Balanced Training Method Explained

← All philosophies

Your weekly hours9h

Fits your week. Coherent from 4 to 25 hours a week.

Where the time goes

Compare with

By time: 80% below LT1, 14% between the thresholds, 4% at LT2, 2% above. Seiler's 80/20 counts sessions, not minutes.

A representative week

Show the week for

Across all three sports.

Mon

Rest

Tue

Bike threshold — 3 × 15 min15 min easy, 3 × 15 min at 225–248 W (90–99% FTP) with 5 min easy between, 10 min easy — 80 minutes. Target set on power, heart rate read for drift, 225 W on the first rep building to 248 W on the third. Every third week this becomes the above-threshold slot: 15 min easy, 5 × 3 min at 270–288 W with 3 min easy, 10 min easy — 52 minutes.

Run 25 min off the bikeOptional, entirely easy at 5:20–5:40/km. Not a brick unless it is written as one.

Wed

Swim 2,500 m aerobic400 easy, 8 × 200 m at 3:32–3:36 (CSS + 6–8 s/100) on 20 s, 300 m drills, 200 easy — 2,500 m, about 50 minutes. Swim goes first when it shares a day.

Run 45 min easy5:20–5:40/km, conversational throughout. Finish with 4 × 20 s strides on 60 s recovery. Deliberately shorter than Sunday: the long run is the longest run of the week, and an easy midweek run that creeps past it quietly becomes a second long run.

Thu

Run threshold — 3 × 1 km15 min easy, 3 × 1 km at 4:20/km (≈163–170 bpm) with 60 s jog, 12 min easy — 43 minutes. 3 km sits under the 3.4 km per-session threshold cap on this athlete's ~34 km run week. Do not add a fourth rep.

Spin 30 minOptional easy recovery ride under 165 W.

Fri

Swim CSS — 6 × 200 m400 easy, 6 × 200 m at 3:22 (CSS + 1 s/100) on 20 s rest, 200 m drills, 200 easy — 2,000 m, 1,200 m of it near CSS. Across the week the swim reads roughly 75/25 by distance, and that is correct — do not force it lower.

Sat

Long ride 3 h160–185 W (64–74% FTP), below the first threshold start to finish, heart rate flat. In the final 8–10 weeks add 2 × 20–30 min at goal 70.3 power (≈200–210 W) in the back half — and when you do, drop Friday's CSS set, because that block is the week's third quality piece, not a free extra.

Sun

Long run 55 min5:20–5:40/km, about 10 km — 30% of this athlete's ~34 km run week, which is the cap. Easy start to finish; a fast finish is a quality session and has to be counted as one. In the final 8–10 weeks, when Saturday carries the race-power blocks and a run off the bike, this comes down to about 40 min or moves to Monday. At this run volume the long run sits at the top of the 25–30% band, so the way to buy a longer one is a bigger run week, never a bigger single run.

Swim 1,500 m easyContinuous and technique, no intervals. The third swim is what keeps the discipline split near 50/30/20.

What Tuesday looks like

Bike, 80 minutes. 15 min easy building to 170 W, then 3 × 15 min at 225–248 W (90–99% FTP) with 5 min easy spinning between, then 10 min easy. Set the target on power and read heart rate for drift — an interval steered by heart rate alone starts too hard and finishes too easy, so the watts hold the intensity and the beats tell you what it is costing. Build within the session rather than emptying yourself: start the first rep near 225 W and finish the third near 248 W, the 10–25 W first-to-last build twelve Norwegian coaches describe, and come away able to have done one more. The failure test runs the other way — if the third rep falls below the first, the session was written too hard, and it should be logged as an above-threshold session and recovered from as one. Every third week, swap this for the above-threshold slot: 15 min easy, 5 × 3 min at 270–288 W (108–115% FTP) with 3 min easy between, 10 min easy — 52 minutes. Optional 25 min easy jog off the bike at 5:20–5:40/km; that is a run, not a brick, unless it is written as one.

Signature sessions

Targets below are for the reference athlete: age-group 70.3 triathlete, ~10 h/week, FTP 250 W, run threshold 4:20/km, run LTHR 172, CSS 1:40/100 m. The reference week holds about 34 km of running, and that number sets every run cap below.

BIKE — THRESHOLD (the session that makes the shape pyramidal)

3 × 15 min at 225–248 W (90–99% FTP), 5 min easy between — 45 min accumulated.
4 × 12–15 min at 220–235 W (88–94% FTP), 2 min easy — 48–60 min accumulated, the Norwegian cycling middle-intensity pattern at 83–87% HRmax.
6 × 8 min at 235–248 W, 1–2 min easy — 48 min accumulated.
Over-unders when durability is the target: 3 × 12 min alternating 2 min at 238 W (95% FTP) / 1 min at 262 W (105% FTP) — 36 min, 24 of them under and 12 over.
Recovery is short relative to the rep and never full — full recovery converts a threshold session into a series of fast reps with a different physiological signature. Set the target on power and read heart rate for drift.

BIKE — ABOVE THRESHOLD (one slot, rotated in every third week)

5 × 3 min at 270–288 W (108–115% FTP), 3 min easy — 15 min at VO2max inside a 52 min session.
10 × 1–2 min at 290–310 W (116–124% FTP), 1 min easy.
Never write long reps at short-rep intensity: 3 × 20 min at 105% FTP and 6 min at 115% FTP are not executable. Sanity-check every target against its rep duration before it goes on the calendar.

RUN — THRESHOLD (governed by the per-session caps)

At about 34 km of running a week the threshold cap is 10% of weekly run volume in any one session — 3.4 km.

3 × 1 km at 4:20/km with 60 s jog (3.0 km at threshold). Running threshold recoveries stay at 30–90 s; metabolic continuity is the point.
Or 14 min continuous at 4:20/km (3.2 km). The classic 20 min continuous tempo is 4.6 km and needs a 46 km run week to sit inside the cap — do not write it at this volume.
Below about 60 km/week, scale the canonical sessions down rather than breaching the cap. Do not add a rep to make the session feel like a workout.

RUN — ABOVE THRESHOLD

Interval work is capped at 8% of weekly run volume and 10 km (2.7 km here); repetition work at 5% and 8 km (1.7 km here).

6 × 400 m at 3:55–4:00/km with 90 s jog (2.4 km).
8 × 200 m fast with full recovery (1.6 km) when the goal race is short.
4–6 × 20 s strides with 60 s recovery appended to an easy run — the cheapest above-threshold work in the plan.

SWIM — CSS

6 × 200 m at 3:22 (CSS + 1 s/100) on 20 s rest — 1,200 m near CSS.
10 × 100 m at 1:40 (CSS) on 15 s rest — 1,000 m.
400/300/200/100 descending at CSS, 20 s rest — 1,000 m.
Long-course version: 5 × 400 m at CSS + 2 s/100 (6:48 each), 30 s rest — 2,000 m.
Re-test CSS every 4–6 weeks and anchor the swim in seconds per 100 m, never to heart rate.

THE EASY WORK — most of the plan

Long ride 2.5–4 h at 160–185 W (64–74% FTP), heart rate flat from start to finish. A 5 bpm drift held for three hours is three hours of the wrong intensity and will break the week's ledger on its own.
Long run 55 min here, about 10 km. The cap is 25–30% of weekly run volume (30% below about 64 km/week) or 150 min, whichever is less, so at 34 km/week 10 km is the ceiling. A 90-minute long run is bought with a bigger run week, not taken out of the cap.
Easy runs 30–60 min at 5:20–5:40/km. Easy swims continuous, 900–3,000 m.

RACE-SPECIFIC (final 8–10 weeks only)

Long ride with 2 × 20–30 min at goal 70.3 power in the back half — about 200–210 W (80–84% FTP) for this athlete over a 2:30–3:00 half-distance bike — followed immediately by 20–30 min at goal 70.3 run pace.
Take goal race power and pace from the athlete's own prior race files. With no prior race, widen the band, say so out loud, and never convert an open-running time prediction into a triathlon run target.
When this session goes in, take a quality session out — it is the week's third, not its fourth.

Who it suits

Most athletes, which is exactly why it is the default rather than a fallback. Specifically:

Recreational and age-group athletes. The 2025 network meta-analysis with Seiler as senior author — 13 studies, 348 participants — found no difference between polarised and pyramidal for VO2max or time trial overall, but a significant subgroup interaction (SMD −0.63, p < 0.05) indicating recreational athletes may improve more with pyramidal while competitive athletes may improve more with polarised. That is the single strongest evidence-based reason for a recreational-athlete product to default here.
Anyone racing at or below the second threshold: half-Ironman, Ironman, marathon, ultra. If the race happens between the thresholds, the plan has to contain work between the thresholds.
Time-constrained athletes at 4–8 h/week who cannot accumulate the easy volume polarised needs for its 80% to mean anything.
Athletes in the preparatory and build phases of any periodised season, whatever the goal event.
Athletes with a high injury or illness cost from hard work — masters athletes, athletes returning from injury, anyone whose above-threshold sessions reliably cost them the following two days.
Triathletes generally, because the swim is structurally threshold-heavy and this is the only option in the set that does not require fighting it.
Anyone who is not sure. It has no volume floor, it is coherent with all four periodisation overlays, and it degrades gracefully at 4 h and at 20 h.

Not for you if

An athlete already chronically fatigued from too much moderate work. The fix is less time between the thresholds, not better-structured time between the thresholds.
Staying here for a whole season if you are competitive and peaking. In a 16-week, four-arm trial of 60 well-trained male runners, sixteen weeks of pyramidal produced the smallest gains of any arm; eight weeks pyramidal followed by eight weeks polarised produced the largest (~3.0% VO2peak, ~1.5% over 5 km). Balanced is a base and build distribution that should hand off to a sharper competitive phase.
Short, high-intensity events — 1500 m, sprint triathlon, criterium — where measured elite practice is more polarised. Balanced will still work; it is not the first choice.
An athlete with no usable threshold anchor. Threshold work run 10–15 s/km too fast becomes above-threshold work: the recovery cost roughly doubles and the intended stimulus is not delivered. Without a test, prescribe by controlled effort and duration — comfortably hard, finish with one or two reps in reserve — not by a computed pace.
An athlete with fewer than about three months of consistent training. Threshold work raises the ceiling; it does not build the floor. They will improve for eight weeks and then stall.
Ironman and ultra athletes in the final specific phase, if the threshold work is displacing durability and race-intensity rehearsal. Move the quality to race pace rather than defending the percentages.

In a triathlon week

DISCIPLINE SPLIT. Start near 50% bike / 30% run / 20% swim of weekly time — Friel's published starting point, and 80/20's own published full-distance plan independently lands at roughly 52/27/19. Shift time toward the limiter, hardest in early base.

SWIM. Counted at 75/25 by DISTANCE, and allowed to read threshold-heavy — roughly 55–75% below LT1, 20–35% at or near CSS, 5–10% above — without being "corrected". The reason for counting metres is practical, not physiological: the pool is prescribed in fixed distances, and 75/25 by metres has been found to land at about 80/20 by time. This is the only option in the set that does not require fighting the pool's natural session structure. Anchor to CSS in seconds per 100 m, re-tested every 4–6 weeks; never to heart rate. When the swim shares a day with another sport, it goes first, so form is not compromised by fatigue.

BIKE. Carries most of the easy volume, and that is its job in a triathlon plan: aerobic development on the bike is what protects the run. Target roughly 80–88% below LT1, 8–15% between LT1 and LT2 including the work at LT2 itself, and 3–6% above LT2. Carry both watts and heart rate: set the interval target on power and read heart rate for drift, and if the athlete can only have one number on the bike, keep heart rate. The middle-intensity session is 4 × 12–15 min off 2 min, or 6 × 8 min off 1–2 min, accumulating 45–60 min. Watch the long ride for drift — that is where the ledger breaks.

RUN. The discipline where the caps bite hardest, because triathlon run volume is low. Per session: threshold work ≤ 10% of weekly run volume; interval work ≤ 8% and ≤ 10 km; repetition work ≤ 5% and ≤ 8 km; long run ≤ 25–30% and ≤ 150 min, whichever is less. At 34 km/week that means a threshold session of about 3.4 km, not five miles, and a long run of about 10 km. Scale the canonical sessions down rather than breaching the caps, and raise weekly run volume if the goal race needs a longer long run.

THREE INDEPENDENT ANCHORS. Bike FTP and bike LTHR, run threshold pace and run LTHR, swim CSS. Never derive one from another. Bike LTHR is commonly a few beats below run LTHR, but the gap is individually variable and is sometimes zero — measure it. Friel, who originally published separate percentage tables for run and bike, has since dropped them: the separation now comes from sport-specific testing, not from a fixed offset.

SAME-DAY ORDERING. Bike and run on one day go AM/PM, or as far apart as possible. A same-day bike and run is not a brick unless it is written as one.

RACE-PACE ANCHORING. Never derive a run target from a triathlon run split, and never derive long-course race pace from an open-race time prediction. Use the athlete's own prior race splits where they exist; otherwise state the uncertainty and widen the band. For this reference athlete goal 70.3 bike power sits near 200–210 W (80–84% FTP) over a 2:30–3:00 ride — a planning heuristic with error bars, not a finding, and always second to the athlete's own file.

WHERE THE QUALITY GOES BY DISTANCE. Sprint and Olympic concentrate the quality at and above threshold and deliberately skip the very top end — the co-author of the triathlon version is explicit that the highest zone is not necessary for any level of racing. Half and full distance shift the quality toward threshold and race intensity, because that is what the race is.

What the evidence says

WELL SUPPORTED

Pyramidal is what trained endurance athletes are actually measured doing. Across 119,452 runners and 151,813 marathons over 16-week blocks, pyramidal was the single most common distribution and was adopted by more than 80% of the runners with the fastest marathon times.
The pyramid is built by adding easy volume, not by adding tempo. In that same dataset, time in the moderate and hard zones stayed relatively stable across performance levels while the easy proportion rose in progressively faster groups, and the proportions of moderate and hard work correlated with SLOWER marathon times. The fastest group also accumulated more than three times the volume of slower runners. This is a description of elite and sub-elite training, not a licence to run a shallow pyramid at low volume.
Elite distance runners keep at least 80% of running volume at low intensity throughout the training year, at 160–220 km/week for marathoners and 130–190 km/week on the track, across 11–14 sessions per week — explicitly pyramidal in shape.
For recreational athletes specifically, there is no penalty relative to polarised and possibly an advantage. The 2025 individual-participant network meta-analysis (13 studies, 348 athletes, 150 recreational and 198 competitive, Seiler as senior author) found no difference between polarised and pyramidal in VO2max (SMD −0.06, p = 0.68) or time trial (SMD −0.05, p = 0.34), with a subgroup interaction (SMD −0.63, p < 0.05) pointing to greater VO2max gains from polarised in competitive athletes and from pyramidal in recreational ones.
The controlled-execution rule. Twelve Norwegian coaches responsible for more than 370 international medals describe interval sessions as high in accumulated volume, progressive in intensity through the session, and controlled rather than exhaustive — building about 0.5 km/h in running or 10–25 W in cycling across the session, with passive recoveries.

CONTESTED

Whether to stay pyramidal all season. A four-arm trial of 60 well-trained male runners over 16 weeks found eight weeks pyramidal followed by eight weeks polarised best (~3.0% VO2peak, ~1.5% over 5 km), with the pure pyramidal arm gaining least. The pyramidal arm's week was 300 / 67 / 24 min of a 391 min total — 77 / 17 / 6. Treat Balanced as base and build, handing off to a sharper competitive phase.
Whether polarised is better on the outcomes that matter. A 2024 meta-analysis found polarised superior for VO2peak (SMD 0.24, 95% CI 0.01–0.48, p = 0.040), driven by interventions under 12 weeks and by highly trained athletes, but null for time trial, time to exhaustion and velocity or power at the second threshold. Note the comparator was polarised versus all other distributions pooled, not versus pyramidal specifically.
Whether elite triathletes are pyramidal at season level. A world-class male triathlete across a 43-week Tokyo build trained 14.74 ± 3.01 h/week at 81.9 / 7.2 / 10.9, described as pyramidal in the initial stages of periodisation and polarised at the end — and the season aggregate is polarised in shape, because the hard share exceeds the moderate share. The honest claim is base and build, not the whole season. Other elite short-course triathlete data from the same research group is polarised at season level too.
How much work between the thresholds is right. Thirteen recreational triathletes preparing for a half-Ironman at about 12 h/week were compared as polarised (84.5 / 4.2 / 11.3) against pyramidal (77.9 / 18.8 / 3.3); the authors concluded coaches should not discard training time between the thresholds for a half-Ironman. But nine Ironman athletes tracked over 18 weeks — training 68 ± 14% easy — showed more easy training associated with faster racing, and the authors warned that a high amount of cycling between the thresholds may contribute to poorer overall performance. Both fit if the moderate work is placed rather than maximised: small in base, race-specific in the last 8–12 weeks, and preferentially on the run and in long-ride race-pace blocks rather than as habitual moderate riding.
The upper bound of the easy share. Definitions of pyramidal in the literature range from about 70–80% easy to about 80% easy with moderate work not exceeding 10%. Present the band, not a single number.

LORE — TREAT AS COACHING JUDGEMENT, NOT FINDING

"2 × 20 at FTP" as the canonical threshold session. Convention. No paper establishes it as optimal.
Every volume floor in this family. The commonly quoted 7 h/week threshold for the 80/20 rule is one co-author's judgement, and his co-author publicly disagrees, saying any minimum is "quite low". The evidence usually cited for 80/20 at low volume — 30 recreational runners over 10 weeks — actually found both groups improved their 10 km time with no significant difference between them, and only a post-hoc best-adherence subset favoured the polarised arm.
"Bike LTHR is 5–10 bpm below run LTHR." No primary source, individually variable, sometimes zero. Measure it.
Race-intensity effort bands for 70.3 and Ironman, including the 200–210 W used above. Widely circulated, poorly sourced, and they differ by 5–10 percentage points between sources. State them as planning heuristics with error bars, and prefer the athlete's own prior race files.
Whether heart rate or power should win on the bike. Form's rule — target on power, drift on heart rate, heart rate kept if only one is available — follows an experienced long-course coaching position, not a trial.
"The distribution label is the training." The same 50 weeks of world-class training classifies as pyramidal by race-pace zones and polarised in some phases by physiological zones. Any distribution readout is method-dependent — report which method produced it, and do not let an athlete re-plan around a label.

How it goes wrong

1. RUNNING THE EASY 80% AT MODERATE. If the easy days are not easy, this is threshold training, and threshold training does not work in trained athletes. The distribution is a diagnostic; the easy days are the programme.

2. THE ZONE 2 COLLISION. Here "between the thresholds" means between LT1 and LT2 — tempo and sub-threshold territory. In five-zone and popular usage "Zone 2" means the opposite thing: the top of the easy zone, below LT1. Hearing "Zone 2 base training" and doing tempo is the more common and more damaging direction of the error. Always name the model or use the physiological anchor.

3. COUNTING QUALITY PER SPORT. Two hard bikes plus two hard runs plus three hard swims satisfies the percentages and is seven hard days on one body. The cap is 2–3 quality sessions per week across all three sports combined.

4. RUNNING THRESHOLD SESSIONS TOO HARD. Probably the single most common execution error in this family. Elite practice is controlled and non-exhaustive with a progressive build; a threshold session finished at 95% effort was an above-threshold session and must be counted and recovered from as one.

5. GIVING RECOVERIES TOO LONG. Elite threshold reps on the run use 30–90 s recoveries on purpose, and the bike sits at 1–2 min for 8-minute reps and 2 min for the 12–15 min pattern; only 15–20 min reps earn 5 min. Recovering fully makes it a different session.

6. LETTING THE THRESHOLD SHARE GROW UNTIL IT IS THE PROGRAMME. Past roughly 25% of weekly time you have drifted into threshold training. Check the shape, not just the presence of tempo work.

7. COUNTING WARM-UP AND COOL-DOWN AS HARD. A 60 min session containing 6 × 3 min of interval work holds 18 min hard and 42 min easy. Getting this wrong makes a compliant plan look non-compliant, and athletes then cut real quality to fix a phantom problem.

8. FORCING THE SWIM TO 80/20 BY TIME. The rule is 75/25 by distance, and the reason is practical rather than physiological: the pool is prescribed in metres, and 75/25 by metres has been found to come out at about 80/20 by time. Applying a time-based ratio in the pool strips out the threshold sets swimming actually needs.

9. READING "80/20" AS POLARISED AND DELETING ALL THRESHOLD WORK. It is not polarised — the threshold work sits inside the 20% deliberately, and for long course it is weighted there on purpose. Athletes who make this error end up with a 70.3 build that never rehearses 70.3 intensity.

10. MIXING THE TWO 80/20s. Seiler's is by session, Fitzgerald's is by time; they differ by roughly a factor of three in hard-work dose. An athlete who reads one and executes the other is doing far more or far less intensity than they believe.

11. TREATING LOW VOLUME AS LICENCE TO TRAIN HARD. Below about 7 h/week the ratio does have diminishing returns, but the rule its own co-author states is an absolute cap: "Any individual can handle 90 minutes a week of moderate to high intensity." The 50% figure attached to it belongs to one worked example of an athlete training 3 h/week. At 6 h/week, 90 minutes is 25%, not 50%.

12. PADDING EASY VOLUME TO HIT THE RATIO. Adding junk easy minutes to make the percentage work is a numerator-and-denominator game, not training.

13. AVOIDING RACE-PACE WORK BECAUSE IT SITS IN AN "AVOID" ZONE. The grey-zone warning is about drift and filler, not about deliberate race-specific work. Long-course race intensity sits right there, and it has to be rehearsed.

14. LETTING THE LONG RUN OUTGROW THE RUN WEEK. The cap is 25–30% of weekly run volume or 150 min, whichever is less — at 34 km/week that is about 10 km, which surprises triathletes who expect a two-hour Sunday. The long run grows when the run week grows, not before it.

Sources: pmc.ncbi.nlm.nih.gov doi.org pubmed.ncbi.nlm.nih.gov pmc.ncbi.nlm.nih.gov pmc.ncbi.nlm.nih.gov pmc.ncbi.nlm.nih.gov doi.org jssm.org pubmed.ncbi.nlm.nih.gov pubmed.ncbi.nlm.nih.gov frontiersin.org pmc.ncbi.nlm.nih.gov pubmed.ncbi.nlm.nih.gov sportsci.org scientifictriathlon.com scientifictriathlon.com scientifictriathlon.com 8020endurance.com 8020endurance.com 8020endurance.com joefrieltraining.com joefrieltraining.com coachray.nz vdoto2.com trainingpeaks.com