Which Events Were Most Correlated to the Final Standings at Crash Crucible?

Many consider the programming at Crash Crucible to be among the best competitions year after year. It is one of the reason why more and more top tier athletes are making the journey to Spartanburg, South Carolina, every October.

This year’s programming featured seven scored events, two of which were back-to-back (Events 4 & 5). Programming of the premier competitions are often evaluated for how well they are balanced and if it is a good overall test.

I like to look at the correlation of events to overall finish to see if any conclusions can drawn to whether a competition had any biases.

Before we get into Crash Crucible, let’s get a refresher on what correlation is and how it’s been applied in this analysis…

What is Correlation?

The Basics

Before we get into the data, you might be wondering what correlation is or how to interpret the table below. Correlation data is between -1.0 and 1.0.

A 1.0 correlation means that two data sets, in this case the overall finish versus an event finish, is exactly the same. That is, the order of athletes from a specific event matches exactly to the overall standings. So 1st in an event is also 1st overall. 2nd in the event is 2nd overall, and so on all the way down to 20th place.

A -1.0 correlation is the opposite. 1st place in an event would take 20th place overall. 2nd place would take 19th overall. In simpler terms, this ‘negative correlation’ means that the best athlete at the end of the day would do the absolute worst in that specific event.

Of course the odds of a perfect correlation, whether 1.0 or -1.0, is nearly impossible. The actual data will end up somewhere in between. In the absolute middle is 0.0. This means there is no correlation between the event and overall standings. In other words, the event finish and overall finish is basically random and that some top athletes did well in the event and some did poorly.

CrossFit-Specific Correlations

When it comes to correlations within a CrossFit competition, almost all events have a positive correlation to the overall standings. I typically define a highly correlated event as one with a correlation above 0.6 and a low correlation is less than 0.3.

So with that, here is the correlation data for the 2024 Crash Crucible.

Event to Final Standings Correlation

The first thing that stands out is that the men’s events were almost all 0.5 or higher. In looking at the event results and overall standings it is clear there was a huge gap between the top and bottom athletes. Of the bottom 5 athletes, there was only one top 10 event finish all weekend (Travis Moore’s 3rd place on Cyclist’s Delight). Conversely, of the top 5 athletes, there was only one event finish 20th or worse (Mathias Porter’s 21st place finish on the same Cyclist’s Delight).

To over simplify, the top athletes were consistently atop the leaderboard, the middle athletes were in the middle and the bottom athletes were at the bottom. There was not a ton of parity in the men’s field as a whole.

For the women, there were four highly correlated events and three very low correlated events. Cyclist’s Delight and Thick-N-Thin were two of the faster events while Midline Sadness required high volume handstand walking paired with power output on the SkiErg.

Top 10 Correlation

After seeing the results and how bifurcated the top athletes were from the bottom, I wanted to see how the correlated looked among just the top 10 athletes.

As you can see, the results are drastically different!

For the men, only Midline Sadness and Mad Moon were highly correlated. Dumb Standard, which was the highest correlation with the entire men’s field was the lowest correlation amongst the top 10 men. It goes back to the fact that the bottom half of the field did not have the requisite strength to be competitive against the likes of Colten Mertens, Taylor Self or even Tanner Balasz.

The two highest correlated events for men were triplets that featured monostructural, gymnastics and weightlifting (if you count weighted GHD sit-ups as weightlifting).

For the top 10 women, the correlation analysis reflects how much movement there was on the leaderboard. For example, Olivia Kerstetter held the leader’s jersey at one point, then dropped outside the top 10 and then clawed back to 6th to finish the weekend. Brittany Weiss, the women’s champion, finished Day 1 outside the top 10.

Pillgatory was the only highly correlated event for the women, an event that featured high volume CTB and odd object thrusters (and some lunges). 

The top 10 women’s correlation showed how balanced the events were and that a weakness in one area would get punished while playing to your strengths would reward an athlete.

Brittany Weiss was the most consistent female with the fewest major weaknesses. She only had two finishes outside the top 10 and her lowest points earned in an event was 39.02 in Dumb Standard.

Kyra Milligan also only had two finishes outside the top 10, but one of those was a 22nd place earning her just 1.63 points. Katelynn Sanders also had two events outside the top 10, but she also never earned more than 87 points in a single event. Weiss, however, had two event wins and 3rd place finish.

All in all, the correlation analysis shows that Crash Crucible was a balanced test in my opinion. To do well, an athlete needed a requisite strength. If the athlete was strong enough, then to be in podium contention that athlete needed the ability to be proficient in gymnastics and combine it with monostructural demands in certain events.

And with P-Score, any weaknesses were amplified against those who had that as a strength.