2026 FIFA World Cup Knockout Bracket Challenge with GitHub Public ESPN APIs

Now that that 2026 FIFA World Cup is completed, congratulations to Spain and all Spaniards.

A month before the FIFA World Cups games started I set out to understand if it was possible to create a competitive knockout bracket based on the historical and current group game stats for using ESPN World Cup data table for all years available (2002 to 2026).

I initially tried to use native KNIME web-scrapping tools to build this workflow but this did not work very well as ESPN sports website actually blocks users from using web-scrapping tools for their websites.

So my next approach was to get input from AI, specifically Claude Fable. The insights/suggestions from Claude were very impressive and instead of attempting to web-scrape from the ESPN website, Claude directed me to a GitHub site that has ESPN public APIs to retrieve the sports data I needed to build my workflow. Sourcing my data from these public APIs proved to be a very reliable and stable data source that I could refresh, based on the python scripts that Claude had build and provided to me.

I then asked Claude to build out the KNIME workflow steps required to effectively build out the Python scripts and consolidate all ESPN FIFA World Cup data from 2002 to 2026.
The link to this workflow is the proof of this exercise and what I had used to build out my knockout bracket.
https://hub.knime.com/s/V_YxfDNaOo00pzx_

So how did I do? Not very well to be honest. I predicted the that either Netherlands or Argentina would win the 2026 FIFA World Cup based on selected stats (i.e., Avg Goal Difference, or power ratings in the dashboard in my workflow.

But I learned so much from this process, gaining deeper understanding of KNIME Analytics Platform and data analytics in general.

The only out-of-pocket costs I incurred is just the subscription to Claude/Anthropic and that was well worth the price for this project.

KNIME Analytics was $0.00

Access to public ESPN APIs (via Git Hub) was $0.00

Knowledge gained from this exercise…priceless.

Special shout out to @Dennis Ganzaroli. If you have not read Dennis’s book “The Visual Data Language” I highly recommend. Dennis has a chapter on sports analytics and when I looked at his sample workflow, I was just intrigued and went deep into the rabbit hole to eventually share the workflow I developed from this exercise/project.

Please have a look and always happy to get insights and constructive feedback.

Cheers, Bernardo