The first one is about NCAA. Check out the original and fully interactive version (you can click on nodes to get details) with description, at http://thepowerrank.com/visual/NCAA_Tournament_Predictions.

The second one is from me, representing click scores vs. odds (for a click) of converting: the higher the score, the more likely the click will produce a conversion, as scores are designed as conversion predictors. Especially useful to predict the value of new keywords (with no performance history) added to a pay-per-click advertising campaign. Produced with R, it's an attempt at representing 4 dimensions in a plane, as well as a new way to segment the click universe:

1. X-axis for the score (average score around 600)
2. Y-axis for conversion odds (1 = average)
3. Size of rectangle represents the number of clicks in a bucket where all clicks have the same multivariate vector and same score
4. Color represents cost per click, called CPC (blue = low, green = medium, red = high)

Related articles:

• One of my R maps (fraud detection)
• R tutorial to produce nice graphs and maps with 256 colors
• Great graphic diagrams
• 14 questions about data visualization tools
• Another cute graph (firework graph): is it useful?
• The top ten worst graphs
• The top 20 data visualisation tools
• Statistical Graphics, Data Visualization
• Sitemap