Four metrics describe curling performance almost completely. How to record them honestly in a club team — and why they are orders of magnitude, not study values.
Four metrics describe curling performance almost completely: shot percentage, hammer conversion, steal efficiency and force efficiency. In the elite band the shot percentage sits at 88 to 90 per cent, a skip at 92 per cent counts as outstanding, and teams below 88 per cent as a team practically do not medal at major events [1].
And now the sentence that governs this entire guide: these figures come from analyses of competition data, not from peer-reviewed research. They are orders of magnitude, not study values — the weakest class of source we work with at all. For tactical orientation they serve. For training, effectiveness or health claims they do not, and they are not used that way here.
Why it is still worth it: a club team argues about impressions after a game. A team with numbers argues about ends. The difference is not accuracy — it is that you can argue about numbers, whereas about impressions you can only go quiet.
Rate every stone by how far it met the call — not by how it felt. The scale does not matter as long as it stays the same across the team. For orientation: elite sits at 88 to 90 per cent, a skip at 92 counts as outstanding [1]. Do not expect 85 at club level — and if you get it, your scale is more likely too kind than your team too good.
The hammer is the last stone of an end; it goes in the next end to the team that did not score, and stays put after a blank end [2]. In competition analytics the team with hammer wins around 60 to 62 per cent of ends at a level score [1]. So note just two marks per end: hammer yes/no, outcome. After five games you can see whether you use your hammer or merely administer it.
A steal is a point scored without hammer — the most honest indicator of pressure. The elite target is a steal efficiency from 0.30 upwards; at one top tournament the median was 0.25, with leading teams between 0.31 and 0.47 [1]. What counts for your team is not the comparison with that range but the comparison with your own last month.
Force means holding your opponent to exactly one point in an end where they have the hammer [2]. In curling that is not a consolation prize but a stage win — you get the hammer back and you prevented the big end. Record the share of your non-hammer ends in which you managed exactly that. Our sources give no robust elite reference for it; here you measure against yourselves.
A metric without an action is decoration. The translation runs through the style of play: an offensive "draw game" with guards, draws and freezes aims at big ends; a defensive "hit game" with takeouts and peels keeps the house clean [3]. With hammer you move stones off the centre and keep the four-foot open; without hammer you play centre to steal or to force a single [3]. Your steal rate tells you whether your non-hammer game creates any pressure at all.
All four metrics depend on your opponent. So a measure that describes only you belongs beside them: session RPE has been examined in Olympic curlers as a sport-specific workload measure and correlated significantly with external loads [4]. Add split and hog-to-hog times as the technical control [5]. Effect on the outside, cause on the inside.
Metrics have an unpleasant side effect in curling: they tempt you into tying training volume to percentages — poor shot percentage, therefore more stones. That is exactly how overuse arises. Volume steering belongs on the workload measure, not on the hit rate [4].
And once more the source limit, because it matters here: no training recommendation follows from competition analytics. Whether more stones, more strength work or more rest is right is answered by training research and your coach — not by a steal rate [1].
No guide replaces coaching, performance testing or medical assessment. What it can do is give you a protocol your team can use to talk about itself more honestly.
For five games, keep only two columns: hammer yes/no and the outcome of the end (point for us / point for them / blank). No shot scoring, no rating of individual stones.
After five games you have your hammer conversion and your steal rate — the two numbers looked at first at elite level [1]. And you collected them without an argument, because nobody's individual stone was judged. Only once those two numbers stand does shot percentage pay off. Start small, measure honestly, and let the number say what everyone in the car home is thinking anyway.
Mainly four: shot percentage, hammer conversion, steal efficiency and force efficiency. In the elite band shot percentage sits at 88 to 90 per cent, a skip at 92 counts as outstanding, and teams below 88 per cent practically do not medal at major events [1]. Those are analytics orders of magnitude from competition data, not study values.
In competition analytics the team with hammer wins around 60 to 62 per cent of ends at a level score [1]. Structurally the advantage follows from the scoring rule: only one team scores per end, and the hammer moves to the team that did not score while staying put after a blank end [2].
The elite target is a steal efficiency from 0.30 upwards; at one top tournament the median was 0.25 and the leading teams sat between 0.31 and 0.47 [1]. For a club team, though, the meaningful comparison is not that range but your own trend across several months.
Through internal measures. Session RPE has been examined in Olympic curlers as a sport-specific workload measure and correlated significantly with external loads [4]; split and hog-to-hog times provide the technical control alongside [5]. Metrics such as steals and forces, by contrast, always depend partly on your opponent [1].
Sliding on the Slider: Learn Balance in Curling