Running band, pebble and water film: how a 20-kilo granite stone glides 45 metres, why it bends and why the physics behind it remains unsettled.
A curling stone weighs 17 to 20 kilos, is made of polished granite and travels some 45 metres across the ice [1] — and the honest answer to why it bends along the way is: science has not settled it [2]. Not "complicated". Not "physicists only". Unsettled.
That is one of the few sentences in sport you get to say with relish. Curling has been played for centuries, televised for decades and studied in high-end journals — and the mechanism that pushes the stone sideways still has no single explanation everyone agrees on [3].
What is settled is how the stone runs. It does not sit flat at all. It rests on a narrow annular running band on its underside [3] — the contact area is a fraction of the stone's footprint. And the ice beneath is no sheet of glass but a pebbled surface of fine, frozen water droplets [4]. A ring riding on a field of tiny domes: there has never been so little contact.
Once you have understood that, you play differently. Not better overnight — but with a completely different eye for the sheet.
A legal stone weighs 17 to 20 kilograms and is made of polished granite [1] — roughly a full crate of water bottles, which nobody lifts. The underside is what matters: the stone stands on a narrow running band, not on its whole face [3]. That tiny contact zone is why such a heavy body slides down a 45-metre sheet at all [1].
Curling ice is scraped and pebbled before every game — fine water droplets are sprayed on and freeze into small domes [4]. The stone runs on their tips. In nipping, ice technicians shave those tips to an even height; the result is faster, curlier ice, whereas oversized pebble gives heavy ice with a lot of curl [4]. The sheet is a workpiece, not a constant.
Now the interesting part. Uniform friction all around the running band cannot produce a systematic sideways impulse; every explanation needs an asymmetry between front and back or left and right [3]. The most obvious candidate — more friction at the back than the front — delivers, even at its extreme, only about half of the sideways deflection actually observed [3]. So the other half is missing.
Standing side by side are, among others, the scratch-guide model (the rough running band cuts slightly angled grooves into the ice depending on rotation direction, and the stone is guided along them), the pivot-slide model, the evaporation-abrasion explanation and the snowplough approach [3]. One elegant experiment favours the scratch guide: using artificially cut scratches, researchers steered stones onto a zigzag path [3]. That is not proof — but it is a strong hint.
Sweeping is the only intervention you have left after release. The broom generates frictional heat that forms a thin water film on the ice; friction drops, and the stone runs farther and straighter [2]. And the dose is measurable: doubling the downward force doubles the heat generated, while doubling the speed gives only a factor of 1.55 [5]. In Olympic sweepers this raised the ice surface by 0.75 to 1.75 degrees (men) and 0.50 to 0.75 degrees (women) [5] — one degree here decides half a metre of running distance.
The amount of curl is a property of today's ice, not of your stone. For orientation: championship ice curls about five to six feet on draws, club ice around four [4]. And it moves during the game: under stones and brooms the pebble breaks down, and the sheet gets noticeably faster from the second or third end [4]. So measure, do not remember.
A stone of 17 to 20 kilos [1] is not a piece of equipment you carry. It gets pushed, pulled and set down — never lifted, never gripped with fingers under the rim. Two stones colliding will trap a fingertip without effort.
Nobody who is not sweeping has any business in front of a running stone. And whoever sweeps has both feet in grippers — the slider's Teflon sole is made for gliding, not walking.
For this whole topic: this guide explains physics, not technique on a human body. Everything touching knees, back and shoulders belongs in coached instruction — and no guide replaces that.
Do the running-band test as soon as someone at the club lets you: tip a stone, run a finger along the band, then run the same finger over the polished surface next to it. The difference is clearly noticeable — the contact zone is rougher than the rest [3].
That roughness sits at the centre of the scratch-guide model [3]. So in two seconds you are touching the very detail physics has been wrestling with for years. After that you look at every sheet differently — worth more than any mnemonic.
A legal stone weighs 17 to 20 kilograms and is made of polished granite [1]. It is not thrown but released out of the delivery and allowed to glide; the sheet is around 45 metres long [1]. It is never carried — pushed and set down.
Because it barely touches. The stone rests on a narrow running band instead of its whole face [3], and the ice below is pebbled — the band runs on the tips of small frozen droplets [4]. On top of that, a thin water film at the contact point lowers friction further [2].
Yes. The microscopic mechanism has not been conclusively explained [2]. The classic account via asymmetric friction delivers, even at its extreme, only about half the observed sideways deflection [3]. Models such as scratch-guide, pivot-slide, evaporation-abrasion and snowplough still stand side by side [3].
The broom generates frictional heat that forms a thin water film; friction drops and the stone runs farther and straighter [2]. Force is the big lever here: double the downward force means double the heat, double the speed only a factor of 1.55 [5]. Measured ice-surface warming reached up to 1.75 degrees [5].
How Dangerous Is Curling? Knees, Back and How Warming Up and Technique Protect You