Am I Rushing or Dragging? A Drummer's Diagnostic Guide

Diagnosis · Timing · Updated 2026-08-16

Almost every drummer has been told they rush. Almost none have been told by how much, in which direction, and where in the bar. And you cannot know it from the inside: rushing feels like the click slowing down, and locking in often feels like being slightly late.

This page separates drift (your placement slides steadily one way) from scatter (your average is fine but every stroke lands somewhere else), because they need opposite fixes and one accuracy score cannot tell them apart. It also builds the gap-click self-test into a ten-rung progression with numbers attached, not one setting you pass or fail.

Rushing, dragging, and the sign convention

Rushing means your strokes land before the grid position they belong to; dragging means they land after it. Here, negative is early and positive is late, so -18 ms is a stroke played 18 milliseconds ahead of the beat. Half the confusion in timing arguments comes from two people using "ahead" to mean opposite things.

Dragging is also not laying back. Laying back is a controlled, stable offset held on purpose: the same few milliseconds late every bar, pulse untouched. Dragging is uncontrolled, usually progressive, and the tempo goes with it. A steady 20 ms behind for three minutes is a style. A 20 ms lag that has become 60 ms by the last chorus is a fault.

Symptom, cause, fix

Most timing complaints are not "I have bad time." They are situational, and the situation is the diagnosis. The third column names the fault type, explained in the next section.

Drift and scatter are different faults

Drift is a sloped mean. The errors march one way: early by 2 ms, then 6, then 10, then 15. Every stroke can be acceptable and the take is still ruined by bar 32, because they accumulate. Scatter is a wide spread. The average is dead on and every stroke is somewhere else: 18 early, 22 late, 30 early, 14 late.

Watch the summary rows. Player A scores 75 percent and Player B scores 62.5 percent, so on a leaderboard A wins. But every one of A's errors points the same way: A is not inaccurate, A is accelerating, and eight more bars at the same slope puts the last stroke near 68 ms out, close to half a 16th note. B's mean is 0.0 ms and B still sounds unsteady, because the strokes never agree with each other. A needs pulse work. B needs the opposite, because scatter at speed is nearly always a stroke problem. Swap their drills and you waste a month.

Why rushing is more common than dragging

Teachers, engineers and click tracks report the same lean: when time goes, it usually goes forward. That is a consistent observation rather than a measured universal, and plenty of players drag instead.

• Arousal raises the felt pulse. Adrenaline, stage volume, an audience, a red record light. Your hands follow the pulse you feel, not the one you set.
• Tension shortens the stroke. A tighter grip means less rebound travel, so the stick returns sooner and the next stroke arrives early. Rushing is often cured at the hands, not at the metronome.
• Early is the safer error in a band. Behind, you are audibly chasing everyone. Slightly ahead, everyone can follow you. Players learn that lean and stop noticing it.
• The excitement lives in the fills. Note density jumps there, and density under tension compresses. Most "I rush" complaints are really "I rush for one bar in four."
• Rushing is self-reinforcing. Once you are 20 ms in front, the click genuinely sounds late, so the instinctive correction is to push harder.

Where dragging actually shows up

Dragging is not rare, it is concentrated. It clusters in quiet passages, where small strokes carry less rebound energy; at slow tempos, where 16th notes at 60 BPM sit 250 ms apart; in half-time and shuffle feels; at the end of long sets; and on the first note after a rest.

The gap-click self-test

The gap click removes the external reference for a known length of time, then shows you where you ended up. Plenty of advice stops at "practice with gaps." One gap setting is a pass or a fail; a progression is a diagnosis. Keep one pattern all the way up the ladder - a single stroke roll in 8th notes works - so the only variable is the silence.

1. Fix the pattern before the tempo: Pick one pattern you can play without thinking and keep it all session. If the vocabulary changes between rungs you cannot tell whether a miss came from the silence or the sticking.
2. Count out loud, then silently, then not at all: Losing the count during the silence is a different failure from drifting through it, and you need to know which one you have.
3. Judge only the return: The one measurement is the first stroke of the bar where the click returns. Early sounds like a flam with you in front, late like a flam with the click in front. The silence itself tells you nothing.
4. Two clean reps before you move up: One good rep proves you got lucky for a few seconds. Two in a row at the same rung is the minimum evidence, which is why a tempo ladder should be gated on accuracy, not on a timer.
5. Log direction, never pass or fail: Write "early" or "late" on every rep. Ten reps that all say early is a diagnosis. Five and five is scatter wearing a drift costume.

Notice rung 9. It runs faster than rung 8 but holds 8.0 seconds of silence against rung 8's 19.2, because a bar at 120 BPM lasts 2.0 seconds and a bar at 100 BPM lasts 2.4. A faster rung is not automatically a harder drift test - it is a harder technique test. Seconds of silence measure your clock; tempo measures your hands. Climb the tempo axis with the rudiment tempo chart.

Turning a gap-click miss into a tempo number

A 16th note at 100 BPM sits 150 ms from its neighbor. So if you return from four bars of silence a full 16th early, you were playing at about 101.6 BPM. Nobody hears 1.6 BPM. Everybody hears where it puts you ten seconds later, and by 16 bars a full beat has vanished. The errors that wreck takes are too small to feel, which is why "am I rushing" is answerable by arithmetic and never by introspection.

Stopping the rush inside fills

• Halve the density first. Play the fill in 8th notes before 16ths. Most fill rushing is a density problem, not a time problem: the hands are asked for more strokes per second than they can place, so they compress.
• Aim at the downbeat, not at the fill. The target is the 1 on the other side: the fill is the approach, the landing is the note being graded. Take the click away where you rush, setting it to beat 1 only or to 2 and 4, since a fill played over a 16th-note click tests nothing.
• Do sixteen reps and count directions. Three bars of groove plus one of fill, logging early or late on each landing. Fourteen of sixteen saying early beats any take you happen to remember, and it slots into the 10, 20 and 30 minute pad routines.

One hand ahead of the other is not rushing

If your right hand lands 12 ms before its grid position and your left lands 12 ms after, your mean timing error is zero. You are not rushing or dragging, you are lurching, and any average-based measurement will declare you fine. It is worst on alternating singles and compounds inside doubles. Test one hand at a time against the click, then check the doubles with how to tell if your double strokes are even. The fix is hand-by-hand isolation, not more metronome time.

The tape method, honestly

Most articles on this subject end with "record yourself and listen back," and they end there because the advice is good. It is also incomplete. A recording tells you reliably that something is wrong and only vaguely what, because the ear catches drift across 16 bars easily and a 15 ms displacement on one stroke barely at all. And if the click is not on the take, you are comparing your playing against your memory of the tempo.

• Print the click on the take. A reference that existed only in your headphones is not something you can measure against afterwards.
• Look, do not only listen. Drop the file into a free DAW at the recorded tempo and compare the attack of bars 1, 5, 9 and 13 against the grid. Four numbers reveal the slope you cannot hear.
• Make two separate passes. One for feel, one purely for the landing after every fill. Judging both at once means judging neither.

What an accuracy percentage can and cannot tell you

If you measure, insist on the window. A percentage without a window is not a number: 85 percent inside 50 ms and 85 percent inside 20 ms describe two different players. A percentage is also just a count of strokes in a box, discarding the direction of every miss and the order they arrived in - the two things that separate drift from scatter. Several apps now listen through the microphone and give per-hit early and late feedback, several of them on device, and drum apps that listen to your playing surveys who does what.

Measure it instead of guessing. Drum Coach₊ from Panda Taps is a rudiment trainer and practice pad coach for iPhone. It listens through the microphone and scores every stroke against the grid, marking each one on-time, rushed, dragged, missed or extra against a default 25 ms window. Per-stroke timing histograms show the shape of your error instead of one opaque number, so drift and scatter look visibly different on screen. A first-run latency calibration stores an offset per audio output route and re-runs when the route changes, so wireless headphones work. Everything is analyzed on device and no audio is recorded, saved or uploaded. Coming soon for iPhone. Learn more about Drum Coach₊.

Frequently asked questions

How do I know if I am rushing or dragging?

Not by feel. Rushing makes the click sound like it slowed down, and locking in often feels slightly late, so introspection points the wrong way. Run a gap-click test where the click drops out for one to eight bars, note whether the returning click lands in front of you or behind you, and log the direction over at least ten reps. Ten reps that all say early is a diagnosis. A five and five split is inconsistency, not a lean.

Why do I speed up when I drum?

Four mechanisms stack up. Adrenaline and volume raise the pulse you feel, and your hands follow the felt pulse rather than the one you set. Tension shortens the stroke, so the stick rebounds sooner and the next hit arrives early. Being slightly ahead is the safer error in an ensemble, so players learn to lean forward. And once you are ahead the click genuinely sounds late, so the instinctive correction is to push harder.

Is playing behind the beat the same as dragging?

No. Laying back is a controlled, stable offset held on purpose: the same few milliseconds late bar after bar, with the underlying tempo untouched. Dragging is uncontrolled and usually progressive, and the tempo goes with it. The test is whether the offset stays constant. Twenty milliseconds behind for a whole song is a feel. Twenty milliseconds that has become sixty by the last chorus is a fault.

How much timing error is actually audible?

It depends on tempo, on context, and above all on whether the error is steady or growing. A small constant offset can pass unnoticed for a whole take, while the same amount of error accumulating bar after bar is obvious to everyone. At 100 BPM, playing at 101.6 BPM is inaudible in the moment and puts you a full 16th note out after four bars. That is why direction and slope matter more than any single figure, and why an accuracy percentage means nothing unless the window it was measured against is stated.

Keep reading

• Drum Timing Accuracy in Milliseconds: How Close Is Close Enough?
• Drum Rudiment Tempo Chart: What BPM to Practice At
• Drum Apps That Listen to Your Playing: What They Actually Measure