April 6, 2016

2016 Vaults in Review

The tyranny of the yurchenko full. Since the dawn of life on earth, we have been hearing about the wicked prevalence of yurchenko fulls on vault and how a lineup consisting of six of the same average, snoozer vaults is super boring. Then, in the year 2015, the NCAA coaches did something shocking. They actually decided to get up from their Rip Van Winkle naps and try to change the rules in order to improve the sport. I know. I'm still not over it. They downgraded the yurchenko full and (theoretically) the yurchenko half to a 9.95 start value with the dual aims of increasing variety on vault and encouraging/rewarding those who are able to show more difficulty. With a season of vaults under the new values behind us, let's look into how it actually worked.



A cursory glance at the lineups of the top teams tells us that we did see a greater variety of vaults this season than in recent years, though usually by about one vault per team. Oklahoma is vaulting three yfulls in 2016 compared to four last year. Florida and UCLA both have four yfulls this year instead of five like last year. LSU and Georgia were already showing a number of higher-difficulty vaults last year and stayed constant at those numbers. Utah showed one 1.5 and five fulls again this year, and Auburn's non-full options remain the same. Alabama and Michigan adapted the most among the top tier, each showing three 1.5s (when Casanova was healthy), compared to one most of last year for both teams.

The change has come about pretty much the way we all expected, with a number of top vaulters who have always been fully capable of performing 1.5s upgrading back to vaults they used to perform, like Kennedy Baker, Haley Scaman, Lauren Beers, Mack Brannan, Talia Chiarelli, and Elizabeth Price to the DTY, providing some more variety in their teams' lineups, as much as you can consider a yurchenko 1.5 instead of a yurchenko full "variety."

We've also seen a few 1.5s make it into vault lineups they weren't making previously from the likes of Breanna Hughes and Pua Hall. Hall is an interesting example of someone who prooooobably wouldn't be making that lineup if the 1.5 and the full were the same value because her landing gets a little 9.750, but the extra .05 in SV makes it worthwhile compared to an average full.

It isn't worth it for everyone, however. We know Bridget Sloan can do a 1.5. We all lived through the "BRIDGET SLOAN IS TRAINING AN AMANAR YOU GUYS I SWEAR" years, but she has stayed with the full all season to get a more reliable score and, primarily, to frustrate Kathy Johnson who always thinks she's going to do a 1.5 and is always disappointed. Not every gymnast who can do a 1.5 has been encouraged to do so by the paltry gift of an extra .05. 

Less common, but still occurring, are Hunter Price situations. These are the pot of gold from the change in values. Price's handspring pike 1/2 has been on the cusp of that vault lineup forever but wasn't quite worth it in the lineup. She has used the .05 advantage to push her way into Oklahoma's six and provide a little extra variety. 

I have not yet addressed those vaults-that-shall-not-be-named that masquerade as arabians and get a 10.0 SV, mostly because I'm done talking about them forever. The rule just needs to change for next year, though certainly teams have been getting an extra boost of 10.0 vaults and alleged variety by exploiting that loophole. This is the one area in which the new vault values have been a clear and indisputable negative, adding controversy where there should be none and confusion where there should be clarity. 

Vault variety is not all the way there yet, and the lineup composition is not too too different than it was in previous years, but it's better. It will take some time for the teams to fully adapt. Next year, they should be more ready for it. For now, the yurchenko full remains the dominant vault, and vault lineups are still a yurchenko parade with only a minority of teams showing any other entries at all. It's going to be very hard to change that at the NCAA level because yurchenko is the entry people perform in JO. Changes would have to come at the JO level if we really want to see equal numbers of handspring and tsuk vaults in NCAA.

Now, I've limited this discussion to the top teams, mostly because those are the teams for which I can recall the vault lineups from last year in order to make a comparison. The coaches really should have to submit their vaulting programs to me so that we all know what everyone has been doing each year. The big concern raised over the years about changing the vault values, however, concerns those other mid-ranked teams. The assumption was that only the top teams would really be able to adapt to the new values by upgrading to 1.5s, leaving the mid-ranked teams to wallow in their lower-scoring fulls, thus reducing parity and the opportunity for upsets.

Did that happen? Let's find out.

For each of the last 10 seasons, I've listed the #1 vault RQS, the #20 vault RQS, and the difference between them to see how the margin has grown or narrowed between the best and the middle.

2016
#1 RQS: 49.445
#20 RQS: 49.050
Difference: 0.395

2015
#1 RQS: 49.560
#20 RQS: 49.065
Difference: 0.495

2014
#1 RQS: 49.550
#20 RQS: 49.140
Difference 0.410

2013
#1 RQS: 49.495
#20 RQS: 49.160
Difference: 0.335

2012
#1 RQS: 49.510
#20 RQS: 49.070
Difference: 0.440

2011
#1 RQS: 49.440
#20 RQS: 49.065
Difference: 0.375

2010
#1 RQS: 49.415
#20 RQS: 49.045
Difference: 0.370

2009
#1 RQS: 49.430
#20 RQS: 49.070
Difference 0.360

2008
#1 RQS: 49.445
#20 RQS: 48.975
Difference: 0.450

2007
#1 RQS: 49.445
#20 RQS: 49.025
Difference: 0.420

We can see that the difference between the #1 vaulting team and the #20 vaulting team in 2016 was smaller than it had been the last couple years, and while we'll need more years with the new values to make a true determination, changing the vault values does not seem to have exacerbated any gap between teams that recruit elites and those that do not. The margin between these teams varies pretty widely from year to year, but 2016 is on the lower end of the variation. The same is true further down the rankings, with the weaker teams facing either a similar or smaller vault gap to the best teams, but not a larger one.

If you were looking for those teams that do show a couple more difficult vaults to be rewarded for that difficulty with a scoring boost over those showing six yfulls, however, you're out of luck. They haven't. For example LSU, our top vaulting team and one that has retained a similar level of difficulty to last year with three 1.5/DTYs, recorded a vault RQS this year .285 higher than that of Arkansas, a team with six fulls. In 2015, LSU recorded a vault RQS .260 higher than Arkansas, when the teams had the same disparity in difficulty but the start values were equal.

Most significantly, we can see from the above RQSs that the change in values has successfully reduced overall vault scoring from the boom years of 2012-2015, which was another likely outcome. No one has six 10.0 vaults this year, so every team should take a scoring hit compared to last year and probably should have taken a bigger scoring hit than actually occurred. If a team is putting up six yfulls, the vault total should theoretically be .250 lower than it was last year, all else being equal, so why is the #20 vaulting team only scoring .015 lower than last year and around a tenth lower than the previous two years? 

Still, while the decrease in scores probably should have been greater than it was given the lower start values, the change has curbed some of the recent hyperactive vault scoring and has returned those scores to levels similar to the pre-2012 era, when yurchenko fulls had a 10.0 SV but scoring across the country was a tad stricter and more realistic than it is now. For whatever reason. Here are the last 10 years of vault RQSs for the top 20 teams in graph form.



This scoring plummet in 2016 is a good thing, and an encouraging step toward saner scores, but a problem arises when we compare the average on vault to the average on the other events, which have undergone no such adjustment in start values or routine evaluation. On these events, the scores have continued to rise, with beam and floor reaching their ten-year highs.







If we take a look at this year's numbers, the average RQS for the top 20 teams on vault is 49.199, compared to 49.280 on bars, 49.253 on beam, and 49.357 on floor. This is the first year since 2007 in which vault has not been the highest-scoring event in NCAA. The other events, particularly floor, are growing too different to the point where it clearly pays more to be good at floor than it does to be good at the other apparatuses. This has been happening for a couple seasons with both vault and floor, separating themselves too much from bars and beam, but now that vault has been adjusted downward, floor is the real outlier.

In themselves, the changes on vault are a positive. They're beginning to encourage more variety, they're bringing the overall vault scores down to a more reasonable level, and they're providing some built-in separation for the judges to use in order to differentiate among various vaults of clearly different quality.

These changes must, however, continue through the other events to bring them into line with vault, which has become the lowest-scoring apparatus this season. It's now much more difficult to get a 10.0 start on vault, so let's make it just as difficult to get a 10.0 start on floor. While the need for more variety may not be as stark on floor as it had been on vault with all those yurchenko fulls, floor composition is just as repetitive and just as in need of help. Double pike/switch side+popa/1.5+layout/rudi is the new yurchenko full. So maybe we'll see a change in 20 years.


2 comments:

  1. How influential was Cals vault diversity in helping them get to nationals?

    ReplyDelete
  2. On floor this year, they added the "up to the level" requirement which requires at least two Ds or am E and a D. This does not really effect the tip tiered teams but effected some of the middle tiered teams. They also added the "up to the level" which required a single bar release. So while the change to the vault start values left a more significant impact, changes were still made to the other events.

    ReplyDelete