2017 Milwaukee Lakefront Marathon Training Cycle
Below is my justification for and an explanation of my 2017 Milwaukee Lakefront Marathon training cycle. The goal at the outset of this year was to get me as close to a sub 3-hour marathon and a BQ on October 1st for the Lakefront marathon.
In coming up with this plan, I decided to look back at the past training schedules:
2016 Wisconsin Marathon training cycle
Total weeks – 15 weeks
Total non-recovery weeks – 12 weeks
Total non-recovery weeks mileage – 662.5 miles
Average # of miles during non-recovery weeks – 55.2 miles
Total non-recovery weeks duration – 97:28:42
Average duration during non-recovery weeks – 8:07:23
# of weeks above 70% intervals hit (including easy) – 12/15 weeks (80% of weeks)
# of weeks over 8 hours training – 8/12 (75%)
# of weeks over 9 hours training – 3/12 (25%)
# of runs at 120 minutes – 1
# of runs over 120 minutes - 4
# of runs at 150 minutes (max duration) – 3
Week at which peak was hit – Week 7 and 12
Scheduled Marathon training pace – 7:52 min/mile
Average Marathon training pace in workouts – 7:55 min/mile
Actual Marathon Race pace – 7:53 min/mile
Final Marathon time divided by 26.2 miles – 7:58 min/mile
2016 Lakefront Marathon training cycle
Total weeks – 20 weeks
Total non-recovery weeks – 17 weeks
Total non-recovery weeks mileage – 1049.4 miles
Average # of miles during non-recovery weeks – 61.7 miles
Total non-recovery weeks duration – 149:53:18
Average duration during non-recovery weeks – 8:49:01
# of weeks above 70% intervals hit (including easy) – 9/19 weeks (47%)
# of weeks above 70% intervals hit (including easy) at the end of the cycle – 1/7 weeks (14%)
# of weeks over 8 hours training – 14/17 (82%)
# of weeks over 9 hours training – 9/17 (53%)
# of runs at 120 minutes – 0
# of runs over 120 minutes - 6
# of runs at 150 minutes (max duration) – 3
Week at which peak was hit – Week 11
Scheduled Marathon training pace – 7:33 min/mile
Average Marathon training pace in workouts – 7:44 min/mile
Actual Marathon Race pace – 7:43 min/mile
Final Marathon time divided by 26.2 miles – 7:47 min/mile
2017 Dopey Challenge training cycle
Total weeks – 12 weeks
Total non-recovery weeks – 10 weeks
Total non-recovery weeks mileage – 564.1 miles
Average # of miles during non-recovery weeks – 56.4 miles
Total non-recovery weeks duration – 73:23:09
Average duration during non-recovery weeks – 7:20:19
# of weeks above 70% intervals hit (including easy) – 11/12 weeks (92%)
# of weeks over 8 hours training – 3/10 (30%)
# of weeks over 9 hours training – 2/10 (20%)
# of runs at 120 minutes – 0
# of runs over 120 minutes - 3
# of runs at 150 minutes (max duration) – 1
Week at which peak was hit – Week 10
Scheduled Marathon training pace – 7:33 min/mile
Average Marathon training pace in workouts – 7:27 min/mile
Actual Marathon Race pace – 7:36 min/mile
Final Marathon time divided by 26.2 miles – 7:40 min/mile
2017 Daniels 10k training cycle
Total weeks – 20 weeks
Total non-recovery weeks – 16 weeks
Total non-recovery weeks mileage – 839.6 miles
Average # of miles during non-recovery weeks – 52.5 miles
Total non-recovery weeks duration – 118:26:40
Average duration during non-recovery weeks – 7:24:10
# of weeks above 70% intervals hit (not including easy) – 17/20 (85%)
# of weeks over 8 hours training – 2/17 (12%)
# of weeks over 9 hours training – 0/17 (0%)
# of runs at 120 minutes – 0
# of runs over 120 minutes - 1
# of runs at 150 minutes (max duration) – 0
Week at which peak was hit – Week 11
Scheduled Marathon training pace – 7:25 min/mile, then 7:02 min/mile
Average Marathon training pace in workouts – 7:21, 7:04, 7:15, 6:55 min/mile
So, tons of information to sift through (understatement of the year, I know). When looking at these cycles I’m looking for similarities between WI and Dopey (two good cycles) and avoiding choices made during the last Lakefront cycle. Daniels cycle was both good and bad (made great progress but certainly faded).
Four total cycles of 15, 20, 12, and 20 weeks in total length with non-recovery training at 12, 17, 10, and 16 weeks. The shorter cycles seem to have been more successful for me.
The average mileage was 55.2, 61.7, 56.4, and 52.5 miles. But more importantly, the average duration was 8 hrs, 9 hrs, 7.5 hrs, and 7.5 hrs. The high duration was not a good choice during the Lakefront cycle.
When evaluating the plans on a weekly interval hit rate %, we get 80%, 47%, 92%, and 85%. So, obviously the cycle goes better when I’m hitting over 70% of the assigned interval paces. When things aren’t going well on a weekly basis, it means the end result also might suffer. Even more important when looking at the last Lakefront cycle is in the last 7 weeks of training, I only hit >70% intervals once (14%). A major sign of problems.
Then I looked at the # of weeks with over 8 hrs of running – 75%, 82%, 30%, and 12%. And also, 9 hrs of running – 25%, 53%, 20%, and 0%. So, it would say to me that going over 8 hrs is neither good or bad at a high rate, but going over 9 hrs too much (50%) is a breaking point for me.
Then, I evaluated the # of runs over 120 minutes – 4, 6, 3, and 1. And the # of runs at max – 3, 3, 1, and 0. Not really sure what conclusions to draw from this.
Next, I looked at when my journal notes appeared to suggest I had hit my peak – Week 7 and 12, Week 11, Week 10, and Week 11 of training. I think this might be the most important stat of them all thus far. My body tends to peak after about 10-12 weeks of training regardless of the training plan used.
So, if you remember some time back, I discussed my last three marathons and the pacing of individual miles (link) (link). The final conclusion was that me running blind to pace was causing a lot of variation amongst the mean. The paces were far from even throughout.
So lastly, I looked at a really interesting set of data. I went back to my training plans and looked up the marathon specific runs by scheduled training paces (training pace), the actual workout training paces (Avg T Pace), the average pace in the marathon (Race Pace), the final marathon time divided by 26.2 miles (R Pace over 26.2), and the ratio of average marathon training pace to final race pace divided by 26.2 (Ratio - which would give me the % off from each other (hence, 0.95 is 5% off pace too fast)). So, I found this new data set incredibly interesting. Marathon actual training pace was within 3 seconds of actual race day pace divided by 26.2 miles for both WI and Lakefront. Disney was obviously slightly slower because of the additional racing, but not too far off. I also used the idea from Hansons that a very well-trained runner may only improve by 1-2% during each marathon cycle. Thus, if my marathon training pace was X, then 98% (or 0.98) of X during the race might be biting off more than I can chew.
View attachment 245633
So, for the WI marathon, I dipped below 98% at mile 2 and was under for 16 of the first 19 miles. Lakefront was at mile 8 and 7 of the first 19 miles. Disney was mile 14 and only 2 of the first 19 miles. So if I ignore the wind in WI and injury in Lakefront, this would really drive home the idea that staying closer to 98-100% of marathon training pace yields the better results. Not terribly surprising actually. Better pacing in the marathon likely equals better end results (and this makes sense completely from a physiological standpoint). My biggest hang-up has always been how to choose an appropriate goal pace on race day (or going completely blind). But this data set really shows that my actual marathon training pace seems pretty predictive of future results.
So, what does this all mean? It kind of shows that whether the training goes well (WI) or goes poorly (Lakefront) that the final results are pretty closely tied to the average pace during training. Although there are confounding factors with WI and the wind after mile 19 and the overtraining/injury with Lakefront (yet things come out equal?). It kind of creates a conundrum:
1) Train at current fitness knowing that I’ll attempt to train at a 6:58 min/mile (3:02:30) and expect the final results on marathon day to be a 7:02 min/mile (3:04:16) (adjusting for greater than 26.2 miles). This would get a BQ, but not eligible to sign up for the race and not a sub-3.
2) Train just slightly ahead of current fitness only on marathon training runs, to train at a 6:54 min/mile (to get a 6:58 min/mile – 3:02:30). This would get a BQ but not sub-3.
3) Train even further ahead of current fitness only on marathon training runs, to train at a 6:48 min/mile (to get a 6:52 min/mile – 2:59:59). This would get both a BQ and sub-3.
Of course, it’s not merely the marathon training paces because all of the scheduled paces of every run are built around themselves. So, would training at just faster marathon pace but all the other paces are based on current fitness actually work? It’s changing a variable which means the outcome may or may not be the same anymore. It also kind of begs the question whether a T+D adjustment would be helpful or harmful. Granted only Lakefront was during the summer and really needed any adjustments.
That’s when it dawned on me, while I don’t have great training data (not nearly nicely summarized) from Lakefront 2015, I do have some data. Lakefront 2015 is still my gold standard for marathon experience. It was perfect conditions (T+D of 100 with clouds from beginning to end), I felt well trained (from memory I was pretty solid on pace and the temps were low throughout summer training), and the race went better than expectations or even my wildest dream. So, maybe I should look at the limited data set from that race and see if I can make similar conclusions.
2015 Lakefront training cycle
Total weeks – 18 weeks
Total non-recovery weeks – 17 weeks
Total non-recovery weeks mileage – 786.25 miles
Average # of miles during non-recovery weeks – 46.3 miles
Total non-recovery weeks duration – 124:27:20
Average duration during non-recovery weeks – 7:19:15
# of runs at 150 minutes (max duration) – 3
Scheduled Marathon training pace – 8:35 min/mile
Average Marathon training pace in workouts – 8:34 min/mile
Actual Marathon Race pace – 8:18 min/mile
Final Marathon time divided by 26.2 miles – 8:21 min/mile
So, since the pace of marathon training runs seem to be a very predictive measure of future race performance, that’s where I started looking first. BTW, very interesting to go back and see all the old styles of keeping track of things. To my surprise, the data was quite rock solid. Scheduled pace and training pace were nearly identical. On race day itself, I ended up with a 2.5% better final time than marathon training went. Was there something I could glean from the individual mile marathon data to suggest how I could repeat a 2.5% improvement from scheduled to result?
View attachment 245632
Well, look at that! The average mile pace during the Lakefront 2015 was within 2% of the training pace until mile 11. I vividly remember the goal in that race was maintain pace until mile 10, and then mile 10 becomes mile 1 because my long run was 16 miles. So, it appears that I made the decision to slowly increase the speed from that point in the race until the end. It looks like it payed dividends as I did not slow down below average pace for the remainder of the race and had the best marathon experience yet (I believe I set 5k, 10k, and HM PRs during the second half of that marathon). So maybe this is repeatable? Stay super tight on the pace for the first 12-14 miles and then slowly increase the effort. If this were doable, and I trained at current fitness of a 6:58 min/mile (3:02:30 marathon), then what would my final marathon time be with a 2.5% improvement (assuming that the average marathon training runs throughout ended up being 6:58)? A 2:57:58 adjusted time based on not running exactly 26.2 miles (but similar distance as other 4 marathons). So, this single race performance says to me that it’s feasible to train where I’m at and not be pinned down to matching that exactly on race day. Yet, if I can appropriately pace myself on race day and have great/ideal weather conditions, then nailing the sub-3 is not completely out. Although, this would state that running with the 3:00 pace group would be out as it wouldn’t fit this model of pacing.
So, when I started developing the 2017 Lakefront training plan, I kept the following in mind:
-Keep the total non-recovery training period short.
-Attempt to aim the peak at 10-12 weeks after the training starts.
-Average duration of between 7.5-8.5 hrs, but most certainly less than 9 hours (even with the shorter length of the plan). In addition, on a weekly basis some weeks of >9 hrs is ok, but keep it minimal.
-If the interval hit rate starts to fall below 70%, then reassess training.
-Train at current fitness and REMEMBER that what I end up with in training for marathon days has a direct relationship with final performance. AND that how I pace myself on race day has a direct relationship as to whether I can slightly improve my final time to reach that ultimate goal of a sub-3 marathon.
I then went back and thought about each individual training plan. What worked? What did I struggle with the most?
Something I really enjoyed during the Dopey plan was scrapping the classic MP Strength (MP – 10 sec) and going with HMP Strength (HMP – 10 sec) instead. After thinking about the physiological basis of MP-10 sec, it was hard to justify it. So, I thought HMP-10 sec made more sense and thought it worked well during those interval sessions. I found a Hansons 60/80 marathon training plan, and was pleasantly surprised to see that plan uses HMP as the longer interval pace. So, I decided to use HMP as my long interval pace.
After reading some of Tom Schwartz material, I decided I definitely wanted to incorporate some critical velocity training in my next cycle. This pace is to fall between VO2max pace and Lactate Threshold pace. If appropriately paced, it offers the opportunity to simultaneously work both VO2max and LT at the same time. But at the same time, this is further from MP than is HMP, which means the Lydiard rule of specificity (paces closer to goal pace for hard workouts = better) states I should choose to do more HMP work than CV work in the cycle especially the closer to race day it gets. I also really liked Daniels R pace (roughly mile pace) and thought it would work if I tacked those onto the end of the CV workouts.
Another person I pulled from was Jeff Gaudette. He wrote an article on runnersconnect about how the main benefits (mitochondrial gains and capillary development) of the long run seem to maximize at about 120 minutes. So, he proposed that after 120 minutes there should be more marathon paced work instead of more long run/aerobic work (since the benefits have been fully realized). He also stated that this would allow even more marathon paced work on tired legs. So, I’ve done this before on a more impromptu basis. I’ve done “fast finish” or “hammer” workouts, but generally never pre-planned and just because the workout was going well. But consistently, these workouts usually led to some fatigue related pain in the proceeding days. Although, I think I may have found a reason. I’ve stated in the past that each pace on the spectrum works the muscles differently. So too much of a single type of pace, puts the muscles at a higher risk for injury because they’ve been over fatigued. So, doing a marathon tempo run on Thursday and then a fast finish/hammer on Sunday causing fatigue pain issues makes sense. Too much marathon tempo work in a small window of days. Interestingly, the Hansons 60/80 plan seems to have the answer. A progression workout proceeds the long run/marathon tempo hybrid run. It appears the progression is about 40% easy work (or slower than long run pace), 30% grey area (between long run and marathon), and 30% hard (MP, then HMP, then 10k pace). So instead of a marathon tempo on Thursday (which is 100% hard), you proceed the hybrid long run with a progression run that is only 30% hard. This makes a lot of sense.
One other choice I made is the following, I spent a lot of time this spring working on speed. I’ve got that now and it’s just about where I want to be. Now I really want to get comfortable at marathon pace. In previous plans, I spent time building to that (and maxing at 10-11 miles of marathon pace). This time around, I’m spending less time building, and more time near the max of marathon pace training. I’m doing more marathon paced work in this plan between the normal Thursdays and now the occasional Sundays. I really want to hammer home that feeling of marathon pace so that it becomes second nature.
Now, things I know I need to avoid from the last Lakefront plan.
I can’t make the easy days hard by extending the duration of them to 90 minutes from 60 minutes. That was a failed experiment. Easy needs to be easy. So, all of my easy days are slow by effort, and will cap at ~60 minutes. No pace goal of any type.
I also need to incorporate step back weeks every 4th week. I work in a science lab that uses the idea of bone remodeling every 4 weeks. So, I really should have heeded that fundamental idea in my previous Lakefront cycle. So, this time, I’m going to incorporate a step back every 4th week of 15-30% of the previous week’s max mileage. This, in addition with the easy days, should help me avoid the bone issue (likely shin splints) I had during the last Lakefront marathon and afterwards.
So, in summary,
-HMP – 10 sec work instead of MP – 10 sec.
-CV work instead of 5k work, but more HMP – 10 sec. overall.
-More marathon paced work, but not two of those types in a week. If Sunday has marathon tempo, then Thursday is an easier progression.
-Easy days stay at 60 minutes.
-Include step back weeks for bone remodeling.
So, with keeping all of those things in mind, I came up with the following:
First off is the pace, which I believe is one of (if not the) most important variable to training. After the last two long runs, I think it’s confirmed to me with a standard T+D adjustment that my marathon projected fitness is a 3:02:32. This matches well with my best-case scenario from Bunny Head Time Trial #1 (19:29) where bad pacing and a lack of competitors/spectators probably held me back just a touch from better. So, with those things in mind, I decided to use the 3:02:32 pacing. My goal is a 2:59:59, but as I always like to say better to train at your current fitness than to reach too far and increase the risk of injury/miss the pace spectrum benefits of training. If I do things right in training, I’ve shown that it’s at least possible to recreate the magic of Lakefront 2015.
View attachment 245621
Although, it does seriously beg the question whether I should or shouldn’t include T+D adjustments. I haven’t previously, so will it change the outcome? Will marathon pace with a T+D adjustment still be a predictive measure of future race performance? I’ll have to think long and hard about this one.
The plan is 15 total weeks in duration.
View attachment 245623
Quick notes:
-Green are nutrition runs (any run longer than 90 min).
-Yellow box is total weekly mileage.
-Red % is the longest run of the week as a % of the total mileage.
-Black box is a weekday morning run
-Total is the total time (duration)
-Easy and Hard % is the weekly breakup of paces above and below the easy threshold (long run pace)
-Green values are the average pace of a run (in total)
-Blue % is the amount of CV pacing relative to weekly mileage
The first two weeks will focus on recovery from the Hot2Trot 10k, and the continued use of only two hard workouts per week. This will help build the mileage up to where this cycle will live. During these two weeks, I’ll reintroduce marathon paced work. I’m hopeful that the success I saw with marathon pace work during the Daniels cycle carries over into this cycle. I’ll also be building my long run back up to its historical higher level than was the Daniels 10k training (which capped at 120 min or 25% of weekly mileage). All runs over 90 minutes will have nutrition (colored green). After the Hot2Trot 10k and the training runs in the heat leading up to it, I realized that doing all my training in the afternoon during the week is making it more difficult for me to hit the necessary paces. So, I looked over my work schedule and put in some morning Tempo runs (on Thursdays) throughout the schedule. They are marked with black boxes. I’ll take those days off from work and focus on an early morning Tempo run in the hopefully cooler weather.
After those two weeks of recovery/building, the real plan will start for the remaining 13 weeks. With the hope that I can capture the same peak at around 10-12 weeks that I’ve seen in the past coinciding with the start of my taper.
View attachment 245626
I’m starting the plan off with more CV than HMP, then transitioning over time to the opposite. Those first three weeks will be the encapsulation of the training ideas. CV+Tempo+LR, then CV+Progression+Hybrid Easy/Tempo, then HMP+Tempo+LR (but a drop in overall mileage). The easy/hard will be kept close to 80/20 throughout. The CV paced work is less than 8% of the weekly mileage and the total duration is aimed to be around 90 minutes. The longest run of the week less than 30% of the weekly mileage. Again, a progression run during the week of 7/10 because of the Sunday marathon paced mileage. The week of 7/17 will be the first step back week at 85% of the previous week’s mileage.
View attachment 245627
The next 4-week block comes with some new tweaks. I’m trying a ladder type CV workout and the first hybrid LR/tempo. The week of 7/24 brings a LR with a fast marathon paced finish. The week of 8/7 needed to be creative because I’ve got a wedding reception in Minneapolis to attend that weekend. So, I dropped the CV/HMP, and went with a progression (which is easier) and then a hybrid easy/tempo. But this hybrid easy/tempo run is special because it will be done as a depletion training run. Thus, no breakfast before the run, and no carbs during the run (electrolytes are permissible). I’ve seen conflicting advice on when to do the depletion training run. I’ve seen some suggest it should occur at the beginning of training as to yield the benefits of it for the remainder of the training. Yet, others propose it’s better to do it in the middle after you’ve already built some of your endurance up. However, both sides agree that you shouldn’t do the depletion run any closer than 8 weeks to the marathon as to allow enough recovery time from it. I figured it was also convenient timing to have the step back week after that run. All these weeks stay near 80% easy and 25-30% long run of weekly mileage. The week of 8/14 is another step back week at 84% of the previous week’s mileage.
View attachment 245630
The next 3-week block has more HMP than CV (specificity). These three weeks represent the peak of the training. There’s a hybrid LR/tempo, long run, and hybrid easy/tempo. The max duration peak is hit twice in back to back weeks (once as a hybrid with easy/tempo and the other as a straight up long run). Even at its peak (66.8 miles), we’re still coming under 9 hours of total weekly duration. In fact, there will be 0 weeks of >9 hrs training, whereas 53% of Lakefront 2016 was >9 hours. I’m hoping that maintaining the balance of easy/hard while keeping the long run at 30% of weekly mileage or less, but not doing >9 hours will be the perfect balance I’m looking for.
View attachment 245631
Lastly, comes the taper. Unlike plans in the past with a 21-day taper, I’m only doing a 14-day taper. When I evaluated my peak from Dopey training, it showed that I lost my peak in about 2 weeks instead of 3. I also felt this during the Daniels training. So, I’m hoping reducing the taper by a week will allow me to maintain the peak through the marathon. In addition, while the 10-day rule still applies (workouts within 10 days of race day won’t have fully realized benefits and increase fatigue) I’ve included a relatively low level long run and a lighter CV workout on the Tuesday proceeding race day. Lastly, the non-marathon mileage of the last week is 28.75 miles, which represents about 44% of the peak mileage. Meta-data research shows the desired % to be around 42-48% of peak in taper week. So, this is within those guidelines.
So, using the same criteria of analysis as the previous training plans, how does this one compare?
2017 Lakefront Marathon training cycle
Total weeks – 15 weeks
Total non-recovery weeks – 13 weeks
Total non-recovery weeks mileage – 876.1 miles
Average # of miles during non-recovery weeks – 58.4 miles
Total non-recovery weeks duration – 116:23:10
Average duration during non-recovery weeks –7:45:33
# of weeks above 70% (including easy) – Hopefully I do well!
# of weeks over 8 hours training – 8/13 (62%)
# of weeks over 9 hours training – 0/13 (0%)
# of runs at 125 minutes – 5
# of runs over 125 minutes - 7
# of runs at 150 minutes (max duration) – 4
Week at which peak was hit – October 1st?
Goal
Only one –
1) Run a sub 3-hour marathon and qualify for the Boston Marathon
Alright, if you’ve made it this far, then!
I spent a lot of time working on this and trying to give myself the best opportunity to succeed.
One last thing, meet “Rainbow Sparkles” (G named it). Steph’s grandfather painted me this original blue and yellow unicorn. The painting will hang right next to my bed so that every morning and evening I can stare at it and visualize the feeling of marathon day. I’m ready to capture the elusive and mythical creature, known as the BQ.
View attachment 245635
Billy, your level of analysis, dedication and introspection are astounding. As you said in your first post...if you want something...prove it by doing what is necessary. Here's to hoping you meet all your goals!
And I love the father's day pictures! Your daughter sure loves the camera and it loves her. She is absolutely adorable.