|Abel Kirui (left) and Moses Mosop (right)|
Renato Canova is a widely-renown coach of some of the most elite middle and long-distance athletes in the world. His runners routinely medal at World Championship and Olympic races and place highly at major marathons. I’ve done a good bit of writing on this blog about his training methods, and those posts are some of the most popular of all of my articles. Unlike many other elite coaches, Renato Canova has no reservations about sharing his training philosophy and the workouts of his athletes. 2011 was a banner year for Canova, as several of his runners won medals at the 2011 world championships, including Abel Kirui, a young runner who won his second marathon World Championship. Additionally, Moses Mosop, a long-time Canova runner with sub-27 10k credentials, made his debut marathon in an earthshaking 2:03:06 for second place at the Boston Marathon, then later smashed the 25k and 30k world records at the Prefontaine Classic meet in Eugene, Oregon. To cap off his amazing season, Mosop won the Chicago Marathon with a course record as well. Soon after this incredible spring and summer, Renato Canova posted the training of Moses Mosop before the Boston Marathon and Abel Kirui before the World Championships marathon on LetsRun.com. I’ve had the training schedules sitting around for some time, but I’ve just now gotten around to copying them to a calendar and translating them into relevant paces.
If you would like to read my other work on Canova, a good place to start is my article titled “Something New in Training:”
And last fall I completed a similar analysis of the training of Canova’s track athletes (1500-10,000m) in the last month before the World Championships:
The Training Schedules
I have transcribed the training schedules of Moses Mosop (four months’ worth of training) and Abel Kirui (two months) onto calendar pages on .pdf files for easy viewing and printing. Two versions exist for each athlete’s training: a schedule with the absolute paces and times for workouts, and a schedule with the paces translated into relative percents of PR pace, which is how Renato Canova discusses relevant training paces. In most cases, the schedules have been “Americanized” by translating kilometers into miles and times per km to times per mile. Some workouts, mostly interval repeats of familiar distances to Americans (i.e. 3ks, 5ks) have not been translated into a per-mile pace.
To download the files in their original PDF form (for better text quality), select “File > Download” on the Google Docs menu.
Remember that all training was done in Kenya, which is at relatively high altitude, and most road runs were over fairly significant hills. However, according to Canova, Kenyans are much better at running close to their sea-level performances while at altitude than Western athletes are (probably because the Kenyans have lived at altitude their whole lives), so I don’t think the schedules need any drastic adjustments to compensate for that.
Glossary of Terms
You’ll need to know the following terms to understand the schedules:
- Fartlek—Swedish for “speed play.” A low-key workout that involves surging and recovering during a run. Usually the efforts prescribed for the surges are general terms like “fast” or “moderate” instead of specific paces.
- Diagonals—A low-key workout popularized by Kenyans. It involves running the diagonal between corners on a soccer field quickly, and jogging along the goal line to the other corner. This is repeated either for a number of diagonals or for a specified length of time.
A Note on Renato Canova’s math
Renato Canova computes many of the paces he prescribes based on percentages of an athlete’s personal record. It’s important to discuss how this is calculated, because it is not straightforward. For example, if we say “run at 90% of 5:00 mile pace,” many people would divide 5:00 by 0.90 and get 5:33.3 minutes per mile. Renato, however, does his math differently. To him, deviations from PR pace are measured quite literally in per cent—parts of one hundredth of the original pace. So, when Renato says “90% of 5:00 pace,” he means “10% slower than 5:00 pace.” To compute this, you do the following:
5:00/100*10 + 5:00 = 5:30 pace
Here are two simple formulas that will aid you in calculating “Renato Canova percents”:
To calculate the resultant pace N by going P percent of race race (RP)
N = RP*(2-P/100)
As you can verify yourself, this formula works equally well if P is greater or less than 100 percent. For 110% of 5-minute mile pace, N = 5*(2-110/100) = 4.5 minutes per mile, or 4:30 mile pace. For 90% of 5k pace, N = 5*(2-90/100) = 5.5 minutes per mile or 5:30 mile pace.
To calculate what percentage P (faster or slower) than race pace (RP) some new pace N is, use:
P = 100*(2-N/RP)
Please note that the analysis of these schedules is only my own interpretation, and that I don’t claim to speak for Renato Canova or anyone else. As I have written before, any scholarly analysis will necessarily involve placing things into a context where they can be understood. Finally, I take responsibility for all typos—I’m sure I’ve made at least a few, so don’t read too far into any one particular day’s workout! Also, for computing Mosop’s relative paces, I used a rough prediction of 2:05:51 based on his 10k PR. Abel Kirui’s percentages are based on his PR of 2:05:04
Observations and Analysis
After a first glance at the schedules, a few things jump out right away. The first is consistency. Barring a disruption in training, Kirui and Mosop train twice a day almost every day. While the workouts will be the focus of most of our analysis, do not forget that these 40-80min runs at an easy to moderate effort, day in and day out, make up the core of their training. These are not so much aerobic training sessions (as they would be in a diet of high-mileage training for a less experienced Western runner), as Kirui and Mosop both have many years of high volume training under their belts, and are thus already in exceptional shape. Rather, think of these as more of supplemental or maintenance work—just like a calculus student has already mastered algebra, yet still uses it on a daily basis to accomplish more complicated tasks, such is the role of easy to moderate mileage in Canova’s marathon training. To Mosop or Kirui, 80min moderate in the morning and 50min easy in the evening is likely no more stressful to them than your regular easy eight-mile loop is for you.
Second, there is very little long easy or long moderate running. Both schedules include one run around the duration of a marathon (2:10 for Kirui and 2:28 for Mosop), but aside from that outing, only a handful of easy/moderate runs are longer than 80 or 90 minutes. To Canova, these long easy runs are not very useful to an experienced athlete, as he writes in a thread on LetsRun.com:
One question : if you want to run a Marathon at 3:19 pace (about 2:20 final time), do you think that running 30k at 4:00 pace can have some connection? If you want running 10000m in 30:00, do you think that the main workouts are 400m on track, and long run must be only easy regeneration?
Instead, the focus of training is in long fast runs, long intervals workouts, and short fartlek-style workouts.
Starting early on in training, and continuing up until three weeks out from the goal marathon, Mosop and Kirui run long fast runs on a regular basis. These are typically between 20 and 40km (12mi-25mi) and are done at least once every two weeks. The pace for the continuous long-fast runs is between 87 and 95% of PR marathon pace. On weeks that do not contain a continuous long fast run, Mosop or Kirui often instead do a broken up or progressive long fast run, for example, 10km at 87% + 12km at 97%. These broken up or progressive sessions allow the runners to do some blocks of continuous running at 95% of race pace or faster after first covering several miles.
Another staple training feature are long interval workouts. The repeats last from 1km to 5k, and sometimes vary in distance and pace. The longer invervals of 3-5km are typically at 95-101% of marathon pace, while the shorter repeats of 1km-1mi are usually at 103% or faster. Very often, the recovery is not a slow jog, but rather is 1km at a prescribed pace. Early in training, this kilometer of recovery is at an easy pace (50-70% of race pace), but as the marathon draws near, this kilometer of recovery gets faster, culminating in the “alternating” workout two weeks out, where the athlete runs 22-24km alternating kilometers at 103% and 89% of race pace. The pace of the long repeats themselves seems to increase over time, from as slow as 86% for Mosop in January up to 96-100% in March.
Short fartlek-style workouts and diagonals are also a common occurrence in Mosop and Kirui’s schedule, and they likely have several roles: breaking up the monotony of the longer continuous running, preparing for the abrupt surges of speed that have become the trademark of top-flight marathons today, and keeping the body and legs fresh between long workouts. Typical sessions involve fairly short bursts of one or two minutes of fast running (~100% of race pace), followed by 30sec to 1min of easy or moderate running. Abel Kirui does a great deal of diagonals, described above, which are quite similar to the short-burst fartlek workouts that Mosop does at least every two or three weeks. These workouts may also serve as “light” versions of the alternator workout which tops off the training schedule. Longer fartleks aren’t unheard of, but are less common than the long interval workouts described above.
As with all Canova schedules, recovery before and after workouts is proportional to the intensity and stress of the workout itself. Relatively easy workouts, like the short fartlek sessions described above, can sometimes be done the day before or after a larger workout. But before and after very long or very intense workouts like the 40km runs and the “special blocks,” there are several days of easy running, often at a reduced volume.
Speaking of intensity, Canova employs Special Blocks every 2-4 weeks (as long as training is progressing nicely) which consist of TWO workouts in a single day, one in the morning and one in the afternoon. These workouts are usually either broken up long- fast runs or long intervals of 1-4km.
Finally, it is well worth noting that there are two, not three, main workouts most weeks. Many western athletes in high school or college who jump at the opportunity to get in a third workout on a non-race week could learn something from Canova’s philosophy of sufficient recovery before and after major workouts.
Moses Mosop races once during his four-month preparation for Boston at the Paris Half-Marathon (six weeks before the marathon). His time is good, but nowhere near the caliber of his eventual performance at Boston. Many people speculate that it’s not possible to be in top-notch half-marathon shape during preparations for a marathon, since the speed of the half-marathon requires a sacrifice in stamina-type training, which is necessary to run a great marathon. This seems to confirm that speculation.
Mosop’s preparation differs from Kirui’s in that Mosop tends to do a good bit more high-speed track training. Mosop comes from a track background and boasts sub-13 5k and sub-27 10k credentials, so perhaps the track sessions of repeats from 200-1200m are playing to his strengths as a runner. Kirui has relatively paltry track PRs compared to Mosop, so he may stay away from high speed track training. But it is also possible that this type of training was planned for Kirui as well, but was shelved because of his knee problems in late July.
As the marathon draws near, the training schedules focus on preparing the athletes to run a championship race. The last major sessions in the final three weeks of training are an alternating continuous run, a short fartlek workout, and one final high-intensity session near marathon pace or faster. Three weeks out from the race, Mosop completes a 15.5mi run at 97.5% of race pace, while Kirui completes a special block session that includes 9mi at 98.5% of race pace. Two weeks out, both Mosop and Kirui complete a fast fartlek session (20x1min fast/30sec moderate and 15x1min fast/1min easy, respectively) and 22 or 24 km alternating between 103% and 89% of race pace. In the final week, Kirui and Mosop are busy with traveling, and take several rest days. It’s unclear what their training is once they’ve arrived at the race location, but Renato does not seem overly concerned with it.
Interestingly, while workout duration does seem to ramp down somewhat in the last three weeks, the volume on the easy days between workouts does not change much. This is in keeping with Renato Canova’s philosophy of easy mileage being a “maintenance” type session, not a stress on the body in and of itself. However, overall volume must decrease significantly in the week before the marathon, given the days off for travel.
Final Thoughts: what’s missing and differences from Western marathon training
One of the most enlightening ways to examine a training schedule is to look at what’s not present. In the case of Renato Canova’s marathon preparations for Mosop and Kirui, what stood out to me was the lack of much sustained long fast runs at around 100% of marathon pace. There are a lot of long fast runs at 88-95% of marathon pace, and a good bit of long interval workouts at 100-110% of race pace, but “staple sessions” to a Western athlete like 12-16mi at marathon pace are quite rare. It’s possible that this might just be a function of altitude and muddy/hilly roads slowing the pace of training sessions. But perhaps Canova believes that the focus should be on extending endurance in long continuous runs at 5-15% slower than race pace—this would certainly be in keeping with his philosophy for the shorter track events. In many ways, the workouts here are quite similar to what a 5k runner would be doing in terms of relative paces. So, a 5k runner’s training would consist of long fast runs at 85-90% of race pace, plus long repeats at 95-100% of race pace, and medium or short intervals at 100-110% of race pace. This is quite similar to the marathon schedules seen here!
This finding illuminates why Canova’s marathon training is so different than the traditional Western approach. When an American coach trains a 5k runner with long continuous runs at 85 and 90% of race pace, it is because he or she believes that these paces are physiologically relevant (because they correspond to the aerobic and anaerobic thresholds). And Canova’s training for a 5k runner looks roughly similar. But, when this American coach is training a marathoner, they train the same PACES—aerobic and anaerobic threshold. But now, aerobic and anaerobic threshold correspond to ~100 and ~106% of race pace! But Renato Canova would train the same PERCENTAGE—85 and 90% of race pace. Seeing the stark differences in how an American might train for the marathon and how Renato Canova has his athletes train should change how we think about preparations for the marathon. Despite the fact that Canova’s athletes are at an extremely high level of fitness and experience, I still think many of these principles can be applied to less-experienced marathon training.
I’m hesitant to make too many recommendations on how to adopt these principles for your own marathon training, as I don’t have a whole lot of marathon experience myself (barring Grandma’s Marathon as a sixteen-year-old!). But I would start by looking to the principles, not the numbers. So instead of focusing on doing 20-40km long fast runs, or 4x5km interval workouts, focus on introducing long-fast runs at 85-95% of marathon pace and long interval workouts at 95-100% of marathon pace into your training. The volume for a 2:20 or 32:40 marathoner will likely be much less than one of Canova’s 2:05 runners. Remember, Kirui and Mosop have a huge amount of mileage and racing experience under their belts, so they require a much greater stimulus from their workouts to improve. And also keep in mind that Mosop and Kirui’s marathon pace is under 3min/km, making a 20km run at marathon pace over in less than an hour.
Finally, despite the eye-boggling workouts that crop up in these schedules, there are some very down-to-earth lessons we can take from them. First, even World Champions get injured! Abel Kirui’s training was disrupted for several weeks because of a knee problem, and he almost didn’t even get to race! And Moses Mosop faced several disruptions in his own training because of travel and a death in the family. Mosop and Kirui did not let these interruptions faze them—they “rolled with the punches” of travel and injury, not trying to force in more workouts despite a busy travel schedule. Second, while Mosop and Kirui indeed ran fairly high mileage, it wasn’t absurdly high—most weeks were in the 120-mile range. But the modulation, or day-to-day variation in mileage, was very high. Mosop would follow up a thirty-mile day with only eight or nine miles of easy running, so he was clearly not afraid to back off and take it easy after a hard effort. Finally, both Mosop and Kirui spend time preparing for the type of race they would encounter. As the Boston and the World Championships Marathons are high-caliber competitions, Canova knew that Mosop and Kirui would have to be ready for abrupt bursts of speed from their competitors, and trained them to be prepared with short fartlek workouts and the “alternating” fast continuous run (much like Canova’s shorter distance charges prepared with interval workouts of alternating speeds). Mosop also did a lot of his long fast runs over hilly terrain, preparing for the ups and downs of Boston.
Here is a brief summary of the important points to take away. In general, Moses Mosop and Abel Kirui’s marathon training programs as designed by Renato Canova consist of:
- Consistency in training: easy to moderate running, twice a day for most days between workouts
- A focus on fast continuous running as opposed to long easy or moderate runs
- Long fast runs of 12.5-25mi at 85-95% of marathon pace at least once every two weeks, sometimes supplemented by broken up long fast runs or progression runs
- Long interval workouts every 2-3 weeks with repeats lasting from 1-5km. Recovery is often, but not always, 1km at an easy to moderate pace. The intervals from 1-2km tend to be 103% of marathon pace or faster, while the longer repeats of 3-5km tend to be at 95-101% of marathon pace
- Short fartleks of 1-2min at marathon pace with short recovery (or diagonals) every few weeks to break up the schedule, to prepare the body for surges in championships races, and to serve as a precursor for the fast alternating run.
- “Special block” workouts with TWO workouts in one day every 2-4 weeks if training is going well, usually consisting of a broken up or progressive long fast run and/or long intervals
- Recovery days before and after workouts proportional to the duration and intensity of the workout
- Two major workouts most weeks
- 9-15mi of continuous running at 97-98% of marathon pace 3 weeks out from the marathon
- 22-24km (13.5-15mi) alternating 1km fast (103%)/1km moderate-fast (89%) two weeks out
- 15-20x1min fast fartlek 2 weeks out
- A focus on PERCENTAGES of race pace over physiologically relevant PACES
I hope you’ve enjoyed this analysis and have discovered something useful. I’d love to hear your thoughts below.
Original threads with Canova’s posts: