After WTC has announced the changes on the Pro qualification system for 2014 (see my post on the KPR changes and their impact), a discussion about the number of Kona slots for the Pro women has started. (Most notably, Rachel Joyce made a case for equal slots in her article on witsup.com about the new KPR system, and the issue has been picked up by a number of others.) I hope that by adding some numbers I can help to further the discussion of this issue.

The case for 35 women slots

WTC is offering 50 slots for the professional men and 35 for the professional women. The main argument for this imbalance is that there are fewer professional women racing. This is in fact the case. Here are a few numbers:

• In Ironman races (2013 qualifying season up to end of July), women athletes constitute 38% of the professional finishes (or roughly 3 out of 8).
• There is a similar breakdown in the athletes ranked in the KPR system: There are 580 male and 308 female athletes ranked (35% female, close to 1 in 3 or about twice as many men as women).
• Jay Prasuhn cites another number in his article on the KPR: “The male-to-female percentage of athletes competing in Ironman and 70.3 events globally is at 67 percent to 33 percent.” (This probably refers to all athletes, pros and age groupers.)

Three different data points, all coming in at a similar result. 

Compared to this ratio, the number of slots (35 for the women compared to 50 for the men) seems reasonable (41% of slots). (Note: WTC has increased the number of slots for women from 30 to 35 for Kona 2013, but hasn’t changed that number when they tweaked the KPR system for 2014.)

The arguments for more slots

Even if the number of pro athletes is different between men and women, there are a few arguments why the number of slots should be the same:

1. “Closed Shop”
2. Comparable Sharp End
3. Equality
4. Why not?

I’ll have a closer look at these arguments and some of the supporting data.

“Closed Shop”

Rachel gives a succinct explanation:

[T]his makes Kona a bit of a “closed shop” for women. … If we assume that the top 10 in Kona will in all probability get there again the following year, that means there are only 25 other spots open for the following year whereas men still have another 40 spots up for grab.

This theoretical observation is backed up by the data for the July qualifiers:

• For the men, there are 21 athletes that finished in Kona 2012, and 22 “new” athletes (51%).
• For the women, the numbers are 18 Kona finishers and 12 new athletes (40%).

Comparable Sharp End

The idea of this argument is that the total number of athletes may be different, but that the “density” of the top athletes (say the top 50 that might be potential Kona racers) is comparable between men and women. I’ll have a look at this theory from a few different angles.

KPR points

• Comparing the KPR numbers of the top ranked male and female athletes, the females have more points up to about #35. For example at #28 the difference is about 700 points. 
• After that, the male athletes have only a few more points until a clear separation occurs at about #70.

My interpretation of this data (taken from my post “Do Women have to race more often than men to qualify for Kona“): There is a discernible draw to qualifying for Kona, and athletes close to the cutoff carefully choose their races in order to maximize points. This occurs both for men and women and I can’t really see a gender difference here.

Time difference between top finishers

I’ve put together a list of the finishing times of this seasons Ironman races and have observed the following:

• The top 3 in a women’s race are usually closer together than the men. (The median time difference between 1-2 and 1-3 are 5 minutes and 11 minutes for the women versus 7 minutes and 12 minutes for the men.)
• After that the men are a bit closer together (for 1-4/1-5/1-6 we have 14/18/22 minutes for the men and 16/22/26 for the women). 
• Even so, the women’s differences are a lot smaller than what could be expected from the “raw” number of athletes. (Half as many athletes usually implies that the time difference should be twice as large.)
• In a women’s race, we also see more changes in the lead between T2 and the finish (in the women’s races, the winner was not leading after the bike in 17 races, in the men’s the lead changed in only 12 races).

To sum up, a women’s race can be more exciting that a men’s race. However, if you want to properly follow the race, this requires a “clean” women’s race and similar coverage which is another issue that should be addressed by WTC and race organizers.

Difference between top TTR rated athletes

My own rating system (“Thorsten’s Triathlon Rating” or TTR) takes all professional finishes over the Ironman distance into account (not just WTC races, and also over a longer period of time than the KPR which just looks at the current season). Here is a graph that show the relative difference (in %) to the top ranked athlete (Andreas Raelert for the men, Mirinda Carefrae for the women):

Again, we have a very small difference. The drop is about the same up to #20. After that, the men’s drop off is a bit slower than for the women, but nowhere close to what would be expected by the fewer number of female athletes. (For the data geeks: The slope between #30 and #95 of the men’s curve is about 0.055 percent per spot, that number for the women is 0.087 percent per spot. The relation between these is a factor of 1.58 instead of the expected number of about 2.)

Equality

Rachel has a clear point of view on this:

We count for half of the world’s population.  We should want to see triathlon be an equal sport in future generations.  What message are we sending out to kids taking up the sport with this disparity?

I would like to add two observations to this:

• The prize money is the same for men and women.
• Adding more Kona slots for the women would make racing as a female pro more attractive, and it should help attract more pro women to WTC races.

Why not?

It’s hard to image that it’s not possible to accommodate 15 more athletes on the Kona pier. After all WTC manages to provide a number of slots to new races on the Ironman calendar (recent examples include Japan, Lake Tahoe, Boulder and Copenhagen).

My opinion

I think the data that I presented here gives a strong indication that just looking at the number of professional athletes does not paint the full picture of the state of professional women’s Ironman racing. Personally, I’d be glad if WTC offers more women the chance to race as a professional in Kona.

WTC has announced some changes to the Pro qualifying system for Kona (known as Kona Pro Ranking or KPR). This post has a closer look at the changes and the impacts for Kona qualification.

Changes to the KPR

On July 5th, WTC announced changes to the KPR system for the 2014 season (apparently after some discussion with athletes). The changes are:

• Kona becomes a P-8000 race (instead of P-6000).
• There are no more P-1000 races, all of these become P-2000 races.
• The points in each race favor the top racers (i.e. the drop off is much steeper, e.g. in a P-4000 race 6th place get 1.670 points instead of 2.240). This also applies to 70.3 races.

Some things will stay the same:

• The number of athletes (50 male, 35 female) and cutoff points (end of July and August) are unchanged.
• The regional championships (currently Melbourne, Frankfurt and Mont Tremblant) stay P-4000 races.
• Point levels for 70.3 races stay the same. (There are still 4 levels of 70.3 races: P-3000, P-1500, P-750 and P-500 races.)

Impact to KPR qualifying

After running a simulation of the new KPR system, the main impact is that KPR cutoffs move down:

• For men, the cutoff moves (at #43) from 3.790 down to 2.820 points (a reduction of 25%).
• For women, the cutoff moves (at #31) from 4.740 down to 3.930 points (a reduction of 17%).

At first this sounds a bit surprising with P-6000 becoming P-8000 and all P-1000 races becoming P-2000 races, but the steeper drop-off results in the total number of KPR points going down. (The total number drops from about 973.000 points to about 807.000 points – a reduction of 17% which is pretty much in line with the KPR cutoff reductions.)

Here are some athletes that would not qualify at the July cutoff:

• Petr Vabrousek (dropping from #13 with 5840 to #48 with 2655)
• Balazs Csoke (dropping from #34 with 4060 to #51 with 2565)
• Thomas Gerlach (dropping from #42 with 3800 to #86 with 1640)
• Mirjam Weerd (dropping from #18 with 5840 to #34 with 3590)
• Mareen Hufe (dropping from #23 with 5280 to #36 with 3405)

Instead, some of these athletes would get a spot:

• Daniel Fontana (moving up from #47 with 3625 to #28 with 3765)
• Marino Vanhoenacker (moving up from #45 with 3720 to #33 with 3450)
• Luke Bell (moving up from #78 with 2230 to #37 with 3230)
• Jennie Hansen (moving up from #38 with 3960 to #25 with 4685)
• Eimaear Mullan (moving up from #40 with 3910 to #28 with 4350)
• Rebekah Keat (moving up from #36 with 4160 to #29 with 4000)

If you want to have a closer look at my simulation results, you can download an Excel sheet with the KPR points under the old and new system.

Strategy implications

Here are a few implications for qualifying strategies that I see:

• after a top 10 in Kona you basically only have to validate (for the men, even a top 13 is probably enough)
• winners and podium finishers are rewarded (e.g. Marino, Luke Bell, Jennie Hansen, Rebekah Keat – another case that would be avoided now is Mary Beth Ellis having to win three Ironman races in order to qualify for Kona 2011)
• the changes make it harder for the frequent racers (e.g. Petr Vabrousek, Thomas Gerlach)
• By sacrificing prize money, you may be able to pick a weaker P-2000 field.

To further nail this down, let’s have a look what is needed to qualify. The July cutoff of 2.820 (men) and 3.930 (women) from up to five results means that an athlete needs on average 564 (men) and 766 points (women). In a P-4000 this corresponds to a 10th and 9th place, in a P-2000 to a 5th and 4th place. (In the old system, you qualified by placing an average 10th or 8th in a P-2000 race.)

It is clear that qualifying by “racing often” is now very hard to accomplish. Instead, athletes have to make sure that they get the majority of points needed to qualify from one very good result. (After winning a P-2000 race, you can probably qualify with a few extra points from 70.3s.)

Where can this lead to?

• I expect more strategic “DNF”s: If an athlete is too far behind during the bike, there is not much “business sense” in trying to move from 12th place to 10th place: Still no prize money, and not enough points for qualifying. Instead, it’s probably better to call it a day and try again in the next race.
• It’ll be interesting to see what happens to the fields in P-2000 races. Originally I thought they would “even out”, but I think that the races with the better prize money (the “old” P-2000 races, typically with a 75k$ prize purse) will continue to attract the better fields – which will make it easier in the old P-1000 races to make some decent points. (As far as I know, most of these races still offer 25k$.) This is a tough choice to make for the professional athletes.
• There will be even more changes to the start lists in the last few days than we already have. (If there’s a strong field, withdraw; if there is a weak field, try to enter late.) Athletes can register for as many races as they want, and can enter or pull out just a few days before the race without any penalties.

In order to address the last point, I’ve been suggesting for some time that WTC should introduce tighter procedures for registering for and withdrawing from races.

Overall Assessment

In my opinion, the KPR system has been working well in the last years in choosing the best athletes for Kona. (There are always some close calls around the cutoff.) I like the increased focus on the top finishers, but I think this is a minor tweak of the part of the system that has been working pretty well.

However, the main point of contention has not been addressed: The doubts whether the KPR system results in the best Kona race, i.e. the best athletes in their best Kona shape. Some questions that should be discussed:

• the qualification requirement of one full Ironman race in addition to Kona (some “pathological” cases: Andreas Raelert walking the marathon in Regensburg 2011 to recover from an injury; lots of athletes following their Kona race with a “going through the motions” late-season IM just to validate)
• qualification races in August (not leaving sufficient time for recovery and a decent Kona build)

I would like to see the discussion around the KPR move to these issues.

There are a lot of long-distance races on the schedule that are not quite as big as WTC-branded Ironman or Challenge races. Probably the biggest of these is evens in the Rev3 series which mainly offers 70.3 distances but also has a few long races. I’ll have a closer look at the Rev3 races soon, but for now I want to have focus on some “unaffiliated” races.

In order to figure out what to do with these races, I had a closer look at three races:

• Norseman: One of the fantastic races that add a great, scenic course to the long distances
• Vineman: Another “pretty race”, but one that also has a long history – it’s probably the oldest Ironman distance race on the North American continent (i.e. not counting Kona)
• Ostseeman: An example of the smaller, non-affiliated races (and very dear do me, as this is one of the IM-distance races I have done and just one hour drive from where I live)

While I like these races, analyzing them with a Pro focus (such as I have for trirating.com) is tricky:

• they don’t have a dedicated “Pro” category (making it hard to identify the Pros, especially new ones or those that have recently changed names)
• very few Pros compete in these races (I already have some “matching” algorithm to deal with slight spelling differences, but even with that I can’t identify too many Pros)
• even if I can identify a Pro, older results may be from back when they were still racing age-grouper (and including these result would skew their “Pro” rating)

In short, I have very few good data points for these races, making an analysis very hard and probably not very meaningful. Therefore, I have decided to focus my work on other races and the questions related to these.

Having said this, by pulling together all “Pro” results from all the years I managed to come up with an “unofficial” course rating for Norseman:

Rating	Swim Rating	Bike Rating	Run Rating
-2:07:49	-06:30	-1:00:47	-0:59:27

As this is not a result of my usual calculation procedure, I will not include this in my list of course ratings. Still, these numbers give a good indication how hard the Norseman course is.