Roller Derby Jammer Conditioning: Training Repeated Pack-to-Jam Accelerations

A jammer takes the whistle already boxed in, wedges a shoulder into the smallest gap in the wall, breaks through, gets two full seconds of open track — and then a blocker rides her hip into the infield, dumps her speed to almost nothing, and she's got ten meters of clear track to find race pace again before the wall reforms ahead of her. That sequence repeats four, five, sometimes seven times inside one two-minute jam. It repeats across twenty-plus jams a bout, on quad skates that never give a stride the free flight phase a runner's does, against four opponents whose entire job is making sure that re-acceleration never gets to finish clean.

Coaches who program jammer conditioning off a generic sprint-interval template keep running into the same surprise: a skater who times well in a straight, uncontested 40 on an empty track still gasses out by the third scoring pass of a hard bout. The straight sprint was never the limiting factor. The limiting factor is sprinting again immediately after getting knocked off your line, over and over, with no way to predict when the next hit lands or how much speed it costs you.

The Skate Mechanics Nobody Trains Around

Quad skates put an athlete in a lower, wider stance than running does, with no flight phase between strides and a much longer ground-contact window per push. Force comes almost entirely from hip abduction and adduction rather than a running gait's hip-flexion drive, and rolling resistance means a stopped or nearly-stopped skater has to re-establish momentum through several full strides before top-end speed is even available again. A repeated-sprint program built from track-and-field research and delivered as running intervals trains a real quality, but it doesn't train the specific mechanical problem of re-accelerating on skates from a dead or near-dead stop.

The Variable No Fixed Interval Controls

A standard repeated-sprint set uses a fixed work interval and a fixed rest interval because that's what makes it measurable and programmable. A jam doesn't work that way. A jammer might get a clean four-second break with no contact at all, or get walled up for eighteen seconds, absorb two hits, and have to find an entirely new gap before attempting the same distance of open track. The recovery a jammer actually gets between acceleration efforts isn't a training variable a coach sets — it's an outcome of how well her blockers are working and how good the opposing wall is that jam. Conditioning that only ever uses predictable, evenly-spaced rest windows leaves that unpredictability completely untrained.

Under WFTDA rules, a jam runs a maximum of two minutes, and jams are separated by roughly a 30-second reset before the next whistle. Lead jammer status lets a skater call the jam off earlier, which means jam length in practice varies constantly from a few seconds to the full two minutes depending on how the pack plays out. A bout runs two 30-minute periods, so a busy jammer can see well over twenty jam starts in a single bout, each one beginning from a dead stop behind the pack line.

Spencer, Bishop, Dawson, and Goodman (2005, Sports Medicine) reviewed time-motion data across field-based team sports and found repeated-sprint sequences typically involve efforts of two to four seconds separated by roughly 20-25 seconds of recovery, with work:rest ratios commonly falling between 1:5 and 1:7 depending on the sport studied. Derby's inter-jam window sits inside that general range, but its within-jam recovery does not — it isn't governed by a whistle or a clock, it's governed by whether the wall holds. The review's own limitation is worth flagging: the underlying studies used inconsistent speed thresholds to define a sprint and relied on video-based time-motion coding, so cross-sport comparisons carry real methodological noise, and none of the source data came from a skate-based sport with an intentional-contact component like derby.

Girard, Mendez-Villanueva, and Bishop (2011, Sports Medicine) reviewed the mechanisms behind repeated-sprint fatigue and found the decline across successive maximal efforts is driven by a mix of incomplete phosphocreatine resynthesis, accumulating muscle metabolites, and measurably reduced neural drive to working muscle on later sprints — not simple cardiovascular tiredness. Their review reports fatigue indices commonly in the range of several percent up to well over 10% depending on the sprint number, sprint duration, and recovery length used in each protocol. The authors are explicit about a limitation that matters directly here: most of the underlying research came from cycling, treadmill, or track-based repeated-sprint models with no mechanical perturbation between efforts, so how those fatigue mechanisms interact with an actual physical hit mid-effort is inferred, not directly tested.

That's exactly where Twist and Highton (2013, International Journal of Sports Physiology and Performance) fill in a piece the pure-sprint literature can't. Working with rugby league players, they documented that collision load produces a distinct, measurable fatigue signature — including depressed countermovement jump output after high-contact periods — separate from and additive to the fatigue produced by running demands alone. Their own limitation applies here too: the data comes from a padded, running-based contact sport, not a skate sport with quad-skate mechanics and a smaller, more concentrated impact zone, so the specific magnitude doesn't transfer directly to derby. What does transfer is the underlying principle: for a jammer, fatigue isn't just what repeated maximal accelerations cost metabolically. It's that cost plus a compounding mechanical fatigue from absorbing hits, and a conditioning block that only trains the first half of that equation is training half the problem.

Why a Clean-Start Sprint Test Undersells a Jammer

A standard flat-out skate sprint test measures a real quality, but it never asks the skater to produce that acceleration from a stopped position after losing momentum to contact — the exact situation a jammer faces on nearly every jam. This is a practical field protocol built to test that specific gap, not a peer-reviewed instrument, so treat the reference numbers below as coaching heuristics to track an individual skater's trend rather than validated pass/fail cutoffs.

Equipment

• Course: a 10m acceleration zone marked with cones on flat track or rink surface, plus a 2m contact zone at the start
• Contact simulation: a training partner applying brief, legal manual resistance at the hip for the first 1-2 seconds, or a resistance band anchored behind the skater
• Timing: timing gates or an IMU wearable at the 2m and 10m marks; a stopwatch works if gates aren't available

Procedure

1. Skater starts nearly stationary at the line, simulating a jammer held up at the pack.
2. On the whistle, the training partner applies resistance at the hip for 1-2 seconds, mimicking a wall hold or hit.
3. The skater must break the resistance and accelerate maximally through the remaining 8m to the timing gate.
4. Record the 2m-to-10m split time. Repeat for 5-6 reps with recovery varied between reps — 15, 30, and 45 seconds — to mirror the unpredictability of an actual jam.

Interpreting the Split-Time Decrement

Sample Session: Contact-Disrupted Repeated Accelerations

Total maximal-effort volume across the session lands around 150 skated meters of true acceleration work, deliberately low — the training stimulus here is the neuromuscular demand of repeated maximal starts under mechanical disruption, not distance covered. A jammer who paces any of these reps rather than skating them flat-out is training a different, less useful quality.

Programming It Across a Season

Run this block twice weekly in-season, separated by at least 48 hours, and once weekly during a heavier bout schedule to manage total fatigue. Re-test the blocked-start decrement every 4-6 weeks rather than every session — it's a near-maximal effort test, and running it too often adds fatigue cost without adding useful trend data.

• Training acceleration only from a clean start. Every drill that skips the contact-resistance element trains a quality a jammer rarely gets to use in an actual jam.
• Borrowing running-based sprint programming wholesale. Quad-skate acceleration mechanics differ enough from running that a program never performed on skates misses the specific stance and force-application demand.
• Using only fixed recovery intervals. A jammer's actual recovery between accelerations is dictated by pack contact, not a stopwatch — training only predictable rest windows leaves the worst-case scenario untrained.
• Measuring only best-effort speed. A fast single rep says little about what happens on rep five after two hits already landed, which is the number that predicts fourth-period performance.
• Retesting the blocked-start protocol too frequently. It's a near-maximal test; running it weekly during a demanding bout calendar adds fatigue without adding meaningfully new information.