
Thirty-Five (35)
Quickly bring out the best ideas from a large group through successive comparative voting.
Thirty-Five is a framework game by Sivasailam 'Thiagi' Thiagarajan designed to elicit the best ideas from a large group. Each participant proposes a response, then the cards circulate randomly before being compared in pairs. In each round, the pairs allocate 7 points between two ideas, resulting in a simple and robust collective ranking. The game facilitates a rapid, dynamic, and participatory selection process, without endless debate.
Walkthrough
- 1
Set the framework and the question
3 minThe facilitator gathers the group standing or ready to stand, then announces the objective: 'We will bring out the ideas deemed most useful by the collective.' They display or state a single, clear, and open question to which everyone must respond with an idea. They specify that the game lasts 20 to 30 minutes and consists of 5 rounds of comparison. Participants listen without yet exchanging ideas, so that everyone starts from the same instruction.
Tip — Formulate the question before the workshop and test it out loud: if it leads to a vague list or multiple topics, narrow it down before starting.
- 2
Write one idea per card
3 minDistribute one card or piece of paper to each participant and ask: 'Write your best answer to the question, one idea only, legible and understandable without oral explanation.' Each person writes silently on the front of their card. Encourage them to keep it short, concrete, and actionable, as the idea will be read by others. The back should remain available for noting the scores from the 5 rounds.
Tip — Walk around the room and identify cards that are too long: suggest rephrasing them into one sentence, otherwise they will be disadvantaged during the quick comparisons.
- 3
Mix the cards without reading them
3 minAsk everyone to stand with their card held face down or with the text not visible. Give the instruction: 'Exchange your card randomly with someone, then repeat several times; do not read the cards during the exchanges.' Participants circulate and exchange quickly, without trying to keep track of their own idea. This step partially anonymises the proposals and prevents the author's reputation from influencing the outcome.
Tip — Demonstrate the exchange gesture yourself and maintain an energetic pace; if the group slows down, announce two more exchanges and then stop abruptly.
- 4
Round 1: compare two cards and allocate 7 points
4 minAt the signal, say: 'Stop. Pair up with the person closest to you.' Each pair reads the two cards they have in hand, then allocates exactly 7 points between them based on their relative value: for example, 4 and 3, or 7 and 0. The score assigned to each card is noted on the back of that card. Participants should not debate for long: they choose the distribution that seems fairest to them.
Tip — Remind that 7 points must be distributed to each pair, no more, no less; this constraint forces a real comparative choice.
- 5
Rounds 2 to 5: reshuffle and re-vote
10 à 12 minAfter the first vote, restart the circulation: 'Exchange the cards randomly several times, without reading them.' At the signal, participants reform a pair, read the two cards, and allocate another 7 points between them, adding the new score to the back of each card. Repeat until 5 rounds have been completed in total. In each round, a card thus receives a score, and its final total can go up to 35.
Tip — Clearly announce the round number before each comparison: 'Round 3 out of 5.' This prevents forgetfulness and facilitates score control on the back.
- 6
Total the scores
4 minAt the end of the fifth round, ask everyone to keep the card they have in hand and to total the scores written on the back. Remind the rule: 'The maximum possible is 35, as a card can receive up to 7 points during each of the 5 rounds.' Participants write the final total visibly on the back. You may ask those with high scores to come closer to prepare for the feedback.
Tip — Allow for a minute of silence for the totals: ambient noise can quickly lead to calculation errors, especially in large groups.
- 7
Read the best-rated ideas
5 minInvite participants to announce the cards with the best totals, starting with the highest scores. Say: 'If you have a highly rated card, first read its total and then the idea aloud.' Note the selected ideas on a visible support, without immediately restarting a substantive debate. The group thus hears the proposals that successive comparisons have highlighted as the strongest.
Tip — Do not attempt to read all the cards: keep the energy focused on the highest-rated ones, then collect the others if you wish to analyse them after the workshop.
- 8
Conclude and connect to next steps
3 minRevisit the best-rated ideas and ask the group what should be done with them concretely. Clarify the next steps: immediate decision, further exploration, additional prioritisation, or integration into an action plan. Thank the participants, emphasising that the ranking comes from 5 successive comparisons and not from a single vote. This conclusion transforms the game's outcome into actionable material.
Tip — Before moving on to the debrief, photograph or collect the winning cards; this is the moment when good ideas can easily be lost.
Variants
- For a very large group, have participants circulate in an open area and use a clear sound signal to stop each round. Keep the same rule: 5 rounds, 7 points allocated between two cards, maximum total of 35.
- To encourage more actionable ideas, impose a formulation on the front such as: action verb + beneficiary + expected result. Do not change the voting mechanism.
- To address multiple questions, play several separate rounds with a different colour card for each question. Avoid mixing questions in the same round, otherwise the comparisons become unreliable.
Debrief guide
- What strikes you about the ideas that came out on top?
- What criteria did you actually use to allocate the 7 points between two cards?
- Was there an idea that surprised you with its score, whether high or low? What does this reveal about the group?
- How does comparing two ideas at a time change the quality of the collective choice?
- Which ideas now deserve to be tested, explored further, or turned into action?
- What limitations do you see in this selection method, and how can they be compensated in the next steps of the work?