There are only a handful of ice tracks worldwide — the push is built on roller tracks. What dry training delivers, where transfer ends and how to work with that.
The honest framing first: worldwide there are only around 16 internationally used ice tracks, four of them in Germany [5]. For the vast majority of bob athletes, push training therefore happens almost all year without ice — on push tracks, also called dry tracks, where a sled runs on rollers along a concrete lane. That is not a fallback, it is the normal case.
In Germany this infrastructure is unusually well developed: push tracks exist in Riesa, Magdeburg and Frankfurt among others, the Riesa facility as a double track since 1996 [1]. That is where the summer builds what decides hundredths in winter.
And now the part training plans often skip: transfer to ice is not one to one. A study using markerless 3D motion capture with active squad athletes shows that on ice compared with land, step frequency rises and stride, stance, swing and flight times shorten, with a smaller vertical centre-of-mass excursion — and most clearly so in the start phase [2]. The authors explicitly warn of a risk of negative transfer if start and acceleration are trained on land only [2]. This work is a preprint — treat it as a strong working hypothesis, not settled law.
A push track reproduces the first metres: block, handle, acceleration of a heavy sled, loading. What it does not reproduce is the friction situation on ice — and that is exactly where the differences in step frequency and contact times sit [2]. Use the track for force transfer and timing, not as an ice substitute.
Propulsion in the push is hip-extensor dominant; what decides is the hip extension velocity of the stance leg at ground contact and the hip flexion velocity of the swing leg [3]. On a dry track you can isolate exactly that because the load stays constant. Your target value is not your 30-metre time but the mean horizontal acceleration of the athlete-sled system per ground contact [3].
On a push track you have a fixed camera position, a fixed lane and repeatable conditions — the best place to capture the angle of attack systematically. In an elite analysis it was the strongest performance predictor across all push phases [4]. You will never get that repeatability in an ice channel.
The differences between land and ice are largest in the start phase [2]. The consequence: what you must take least literally on land is precisely the first ground contacts. Train them on land for strength and movement pattern — and correct them on ice as soon as you get there.
Some trials test single sled pushes alongside the classic combine battery [6]. A push track is the only place where you can practise both in combination: sprint values and sled work on the same day, under the same conditions.
Date, push split time, angle of attack. After one summer you will see whether your progress came from strength, technique or merely better weather. Without a log, dry training is a feeling, not a training process.
On a push track you accelerate a heavy sled on rollers — and then jump into it. Loading is the most delicate moment here too and must be practised exclusively under qualified coaching, never alone and never as a dare.
Dry starts are maximal sprint work against load. The typical costs are muscle injuries to the hamstrings and adductors plus lumbar spine overload. Increase volume and repetition count across weeks, not days, and end the session at the first clear drop in quality rather than at the last planned attempt.
This guide describes the training environment and the transfer situation, it does not replace coaching. Push track, sled and ice channel all run through club and federation.
Keep two separate leaderboards — one for land, one for ice — and never compare them. It sounds pedantic but it follows directly from the data: on ice the same person runs with higher step frequency, shorter contact and flight times and a smaller centre-of-mass excursion than on land [2]. Mixing the lists makes you read a surface change as a loss of form and correct in the wrong place. Two lists, two trends, two decisions.
A fixed facility where a bob is accelerated on rollers along a lane — the push is trained without ice. In Germany such tracks exist in Riesa, Magdeburg and Frankfurt among others; the double track in Riesa has been running since 1996 [1]. It is the standard environment for summer training.
On a push track, supplemented by sprint and strength work. Worldwide there are only around 16 internationally used ice tracks [5], so ice time is scarce. Training is hip-dominant: forward acceleration is carried by the hip extension velocity of the stance leg at ground contact [3].
The base transfers well, the fine structure does not. On ice, step frequency rises while stride, stance, swing and flight times fall, with a smaller vertical centre-of-mass excursion, most clearly in the start phase [2]. The authors warn of negative transfer from land-only training for start and acceleration [2]. That work is a preprint.
Split times and the angle of attack. In an elite analysis the angle of attack was the strongest performance predictor across all push phases [4], and the push track provides the repeatable conditions that make a clean measurement possible in the first place. The performance variable is horizontal acceleration per ground contact [3].
Yes, for the fundamentals. Sprint speed and heavy hip extension work can be built anywhere, and the combine events that predict push ability need no track [6]. What is missing without one is work on the sled itself — handle, timing and loading. That is exactly why training camps at a site with a track are worth it [1].
Push Harder: Sprint Power, Horizontal Force and Combine Preparation