What's Coming In 2025

TL;DR

  • New Data Sources
  • New Research
  • Richer Content

The UFC is starting strong in 2025 with their first full effort event happening this weekend with UFC 311. Similarly WolfTickets.AI needs to keep up so I’m starting with a quick update blog about a few recent improvements and exactly where the site is going in the near future.

New Data Sources

Last year there were a number of updates to the crawlers for historical performance data, industry analyses, and odds data.

This all helped me to reduce the time it takes to update the site’s data weekly from several hours to around 45 minutes end to end. I still have a few more pipelines to fully automate but the workload is manageable now and is letting me get predictions out along with the insights out sooner(often by Tuesday of the week of an event).

The biggest improvement came over the holiday break for 2024 with me finally cracking the process needed to obtain round by round raw data for each fight in the UFC.

New Research

This data is going to take me a while to fully digest into how to best fit it to models but I’ll be writing blogs that explain the Exploratory Data Analysis(EDA) process, so stay tuned.

Questions I’m hoping to answer:

  1. What percentage of fights have data for rounds(1 -> 5)?
  2. How does that compare by weight class?
  3. Are there trends in specific fighter performance round by round?
  4. Do those trends have predictive powers?
  5. What does it look like to see one fighter’s output impact another’s round by round?

With these answers, new features will be created, then new models to further improve the accuracy of the predictions.

Richer Content

Last year I released:

  1. Sitemaps for the entire site that map to all the historical results with friendlier names in the URL. (Improved Google and others crawling the data).
  2. Multiple versions of the write-ups for each match, this will remain an iterative process but I’m really happy with the differences from my first to the latest .
    1. If you’re not a subscriber you can check out a sample for this week here: https://gist.github.com/chrisking/2e91a3a203dab576aa381208a7c42e2c
  3. Weighted scoring to shore up some of the gaps in the models around rapid decline of fighters, lack of data for new fighters, and more. These have helped scope how risky a prediction is and fed into the betting system.
  4. Structured betting, lots of ups and down here but the approach has stabilized for selecting the bets weekly and allows me to ship out a consistent collection of parlays for each event.

While great, that still leaves a lot to improve and coming soon I’ll be adding the following:

  1. Fighter pages - A resource to see all the predictions, write-ups, fight histories, and stats for each fighter. Will also be searchable.
  2. Some content before events for everyone. This won’t have the prediction breakdowns, recommended bets, or other items but it’ll help with SEO and provide a preview to others.
  3. Improved historical reporting / performance. I’ll always be working to improve this but there a few years of data from WolfTickets now that needs to be surfaced along with new features, models, and other improvements.

Conclusion

Thank you for all the support and feedback as development on WolfTickets.AI, here’s to a 2025 of outstanding fights, improved models, and interesting insights!

Remember, this week’s fight predictions, write-ups, and bets are here: https://wolftickets.ai/events/ufc-311-makhachev-vs-tsarukyan-2/3700103729472585463/

Good Luck! — Chris@WolfTickets.AI

PS - I’m working on a entirely different application that focuses on improving the planning and cooking process for meals at home(entirely different interest of mine). I’ll be opening this up for private previews in the next few weeks, reach out if you’re interested!