First, three things to know
- A percentage counts only the people who voted on that statement. The number in brackets tells you how many people that is.
Example: “96% (28)” means 28 people voted on it, and 96% of them agreed. - The report counts only people who are in a group. 85 people voted. Pol.is put 50 of them into groups. The tables in the report use only these 50 people.
- Small groups give big percentages. Group A has only 10 people. On many statements, only 6 to 9 of them voted. So “100%” in Group A can be just 6 or 7 people.
1Overview
Who took part?
Words you need
- Participant: a person who votes or writes a statement.
- Statement: one sentence that people vote on. Each statement gets a number.
- Opinion group: people who voted in a similar way.
The six numbers
| Number | The report says | It means |
|---|---|---|
| 85 | people voted | 85 people voted at least once. |
| 50 | people grouped | Pol.is put 50 of these people into opinion groups. |
| 1,271 | votes were cast | All the votes together. |
| 51 | statements were submitted | All statements, including some that the organisers removed. The report shows 46. |
| 14.95 | votes per voter on average | Each person voted on about 15 statements. |
| 3.83 | statements per author on average | This is an average. In fact, one person wrote most of the statements. The other writers wrote 1 or 2 each. |
2Raw Data Export (Anonymous)
Where can I download the data?
This part is for people who want to study the data on a computer. It has files to download and the licence. You can skip it today.
3How divisive was the conversation?
Did people agree, or did they split?
How to read it
- Each dot is one statement.
- Left side: people voted the same way. Almost all agreed, or almost all disagreed.
- Right side: people split. Some agreed, some disagreed.
What we see
Most dots are on the left. Only a few dots are on the right. So people agreed on most statements. They split on only a few.
Try it
Move your mouse over a dot. The report shows the statement. The dot on the far right is the statement that, by Pol.is's calculation, split people the most.
Here it is statement 29: “Companies already protect the environment through social programmes.”
10% agreed, 83% disagreed, 6% passed. The (30) at the end means 30 people voted.
Look at the numbers, not only the dot. Here 83% disagreed, so most people voted the same way. The place of the dot comes from Pol.is's own calculation.
Ignore the “OVERALL 0” above the bar. It is a small display error in the report. Other tables show “OVERALL 50”. The real vote count is under the bar: (30).
4Majority
What do most people think?
How to read it
- This list shows statements where 60% or more of people voted the same way.
- “The same way” can mean agree or disagree. So the list has statements that most people support, and statements that most people reject.
- Each row has three bars: OVERALL (all 50 grouped people), Group A and Group B.
- Colours: green = agreed, red = disagreed, grey = passed, white = did not vote.
What we see
- Most people agree: statement 30, “There should be a public list of companies found clearing land with fire, open for anyone to check.” 96% agreed. 28 people voted.
- Most people disagree: statement 31, “Farmers are the problem, not companies.” 79% disagreed. 24 people voted.
5Opinion Groups
Which people voted in a similar way?
How to read it
- Pol.is puts people who voted in a similar way into the same group.
- The computer makes the groups from the votes only. It does not know who you are, your age or your job.
- Under each group, the list shows the statements where this group voted most differently from the other groups. These are not the topics the group cares about most.
What we see
- Group B (40 people) agreed with almost all the actions. Example: statement 19, “Companies' operations should be suspended until the investigation is over.” In Group B, 91% of the 24 people who voted agreed.
- Group A (10 people) said no to some actions. On statement 19, nobody in Group A agreed: 55% disagreed and 44% passed (9 people voted).
- Another example: statement 37, “Bring those who caused the haze to the international court.” Group B: 96% agreed (25 people voted). Group A: 55% disagreed (9 people voted).
Be careful. Group A is small. Many statements have only 6 to 9 votes from it.
Say: “A small group voted differently on some statements.”
Do not say: “Group A is against taking companies to court.”
6Areas of uncertainty
Where are people not sure?
How to read it
- This list shows statements where more than 30% of people pressed “pass”.
- “Pass” can mean: “I don't know”, “I don't understand”, or “It depends.”
- The report says these statements are good places to explain more and talk more.
What we see
- Statement 21, “The government must have more frequent cloud seeding.” 46% passed. 26 people voted.
- Statement 33, “Sustainability certifications like MSPO are mostly greenwashing.” 35% passed. 28 people voted.
- Statement 46, “Malaysians simply don’t care.” 60% passed. Only 5 people voted, so this one says little.
Many people are not sure about cloud seeding, or about certifications like MSPO.
7Graph
Who thinks like whom?
How to read it
- Each number is a statement.
- The letters A and B show where each group is.
- If people voted on two statements in a similar way, the two numbers are close together.
- A group is close to statements it agreed with. It is far from statements it disagreed with.
- Look only at near and far.
What we see
Most statement numbers are near the centre, close to B. A is further away, on the left. So the big group (B) is close to most statements. The small group (A) is further from them.
Try it
Click a number. The report shows that statement.
8All statements
What happened to each statement?
How to read it
- This is the full list of the 46 statements. Each row reads like the Majority table: OVERALL, Group A, Group B.
- The list starts in number order (“Statement Id”).
- Use “Sort by” to change the order. Two easy ones:
- Number of votes: statements with the most votes come first.
- % Agreed: statements with the most agreement come first.
- Be careful with “% Agreed”. A statement with only 1 vote can be at the top. Always look at the number in brackets.
To finish
Pol.is does not pick a winner. It shows us two things:
1. Where we already agree.
2. Where we still split, or are not sure.
In this workshop, we start from the places where we split.
Run your own Pol.is conversation
How did we set up this conversation? Nine steps, with pictures from our own haze conversation.
Only the organiser needs an account. People who vote just open the link.
Step 1 · Make an account
Go to pol.is/home and click Sign in. New here? Click Sign up, or use Continue with Google.
Step 2 · Create a conversation
After you sign in, you see All Conversations. Click Create new conversation.
Step 3 · Write the topic and description
You are now on Configure. Topic is your question. Description says why it matters and what people should do.
Step 4 · Add seed statements
Seed statements are the first statements people vote on. Type one and click Submit, or upload a CSV file with one column called comment_text.
The Pol.is guide suggests 15 to 40. Our haze conversation has 34. Write one idea per statement.
Step 5 · Choose the settings
Further down on the same page. We ticked “No comments shown without moderator approval”: a new statement appears only after we accept it.
Step 6 · Share the link
Open Distribute. Copy the Share link and send it out, or use the Embed code to put the conversation on your own web page.
Step 7 · Check new statements
Open Moderate. New statements wait in Unmoderated. Accept or reject each one. Reject spam, repeats, and statements with more than one idea.
Step 8 · Watch who takes part
Open Monitor to see how many people voted and wrote statements.
Step 9 · Make the report link
Open Reports and click Create report url. Anyone with the link can read the report. It updates as people vote, so you only need one link.
Pictures from our own admin pages, 26 October 2026. Steps from the official guide: compdemocracy.org/Quickstart and compdemocracy.org/Moderation.