Each card describes a narrative device that serves a specific intent, such as building an argument, setting the pace, or drawing the audience in, with a published example. Use them on their own or in combination, to come up with new data stories or to analyze existing ones.
also: Incorporating the audience
Speak to the audience directly and make them part of the narrative.
Increases empathy and invites personal comparisons. Makes the data more personally relevant and lets readers use themselves as a measuring device.
The narrator or subject unexpectedly addresses the viewer directly.
Creates a direct connection with the viewer, exposes the artificiality of the presentation and challenges the objectivity of the observation, making clear that this is one interpretation. The surprise grabs attention.
Start by presenting a problem that needs solving or a situation that needs changing. End with a call to action.
Provides motivation, offers an outline for the narrative and gives a sense of purpose.
also: Juxtaposition
Show two or more datasets, categories or points in time side by side, or change a single graphic over time, and draw conclusions from the difference.
Highlights contrast, difference, the absence of difference, or change over time. Leaves room for readers to explore minor differences on their own.
Represent abstract quantities with concrete, familiar objects, often one visual object per data point, as in Isotype.
Makes large, otherwise abstract numbers easier to grasp and more memorable.
Use or establish a graphical convention, then break it.
Creates surprise, shows the extraordinary nature of the data, and questions the convention and the assumptions behind it.
Present something known and familiar in a novel, unexpected way.
Confronts and challenges viewers’ expectations, highlights implicit assumptions, and forces a fresh look at a well-known case.
Hand control to the audience: let them explore the data freely at the end of a guided narrative, or through simulations and games in which they set the parameters themselves.
The highest degree of active engagement. Readers can test the story’s claims, look up their own cases and reach their own conclusions.
also: Familiar setting
Embed the entry point of a narrative in a setting the viewer already knows, for example their own region.
Lowers the entry barrier and makes the information personally relatable.
also: Gradual visual reveal
Reveal elements of the data progressively, through visual hierarchy, paging or animation, building up to the whole picture.
Lets the story unfold in the viewer’s mind and chunks the material so it is easier to absorb.
Show that there are concrete entities, such as individual people, behind or at the heart of the data.
Makes abstract data more relatable and can establish an emotional connection between the viewer and the fate of the individuals.
Ask viewers to guess what they will find before showing it. Without interaction, interrupt the narrative with a question and give the answer afterwards. With interaction, prompt readers to enter or draw their guess.
Stimulates curiosity, engages readers in causal reasoning about the phenomenon, and can lead to a game-like experience.
also: Meaningful use of space
Use space and direction of movement to reinforce positive or negative feelings. Frequently, up = good, down = bad, left = retreat, right = progress.
Supports the metaphor, helps viewers construct their own, and uses familiar conventions to ease understanding.
also: Again and again
Present a phenomenon several times, changing the main variable, the time frame or the animation while everything else stays the same.
Shows that the same phenomenon happens over and over, and strengthens the narrative through rhythm. A constant frame of reference makes changes and differences stand out.
Pose a question, often in the title, that sets up the point the piece will make.
Connects narrator and audience, triggers the wish to answer, and implicitly steers readers towards the takeaway of a more open-ended piece.
also: Hidden data
Deliberately leave out or hide certain data, like a pause in music.
Emphasizes what is shown by de-emphasizing the rest, creates rhythm, and invites the audience to infer what is missing.
also: Speeding up
Change the pace at which data is presented: accelerate over time, or slow down and pause at key moments.
Speeding up conveys acceleration, momentum and urgency, and can shift from the individual to the group. Slowing down focuses attention on one moment and gives time for reflection.
Let viewers locate themselves in the data, for example by entering their location, age or body measurements.
Anchors viewers in the piece and makes the data personally relevant. Readers can use themselves as a measuring device.
The cards came out of the Dagstuhl Seminar on Data-Driven Storytelling in February 2016. We looked at how data-driven stories work, and how they differ from other kinds of narrative and from other kinds of data visualization.
We define a narrative pattern as a low-level narrative device that serves a specific intent. Starting from around 40 candidates found in the literature and on the web, we narrowed the collection down to a core of 18 patterns and sorted them into five groups. Many patterns serve more than one intent, and they often work best in combination.
To create: draw a few cards at random and sketch how each pattern could shape a narrative for your story. Draw again to produce alternatives and start a discussion.
To analyze: look at a published data story and identify which patterns it uses, how they work together, and what intent they serve. This works well in teaching.
The collection is not definitive. We welcome your feedback, experience reports and suggestions for new patterns at mail@napa-cards.net.
Benjamin Bach, Moritz Stefaner, Jeremy Boy, Steven Drucker, Lyn Bartram, Jo Wood, Paolo Ciuccarelli, Yuri Engelhardt, Ulrike Köppen, Barbara Tversky. Narrative Design Patterns for Data-Driven Storytelling. In N. H. Riche, C. Hurter, N. Diakopoulos, S. Carpendale (eds.), Data-Driven Storytelling, pp. 107–133. A K Peters/CRC Press, 2018.
The preprint is the version submitted in February 2017. Please cite the published chapter.
@incollection{bach2018narrative,
author = {Bach, Benjamin and Stefaner, Moritz and Boy, Jeremy and
Drucker, Steven and Bartram, Lyn and Wood, Jo and
Ciuccarelli, Paolo and Engelhardt, Yuri and
K{\"o}ppen, Ulrike and Tversky, Barbara},
title = {Narrative Design Patterns for Data-Driven Storytelling},
booktitle = {Data-Driven Storytelling},
editor = {Riche, Nathalie Henry and Hurter, Christophe and
Diakopoulos, Nicholas and Carpendale, Sheelagh},
publisher = {A K Peters/CRC Press},
year = {2018},
pages = {107--133},
doi = {10.1201/9781315281575-5}
}Thanks to Schloss Dagstuhl for hosting the seminar, and to Sheelagh Carpendale, Nicholas Diakopoulos, Nathalie Henry Riche and Christophe Hurter for bringing us all together.
Card images are screenshots of the cited examples; rights remain with their owners. Examples marked archived are no longer online in their original form and link to the Internet Archive. The pattern texts are also available as JSON.