How to select tracks for a DJ set

Technique keeps the music playing. Selection determines whether the room cares. A DJ with average technical skills and excellent taste will almost always outperform a technically precise DJ playing the wrong tracks in the wrong order for the wrong room.

This is not an argument against developing technical skills. It is an argument for giving selection the same deliberate attention most DJs give to their mixing. Most DJs spend hours on transition technique and five minutes on their crate before a gig. The ratio should probably be reversed.

The DJs whose sets people remember are almost never remembered for their technical execution. They are remembered for the moment when a particular track landed at a particular time and the room changed. That is selection. It is learnable. It requires attention, preparation, and willingness to adapt.

Before you select tracks for a gig, you need to know what kind of gig it is. Not just the genre. The demographics, the time slot, the promoter's expectations, the room's physical acoustic characteristics, whether it is a seated venue or a standing crowd, whether people are there to dance or there to socialise.

If you have played the room before, you have data. Which tracks got a response? Which died? Where was the energy highest? What time did the floor fill? Use that information to anchor your crate for the next time.

If you have never played the room, find a recording of someone who has. Hear what they played and when. You will learn more in forty-five minutes of listening than in hours of guessing.

If no recording exists, ask the promoter or another DJ who has played there. What bpm range does the room respond to? What time does the floor typically fill? What genre or era of tracks has historically worked? These are not secrets. Most promoters want you to succeed and will tell you.

Every track you play either raises the room's energy, holds it steady, or drops it. You need to know which of these a track does before you put it in your set. Discovering this mid-set, while it is already playing, is too late.

Energy matching does not mean energy must always go up. A set that constantly escalates without relief is exhausting to be in. Strategic drops, plateaus, and rebuilds are what make an arc feel alive rather than just loud.

Once you have a rough energy arc in mind, use key and BPM as filters to narrow your options. You are not looking for a single rigid path. You are looking for clusters of tracks that could plausibly follow each other without jarring transitions.

A practical approach: tag your tracks with Camelot key positions and BPM. When you are building a crate for a set, filter by BPM range first to establish a tempo corridor. Then filter within that corridor for tracks in compatible Camelot positions. You will end up with a shortlist of tracks that can transition smoothly between each other, even if you do not know the exact order yet.

This is not about making every mix perfectly harmonically compatible. It is about having options. When you are in the mix and you need to pivot, you want to reach for something you already know will work, not discover on the dance floor whether it does.

Selection is an editing process. You are not adding tracks. You are removing them. Start with more material than you need and cut aggressively until what remains is the tightest possible version of what the room needs.

If you have 200 tracks in your library that could work for this gig, your crate should probably contain 40. Of those 40, you might play 20 to 25. The other 15 are backup options: tracks that work if the floor needs something different than you anticipated.

The discipline is in the cutting, not the collecting. Adding tracks is easy. Removing the ones you love but that do not serve this room on this night is the actual skill.

The tracks to leave out are usually obvious in retrospect and invisible in advance. These are the signals: tracks you are playing because you want to hear them rather than because the room needs them; tracks that are so new you have not tested their energy in a room before; tracks that are technically perfect but emotionally wrong for the moment; tracks you are playing as requests when they would destroy the arc you have been building.

On requests: a well-framed request tells you something about what the room wants. The specific track requested rarely matters. If several people want something faster and harder, that is a signal. Respond to the signal, not necessarily the specific track.

Track selection errors are the hardest thing to identify in a live set but the easiest thing to hear in a recording. When you listen back, pay attention to the moments where the crowd response dropped, where the energy plateau set in, where the floor thinned. Those moments are almost always selection failures, not technical ones.

A track that killed momentum in a recording will usually do the same again. Build a list of tracks that did not work in the rooms you have played. Not because they are bad tracks, but because they did not serve those rooms. That list is as valuable as your list of guaranteed bangers.