2013 WFTDA Bracketology #2: Division 1 and the High Cost of Inflexibilty


With WFTDA rankings locked-in and the participating teams seeded into four cities across the United States, the first-ever WFTDA Division 1 playoff tournaments are set to kick-off this September.

As the growth of the modern game continues to seek a clear direction, the WFTDA is heading into new territory this postseason. Besides overhauling the ranking system to eliminate biases inherent in an opinion poll, the governing body has also made a significant change to its playoff format.

Out are the four distinct regional tournaments, which have been replaced by what is effectively one big championship playoff divided into four equally-seeded qualifiers. This method was selected by WFTDA member leagues to, according to them, allow for “more competitive play within and across” the whole of the playoffs, and ensure that “the best teams are eligible for Championships,” which this year will happen in Milwaukee the weekend of November 8.

In this installment of WFTDA Bracketology, it’s time to pick apart the Division 1 brackets and see how the WFTDA attempted to meet these and other goals, whether or not the methods it selected were the best way of meeting them, and if the concessions it made to do so were really in the best interests of its member leagues, and for roller derby as a whole.

In case you missed it, check out the first WFTDA Bracketology post wherein we discovered some issues with the Division 2 bracket and took observation at an alternate look at the Division 1 tournament, two things which will come in very handy for what you’re about to read here.

The 2013 WFTDA Division 1 Playoffs

The five events that make up the run for the Hydra championship trophy are no longer called “The Big 5” by the WFTDA, but the scale of the tournaments that now comprise the Division 1 playoffs are still pretty damn big. Forty teams are eligible regardless of what people are calling it these days, so let’s give “The Big 40” their due and see who is playing where this year.

Fort Wayne, Ind. – September 6-8


2013 WFTDA Division 1 – Ft. Wayne

WFTDA Tournament Page
Official PDF Bracket

    1. Denver (2)
    2. Montreal (8)
    3. Arch Rival (11)
    4. London (16)
    5. Rose City (19)
    6. Ohio (24)
    7. Wasatch (27)
    8. Cincinnati (32)
    9. Bleeding Heartland (35)
    10. Grand Raggidy (40)

Richmond, Va. – September 13-15


2013 WFTDA Division 1 – Richmond

WFTDA Tournament Page
Official PDF Bracket

    1. Angel City (5)
    2. Texas (6)
    3. Naptown (13)
    4. Philly (14)
    5. Kansas City (21)
    6. No Coast (22)
    7. Tampa Bay (29)
    8. Jacksonville (30)
    9. New Hampshire (37)
    10. Terminal City (38)

Asheville, N.C. – September 20-22


2013 WFTDA Division 1 – Asheville

WFTDA Tournament Page
Official PDF Bracket

    1. Gotham (1)
    2. Rocky Mountain (9)
    3. Windy City (10)
    4. Minnesota (17)
    5. Steel City (18)
    6. Madison (25)
    7. Houston (26)
    8. Nashville (33)
    9. Oklahoma (34)
    10. Columbia (41)

Salem, Ore. – September 27-29


2013 WFTDA Division 1 – Salem

WFTDA Tournament Page
Official PDF Bracket

  1. Bay Area (4)
  2. Atlanta (7)
  3. Rat City (12)
  4. Detroit (15)
  5. Charm City (20)
  6. Boston (23)
  7. Sacred City (28)
  8. Victoria (31)
  9. Chicago Outfit (36)
  10. Toronto (39)

Again, the numbers in parenthesis indicate each team’s respective June 2013 WFTDA ranking which determined who got in and who went where in the playoff seeding.

Eagle-eyed readers will spot that Columbia, ranked #41 in the WFTDA, got into the tournament despite not being in the top 40. This is because—and this should not be a surprise to you unless you have been living under rock and key for the past few months—the current #3 ranked team, three-time finalist, and 2009 WFTDA Champion Oly Rollers did not meet playoff eligibility requirements. Their vacated spot created an opening at the bottom of the table, which the Quad Squad claimed.

The absence of Oly has been the subject of much discussion and a lot of speculation, some of it ridiculous and/or spiteful—and false, in both cases. The fact is that Oly made a team decision to not participate in any sanctioned WFTDA-rules games this year, except for the two they previously agreed upon in 2012. (London and Jet City, both Oly victories.) This, by extension, meant that they willingly chose to skip the 2013 WFTDA playoffs.

A major factor in their decision likely stems from the obvious friction between Oly and some (that’s some) in the roller derby community about how a handful of their players conduct themselves, despite playing by WFTDA rules on and off the track. It has also since come to light that Oly’s head NSO, Miss Nomer, was allegedly secretly investigated and legally threatened by the WFTDA  for reasons related to her and other officials’ independent connection to USARS roller derby and its efforts.

And oh-by-the-way, the Oly Rollers have doubled-down on USARS play in 2013 after capturing the inaugural USARS championship last season.

Whatever the real reasons for Oly’s withdrawal from WFTDA play this year, the end result is that there is one fewer uber-powerful team in the WFTDA playoffs. Not having the 2009 champion playing in the 2013 championship tournament is a pretty significant mark against it. Not the least of which because there is now one less rock in the sling to try and have someone beat the Goliath that is the Gotham Girls of New York, the perennial Hydra favorite.

Despite Gotham coming in as the #1 ranked team in the WFTDA, however, the new bracket system will make their probable path to the final a somewhat tougher test this year. Playoff teams were segregated by region in previous years, which packed together many championships-deserving teams and created situations where they got knocked out of action before their time. It was not their lack ability that did them in, but the regional format’s lack of space for them in the WFTDA finals.

Teams have been equally distributed throughout the divisions and brackets this time around, as determined the an S-curve distribution we explained here last year. Once all 40 playoff teams were slotted into their spots, the four resulting brackets were assigned a playoff site by the WFTDA based on the best possible combination of reasonable driving distance and concentration of participating teams within proximity of each host city for each group of ten teams involved.

From a competition standpoint, the bracket-balancing act that the WFTDA is using will absolutely ensure that the best teams have a fair chance to play their way to the Championships without them having to worry about facing each too early on in the tournament. Taken in vacuum, this is a massive improvement over previous years.

However, the way the WFTDA went about doing this may have turned out to be an overreach. By insisting on a rigid seeding despite it being based off of an inexact and unproven ranking system, and despite having a player population that is made up almost exclusively of part-time amateur athletes of limited resources, the WFTDA may have lost a lot more than they gained with the new divisional format.

To see why requires an understanding of how an inexact and unproven ranking system can provide some much-needed flexibility in such a scenario.

The High Cost of Inflexibility

One of the WFTDA goals for the new league ranking and divisional playoff system, as stated in its white paper (PDF) on the matter, is this:

The best teams at Championships. Teams are seeded evenly across Playoffs based on rankings, removing equal representation from geographic Regions. This allows for more competitive play within and across the WFTDA Playoffs and ensures that the best teams are eligible for Championships.

A fair goal, and on the face of it, one that is hard to say has not been met in 2013. However, spreading out teams evenly across playoff sites is just part of the equation; where to put them after the brackets have been filled? The WFTDA thought of that, too, explaining where the preset brackets would land based on two criteria:

• If a hosting league qualifies for Playoffs, the bracket they are seeded into will be matched with their location.

• Absent a host league, WFTDA’s tournament team will look at the concentration of teams from a given area in close proximity to one of the Playoffs locations. Each bracket will be assessed for leagues’ driving proximity to each Playoffs location, and the bracket with the highest concentration of teams within proximity will be assigned to a given location.

The only team that hit the bracket jackpot and will play in their own tournament is Killamazoo, the league host of the Michigan-based Division 2 playoff site. For everyone else, it was a spin of the slot machine and a  hope that they would land a playoff site that was nearby.

But with how chips fell and where teams ultimately landed in relation to their seeds and nearby playoff sites, there turned out to be a lot of teams, players, and fans disappointed with where they wound up, creating a “would have, could have, should have” situation for many.

The four Division 1 tournaments are mapped out below, with red pins on each map showing the teams travelling to each of the four playoff sites. The purple pins on all the maps are the four playoff sites, for reference. The darker purple line represents the teams with a travel distance that just-so-happens to match up with the one-way average of the other North American teams participating in their respective tournaments.


Looks as if a lot of teams have to fly all the way across the country just to participate in a divisional tournament. What happened to “driving proximity?” (*London/VRDL not included. It’s a bit hard for them to drive across the ocean.)

A  breakdown of the travel involved for some of the teams to many of the tournament sites paints a rather alarming picture. Although the Asheville playoff site seems to have a reasonable set of itineraries for the 10 participating teams, things start to get less amicable as we go through the other locations.

Fort Wayne has a good cluster of five teams very close by, but this is contrasted by three domestic teams (four if you include London) that need to travel  more than a thousand miles to start their playoff runs. Even worse: Rose City is less than 50 miles from the playoff site in Salem, yet is being forced to fly 3,800 miles to Indiana and back. To the Portland roller derby community, to see their team go so far away like that has got to be a major league bummer.

As we head further east to Virginia, the travel distances of those going there start to become a noticeable burden. Five teams at the Richmond site must travel around or at least 1,000 miles one-way; two, Angel City and Terminal City, tip the scales at plus-or-minus 2,300 miles one-way. The teams nearby and the teams far away cancel each other out in the average, which remains at the perilously high thousand-mile marker and pushes further away the “driving proximity” pipe dream.

And then we come to Salem, Ore.

Oh, Salem.


Here’s hoping the WFTDA has a travel agency sponsorship. Yikes.

Look at this map. Look at it! Six teams are being made to fly across the country and pay for all the logistics involved in doing so, despite all of them being much closer to the other three playoff sites. At a soul-sapping 5,100-mile round trip, Boston has the furthest to go of all the playoff teams on the continent.

The average travel distance to Oregon (excluding VRDL’s 8,000 mile journey) of 1,573 miles, about the distance between Salem and Chicago, makes me wonder how so many skaters would be able to cope with such a big financial strain on their travel budgets.

Actually, some skaters have wondered the same thing. This was thought aloud by one of those very skaters potentially coming to Salem from Chicago back when the WFTDA originally announced its new playoff format last year:

Why is this necessary? Why are so many part-time amateur athletes being asked to dump so much money into large travel expenses like this? Why do so many local fans, volunteers, and sponsors need to be separated from the teams they want to support in person? Why is it that some teams might have to make such a large round trip twice in the span of two months, should they qualify for Championships? (God save the London Rollergirls.) Shouldn’t a huge-large roller derby tournament trip be avoided unless it’s something a league has ample time to plan and raise funds for, or is an expense that is deemed absolutely necessary?

That last question is the key to this travel mess. Is it absolutely necessary that the WFTDA rankings be adhered to so strictly that it causes all of these burdens on this many leagues? Over one-third of Division 1 and Division 2 teams have trips of 1,000 miles or more to plan and pay for, pulling all of that money out of the derby economy and the coffers of leagues who may be better off using that dough to stage events closer to home.

Yes, this self-imposed travel nightmare was voted for and agreed upon by a majority of WFTDA teams. But did a majority of the 60 playoff teams want this? The number of teams in the playoffs is (and was, at the time of the playoff vote) a smaller number than the total majority in the WFTDA. Further, did a majority of the 20 Division 2 playoff teams agree to take on the potential extra travel expenses?

Even if a proper majority said that this was the way to go, the fact is that holding a tournament of this particular format and scale requires a potentially big financial commitment from all of the participating teams, not just a majority of them. So doesn’t it make sense to wait until 100% of leagues agree to this, rather than just the 51%?

Because it seems as if a lot of D1 and D2 teams have been suddenly taking to the Internet to ask for financial donations to help cover playoff travel costs. Like this one, for example. This one, too. And this one. And this one. And this one; also this one, with this one along with this one alongside this one in addition to this one, plus this one, not forgetting this one. And of course, this one. (And this one.)

The “please give us free money so we can participate in the playoffs” model, even if a majority voted on it, is probably not something that’s realistically sustainable in the long run. Surely, there must be a better way of meeting the WFTDA’s goal of an evenly-seeded playoff tournament without putting so much financial pressure on so many leagues.

There is, actually. It is totally possible to get the competitive benefits of equally distributed brackets without sacrificing regionally-accessible tournaments, setting up a best-of-both-worlds scenario where far more teams—and their fans and sponsors—don’t have to be taxed to help fund their leagues and then be told they can’t personally represent them in their most important games of the year.

As it turns out, the WFTDA has already figured out how to do this. It just appears as if it hasn’t quite realized it yet. Or more likely, is as of yet incapable of making the concessions necessary to be flexible enough to implement it.

In the same news announcement where the WFTDA explained who plays where, there were listed two “exceptional circumstances,” the only scenarios under which the perfectly equal, totally inelastic, we-have-to-do-this-to-meet-our-goal S-curve seeding distribution would be deviated from:

• Multiple host leagues seeding into the same bracket: If two Playoff hosts are seeded into the same bracket, the lower-seeded team will swap places with another team seeded into the same bracket spot. For example, teams ranked 9 to 12 are all seeded into the #3 spot in their respective brackets. The team that would gain the most geographic advantage is given preference for the swap.

• At the discretion of WFTDA Games Oversight, an exception can be made for league with a demonstrated, exceptionally large and dedicated fan base in close proximity to one of the Playoff locations.

Luckily, no bracket found itself in double jeopardy. But more to the point, the WFTDA deemed that no league showed a “demonstrated, exceptionally large and dedicated fan base in close proximity” to a site to see the need to move them out of their preordained bracket, since all four brackets follow the S-curve exactly.

This makes sense once you realize that teams like Detroit (travelling 1,990mi to Salem instead of 163mi to Ft. Wayne), Chicago Outfit (1,780mi to Salem vs. not 130mi to Waynetown), and Rose City (1,900mi to Indiana rather than 50 miles down the road) don’t have large or dedicated fan bases.

What’s that, derby fans of Detroit, Chicago, and Portland? You say you are large and dedicated? Don’t be silly. The WFTDA clearly decided that you are not. It would have made an exception for you if you were! Suck it up whiners, this is a democracy.

So yeah, there have been some questionable decisions and non-decisions made as to why no concessions were made for at least these teams, especially when so many others had to give up a cheap hours-long drive for an expensive hours-long flight.

And for what?

The one and only possible justification that the WFTDA has in doing it this way is that its new ranking system is so good at determining the true strength of ranked teams, one of them would be unfairly advantaged or disadvantaged if it played against an opponent that was one or two ranking spots above or below the regular S-curve assignment.

The WFTDA apparently feels that it is crucially important for bouts set by the actual S-curve, like #16 vs. #19 and #15 vs. #20 to happen; rather than bouts that slightly deviate from it, like #16 vs. #20 and #15 vs.#19. This is despite the fact that the latter match-ups would save several thousand miles and (crucially important) dollars worth of travel costs and allow the local fans of the #19 and #20 ranked teams (Rose and Charm) a much more realistic opportunity to see their teams up close (in Salem and Ft. Wayne, respectively).

This comes back to the proposed deviations that would have been made to the S-curve. Funnily enough, the WFTDA explanation of its “exceptional” exceptions is exactly what proves that they are anything but.

The Pros (and no Cons) of Flexible S-Curve Seeding

Taking the WFTDA example, here are the four teams ranked 9-12 (10-13, adjusting for Oly’s absence), which eventually became the 3-seeds in the four Division 1 brackets. Look at how many ranking points they have in comparison with their neighbors:

309.93 – Windy City (10)
307.67 – Arch Rival (11)
305.82 – Rat City (12)
301.80 – Naptown (13)

There isn’t much of a gap between these four teams, is there? Two points here, two points there, four points over yonder. That’s less than a 1% difference compared to how many ranking points they have. The same case can be made among the teams ranked in the 21-24 (22-25) zone, the 6-seeds which will be playing against the 3-seeds listed above:

267.49 – No Coast  (22)
265.49 – Boston (23)
263.94 – Ohio (24)
262.50 – Madison (25)

Click here for the current WFTDA rankings.

Click here for current WFTDA rankings.

The closeness of the rankings among any four teams within a tight group, such as these two, suggest that they are roughly of equal rank. Okay, Boston is 1.55 ranking points higher than Ohio. But what does that even mean? If the two played each other on the track, their rankings would suggest their match-up would be very close. But that’s irrelevant here, since they will be in different brackets altogether.

However, if either of the two played Windy City, the difference in their rank strength would suggest a Windy City win either way. In fact, Windy City would (in all probability) beat any one of the four teams in the second group quite easily, regardless of their exact ranking points difference. So why is it that the team they wind up first facing in the playoffs absolutely must be the Mad Rollin’ Dolls of Madison?

The differences among a group of four teams in proximity is, for the most part, negligible. This difference becomes even more irrelevant as the high-low gap between ranking groups grows. Actually, this gap will always show up in tournament seeding, because seeding is designed to spread out the ranking of teams so that they start far apart and end up closer together.

We can take advantage of this to create some “wiggle room” within each group of the rankings table and make harmless adjustments to the S-curve, where possible, when it is advantageous to do so. In other words, we can have “competitive play within and across the WFTDA playoffs” without completely sacrificing “equal representation from geographic regions.” This is one of the rare cases where the WFTDA can, for the most part, have it both ways.

This is a task that can be achieved by changing the priority order in which the playoff brackets are populated. For the 2013 tournament, the WFTDA appeared to break things down in the following order:

  1. Calculate team rankings
  2. Simultaneously fill all brackets based on rankings, using a rigid S-curve distribution
  3. Assign seeds to teams based on positions with each bracket
  4. Assign brackets to playoff sites based on best geographical proximity of participating teams

This method assumes that step #2, the preservation of ranking order, is more important than where teams ultimately wind up. But we have already determined that is a misplaced priority. A difference of 0.09 ranking points is a statistical tie, yet Atlanta (#7; RP: 324.68) was made to travel an extreme distance across the country while Texas (#6, RP: 324.77) got a relatively average travel distance.

Both of these teams, effectively of equal rank, could have had an easier trip. All the WFTDA had to do was instead populate the playoff brackets in this order:

  1. Calculate team rankings
  2. Assign seeds to teams based on rankings, using a flexible S-curve distribution
  3. Assign playoff sites to teams based on geographical proximity of equally-seeded teams
  4. Independently fill brackets based on seeds

For example, if we figure that Texas (6), Atlanta, (7) Montreal (8), and Rocky Mountain (9), are all of roughly the same strength when compared to Houston (26), Wasatch (27), Sacred City (28), and Tampa Bay (29), we can generally ignore the precise ranking of each team and just assign them an equal bracket seed. This means any potential bout combination of 2-seed (rank 6-9)  vs 7-seed (rank 26-29) will probably end the same way regardless of the precise rank vs. rank setup: The 2-seed will win the vast majority of the time.

Knowing this, we can forget about sending #7-ranked Atlanta to Oregon and #9-ranked Rocky Mountain to North Carolina because of how the other brackets were locked-in. Instead, we can send 2-seed Atlanta to Virginia and 2-seed Rocky Mountain to Oregon, since it is far more geographically advantageous and negligibly competitively disadvantageous to do so.

This is how the S-curve becomes “flexible.” By broadening the curve to four-team ranking chunks, a multitude of placement options open up when it comes time to determine which one of the four X-seeds, regardless of ranking, goes to which playoff site.

Expanding this methodology to the whole Division 1 tournament field, giving placement priority to higher-ranked teams within the same seed group and avoiding unbalanced or unreasonable travel distances wherever possible,  eventually creates four brackets at four tournament sites that might, in one example, look something like this:

Division 1 Playoffs Example with Flexible S-Curve Seeding

Fort Wayne, Ind.                  Richmond, Va.
 1. Angel City (5)                 1. Gotham (1)
 2. Montreal (8)                   2. Atlanta (7)
 3. Windy City (10)                3. Naptown (13)
 4. Detroit (15)                   4. London (16)
 5. Kansas City (21)               5. Steel City (18)
 6. No Coast (22)                  6. Boston (23)
 7. Wasatch (27)                   7. Tampa Bay (29)
 8. Cincinatti (32)                8. Jacksonville (30)
 9. Chicago Outfit (36)            9. New Hampshire (37)
10. Grand Raggidy (40)            10. Toronto (39)

Asheville, NC                     Salem, Ore.
 1. Denver (2)                      1. Bay Area (4)
 2. Texas (6)                       2. Rocky Mountain (9)
 3. Arch Rival (11)                 3. Rat City (12)
 4. Philly (14)                     4. Minnesota (17)
 5. Charm City (20)                 5. Rose City (19)
 6. Ohio (24)                       6. Madison (25)
 7. Houston (26)                    7. Sacred City (28)
 8. Nashville (33)                  8. Victoria (31)
 9. Bleeding Heartland (35)         9. Oklahoma City (34)
10. Columbia (41)                  10. Terminal City (38)

This distribution of teams does several things, all of them beneficial. The first and most obvious is the massive amount of travel savings enjoyed overall. By trying to hold down high travel costs where possible instead of just dealing with the consequences of slotting in exact ranked-based brackets, the average distance for many, many more teams to reach their playoff site comes down to comfortably reasonable levels.


By prioritizing team placement by seed and location, rather than by rank and bracket, fewer teams need to be displaced from their region, fewer cross-country flights are necessary, and a nearly 40% fewer miles need to be traversed overall.

(The comparison is even more pronounced when you compare the two maps directly in this animated GIF comparison of the two seeding distributions.)

Under the WFTDA rigid S-curve placement model, the total one-way distance traveled for the 38 North American teams to reach their playoff sites is 36,726 miles, an average of 966 miles per team. Under a flexible S-curve model, of which one way of following is shown above, the mileage total is 22,578, or 594 mi. per team. This is nearly a 40% reduction in travel, or more than 14,000 miles—which, coincidentally, happens to be the total distance (14,587 mi.) the 20 Division 2 playoff teams need to cover to get to their two playoff sites.

Basically, WFTDA leagues are potentially wasting so much on travel in Division 1 this year, they could have cashed in their frequent flyer miles to get two divisions’ worth of travel for the price of one.

Another benefit is that it preserves a “regional” tournament feel at each of the four playoff sites. Better than a regional, actually, since it gives local fans both a nice mix of familiar teams in their region to root for/against, plus a few extra teams coming in from other regions that they would not normally get to see in person.

Westerns would become even more Besterns with most of the area teams also playing versus Minneosta and Madison, for example. This is in contrast to how “westerns” are now, with six teams coming in from far from the east coast and two not attending from the immediate area—or in some cases, the very immediate area.

But all this talk about cutting down on travel expenses and preserving the regional flavor of the brackets is secondary to the reasons why these principals were abandoned in the first place. The WFTDA wanted to make sure all of the teams involved would have a fair shot at getting to Championships and that all tournaments would be equally competitive from top to bottom.

So the big question remains: Does flexible S-curve seeding, with all its advantages, still meet the same goal as rigid S-curve seeding the WFTDA decided was the best path to take?

Below are two lists. The one on the left shows the teams predicted to make it to the WFTDA Championships this year, using the Flat Track Stats predictor tool (as calculated the week of July 15) to determine who will probably make it to Milwaukee under the rigid S-curve. The one on the right lists the teams who will probably to make it to Champs under the flexible distribution.

Take careful note that the initial seeding distribution of teams under the flexible model proposed above only preserves 14 of 40 teams that were initially distributed rigidly, making the starting brackets very dissimilar from each other.

But as for the ultimate result? Check this out:

Probable WFTDA Championship Participants
Rigid (Current WFTDA)            Flexible
 Gotham (1)                       Gotham (1)
 Denver (2)                       Denver (2)
 Bay Area (4)                     Bay Area (4)
 Angel City (5)                   Angel City (5)
 Texas (6)                        Texas (6)
 Atlanta (7)                      Atlanta (7)
 Montreal (8)                     Montreal (8)
 Rocky (9)/Windy (10)*            Rocky Mountain (9)
 Rat City (12)                    Windy City (10)
 Philly (14)                      Philly (14)/Ohio (24)*
 London (16)                      London (16)
 Minnesota (17)                   Minnesota (17)/Rose (19)*

The three asterisks indicate a third-place game or sooner where the FTS numbers indicate that either of the two listed teams could probably win, effectively giving each of them a 50/50 chance of making it in. The three italicized teams out of the combined 27 listed here are the only teams that don’t show up in both lists. But more importantly, these differences do not affect 10 of the 12 probable championship contenders—also known as some of the best teams in the WFTDA.

Expanding upon this even further, the 16 divisional semifinal bracket slots give the teams populating them two chances to make it to Championships, by either winning and advancing to the divisional final; or losing but eventually winning in the third-place game. If there was really difference between the two models, there would be a noticeable disparity between the 16 teams who got there via rigid assignment and the 16 teams who earned that double-chance with the benefit of flexibility.

But again, the difference isn’t really there. Of the 35 combined teams that make up those bracket spots (again adding in a few extra teams for games that are too close to call) there are only four total deviations similar to those above. More importantly, at this stage all 10 of the top-ranked teams and 13 in total have their probable chances completely unaffected by the change in format. This still suggests, and very strongly at that, that “the best” teams in the WFTDA are still getting assurance that they would have a fair shot at playing for the Hydra, regardless of how the brackets were initially built.

Therefore, we can only come to one conclusion: A flexible S-curve seeding distribution can not only achieve the same goal as a rigid S-curve, but it can do it with significant travel cost savings and allow more fans, volunteers, and sponsors support their teams closer to home.

– – – – – – – – –

The WFTDA did not come to the same conclusion.

The assurance of fair and equal seeding was the entire reason why its member leagues voted it necessary to completely dump the regional aspect of its playoff tournaments. However, it is extremely difficult, and probably impossible, to argue that the small theoretical benefits gained with remaining true to the exact rankings are worth taking over the very large and very real benefits of reduced travel/travel expenses and higher local fan/sponsor participation at each of the playoff sites that flexible S-curve seeding provides, when the same ultimate goal can be achieved either way.


Note the RPI rank/rating and assigned seed.

When the WFTDA revealed the tournament brackets a few weeks ago, they noted that the S-curve seeding model it used to make the brackets was “similar to what is used for the NCAA basketball tournaments,” as if that were a justification for their methodology. Unfortunately, what the WFTDA is saying is not matching up with what it is actually doing, as the way the WFTDA did its tournament seeding is not at all similar to how the NCAA does it for its tournament.

The chart to the right lists the teams that qualified for the 2013 Men’s Division 1 Basketball National Championship tournament, or March Madness. The specifics behind the calculation of the NCAA ranking system, Ratings Percentage Index (RPI), is not of concern to this comparison. What is, is the distribution of tournament seeds. Look over the chart and notice there is no absolute connection between team rating and assigned seed. But there is a general connection, in that the higher the team rating , the higher its assigned seed will generally be. But not exactly.

While the NCAA ultimately starts with a giant list of teams ranked 1-64 (or 1-68), the calculated order of these ranked teams is never the exact order they are distributed throughout the brackets. This is because the NCAA must weigh other factors, particularly the geographical proximity to playoff sites for higher-ranked teams or smaller schools. Although distinctions can be made between a “stronger” 3-seed and a “weaker” 3-seed, there is still wiggle-room to shuffle the distribution to where it is optimal for all parties involved, without disrupting the overall equality of the brackets.

That is to say, the flush-wish-cash NCAA has no problem being flexible with seeding and being reasonable where they can be. For them, it is standard practice, because it only makes sense to do it that way.

With the DIY, donation-dependent WFTDA, flexible seeding is an “exceptional circumstance.” But for the body, it is standard practice to ask so many part-time athletes, fans, volunteers, and sponsors to pull even more money out of their pockets, when it turns out that money could be being spent unwisely—because that is what the member leagues voted for.

Which is why it’s troubling that the same member leagues that voted to unnecessarily spend on playoffs travel will be asking for further donations to assist with the WFTDA.tv Division 2 webcasts.

Even if the current playoff format works out to be a success on the track, which all indications say it will probably be, what the WFTDA had to sacrifice to get there could be interpreted as a clear mistake. Or at least, very poor judgement. And even its member leagues come around next year and choose to be more flexible with placement of teams into playoff locations, that’s not going to be of any help to the member leagues that are apparently scrambling for additional funds this year.

Moving forward, the WFTDA will need to learn to be flexible enough to help ensure everyone can keep pace with—and pay for—the growth of the game, not just the majority of it. Let’s hope the WFTDA will realize this soon enough to avoid it potentially costing everyone more than it has to.

18 responses to this post.

  1. Posted by Jerry Seltzer on 21 July 2013 at 4:34 pm

    first of all, I think you meant in the case of Miss Nomer “threatened with legal action” as opposed to legally threatened……and the following is most impractical, but the gate receipts at these tourneys depend upon the teams traveling to them, and it would certainly relieve the hardship if a certain amount of money by the tournament promoters could be used for travel expenses by leagues coming more than a certain distance or something to mitigate these very high costs.

    I see a number of leagues are on various fund me sites and not doing that well.

    I fully agree with you that since the ranking system is at best imperfect, why not just schedule teams as close to their home base as possible.

    waiting for immediate action on the above………..stop smiling, Windyman.


    • With the POSSIBLE exceptions of Windy City and MAYBE Rose City, VERY FEW derby fans travel to go see the regionals, to the point where IMO WFTDA would be better off using the CBI/CIT/NIT tournament format.


  2. One thing I’ve often said about the WFTDA is that the beauty of how the organization is set up will also be it’s downfall; seemingly everyone has a voice, but there doesn’t seem to be one person where the buck ultimately stops. When you ask fans and participants of other sports who the Commissioner/President is of it’s highest possible league is, just about everyone knows:

    NFL = Roger Goodell
    MLB = Bud Selig
    NBA = David Stern (soon to be Adam Silver)
    NHL = Gary Bettman
    NASCAR = Brian France
    WNBA = Laurel Richie
    PGA = Tim Finchem

    Now, I’m sure since everyone is reading this on the internet, chances are you’re about to (if you haven’t already) start scrambling to find out who the President of the WFTDA is. And I’m sure I’ll probably get responses saying, in effect, “everyone knows that.” However, over the last year or so while having conversations with people in person at various derby events, when one subject or another about the latest brew-ha-ha the WFTDA has spinning around them, I always get around to asking whoever I’m talking to “who is the President of the WFTDA?” Sitting here thinking off the top of my head, I’ve asked about 15 people that question. Believe it or not, only one person has ever been able to tell me who the current WFTDA President is (and she even revealed to me after getting it right that she only knew it because she had just attended the annual WFTDA meetings).

    Of course, I’m sure that EVERYONE reading this associated with the WFTDA to one degree or the other absolutely, positively knows that Alassin Sane of the Atlanta Rollergirls is the current WFTDA President. But with the way they conduct business and how seemingly almost nobody I’ve actually talked to in person without the luxury of being able to look it up on the internet at the time knows who the WFTDA President is, the league seems to be led by Kaiser Soze’s kid sister or something; just a bunch of people acting on the orders of someone from somewhere, they’re just not exactly sure who or where these orders came from originally.

    What I’m getting at is that, personally, I think the WFTDA needs a “face” of the organization that everyone within the derby community is or will become familiar with (like the Commissioners/Presidents listed above) and be a leader when it comes to addressing different issues (both easy and complicated) and making sure they’re addressed in a timely yet thoughtful manner. They don’t have to be a dictator over the sport (which the France family basically has been with NASCAR since it’s inception), nor do they have to push their own personal agendas. Just someone who can look at something, say “hey, this seems broken” or “hey, this could probably be done better” and help accelerate different changes (moderate to extreme) or come to the conclusion that everything is fine and should carry on as is. In other words, someone at the top who can say “I hear ya, I’m on it, we’ll have a conclusion soon” whenever something pops-up and can both follow-up and be transparent about the issue as it reaches is resolution.

    The WFTDA is a well-intentioned organization that has done wonders for the growth of derby around the world. However, it often seems like an organization without any real direction to it. And I honestly think that if there was someone at the top with some real teeth in every matter, then stuff like this that makes seasoned sports fans/people like myself and Windy Man go *facepalm* often would, if not go completely away (because little issues will never go away…just ask anyone on that list above), then they would be at least addressed and dealt with reasonably quickly and not drag-on for 2 years (like the whole NOller derby issue) or cost people who aren’t exactly made out of money more cash that actually HAS to be spent (like this travel subject here) among other things.

    I applaud the WFTDA for this new playoff format. No doubt that, competitively, it should prove to be really good. However, although I definitely wouldn’t call the entire thing a bad idea, I will say that I don’t think it was thought-out as well as it could have been. Hopefully the powers-that-be will take a good, long look at this article and say “you know, he has some good points here. It’s too late to do anything about it this year, but once the tournament season is over, let’s take a closer look as to how we can make this both an awesome tournament while also making it more financially efficient for as many of those involved as possible.” Hopefully someone with some sort of power within the organization will step up and say “DO IT!”


    • Posted by Jerry Seltzer on 21 July 2013 at 7:39 pm

      OK, Elwood, I would be perfect for the job except I did that 40 years ago and would never do it again.

      And WFTDA is not crazy about me.

      You are so right on that it is surprising that others haven’t caught on.

      Until there is one face of the game where the buck really stops,,it will not get beyond where it is now in the public perception.

      Modern Roller Derby is like the bumblebee; science tells you that it should not be able to fly, but it continues.


    • I would actually like to see a former WFTDA skater take those reigns. Someone who’s sole job is to be the President and nothing else (skating, home league duties, etc). They have to love the sport (something that us hockey fans have frequently bellyached that Gary Bettman doesn’t have this true love for hockey…just the paycheck and prestige that comes with being a Commissioner of a major sport), yet look at it with a critical eye when needed. Take both the praise and criticism with equal weight. To always be on the look-out for what’s best for their product. I’m sure it’ll happen someday…we just gotta wait for a while for it to actually happen.


    • The WFTDA is a well-intentioned organization that has done wonders for the growth of derby around the world. However, it often seems like an organization without any real direction to it.

      This is exactly true. Unfortunately, I fear that the WFTDA, collectively, doesn’t yet realize this.

      Although roller derby is growing, in due part to the WFTDA being the first to standardize a ruleset and whatnot, the WFTDA itself is growing sporadically and focusing on the wrong things at the wrong time. For example, it’s trying to have a model playoff system like the NBA or the NCAA basketball, but its rules are stuck in the 1950’s before there was a basketball shot clock. They don’t fully understand the game they are developing, yet they are trying to develop everything around it as if they already do. I find this reckless.

      There really isn’t much the WFTDA needs to change to start doing things the right way. For one, the president and the rules committee—two positions that are voted-in by member leagues—should have veto power over any vote the member league population does, in much the same way our own democracy has a checks and balance system. You can still have a leader who dictates direction and be a democracy; actually, a democracy isn’t one without a leader who is given (revocable) power by those he or she governs over. It’s an oligarchy, otherwise.

      Just because the WFTDA is a democracy doesn’t mean everything they do is automatically correct or the best way of doing. I see WFTDA people on Facebook all the time saying things like, “well, the WFTDA democracy agreed on this,” or, “it’s a good thing we don’t have someone telling us what to do,” with almost a cult-like tone. Well, I’d rather be governed by a dictator that knows what he is doing and truly cares about and acts for all his subjects, than a democracy that doesn’t know quite what it is doing and only caters to a small group of people.


    • “Of course, I’m sure that EVERYONE reading this associated with the WFTDA to one degree or the other absolutely, positively knows that Alassin Sane of the Atlanta Rollergirls is the current WFTDA President”.

      I agree. I left Atlanta four years ago where I began my derby photography career. I was acquainted with Alassin Sane, yet I did not know she is the current president.


  3. Posted by The Contrarion on 21 July 2013 at 8:22 pm

    Seeding by location would mean that the locations picked by the WFTDA would pre-determine most of the teams to play in those brackets. You’ve got things backwards.


    • I don’t think he’s really suggesting seeding by location. From what I gathered, he’s recommending that travel distance be weighed a little more when determining who goes where.

      For instance, the Rose City situation he brought up. Looking at the brackets and overall seeding, I can see a way for not only Rose City to be able to save a lot of money, but three other teams ranked closely to them also be able to save quite a bit of dough.

      You have these three teams:

      Rose City (19th)
      Charm City (20th)
      Kansas City (21st)

      With the way things panned-out, you have Rose City (Portland, OR) going to Fort Wayne, IN, Charm City (Baltimore, MD) going to Salem, OR and Kansas City going to Richmond, VA.

      Now, they’re all within 2 spots of each other. Which means, in theory, they’re very closely matched. What I believe Windy Man is saying is that you can soften the hard S-curve distribution and allow Rose City to play in Salem, Charm City can drive a couple hours south to Richmond and Kansas City can head to the much closer Fort Wayne. All three teams involved in this will save TONS AND TONS of money and the venues which would be hosting them would do much better considering so many more fans of each of these three teams would be making the trip, thus more money spent in both the arena on various things (everything from merchandise to concessions) and, not to mention, local businesses would do that much more business…warming them to the roller derby crowd.

      But, also looking at the seedings, there will be some inevitable cases where a team will have to dig deep and spend a lot of money to travel to their tournament. For this year, I can sympathize with Wasatch (Salt Lake City, UT…ranked 27th). They’re slated to go to Fort Wayne (which is still closer for them than Richmond or Asheville). But when I look at the teams ranked within 2 spots of them either way (25-29), I see teams like Houston and Madison (who are both going to Asheville), Sacred City (Sacramento…ranked 28th…who drew the lucky straw and only have to drive up I-5 about 7-8 hours or so to Salem) and Tampa Bay (who gets to stay on the east coast in Richmond). So, without disrupting the seedings too much, you just have to pat Wasatch on the shoulder and say “sorry, you gotta get to Fort Wayne.”

      One thing Commissioners do in other sports is look out for the owners and franchises and do everything they can to help each team be as successful financially as they possibly can. This is something that the WFTDA can do in some cases (like with Rose/Charm/KC), but sometimes can’t in other cases (like Wasatch). But with the way it’s set up this year, there are going to be a lot of people spending a lot more money than they really should be to play in these events (or even attend them to watch). Just something for the seeding committee to take into consideration while planning the 2014 tournaments.


  4. Posted by N8 on 21 July 2013 at 8:59 pm

    I completely agree with your analysis of the problem. I don’t completely agree with the solution. One important aspect that you overlook is by taking the simplistic view that the better team will win. Reality, and FTS’s algorithm as well, is more robust than that. Team A may be ranked higher than Team B, but if they played each other a hundred times Team A wouldn’t come out with a hundred wins. For any bout there’s a likelihood that you’re going to have one of those games you win. The point being that as you switch around those ratings, you’re either changing someone’s opponent to one they have a smaller chance of beating or to one they have a larger chance of beating; either way someone’s getting an unfair deal. Even if both move that just means someone somewhere else is getting the bad end of the deal.

    For all the unfairness that the S-curve has for costs it’s at least fair in strength of opponent. All brackets will have the same average ranking within the bracket (once you discount teams that were ranked but unable to attend). Furthermore, it eliminates the opportunity for bias that occurs when you have some individual/committee hand reorganizing the brackets for whatever metric of “closeness” is chosen.

    The only fair solution I’ve seen that solves both problems is to go back to geographic regions.


    • For any bout there’s a likelihood that you’re going to have one of those games you win. The point being that as you switch around those ratings, you’re either changing someone’s opponent to one they have a smaller chance of beating or to one they have a larger chance of beating; either way someone’s getting an unfair deal. Even if both move that just means someone somewhere else is getting the bad end of the deal.

      I understand what FTS ratings are designed to do, hence why I kept emphasizing that the listed teams would probably advance and get to Championships, including taking care to note which match-ups in my bracket breakdowns where there was not an obvious probable winner either way. Regardless, the prediction model used or the prediction itself is not the emphasis of my analysis, but the fact that the same probability figures tabs two completely different ranking distributions in the brackets as ending up with the same probable result, even if the exact probability figures along the way are slightly different.

      Plus, when the top seeds in each bracket advance to the point where these small probability differences might become significant, they’ve already qualified for Championships, haven’t they? If the goal was to get the top teams to Champs, it doesn’t matter if a divisional final of #2 vs. #8 is more optimal and fair than one that is #2 vs. #6, because both teams in either case have already made it through, meeting the set goal.

      Regardless, it’s interesting how you point out that someone is getting an “unfair” deal by a slight ranking/seeding adjustment causing a downward shift in their given probability numbers through each stage of the tournament, which of course will happen. The whole point of this post was to demonstrate that this slight unfairness in the bracket can be, and really should be, cancelled out by the unfair deal many teams would avoid by not having to pay for such extraordinary travel arrangements.

      A handful of teams would be slightly disadvantaged by minor imbalances in the seeding, yes. But it doesn’t make sense to financially disadvantage the entire WFTDA to avoid those imbalances, especially at this stage of the modern game’s growth. (Probably.)


    • Posted by N8 on 21 July 2013 at 10:26 pm

      It matters a whole heck of a lot to teams in the 4/5 slots, if nowhere else. But, especially considering how WFTDA’s rankings missed the mark so egregiously on several teams, if there was any room for fudging then there’d be room for people to pick their preferred matchups when they’re fudging and no amount of oversight would ever make anyone happy.

      “But it doesn’t make sense to financially disadvantage the entire WFTDA to avoid those imbalances, especially at this stage of the modern game’s growth.”

      Not at the expense of taking a sport that continues to fight the “rigged” stigma and then throw in a process that rigs it, even if it’s for average financial benefit. Some teams are going to get more financial benefit than others, and then we have someone picking favorites again. On paper it should be great, but it’s a situation where someone is always going to get pooped on, so the best we can do is make it equally likely that anyone gets pooped on without any intervention with someone who might have special interests.


      • But, especially considering how WFTDA’s rankings missed the mark so egregiously on several teams, if there was any room for fudging then there’d be room for people to pick their preferred matchups when they’re fudging and no amount of oversight would ever make anyone happy.

        No amount of oversight would make people happy. No oversight at all will make people happy, either. So if you’re going to have some unhappy people no matter what the situation, better that the ultimate result is the best for the health of the game in the long run.

        I like to say that there will come situations where what’s best for the WFTDA voting members and what’s best for the growth of roller derby will not be the same thing. This may be one of those situations, and how they choose to resolve such a conflict will say a lot about how the WFTDA itself has matured.

  5. Posted by The Contrarion on 25 July 2013 at 7:48 pm

    Hasn’t Oly been running a “Get us to USARS Nationals” campaign for the last bunch of months?


    • They sure have. However, Oly is taking three or four months to plan and fund-raise for the one big travel event it’s doing all year, which is not an unreasonable schedule to solicit reasonable donations from people. Also note that USARS provides financial assistance to travel to its national tournament, in the form of regionals prize monies. (Winners get their checks as soon as they submit their nationals application.) So what Oly is doing on the USARS side is not unreasonable, for where modern derby is at the moment.

      Contrast this with WFTDA, where they give 19 teams (21 if you count VRDL/London) as little as five weeks to make trips of over 1,000 miles. Not only may these teams not be financially prepared for such expenses (Tucson and Silicon Valley may not have been), as I demonstrated, they are expenses which are wholly unnecessary. I’m not sure if the WFTDA pays out to teams making it to Champs to help them there, but I know for certain they don’t assist teams going to the divisional tournaments despite knowing that many of them would have a long, looooooooooooong way to go to get them.


  6. Posted by Kraken on 26 July 2013 at 5:14 am

    I’m pretty sure that the WFTDA member leagues who thought they had chances to make the rankings have been planning for these expenses since this thing was voted in. If you have any foresight at all, then you should’ve been fundraising since last year when this all started to be financially prepared to go to any of the locations if you qualified.


  7. I’m not encouraged by this…. http://wftda.com/tournaments/2014


  8. […] more accurately, on WindyMan’s Roller Derby Notes, the blog that was the precursor to this […]


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

%d bloggers like this: