What Stats Matter in Pegaxy?

As of March 30 2022, Pegaxy implemented a change to make stats matter. This blog post is a deep dive into what stats return the highest and lowest VIS per race. For future updates and analysis follow me on Twitter @BondiCrypto.

Too Long, Didn’t Read

• I analyzed 1.2 million races by 4,802 unique Pegas.
• Now that stats matter, the majority of Pega earn under 5 VIS per race. But the top 25% are earning 11.5 VIS per race.
• There are some clear stats combinations that can improve a Pega. High speed and low strength is the best combination.
• You can check the predicted VIS per race of a Pega using this simple tool I built: https://tinyurl.com/yd4xvbf6

Overview of the data used in this post

The data used in this analysis was downloaded from the Pegaxy API. It comprises a sample set of 1.2 million races by 4,802 unique Pegas. The analysis focuses on the average VIS won per race per Pega. This is a better measure of value than win rate (which is what is shown in the Pegaxy UI) because we ultimately want to optimize for VIS, and many Pega are boom or bust which means they can be high earners while showing a low win rate.

On average, a Pega will earn 6 VIS per race now that stats matter. But the earnings are skewed significantly. The highest earner in the sample earns 52.5 VIS per race while the worst earners return <0.2 VIS per race. The top 25% of Pega are earning an average of 11.5 VIS per race while the bottom 25% of Pega are earning 2.3 VIS per race. A distribution of earnings is shown below.

Which Stats Combinations are Best?

To test which stats are good combinations, I ran several regressions to predict VIS per race. By far the largest prediction of positive race results is high speed and low strength. But several other combinations are also influential, I have summarized in the diagram below.

So what should I do?

Ultimately, the combination of 2 stats is just a heuristic and is not sufficient to predict VIS per race. An analysis of all 6 stats together is necessary to determine which Pega will be successful. This should provide stat combinations to aim for in breeding and stats combinations to avoid.

To arrive at a model to predict VIS per race, I grouped all 6 stats into 3 categories where high is a stat over 7, medium is a stat between 3 and 7 and low is a stat below 3. For my sample data set this created 695 combinations of stats. The top 20 best and worst combinations are shown below. To rank your own pega stats and receive a prediction on their VIS per race, follow this link: https://tinyurl.com/yd4xvbf6

Parting Thoughts

The above was a quick summary of my analysis of Pega stats. If you found it useful and would like to read more, let me know on Twitter @BondiCrypto. Comment on this post what you’d like to see next or any thoughts and feedback you have of this blog post. Happy racing Pegaxians ❤