Geek Out on Some Stats from Week 1 of Regionals

Now that we have final results from the East and South Regionals we can begin to geek out on some stats. What is great about the Regional format is that we can look at the data from Week 1 and then apply it when watching the action unfold in Weeks 2 and 3.

So let’s waste no more time and look at several observations from Week 1.

Time Caps

There appears to be a sizable gap between the number of male athletes who were time capped this weekend versus the number of female athletes. Out of the 490 times a male athlete took the floor over the six events, just 76 times did a male competitor run out of time. On the women’s side, however, 213 times a female athlete did not finish the work in the allotted time out of 496 times on the floor. In percentage terms, 16% of male scores were time capped versus 43% of women’s.

The biggest culprit of the variance was Event 2 where 6% of men hit the time cap whereas 46% of women did so. The difference was due to the ring dips. While men who were not tearing a pec were ripping through 21 ring dips unbroken or in two sets, the women were breaking early and often.

Event 3 saw an 48% versus 79% disparity thanks to the rope climbs. Event 4 was 11% versus 37% (heavy kettlebells) and Event 5 was 24% compared to 76% (thanks to 45 ring muscle-ups).

Of course, the top athletes for both men and women had no trouble with the time cap. But the disparity begs the question…is there a bigger gap of ability between the top athletes and the bottom athletes for the women compared to the men, or is Castro’s programming harder for the women?

The next section might answer the question.

Time Competing

We already mentioned it on the 5 Things We Learned From the First Weekend of Regionals article we published this morning, but Mat Fraser spent very little time on the competition floor this year compared to last. What we did after that, though, was to look at the 1st place finisher compared to the athlete who did all six workouts and see how long each was on the floor.

Here’s a quick table showing the times.

What’s interesting is that the bottom athletes were very comparable in their time on the floor. But at the top of the leaderboard, the male athletes were on the floor significantly less than the female athletes.

So to answer the question we posed in the last section, “is there a bigger gap of ability between the top athletes and the bottom athletes for the women compared to the men, or is Castro’s programming harder for the women?”, we believe Castro’s programming is harder for the women.

Here’s a table showing the current event records after Week 1 for both the men and the women.

In every event, the men have walked away with the fastest time. For Event 5, the women’s fastest time is 1:47 slower than the men’s fastest time. Why? Because of 45 ring muscle-ups. How about 59 seconds on Event 6? The Air Assault Bike. The fastest men finished the bike in 25 seconds. The fastest women…close to 45 seconds.

Lighter loads on barbells and dumbbells make sense, but it might be worth evaluating if adjusting reps on bodyweight movements is the right thing to do.

Correlation Between Events

There was a lot of commotion about Events 2 and 5 being too similar leading up to the start of Regionals. Both involved rings, a dumbbell and a pushing/pulling movement. On the live broadcast Dave Castro commented that they were two very different events.

Well, let’s take a look using some statistical analysis.

Event 2 vs Event 5 Correlation:
South (Women): 0.492
East (Women): 0.462
East (Men): 0.409
South (Men): 0.037

None of the four competitions had a correlation above 0.5. While a correlation in the 0.4 is pretty strong, it wasn’t as high as you might expect based on people’s belief that they were too similar of workouts.

The next two data sets look at the highest correlated events and the lowest correlated events. The results are interesting. For a Regionals competition that doesn’t feature a 1-rep-max or super heavy weights, the events that require muscle endurance and an engine are highly correlated. So Event 1, 3 and 5 are commonly seen as highly correlated between the regions.

Highest Event Correlations:
East (Women): 0.668 (Event 3 & 5)
South (Women): 0.678 (Event 1 & 4)
East (Men): 0.686 (Event 3 & 6)
South (Men): 0.570 (Event 1 & 5)

Lowest Event Correlations:
East (Women): 0.183 (Event 5 & 6)
South (Women): 0.363 (Event 5 & 6)
East (Men): 0.119 (Event 5 & 6)
South (Men): -0.028 (Event 5 & 6)

On the flip side, the lowest correlation between events was Events 5 and 6 for every region. The muscle endurance required for 45 ring muscle-ups is very different than the explosiveness for the Air Assault Bike sprint and Sandbag Cleans.

Events Correlating to Final Standings

If you were going to look at the results of a single event and try to predict the final standings at the end of the weekend, which event would be the best indicator? Any guesses?

Highest Event Correlation to Final Standings:
East (Women): 0.823 (Event 3)
South (Women): 0.846 (Event 4)
East (Men): 0.811 (Event 3)
South (Men): 0.731 (Event 4)

Lowest Event Correlation to Final Standings:
East (Women): 0.551 (Event 6)
South (Women): 0.482 (Event 6)
East (Men): 0.561 (Event 1)
South (Men): 0.579 (Event 5)

The best predictor of final standings based on one event would be Event 3 or 4. Both events are on Saturday, or what CrossFit calls ‘Moving Day’. Interestingly enough, Event 3 was the highest correlated for men and women in the East while Event 4 was the highest correlated for men and women in the South. Either way, as you watch the Regionals the next two weekends keep an eye on who does well on Saturday because there’s a good chance you will see them on the podium Sunday night.

The lowest predictor of overall success at Regionals is Event 6 for the women and technically Event 5 for the men. While the data shows Event 1 for the Men’s East Regional has the lowest predictability, it is dragged down by the athletes who injured themselves during Event 2 after doing well in Event 1. If you remove those athletes from the calculation, Event 5 is also the lowest correlation compared to overall standings.

What this means is that you will likely see some separation of the leaders on Saturday, but depending on the size of the point spread, the athletes just outside the qualifying cut line could make up some ground on Sunday. Whether there are enough points to do so is what can make Event 6 exciting.