Tuesday, January 23, 2018

Regulatory Re-Writing via Arbitration? Chand at CAS

Last week the Court of Arbitration for Sport provided an update on its arbitration proceedings in Chand vs. CAS (here in PDF). This is the case that involves the IAAF testosterone regulations governing eligibility of female athletes. I won't go into the (substantial) backstory here, but if you are interested see my recent paper on "sex testing."

In this post I argue that CAS should not allow IAAF to re-write its regulations through this arbitration case.

This is an odd arbitration. Several years ago CAS had suspended the IAAF regulations with a request for IAAF to return with evidence in support of them (if this seems backwards, it is). IAAF commissioned research, presumably to support their regulations, which were published last year (Barmon and Garnier 2017).

The IAAF data do not support the IAAF regulations. Everyone seems to agree with this, as IAAF has asked CAS to present new regulations as part of the Chand proceedings. CAS agreed to this request and in response IAAF submitted to CAS "draft revised regulations that would only apply to female track events over distances of between 400 metres and one mile."

This is an unusual arbital decision. IAAF is engaged in regulation re-writing as part of a CAS proceeding.

Even more odd, the evidence from IAAF in Bermon and Garnier (2017) shows that as the 100m distance -- that Chand specializes in -- high testosterone women are slower than low testosterone women, shown in the figure below.

As a matter of the CAS arbitration in this case, that should be the end. Even if one were to privilege testosterone as a proper object of regulation of female athletes (which to be clear - I do not), the IAAF data does not support such regulation in the case of Chand. End of story. Case closed. Send IAAF back to the regulatory drawing board.

But for whatever reason (and I'm not sure, but I'd guess the reason's initials are CS), CAS is allowing IAAF to re-write its testosterone regulations via the Chand case. You've heard of "activist judges"? Well, these are "activist arbitrators" working on behalf of those in IAAF who have expressed a desire to ban certain women from sport.

But here as well the IAAF data of Bermon and Garnier (2017) don't support the proposed regulations of testosterone in women at distances of 400m to one mile. Consider the figure below:

These IAAF data (pink bar) indicate that over distances of 400m, 800m and 1500m high testosterone women are on average 1.1% faster than their low testosterone counterparts. Unfair, IAAF might scream.

But look at the data for men at 400m and 1500m (blue bar). These data indicate that high testosterone men are on average 1.1% faster than their low testosterone counterparts. Surely if high T in women in selected events where performance differs is to be regulated, then high T in men in selected events where performance differs is also to be regulated?

If IAAF responds that the T standard applies only to women but not men based on performance data, then this is the very hallmark of sex discrimination. This only scratches the surfaced of flawed T regulation (see my paper for more nails into this intellectual coffin).

T regulation is simple in theory. But when theory meets the real world, T regulation fails comprehensively, taking IAAF's own data at face value. Dig a little deeper the IAAF study falls apart (e.g., it includes dopers).

Bottom line: CAS should not be engaged in regulatory re-writing via the Chand case. Chand should be permanently reinstated immediately. IAAF T regulations should remain voided. If IAAF wants to keep trying to regulate women's bodies, then they should start again.

Monday, January 22, 2018

Caveat Lector: How Kathy Carter Wins the US Soccer Presidential Election

I'm not an insider. Far from it. I have nowhere near the insight and knowledge of long-time observers like @duresport or @pkedit (and of course there are others). I'm just a professor with a blog. I used to work at 538 for Nate Silver so I do have a license to speculate irresponsibly about elections.

So with that caveat lector out of the way, here are some scenarios for how the US Soccer presidential election might play out, with my expectations for how things will occur.

This post will make the most sense if you have a bit of background in the election and its candidates (here is a good place to start). There are at present eight candidates who I'll divide into three categories (and each has impressive resumes, I just list their primary present occupation below):

Status Quo Candidates
  • Kathy Carter, president of Soccer United Marketing
  • Carlos Cordeiro, USSF vice president 
Reform Candidates
  • Kyle Martino, TV analyst
  • Eric Wynalda, TV analyst
Others
  • Steve Gans, attorney
  • Hope Solo, professional player
  • Michael Winograd, lawyer 
  • Paul Caligiuri, coach
Conventional wisdom (i.e., what you'll see in the MSM and on Twitter) holds that the "others" don't really have a great shot at winning, but could have an impact. This seems right to me.

The electoral process and math are somewhat complex and opaque (described a bit here and formally here in PDF). In a nutshell, the goal of the election is for a candidate to secure half of the vote, defined as "a majority of the weighted vote of eligible votes cast in that round of balloting." This may be important in a close election, as it means that abstentions reduce the number needed to attain a majority. (Oddly, abstentions are not recorded votes, but non-votes: "members who wish to abstain from voting should refrain from voting and not press any number on their electronic keypad vendor.")

In the event that no candidate secures a majority, there is a 10 minute break and the vote is re-run. After this happens three times with no majority (if delegates decide to do so in advance) the lowest vote-getter will be removed from the listing, and can still be voted for as a write-in. Candidates may also decide to voluntarily drop out. I don't think they get to that contingency.

If all that makes sense, well, you are doing better than me. Let's now look at how things might play out in the election.

Round 1

I'm not convinced that there will still be 8 candidates still running when the voting starts. Candidates might yet choose to drop out ahead of the vote. But let's assume that there are all 8.

MLS interests (for lack of a better term) will have more than 20% of the overall vote and perhaps, speculatively, as much as 35%. Conventional wisdom, which I won't get into here, holds that MLS interests favor the status quo candidates. 

It is perfectly conceivable that one of the status quo candidates (Carter or Cordeiro) wins outright in a first ballot, but probably unlikely if both are in the mix. Watch out if one drops out ahead of the vote -- a sure sign that they have the votes for a 1st round victory. 

But let's say both stay in and give the two of them 25% of the vote up front, based solely on the votes held by MLS interests.

That leaves 3/4 of the vote, which we can split 3 ways equally in the absence of further information to the contrary: 1/3 to Status Quo, 1/3 to Reformers and 1/3 to the others. That would give a first round result of:
  • Carter/Cordeiro - 50%
  • Wynalda/Martino - 25%
  • Field - 25%
I have more confidence (gut feeling, or maybe indigestion, whatever) in the 50% for the Status Quo than the breakout of the other 50%. Some of the vote for Others might also be tactical in the sense of seeing the lay of the land in Round 1 and encouraging a Round 2 (losta game theory dynamics here for sure). But even if each of these guesstimates of mine is off by +/-10% I don't think it changes the calculus (political and electoral) that follows.

Horsetrading Post-Round 1

I'd guess that most if not all of the Others leave the race at this point. They made their points, had their impact and see things are getting real. I'd further speculate that most of this support then goes to the Reformers- why would you vote for a dark horse if you wanted the Status Quo?

The most important question now is whether Carter or Cordeiro drops out in favor of the other. Cordeiro is already USSF VP, and electing Carter locks in the Status Quo candidates in the top two positions. What happens if Cordeiro gets voted in and the Board has to install a new VP during the next year? Politically, that VP might need to be the losing Reformer - Wynalda or Martino. Its a risk for the Status Quo to open Cordeiro's seat. 

On the other hand, Carter has drawn a lot of criticism due to her role at SUM, among other things. From a purely political tactics perspective, her dropping out in favor of Cordeiro would surprisingly eliminate these negatives and maybe push him over the top. Soccer politics isn't really known for its cleverness, so I am going to say that greed wins and Cordeiro drops out in favor of Carter.

Either way, if Carter or Cordeiro drop out, this would necessitate a similar deal between Martino and Wynalda, with very much a similar calculus to be made. Wynalda has run the higher profile campaign, but also has higher negatives than does Martino. I'd guess that this decision would be based on who gets the most votes in the first round. But I don't know if both or either would set aside ego for the other. Regardless, if the Status Quo center on a candidate the Refomers will have to as well.

Round 2

So let's say Round Two starts out with one of Carter and Cordeiro versus one of Wynalda or Martino, with the possibility of a straggler Other still hanging around. I don't think it matters which of the Status Quo candidates or Reformers are on the ballot at this point. The lingering uncertainty will be whether the Status Quo candidate can get over 50%. I think it'll be close, but ultimately the Status Quo will win.

As long as I'm making all this up, I can be even more precise. There will be a significant protest vote in the form of non-votes (which is actually why a remaining Other might be important). The resulting smaller number of delegates who vote will allow Kathy Carter to win the presidency with more than 50% of the vote but less than 50% of those eligible.
  • Carter - 47.34% of eligible votes
  • Martino - 43.86%
  • Other - 3.61%
  • Non-votes - 5.19% 
There you have it. Two decimal places. I hope it adds up to 100%.

Irony Alert

For reformers, a Carter/Cordeiro regime would not be all bad news, as it would all but guarantee a far more significant reform agenda would get pushed through USSF in the years to come. The reality is that this election is only the start of change coming to USSF. Buckle up.

Monday, January 15, 2018

Introduction to Sports Governance: Spring 2018 Syllabus

My spring syllabus for Introduction to Sports Governance (ETHN 3104) at the University of Colorado is now available here as a PDF.

This spring we have another impressive list of guest speakers lined up: Hope Solo, Vitaly and Yulia Stepanov, Kara Goucher, Casey Malone, Mara Abbott, Jim Trotter, Solomon Wilcots, Ceal Barry, Jay Smith, Phil DiStefano, Rick George, Jor Jupille, Mike Macintyre, Tad Boyle, JR Payne, Valerie Simons and several possible surprise guests.

For locals, we will have some evening events associated with the class open to the public, so stay tuned.

We are reading three books:
  • The Edge (by me)
  • The Sports Gene (David Epstein)
  • Cheated (Jay Smith and Mary Willingham)
Lots of other readings - academic, journalism, official reports, etc. We are also watching several films, including Icarus, Venus Vs. and Let them Wear Towels.

It will be a great class - the best ISG so far. I am really excited for the semester. 

Comments welcomed!

Wednesday, January 10, 2018

Making Sense of USSF Electoral Math

OK, I'm diving into the challenge of trying to understand that incredibly arcane mathematics of the upcoming US Soccer Federation presidential election. As a reminder and disclaimer, I am the co-President of the Board of Directors of FC Boulder, a member of the Colorado Soccer Association, which in turn is a member of USSF. This analysis however is done in my professional capacity as a professor who studies arcane things like US Olympic NGB elections.

The below analysis is preliminary and I welcome corrections or clarifying information. I'll update this post as new information becomes available. Some details on the election process can be found in the Office Election Abstract (here in PDF) prepared by US Soccer. This post by Anthony DiCicco is also a useful resource. Paul Kennedy has some similar data from the 2017 USSF AGM here.

Those eligible to vote in the upcoming USSF election are called "delegates" as members of the USSF "National Council" as defined by the 2017-2018 USSF Bylaws. In 2016 USSF identified 534 delegates eligible to vote, as indicated in the memo reproduced below. These numbers will no doubt be slightly different for 2018, but they give us a basis to explore weightings.

US federal law (Ted Stephens Amateur Athletics Act 1978) under which USSF operates as an Olympic sport governing body requires that athletes have at least 20% of voting authority. To fulfill this requirement USSF employs weights to the votes of its delegates. Not every individual delegate's vote is equal to another's - a shocker, I know.

There are five categories of delegates, listed below with their allocated 2016 delegates, each represents a person:
  • Youth Council (291)
  • Adult Council (191)
  • Professional Council (14)
  • Athletes (4)
  • Other (34)
These totaled 534 delegates. 

Because athletes have only 4/534 of the delegates (0.75%) USSF in 2016 implemented a weighting system that gives 20% of the vote to the athlete delegates as follows (USSF uses two decimal places so I will also). Each number below represents a vote:
  • Youth Council (291 - 25.64%)
  • Adult Council (291 - 25.64%)
  • Professional Council (291 - 25.64%)
  • Athletes (228 -20.09%)
  • Other (34 - 3.00%)
After the 2016 weighting, there were 1135 possible votes. The three councils each get 25.64% of the vote and are tasked under USSF Bylaws with determining how each allocates the votes of their delegates. 

For the Professional Council's 25.64% share, the vote weighting inside the Pro Council is interesting:
  • MLS 9 - 64.29%
  • NWSL 3 - 21.43%
  • NASL 1 - 7.14%
  • USLPRO 1 - 7.14%
The MLS and its minor league USL together get 71.43% of the 14 votes that it uses to determine the overall Professional Council vote. This means that MLS/USL is responsible for 18.31% of the overall presidential vote (that is 71.43% of 25.64%). The NWSL (the professional woman's league) gets 5.49% of the overall presidential vote (a bias pointed out by Anthony DiCicco among others).

We can do some rough math as to how many other votes are possibly controlled by MLS. 

Consider:
  • Among the "Other" votes (which include USSF Board members, past presidents, life members etc.) MLS clearly has 3/34 and could easily have half or more of the total, or between 0.26% and >1.5% of the overall presidential vote.
  • Each of the 4 Athlete delegates is responsible for about 5% of the overall presidential vote. There are 8 of 20 members of the USSF Athlete Council who played in the MLS pyramid. While it is not clear how the council allocates its votes or who (or how many) the delegates will be (or who they might vote for), it is reasonable to assume that half or more of the delegates (maybe as much as 15% of the overall presidential vote) will come from former players in the MLS pyramid. It is not difficult to imagine an interest among some in supporting MLS priorities.
So without even considering the Youth or Adult Councils we can estimate that MLS interests control 18.31% + 1.5% + 15% or about 35% of the overall vote (on the lower end this is about 29%).

If so, this would mean that MLS may only need 24% of delegates among Youth and Adult Councils to secure a 50% majority for its preferred candidate. Put another way, 75% of the Youth and Adult Councils could vote against MLS interests and still lose the election. That is just math.

Some initial conclusions:
  • The USSF election procedure is ridiculously and unnecessarily complicated;
  • There is a huge bias against women;
  • There is a huge bias in favor of MLS;
  • Some of the potential problems in the arcane process could be mitigated with open (not secret) ballots. Let's see who everyone votes for;
  • I'll be surprised if MLS does not get their preferred candidate.
Comments welcomed!

Friday, December 15, 2017

The Case for a Lessons-Learned Review of WADA & Russia

As we move toward the denouement of the IOC's sanctioning of Russia leading up to the 2018 Pyeongchang Winter Olympics, I'd like to raise a broader, and arguably even more significant issue. That is the role that WADA (and IOC, IAAF and other IFs) has played over the past decade with respect to the allegations of institutionalized, systemic or even state-sponsored doping in Russia.

WADA and sport will be improved by an independent look back at what went right, what went wrong and what can be learned.

Consider this partial timeline:
The focus of the Russian doping scandal has been, understandably, on Russia. After the 2018 Olympics it is time to take a step back and take a look at WADA and the sports organizations that it supports. There are lessons to be learned here ... if we actually want to learn them.

Tuesday, December 5, 2017

Update on NGB Evaluation Project: SGO Criteria and Next Steps

Earlier posts on this project can be found here and here. Now some updates . . .

We employed 36 criteria from the Sports Governance Observer (2015) framework (available here in PDF), to an initial set of 22 US Olympic national governing bodies to arrive at a preliminary governance scorecard for these organizations. You can find our aggregate rankings here and a detailed breakdown by criteria here.

To facilitate understanding the methods and criteria, below please find a summary set of images with the 36 criteria briefly listed (please consult the full SGO 2015 report for considerably more details on the criteria and their application). Each organization is given a score of 1 (poor) to 5 (state of the art) for each of the criteria, which are then aggregated. Each dimension is equally weighted under the SGO scoring methodology.

To get a sense of the magnitude of the evaluation task, consider that there are 72 US NGBs (39 summer, 8 winter and 25 Paralympic federations). Thus, evaluating all of them requires assigning a score across 72 * 36 categories, or 2,592 individual scores. Our methodology requires that two people independently score an organization, and a third performs a final check.

If that process takes, conservatively, an hour of effort, then creation of our overall scorecard is the result of about 2,600 hours of work - or 1.2 years of people-effort. It is a huge task. In the spring will will roll out a website and a mechanism for people (including NGBs to) contribute to or scoring by proposing scoring changes based on evolution in governance. COnsequently, any such scorecard is a snapshot and governance is fluid. So score will (and should) change over time -- ideally towards better governance.

We've been really encouraged by the positive reactions to our project from across the NBGs. In coming weeks we will be announcing additions to the research team, more scores and the dedicated website.



Saturday, December 2, 2017

Full Preliminary SGO Rankings: US NGBs

There has been a tremendous amount of interest in our scorecard of the governance of national governing bodies for Olympic sports. This interest has been overwhelmingly positive.

To share further details of our rankings, below is a figure showing the full set of preliminary rankings of 22 national governing bodies for Olympic sports. You can read more about the details of our methods and see my presentation at the Play the Game conference at this post.
You can click on the figure above to obtain a higher resolution image. If you'd like the spreadsheet with those data, that can be found here in XLS. To interpret the rankings, you will need to cross-reference the 36 criteria under the 4 diemnsions (listed by number in the left-most column) developed in the Sports Governance Observer, available here in PDF.

Do note that these are preliminary rankings. In 2018, we expect to complete rankings for all NGBs and develop a comprehensive website with all of the details. It is our hope to produce such rankings periodically to aid in evaluation of governance.

Please share questions, comments, suggestions in the comments below.