Thursday, March 5, 2020

Caroline Bloom 2020 Ultraman Florida Training and Race Analysis


Caroline Bloom just finished second at the 2020 Ultraman Florida event in the largest field of women ever assembled there.  She put up the 8th fastest swim, 8th fastest day 1 bike, 8th fastest day 1 combined time, 7th fastest day 2 bike, and 8th fastest run which was all good for the 6th best combined time ever.  Below is an analysis of the process we took over the past 28 months.  In the next week I will post a separate analysis of her performance at Ultraman Florida.  I’ll describe our starting point and touch upon all the major considerations we made, that guided that process, without going too far down the rabbit hole in any one direction.

When Caroline and I met at Ultraman Florida in February 2017, where we were both crewing for athletes, she was less than a year removed from her first sprint triathlon.  She had done a few running relays and a couple running road races over the previous 2-3 years, but had absolutely no athletic experience prior to that. Her 5k PR was 24:58; her half marathon PR was a 1:56:10 from a couple weeks earlier.  From a technical standpoint, she couldn’t drink from a water bottle without stopping her bike.  She wasn’t super comfortable in the swim, but was proficient enough that it wasn’t a concern for the sprint tri’s she had done or had coming up.

She was damn determined though.  In the wake of seeing Ultraman she signed up for her first half iron.  Some of the stats from that race are below.

Kerrville 2017:
Swim: 44:45 (2:02 per 100y/2:13 per 100m)
Bike: 3:13:23 (134w avg./146NP)
Run: 2:24:58 (11:02 per mile; 6:51 per km)

We started working together the week after that race.  Ultraman 2020 was the end goal; it had been her end goal since she had seen the race in February – I like to encourage the athletes I work with to dream big and to have patience with the process and let progress happen naturally, so part of that went into setting the exact year she was aiming for (full disclosure, and probably the biggest insight into where our starting place was, is that with most athletes I am trying to get them to consider taking a little more time depending on their goals, where Caroline kept asking if 2020 gave her enough time just to be able to finish the event).  What that means typically is having an A+ race or event we are building to (3-5 years down the road) and then there are A races (true peak/taper) 1-2 times per year with other B and C priority races sprinkled in (typically B races will involve short rest and are on the calendar well in advance; C races can be inserted at any point and are never prioritized with rest).  So with all that in mind, we mapped out that her schedule to get to UMFL (A races in Bold):

Kiawah Marathon – December 2017
70.3 Galveston – April 2018
24 hours in the Canyon - June 2018
Kerrville Half – September 2018
Ironman Arizona -  November 2018
*Cherry Creek Half Marathon – April 2019
*24 hours in the Canyon - June 2019
*Kerrville Sprint and Half – September 2019
*Denver Half Marathon – October 2019
Ultraman Florida February 2020

*races added after Ironman Arizona, all B priority


In the first month we did a few testing sessions: effort based on the run and technical based in the pool.  The bike wasn’t a huge priority with the marathon coming up, so we just kept those sessions easy.  To be honest, all was pretty easy with recovery in mind from her first big, long race at Kerrville; her weekly volume ranged from 4-8.5 hours that first two months prior to the marathon.  In there she performed a 2-mile run time trial.  She averaged 819 pace per mile (5:10 per km); average HR was 169bpm and max HR was 177bpm and it was even paced.  On the swim I prescribed a number of easy to high effort swims and had her track stroke count per lap.  Stroke count went as low as 23-24 strokes and up toward 29-30 depending on effort and distance. 

Taking into consideration all of the above we set the following priorities:
1.       Bike technical skills
2.       Swim efficiency (focus on distance per stroke)
3.       Aerobic base
4.       General strength and mobility

Much of those four came from her lack of experience.  In order to train any serious volume she needed to be strong to prevent injuries and adapt to the workload, the aerobic base was important since she had no prior history in endurance athletics and the events she was doing were highly aerobic, and a lack of technical skills on the swim and bike (again, mostly due to inexperience) were already limiting her performance even before she gained a high level of fitness.  Those four things remained priorities throughout the entire 28-month process to Ultraman, though some of the specifics changed. 

In the beginning Caroline was practicing left, right and U turns in empty parking lots for 10-15 minutes at the beginning of most outdoor rides as well as practicing sipping from/removing her bottle during rides.  By December 2019 she was forcing herself to stay aero as much as possible on the Queen K in windier conditions.  And progress was not linear.  There were rides right up until two weeks before the race that were overwhelming from a conditions standpoint, but she kept getting out there and kept increasing her experience and slowly but surely is gaining greater and greater skills.  On many of these rides we throughout any specific effort targets and just let the skills/conditions acquisition be the primary focus.

On the swim, much of the first year was based on achieving technical proficiency based on strength and distance per stroke.  As someone who was inexperienced her stroke count would often jump around 3-4 strokes per length in a workout without her being able to realize what it was that allowed her to take fewer strokes, which was frustrating for her.  As she progressed through those workouts and gained experience the range decreased, she became more aware of what allowed her to take fewer strokes and she cemented the movement patterns that were efficient.  As she got more efficient we started adding pace specific work (to build specific strength), then in the final 6 weeks we added some rate work which really allowed her to maximize her fitness in different conditions. 
We addressed aerobic base from both an intensity and volume perspective.  And important to note we did not use the event distances to dictate what we needed to be doing with that volume.   Below are some of the totals by year for the 28 months leading up to Ultraman Florida.  The majority of that was at very low intensity, based both on heart rate and pace/power.  Early on heart rate was the primary metric we used, but as aerobic strength and efficiency increased that was largely replaced by pace/power in the last 6-8 months.

Year 1 (Oct 1 2017-Sept 30 2018)
420 total hours (8 hrs per week)
1.42 hrs per day active (70 days off)
Peak week: 21 hrs (14 weeks above 10 hrs)
Swim: 185,000yds/169,000m (15% volume)
Bike: 2850 miles/4590km (50% volume)
Run: 608 miles/980km (28% volume)
Strength: 5.5% volume by time

Year 2 (Oct 1 2018-Sept 30 2019)
570 total hours (11 hrs per week)
1.84 hrs per day active (55 days off)
Peak week: 20 hrs (12 weeks above 15 hrs)
Swim: 279,000yds/255,000m (15% volume)
Bike: 4000 miles/6440km (48% volume)
Run: 553 miles/890km (21% volume)
Strength: 13% volume by time


UMFL Specific Build (Oct 1 2019-Feb 2 2020)
245 total hours (14.4 hrs per week)
2.02 hrs per day active (6 days off)
Peak week: 21.5 hrs (3 weeks above 19 hrs; 6 weeks between 9-12 hrs)
Swim: 154,000yds/141,000m (19% volume)
Bike: 1700 miles/2735km (48% volume)
Run: 305 miles/491km (23% volume)
Strength: 8% volume by time

Volume only tells a piece of the story and usually not the most important piece.  We took the position that consistency is a bigger and more important piece of success.  So, we started at a very modest weekly volume, as mentioned above.  There were greater and lesser volume weeks, but the majority of weeks were concentrated within a fairly tight range.   These numbers show that as time went on, overall volume and consistency increased.  Keep in mind there were some mandatory off days after A races as well as those built in to the schedule at different points in the year, that skew the yearly numbers in comparison to the last 4 months, but a good number of those off days early on came from sickness, oftentimes associated with travel.  We got better over time at reading the warning signs, limiting back-to-back travel, and/or scheduling many light days on either end of that travel to limit getting run down.  We used heart rate variability, subjective metrics, and descriptions of perceived exertion on all workouts to help track that recovery.  There was a fairly large volume jump in her daily average (1.42hrs to 1.84 hrs) from years 1 to 2 but that did not really increase much in the final build (2.02 hrs).  Where we found the majority of the volume increase was in the consistency of taking fewer days off, and with fewer massive days that required excessive recovery we were able to keep progress and fitness moving forward consistently.  There were bigger days but we took a much more gradual approach, looking at big weeks and big blocks of training stress vs. huge single days of stress.

To expand on what was mentioned above, the majority of her work was done at very low aerobic intensities (approximately 90% of volume by HR and 80% of volume based on pace and power). For Caroline that meant approximately 80% of her volume was below 138bpm on the bike, 10% between 139-147bpm and 10% over 148bpm.  On the run she was approximately 80% below 150bpm, 10% between 150-159bpm, and 10% above 160bpm.  Paces and powers changed throughout so that is a bit more difficult to pinpoint here, without making this article more unwieldy than it is.  We emphasized these intensities because they were key for metabolic and aerobic development, and that targeting those areas, at this time, were what would lead to the greatest success not only at Ultraman but also for any distance and future race she wished to pursue post-Ultraman.

Following her marathon in December 2017, we shifted the focus to the bike.  This was due to her cycling only A race, but also worked with the bigger picture because it allowed us to ramp up volume with lower impact exercise.  Volume and frequency remained high until 24 hours in the Canyon.  She had done the race in 2017 and had slept through a good portion of the night, but still had gone just over 200 miles.  In the leadup we had identified a couple simple goals to improve cycling efficiency: one was to keep pedaling (limiting time at zero power and zero cadence), and then to produce a moderate amount of power at all times when pedaling.  There were rides early on where 15-20% were at 0-20rpm and 0-20w, due to weather conditions and discomfort with traffic.  The stated goals for Canyon were to pedal when on the bike and to ride for 24 hours (limiting time off the bike to under 90 minutes, as opposed to 7 hrs the previous year).  These goals were met, but Caroline was still unhappy with the 224-mile result (360 km).  We discussed the mental side of endurance athletics regularly in training, regarding how she approached some workouts or conditions.  We explored her mental state a great deal in the analysis of the effort in Canyon: we discussed the differences in conditions she faced between the two years (much more difficult that year), how she mentally approached the tough points in the middle of the night, and why she was feeling the way she was when she’d technically accomplished the goals we had set forth.  This conversation evolved into what her goals truly were, beyond what were stated; it was a conversation that frustrated both of us, was uncomfortable at times, and took quite a bit of time.  These are conversations that can’t happen without significant investment of time and energy and require a deep level of respect and trust in each other because growth isn’t easy.  This conversation continued and evolved over the next year, particularly through the following races:

Kerrville – September 2018:
Swim: 37:17 (1:37 per 100y/1:47 per 100m, 139 avg HR/149 max)
Bike: 3:09:16 (138w avg./146NP, 143 avg HR/164max)
Run: 2:20:40 (10:42 per mile/6:38 per km, 158 avg HR/173 max)

Ironman Arizona – November 2018:
Swim: 1:16:59 (1:42 per 100y/1:52 per 100m, 140 avg HR/154 max)
Bike: 6:32:19 (121w avg/127 NP, 143 avg HR/158 max)
Run: 5:07:10 (11:42 per mile/7:16 per km, 150 avg HR/169 max) 2:38-2:29 negative split

We approached Kerrville with the goal of a negative split on both the bike and run.  We discussed executing a solid race with a PR, but also wanted to ensure a quicker recovery so we could keep the build going to Ironman Arizona.  Overall, the race was a giant success and was significantly faster than the year prior in all disciplines.  After we looked at the race it was obvious that she would be able to race close to those numbers at Ironman Arizona.  Part of what helped us make that determination was the continuation of conversations we had about her perceived effort during the race and the difference between her motivation and ability to go harder, and why she made the choices she did. 

Approaching IM Arizona I had an idea that her aerobic training zones were slightly low and that her potential was slightly above some of the targets, but without reliable data to the contrary was happy to keep zones slightly lower. We compensated for this by expanding her target zones for the race, and then leaving an option to push as hard as perceived exertion allowed last loop of the bike and second half of the run.  She spent most of the bike and run at the upper end of her target HR ranges.  Power faded a bit through the 3 loop course and then she negative split the run.  Between this field data and the lab testing she would have done in August 2019, we confirmed that during that first year her zones were set approximately 5bpm low, which I think actually proved to be an advantage, as she absolutely never drifted above her aerobic targets (easy was truly easy all the time).  Again, some of what helped us draw some of those conclusions, prior to the lab results, were the conversations regarding her decision making: when she chose to push and to what extent.  We approached it not from a point of criticism on the race (it was a huge success on all fronts) but as a learning process for how we could structure the next year of training, both from a mental and physical perspective, and allow her to race even stronger at Ultraman.

As mentioned above, in August 2019 she was able to get some metabolic testing done.  We had a pretty good nutrition plan dialed in by this point, but the data provided some great information to cement exactly how she was burning fat and carbohydrate at different effort levels.  This allowed us to have greater confidence in the nutrition plan as well as confidence in the effort we thought she could push as well as how hard she could surge (and for how long).  And we took that information into her third run at Kerrville.  Because Ultraman is a stage race, we decided to use the whole weekend: she would race the sprint distance on Saturday and the Half Iron on Sunday. 

Kerrville 2019

Sprint Split
Power/Pace
HR
Half Split
Power/Pace
HR
Swim
N/A
1:34/100y
1:44/100m
137 avg
145 max
42:00
1:47/100y
1:57/100m
Didn’t record
Bike
N/A
170 avg
177 NP
147 avg
156 max
2:55:45
174 avg
182 NP
148 avg
159 max
Run
N/A
10:16/mile
6:22/km
157 avg
166 max
2:59:10
2 miles @
10:16/mile
6:22/km
2 miles @
150 avg
157 max

The plan was to execute the sprint at goal Half effort and then hit the same power and paces the second day in the Half.  I’m not including splits for the Sprint since they are random distances.  The HR data and pace for the Half run only includes the first 2 miles/ 3 km because the remainder included a large amount of walking due to extreme abdominal pain.  We explored a variety of possibilities for this occurrence but was tough to definitively say.  It did make us cautious about over hydrating on the bike (Caroline took in some extra, and very quickly early in the bike), and second-guess a few things we thought we knew about the nutrition plan.  Ultimately, we drew confidence from having executed the plan in workouts prior to this and then re-affirming that in workouts following Kerrville.  We also considered that at different points in her menstrual cycle hydration needs change.  We thought we had some of those things figured out, but ultimately decreased fluid intake even further when she was in the high hormone phase of her cycle. 

For over a year we had structured Caroline’s training around her menstrual cycle, using the research to create a situation that would lead to the greatest adaptations.  We had identified from a year out that Ultraman was likely to coincide with the final week of her cycle.  Beginning with Kerrville, we structured her training so that her more intense session were early in her cycle, then more aerobic and race-like in nature during the last 10 days of her cycle.  We played a bit with some intensity in those later sessions to simulate any potential surges or race moves she might want to make.  What we found is that there were really no limitations on intensity of those surges in an aerobic situation.  Building on what we had learned earlier in the year, as long as her body gradually acclimated then she could build to moderate and high intensity. We never tried anything super intense, above threshold, but she was able to perform well on longer threshold intervals following a long easy start to the ride.

Since running had not been a high priority in 2018, we added half marathon races in April and October.  These would provide good long hard runs in a race setting and provide a good comparison to see how fitness was improving.  Overall, it showed exactly what we wanted to see.  Aerobic strength and efficiency improving and pace improving dramatically.  The October race also provided a nice confidence boost following Kerrville.


Time
Avg. HR
Max HR
Pace
April 6
2:09:36
156bpm
166bpm
9:03/mile 5:37/km
October 20
1:57:07
155bpm
170bpm
8:40/mile
5:23/km
*both races had 300 ft/ 91m of elevation gain/loss

At this point, I think it is important to note that the original plan was to ramp up volume between August and November, through this entire time frame.  There would have been a recovery block leading into both Kerrville and the half marathon but overall the volume was planned to trend up towards 17-19 hours per week, setting up a three week training block in Hawaii in December.  That did not happen.  For 8 weeks from mid-August to mid-October all weeks were between 8.75 hours and 13.5 hours.  This turned out to be a high stress time for Caroline, and without going into detail, we backed off the training stress to accommodate the additional life and work stress.  As things began to stabilize we gradually ramped up toward where we wanted to be training.  We didn’t try and jump up to any preconceived workouts of what we should have had her performing in November, but gradually added stress and ended up stabilizing at 15 hrs per week for most of late October and November, with an emphasis on strength (both sport specific and gym-based sessions).

In late November Caroline traveled to Hawaii and crewed at Ultraman World Championships.  Taking into consideration what we had learned over the previous years, and the high stress period she had just come off, the week prior to travel was a deload week of 10 hrs (with no complete off days), and the week of travel/crewing was super light with 5 hrs of volume (3 complete off days while crewing).  She came off this rested, and with only training and recovery to pay attention to was able to put up weeks of 21.5, 18.5, and 20 hours.  The longest ride of the block was 7 hrs (85 miles/137km with 8000 feet/2400m of gain and 4500 feet/1300m of loss) on the third day.  Originally I had planned 3 longer rides targeting harder climbing sessions to really start to take Caroline out of her comfort zone on longer sustained intervals, but early in the second week we were out on the Queen K and the winds and traffic rattled her to the point that effort was not the limiting factor. The biggest reason we instituted this change to the plan is because it shifted the focus from a secondary priority (energy system/threshold development) back to a primary priority (cycling skills acquisition). We were still able to attempt the energy system development out on the Queen K, though it was more difficult for her to sustain that speed and effort amid uncomfortable surroundings, but that was a trade-off we were comfortable making in the context of long-term development.  Taking into account there may be winds in Florida, and there would certainly be traffic, we made the decision to change the plan and gain exposure in those conditions as much as possible, so she ended up doing 7 x 3-4 hr rides out on the Queen K over the final 15 days.  We operated under the assumption that the volume of aerobic work was what was important and not necessarily getting up to 7-8 hrs frequently, and that her muscular endurance came from that weekly and monthly volume more than from any specific extremely long ride.  Speaking to her mental toughness, she got frustrated with me multiple times through this process and each time she would take her space and quietly get tougher on the highway, staying aero more and more and pushing at intensities she had the fitness for more and more.

Following her block in Hawaii, and just before New Year she went to Arizona.  While she was there she did a 10k swim on Saturday and her longest ride of the buildup to Ultraman (an 8hr, 120 mile/195 km aerobic effort), on Sunday.  The only reason this is of note is because of the conditions in the lake.  The water was 54 degrees (12 Celsius) and the wind was super strong which created a ton of waves the second half of the swim.  Caroline really struggled in the waves at first.   Eventually she settled but was far from effective with a low stroke rate.  It wasn’t that it was terrible, but significantly slower than if conditions were different.  We took that information and decided to combine some rate work to her remaining training.  With all the experience she had in the water (and experience thinking about what she was doing), she took to these changes quickly and was able to operate at a higher or lower stroke rate without spiking her effort. 

When she returned from all her travel, the plan was to recover a week and then hit one last solid training block with three longer 5-6 hr rides.  Her recovery metrics and perceived exertion were a little off over that week, so I put some restrictions on the first long ride – if the power targets couldn’t be hit, rather than fighting it out or turning it into a long slog on the trainer that still felt difficult, we’d take it that she needed more recovery and just call it a day.  That is exactly what happened, and that turned into a 96-minute ride that failed halfway through the first hard interval.  Three days later she performed a 3-hour ride with less perceived exertion and at similar intensity, and the remainder of the block went exceptionally well.

The last piece of the build that I will speak to is the progression of long runs, or rather absence of super long runs.  In the last four months Caroline performed 13 runs between 90-125 minutes.  Basically, this broke down as three per month, and four in January (3 of those were between 120-126 minutes).  We approached those runs with the same perspective that we approached most of the other volume, that consistency was more desirable than excessive length, especially due to the longer volume.  Over that time the runs gradually increased from more near the 90-minute end and entirely aerobic, to more at the 2 hr end as well as including more intensity.  As much as the long runs I look at two 5k races that we added last minute, on fatigued legs. Both were PRs and both showed a high degree of fitness, pacing and strength, all of which led to some good run confidence without pounding her with excessive mileage that may or may not have been difficult to recover from.


Time
Avg HR
Max HR
Pace
December 22
23:48
162bpm
178bpm
7:39/mile
4:45/km
February 1
22:45
168bpm
178bpm
7:16/mile
4:30/km

RACE ANALYSIS

For all disciplines there is a simple equation for how fast you go:
force (distance per stroke, stride length or gearing) x rate (stroke rate or cadence) = speed. 
I’m of the philosophy that if you really want to make a change you have to really focus on one (maybe 2) thing(s).  For Caroline, in the swim, as mentioned above, that was on the force she applied and for new or inexperienced swimmers oftentimes that can be accomplished with a reduction of effort.  Over two years it was obvious that she was making gains in her efficiency and maintaining stroke count and that she had translated this well to longer aerobic swims both in the pool and open water.  With her aerobic base from her work in other disciplines, combined with a lower stroke rate she was oftentimes unable to raise her heart rate in the pool.  Following her metabolic testing, a couple of max effort time trials, and knowing where we wanted her heart rate on the bike it became obvious that she could swim as hard as she wanted and she would not be going too hard, and we discussed that. 

Then she swam the 10k in Arizona and we decided to work a bit on her rate to give her the better tools to swim in any condition.  A week and a half after that her shoulder started bothering her so we largely backed off the swim.  We kept in a few short swims and focused on swimming easy at higher rate.  Since nothing was long and nor super hard we didn’t have a good idea about how this translated to the race, though I was skeptical that there would be a massive change based on a few thousand yards spread out over a month.

I misjudged.  There were signs based on her subjective feedback (lower RPE combined with slightly higher speed) but it was mostly on very short intervals (20 seconds) with almost 1:1 work to rest, so I dismissed the impact it could have.  She entered the race motivated and thinking she could go as hard as she wanted.  The data from the three 10k swims she did over the last 5 months are below:


Time
Pace per 100y/100m
Stroke Rate
Avg. HR
Max HR
September 8
3:55:25
2:07/2:18
30
128
139
December 28
3:55:39
2:08/2:19
29
121
141
February 14
3:13:47
1:43/1:53
34
138
150

If we break down the swim even further some things pop out.  Her peak stroke rate for 10 minutes was 36rpm was before the conditions got rough, and her peak HR was hit 20 and 60 minutes into the swim (both before the conditions got rough).  What this shows is that we did not teach her to raise her stroke rate based on conditions, we raised it overall, which also led to an increase in effort.  It did however, have the intended effect of making her much stronger when the conditions got worse.  Her peak pace for 10 minutes, with a big current and tailwind, was 1:32 per 100y (1:42 per 100m).  Her slowest pace, into the wind and waves, again for 10 minutes, was 1:56 per 100y (2:06 per 100m), and her peak pace immediately followed her slowest pace, in and out of the 4th turn buoy.  We did not anticipate that and got lucky that her effort didn’t spike too high to affect the rest of her race.  As it turned out the effort she was able to exert was optimal for the distance she was racing and translated well to her effort out onto the bike as well as for the first day of a three day race, and put her in second place woman, coming out of the water.

The plan each day was to finish stronger than she started and to get relatively stronger each day, so that the real “racing” was done in the last 30km of the run.  From her cycling training her top 18 peak 60-minute power marks were set since the beginning of August.  The other two top-20 marks were threshold effort workouts that were both 162w for 60-minutes, set in March of 2019.  Her peak 3 hour power all time was set at 157w during a 70.3 relay on August 3, 2019.  This didn’t end up being a total max effort because the course was crowded and she didn’t always feel comfortable passing, but it was still a high end effort (141 avg. HR/164 max), though if she was maxing out her HR likely could have been 12-15 bpm higher (peak 60-minute HR was 151bpm, was also her peak for the year at that point, and which corresponded to a new 60-minute max power at the time).


Time
Avg. Power
Normalized power
Variability
Index
Avg. HR
Max HR
Day 1 Bike
4:49:13
173w
179w
1.03
138bpm
155bpm
Day 2 Bike
9:21:37
175w
185w
1.06
134bpm
159bpm


Peak Power

1 minute
5 minute
20 minute
60 minute
3 hour
Day 1 Bike
222w
194w
186w
180w
177w
Day 2 Bike
274w
216w
194w
188w
182w

Those peak powers represent 10th best all-time 1-minute power, 7th best all-time 5-minute power, 16th and 6th all-time 20-minute power, 7th and 4th all-time 60-minute power, and 1st and 2nd all-time 3-hour power.

On day 1 all of those peak powers happened in the middle of the bike, while on the second day there was a split.  The 1 and 5-minute max powers came in the second hour of the day on two of the biggest hills on the course.  The 20-minute, 60-minute, and 3-hour peaks all happened in the second half of the day once the terrain flattened out.  If we break down her ride on day 2 further we can see how the hills on the first half compare to the flatter terrain in the second half, and further compare to the first day’s ride.


Time
Avg. Power
Normalized power
Variability
Index
Avg. HR
Max HR
First
81 miles
130 km
4:38:37
174w
186w
1.07
130bpm
155bpm
Last
91 miles
146 km
4:43:00
178w
183w
1.03
138bpm
159bpm

While it was always part of the plan to push harder in the second half, one of the things we discussed after the first night was while her power numbers on day 1 were better than she’d ever done, she had stayed in the effort (heart rate) range that we were targeting and her efficiency factor was lower than it had been in many of her training rides.  So, with the assumption that the harder swim had negatively affected her ride to some extent, we discussed that day 2 could likely be ridden as strong or stronger than day 1 and sure enough that panned out.  Caroline’s Efficiency Factor was 1.30 on day 1 and was 1.38 on Day 2 (1.43/1.33 by half).   

On the run the plan was to start relaxed, keep HR low, and then be able to push the last 30km.  After the first two days we knew that Caroline was in contention for second place, as she was only 16 minutes behind.  Knowing how much the increasing heat of the day combined with the cumulative pounding of hours on your feet factors into the difficulty of the second marathon, it was super important to keep the perceived exertion low in particular. 

It’s impossible to give exact splits for the half marathons, but they were fairly even and all between 2:19:30-2:22:30; the other statistics by quarter are listed below as well as peak pace data:


Avg/max HR
Cadence
Pace 
Elevation +/-
First 13.1 miles
21 km
131/147
84
10:39 per mile
6:37 per km
350/360 feet
106/109m
Second 13.1 miles
21 km
138/154
84
10:41 per mile
6:38 per km
450/460 feet
137/140m
Third 13.1 miles
21 km
143/158
84
10:48 per mile
6:42 per km
300/330 feet
91/100m 
Fourth 13.1 miles
21 km
142/166
83
10:54 per mile
6:46 per km
190/130 feet
58/40m


1/2 mile
800m
1 mile
1600m
5km
Half marathon
Marathon
50km

Peak Pace
8:56/mile
5:33/km
9:13/mile
5:43/km
9:50/mile
6:06/km
10:18/mile
6:24/km
10:40/mile
6:37/km
10:39/mile
6:36/km
Location
Finish
mile 39
km 63
mile 14-17
km 22-27
mile 5-18
km 8-29
mile 0-26
km 0-42
mile 10-41
km 16-66

Those peak paces include the third faster all-time Half marathon, best ever marathon, and only 50km.  What is interesting to note is the faster pace for the 50km split vs. the marathon, and that there is a significant separation between where the two occurred.  There were walk breaks throughout, by design, except for near the end when outside factors started to become overwhelming.  Caroline had been experiencing patella pain starting 5 miles into the Day 2 bike and ½ mile into the run.  We got her a strap to put on her knee around 3 hours into the day, and that helped for a while.  By the time the effort of the day and a blister situation developed in the last hour of the run, motivation waned and walking became a bit more frequent.  But this was really relegated to miles 47.5-51.5.  If we look at the run a little differently, we can see how the day really developed:


Avg/max HR
Cadence
Pace 
First 34 miles/
First 55km
136/154
84
10:47 per mile
6:42 per km
Miles 34-47.5
Km's 55-76
146/158
84
10:26 per mile
6:28 per km
Miles 47.5-51.5
Km's 76-83
136/147
80
11:54 per mile
7:23 per km
Last .9 miles
Last 1500m
150/166
88
9:03 per mile
5:37 per km

Breaking things down that way, really shows how well she executed the plan.  The first 34 miles/55 km is relatively relaxed by HR, and based on her feedback throughout her perceived exertion was low for at least 25 miles but certainly rose following that mark.  However, as we discussed when I started pacing at mile 34, her HR was low, she wasn’t overheating, and her leg muscles felt good enough to push – I reminded her that perceived exertion was higher simply because she was at a point she had never been before.  She didn’t necessarily feel like she could go harder but recognized she could based on the metrics.  So we broke things up into smaller chunks (1/2 mile/800m to 2 mile/3km intervals at 915-945 per mile/ 545-6:00 per km pace with 60 second walk breaks) and that worked for a long time.  It’s not easy to push from that far out – it takes extreme mental strength and confidence.  And she accumulated a lot of mileage in that last 30km at that faster pace.  Because she was willing to push like that and because she was willing to push again over the final mile we can really see how things could have been different over miles 47.5-51.5/ Km's 76-83, and how she can probably maintain that as a steady effort the next time, which will inform our training strategy and goals when that time comes around.