Predictive graphics, which require users to think and deduce before they can access the result, broaden possibilities for media to attract their audience. Recent research suggests that visualizing one¿s predictions improves recall and comprehension of data. Through a qualitative study of interactive infographics, the present article identifies two types of predictive infographics in digital media: users´ prediction-based graphics and calculation-based predictive graphics. How to challenge users and interaction strategies are analyzed, describing models that could be useful for both researchers and professionals from the field.