IM Arizona

The current algorithm applied by Ironman to allocate Pro slots between and women (see my previous post “Kona Pro Slots – Part 1: Reverse Engineering The Assignment Algorithm“) is working based on the size of the field. This post describes two different approaches: Strength of the Field and Performance. I’ll discuss the Pros and Cons of these approaches in my next post.

Strength of Field

Measuring Strength of the Field

Measuring the quality of a field isn’t straightforward: I have tried a number of approaches that either didn’t work or were too complicated before settling on a relatively simple “points” system when showing in my seedings how good a field is. (The details can be found in my March 2017 “Strength of Field” post.) The base approach is to look at which starters have been racing Kona in the two previous years. As we’re looking for a system to determine how many athletes in each category to qualify for Kona, this seems to be a reasonable approach for assigning slots.

Basically, the points system determines how many “Kona athletes” are in the current race. Here’s how the system works:

• 1 point for each athlete that has raced the previous Kona race (so for the current 2019 qualifying season athletes that have raced Kona 2018),
• 0.5 points for each athlete that hasn’t raced the previous Kona race but the year before (athletes that haven’t raced Kona 2018 but Kona 2017),
• 1 bonus point for each athlete that has won Kona in the past,
• 0.5 bonus points for each athlete that has finished on the podium in Kona before

If you apply these numbers to the male and female fields, you get two numbers that you can then use as the base for the Jefferson Method as determined in the previous post. Please note that this system still favors the male athletes (as the male Pro field in Kona has been larger and there are therefore more male Kona Pros that can contribute points).

Example 1: IM Arizona

First, let’s have a look at the female field in Arizona:

• Kona 2018 athletes: Annett, Jackson, Kessler, Lester, Robertson, Smith (6 points)
• No Kona 2017 athletes that haven’t also raced in 2018 (no points)
• No previous Kona winners (no points)
• Kona podium: Jackson (0.5 bonus points)

This means the female fields “scores” 6.5 points. Next up, the men’s field:

• Kona 2018 athletes: Dreitz, Plese, Skipper (3 points)
• Kona 2017: Llanos (0.5 points)
• No previous Kona winners (no points)
• Kona podium: Llanos (0.5 bonus points)

The men’s field has a strength of 4 points. Running these number through the Jefferson Method (for unassigned slots), we get the following grid:

Arizona	Strength	1	2
Male	4	4	2
Female	6.5	6.5	3.25
Total		2 slots

The result would have been an even split of slots.

Example 2: IM Mar del Plata

Here’s the strength of the female Pro field in Mar del Plata:

• Kona 2018 athletes: Carfrae, Cheetham, Crowley, Lundstroem, Piampiano (5 points)
• No Kona 2017 athletes that haven’t also raced in 2018 (no points)
• Previous Kona winner: Carfrae (1 bonus point)
• Kona podium: Carfrae, Crowley (1 bonus point)

This means the female fields “scores” 7 points. The men’s field:

• Kona 2018 athletes: Chrabot, Hanson, O’Donnell, Potts, Weiss (5 points)
• No additional Kona 2017 athletes (no points)
• No previous Kona winners (no points)
• Kona podium: O’Donnell (0.5 bonus points)

The men’s field has a strength of 5.5 points. Here’s the resulting grid for these numbers, taking into account the extra slots:

Mar del Plata	Strength	1	2	3
Male	5.5	Auto	2.75	1.5
Female	7	Auto	3.5	2.33
Total		4 slots (1 each minimum)

Once again, we’d get even slots.

Performance

On the “If We Were Riding” podcast, Kelly O’Mara suggested to not “fix” the number of slot assignment before the race but to use the performance on race day. While her full blown version of “qualifying standards” is much more complicated and probably not workable given the differences between courses and conditions on race day, here is a suggestion on how to compare the performances between the male and female Pros in a given race.

Comparing Finishing Times

When looking at the best finishing times between the men and women, a conversion factor can be calculated. As a start, the average Top 10 finishing time of the male Pros in Kona this year was 8:04:02, the average Top 10 time for the females was 8:48:05. Dividing the female average by the male average, we get a factor of 91.66% that can be used to convert the female times into male equivalents and thus compare their performance. (Obviously, this factor will need to take more races into account if this approach is going to be used, but it seems to reasonable to calculate a factor after each Kona race and then use it for the remainder of the qualifying season.)

Example 1: IM Arizona

When applying the conversion factor to the female times, we get the following order of performances:

1. 7:55:59 Heather Jackson (8:39:18 * 91.66%)
2. 8:00:29 Carrie Lester
3. 8:04:24 Eneko Llanos (no change to his finish time)
4. 8:07:09 Jen Annett
5. 8:08:41 Clemente Alonso

As we have four slots available, three would go to the women (Heather, Carrie and Jen) and only one to the men. (This would also “observe” the minimum of one slot for each gender.)

Example 2: IM Western Australia

For Western Australia we get the following performance order:

1. 7:56:00 Terenzo Bozzone
2. 7:57:40 Cameron Wurf
3. 8:05:34 Caroline Steffen
4. 8:07:18 Matt Burton
5. 8:09:43 Luke McKenzie

IM WA also has four slots, in this case they would be assigned to three men and one woman.

Example 3: IM Mar del Plata

The conversion for Mar del Plata gives the following ranking:

1. 7:30:23 Michael Weiss (already qualified)
2. 7:38:33 Sarah Crowley
3. 7:39:47 Matt Hanson
4. 7:41:39 Susie Cheetham
5. 7:42:22 Mario de Elias
6. 7:46:30 Minna Koistinen
7. 7:46:39 Jesper Svensson (already qualified)
8. 7:48:09 Lukas Kraemer

With six available slots, we’d have an even split with three male and three female slots. (In addition, Mirinda Carfrae validates her Kona Winner slot.)