, people paid money to nominate, so I wanted to make sure that every song got the fairest chance I could give it.

The solution I’ve come up with is to create brackets based on Spotify streams as a proxy for “well-known-ness”. It’s an imperfect proxy – famous older tracks will tend to have less streams than famous newer ones – but it broadly works, and means that songs will be competing (until the final rounds) within their “weight class”.

I ended up with 6 brackets of 32 tracks each, from the whales (75 million streams and up) to the minnows (300k streams or less). For the six tracks only on YouTube, I took the views of their most-viewed upload. It sort of works! I hope.

Having sorted that out I went through each bracket to put it into eight groups of 4, that being the number allowed by the sacred geometry of Twitter polls. This was harder – and more enjoyable. The aim was to make sure every group would be at least somewhat competitive, if not for winner then for second place.

My ground rules were:

  1. Rough similarity of style, era or ‘feel’.
  2. No tracks by the same artist in the same group.
  3. No tracks nominated by the same person in the same group.

(If I’ve failed on the third, my apologies – there were lots of nominations and JustGiving is – entirely fairly – not set up for going through comments)

So I’ve ended up with 48 groups, most of which have an internal logic which works (for me). I’ll be putting them up on Twitter over 8 days. Each day we’ll start with the whales and move down to the minnows, and I’ll accompany each poll with brief descriptions and links (so people can check out the obscure tracks).

I still don’t know how well this will work! There’s no unifying theme – which is great, it’s just a lot of tracks people passionately love! But might make for a weird poll. I guess I’ll see how things shake out after the first round.

Good luck to everyone who nominated – and I hope you enjoy it.