The Dot Collector is an app used in meetings that allows people to express their thoughts and see others' thoughts in real time, and then helps them collectively reach an idea-meritocratic decision. It surfaces people's thinking, analyzes it, and uses the information to help people make real-time decisions better in a few ways. Specifically:
- Participants continuously record their assessments of each other by giving them "dots," positive or negative, on any number of several dozen attributes. These dots are laid out in a grid that updates dynamically, so that everyone in the conversation can see one another's thinking as the meeting progresses. Doing this helps people shift their perspectives from being stuck in their own heads with their own opinions to looking down on everyone's views. Seeing things through everyone's eyes naturally causes most people to adopt the higher-level view in which they recognize that their own perspective is just one of many, so they ask themselves which criteria are best for deciding how to resolve the issue at hand. In this way it promotes open-minded, idea-meritocratic, collective decision making.
- It helps people make better decisions by providing advice in the same way a GPS does. By taking data on what everyone in the room is like, the app is able to give people individualized coaching, which is especially important when their own opinions are unlikely to be right. We have found that helping people through such times can be invaluable.
- The Dot Collector highlights what we call "nubby questions"-- cases where the pattern of answers and attributes of people on different sides of an issue suggest that there's an important disagreement to be resolved. For example, it will alert you automatically if you disagree with the believability-weighted majority on a given issue and give you guidance on the appropriate steps to take to resolve that disagreement in an evidence-based way.
- It enables believability-weighted voting. The Dot Collector provides both a polling interface where people can vote yes or no (or provide a numerical rating) and a back-end system of believability weighting, which allows us to look at vote results on both equal-weighted and believability-weighted bases, not as just simple majorities but also based on which way the people whose views have the most merit voted. While this may sound complicated, it's simply a way of helping people keep track of believability without having to remember who is more believable at what.