Team statistics: take them, read them, use them

1. Agree on the job before the throw

Choose one observer, ideally a coach or teammate who is not playing. Before delivery, record the player, shot type, turn, requested weight and intended finish. Include acceptable alternatives: does a hit have to stay, roll behind cover, or remove both stones?

World Curling’s statisticians follow the called task and aim to judge consistently. The observer needs to understand the skip’s intention. World Curling: how statistics are collected.

In practice, draw the target and agree what earns each grade. A stone touching the eight-foot might fulfill one call and miss another. Avoid redefining success after seeing the result. If observers change, have them independently grade the same practice shots and resolve differences before combining their records.

2. Give the shot a grade

Use these HouseCall starter anchors:

• 4: The intended job was achieved.
• 3: Most of the job was achieved, with a small shortfall.
• 2: A meaningful part of the job was achieved.
• 1: Only a limited part of the job was achieved.
• 0: None of the intended job was achieved.

Decide the boundaries for the particular shot. These grades describe fulfillment, not inches from a universal target. Discuss uncertain calls afterward using the original target and, if available, video.

The World Curling/CURLIT manual uses 0–4, with four worth 100%, and gives separate instructions for different shots and exceptions. Official 2025–26 statistics training.

3. Calculate execution percentage

Execution percentage = total grade points ÷ (4 × graded shots) × 100.

Fictional example: eight grades of 4, 3, 2, 4, 1, 0, 3, 3 total 20 points. The possible total is 32. Execution = 20 ÷ 32 × 100 = 62.5%.

Only two shots received a 4, so the fully achieved rate under these agreed targets is 2 ÷ 8 = 25%. The two percentages answer different questions. Execution percentage includes partial credit; it cannot tell you the chance that the next shot will fully succeed or win the game.

Count zero grades. Leave an unobserved shot ungraded and record why. Keep excluded shots in the log with a reason; never quietly remove ordinary misses. World Curling excludes deliberate throw-throughs and specified incidents such as debris picks or burned stones, while delivery and protected-stone violations receive zero. Official exceptions.

For a team or several games, pool grade points and graded-shot counts. Do not average percentages when their shot counts differ. If no shots are graded, show “no graded shots”.

4. Split the numbers into useful groups

HouseCall’s simplified categories are:

• Draw family: positioning shots, such as an open draw, guard, freeze or soft tap.
• Hit family: removal shots, such as a takeout, peel, double or hit-and-roll.

Keep the specific shot name too. A draw behind two guards asks something different from an open draw. For an unusual shot, write the task rather than guessing a category.

Record clockwise or counter-clockwise rotation, viewed from above. In-turn and out-turn depend on the throwing hand; clockwise/counter-clockwise avoids that ambiguity. Curling Canada’s explanation.

Start with each player’s overall percentage, draw/hit split and rotation split. Put the graded count beside every percentage. Keep practice and game results separate.

5. Read the context

Add a short difficulty note before the throw: open target, narrow port, precise roll, multiple contacts. Use difficulty to compare similar tasks; this starter system adds no bonus points or invented difficulty multiplier. Renée Sonnenberg’s coaching handout pairs shot tracking with difficulty, weight, error type and video. Coaching handout.

Small groups swing easily. Over eight graded shots, changing one grade from 0 to 4 changes the percentage by 12.5 points. A few attempts can suggest a practice question, but cannot establish a stable player ranking. Check whether the pattern returns over comparable sessions with the same grading approach.

The recorded grade follows the shot. Delivery, sweeping, communication, broom placement and ice reading all contribute. Keep observations such as “early weight call” or “late change to Plan B”; avoid assigning blame from the number alone. Curling Canada stresses all four players’ role in shot communication. Team communication.

6. Turn one pattern into practice

After the game, pick one strength and one repeatable question. “Six of ten draws finished heavy” leads to a weight-control drill and another comparable recording session. Review the evidence with the players before deciding why it happened. Combining statistics with video can challenge a team’s first impression. Curling Canada: Back to Basics.

For Skip’s School’s decision calculators, separately count the actual outcome branches in comparable situations, such as score two, score one or concede one. Do not paste execution percentage into a full-make or game-winning probability field. A successful call and a successful delivery are different questions.

Use the digital tracker in practice or after a game. Check your competition’s rules before any live use.

Copyable shot-log fields

Session/date; game or practice; observer; player/position; end and shot; score before shot; hammer; shot family and subtype; clockwise/counter-clockwise; intended finish; requested weight; difficulty/context; grade 0–4 or ungraded reason; actual finish/outcome; observed weight/line issue; sweeping/communication note; video time.

End the review with: “Keep doing ___; practice ___; check again when ___.”

From a record to a decision

Explore the Moneyball examples in Skip’s School. Use observed outcomes from comparable shots; execution percentage is not a win probability.

Keep the review constructive with the mental-game guide.