Methodology · The glass box

Physio guard-rails: the limits TriPaced refuses to cross.

Athletes break themselves following blind algorithms. TriPaced has hardcoded physiological limits the planner cannot violate, even when you ask it to. They come from the sports-science literature, not from product opinions.

Why hardcoded limits

"Adaptive AI" sells well. Adaptive without limits is an injury machine. A plan that will do anything you ask is not a coach — it is a yes-man with a calendar.

Good human coaches enforce limits by intuition, but only when they are watching. TriPaced enforces them by code, so they apply 100% of the time. A guard-rail is a hard refusal: the planner cannot generate a week that breaks it. Ask for the illegal week and the engine returns the closest legal one, with the rule that fired written next to it.

The twelve guard-rails

#Guard-railLimitWhy
1Weekly CTL ramp≤ 8 TSS pts/wk sustainedAbove this, injury risk climbs sharply (Soligard 2016)
2TSB floor≥ −25Below it, recovery debt and infection risk spike
3Race-week TSS reduction≥ 25% vs build avgTaper science (Mujika 2007)
4Race-week run intensityno Z4+ block past T−7Avoid trashing the legs you need for the marathon
5Long ride / long run spacing≥ 36h between peaksJoint and connective-tissue recovery
6Daily TSS ceilingdiscipline-specificCaps acute single-day overload
7Strength density≤ 2×/wk, 0 in race weekAvoid residual neuromuscular fatigue
8Breakthrough density≤ 1 / 7 days, 0 race weekAdaptation needs recovery between hard efforts
9Swim CSS deviation≤ ±3% per weekKeeps pace progression physically realistic
10Bike FTP detection ramp≤ 5% per 6-wk cycleFlags outlier test results instead of trusting them blindly
11Polarized ratio≥ 75% Z1–Z2Polarized-training consensus for long course (Seiler 2010)
12Recovery week TSS ratio≤ 65% of prior 3-wk avgGuarantees the recovery week is actually recovery

The numbers above are the defaults for the standard athlete tier. They are not arbitrary round figures; each maps to a published finding or a coaching consensus, listed in the sources.

What happens when you push

You ask: "Skip the recovery week, I feel great." TriPaced refuses and returns the legal alternative — usually compressing the build by a week instead — and shows you guard-rail #12. You see the trade, not just a "no".

If you force-validate a week anyway, the warning stays visible on it, the event is logged, and the next plan adapts around the choice you made. That is the difference between a plan that looks smart on race day and a coach that says no during the build, when saying no actually matters.

How we set the thresholds

Roughly 60% peer-reviewed literature (Banister, Foster, Mujika, Soligard, Coggan, Seiler), 30% pro-coach interviews and long-course consensus, and 10% TriPaced dataset retrospectives — when we look at athletes who did not finish, which guard-rail had they been quietly ignoring four to eight weeks earlier?

The thresholds are reviewed quarterly and we accept athlete-submitted counter-evidence. When a limit changes, the diff is public. A guard-rail you cannot inspect is just another black box, which is the thing we are trying to replace.

Frequently asked

Can I disable a guard-rail?

No. They are hardcoded by design. You can ignore the planner's output, but the planner will not generate the illegal week — it returns the closest legal one and names the rule.

What if my coach disagrees with one of your limits?

Send the citation. We update a threshold when the literature supports it, and the change log is public.

What if I'm an elite who needs to push harder?

Elite-tier guard-rails differ slightly — higher ramp ceiling, lower TSB floor — but they are gated behind review, not a toggle. Apply through the beta.

Why is polarized 75/25 hardcoded when some coaches do pyramidal?

Polarized is the long-course consensus (Seiler 2010; Stöggl & Sperlich 2014). Pyramidal fits shorter distances, and the distribution panel supports it — the guard-rail protects the 75% easy floor athletes most often break.

The CTL ramp seems low — other apps push 10 to 12.

Some apps optimise for fast fitness gain over injury rate. We optimise for getting you to the start line healthy; 5 to 8 points per week is the IOC consensus band (Soligard 2016).

Keep reading

TriPaced builds your long-course plan inside these twelve limits automatically — and tells you, in plain language, every time one of them shapes your week. The beta is open and reviewed by hand.

Apply for TriPaced beta →

Not ready to apply? The beta is free and open — see your real CTL, ATL and TSB, with the reasoning shown behind every change.

See the glass box for yourself →

Sources. Soligard, T. et al. (2016). How much is too much? IOC consensus on load in sport and risk of injury. British Journal of Sports Medicine. · Mujika, I. & Padilla, S. (2007). Tapering for competition: a review. Medicine & Science in Sports & Exercise. · Seiler, S. (2010). What is best practice for training intensity and duration distribution in endurance athletes? International Journal of Sports Physiology & Performance. · Stöggl, T. & Sperlich, B. (2014). Polarized training has greater impact on key endurance variables than threshold, high-intensity or high-volume training. Frontiers in Physiology. · Foster, C. et al. (1998). A new approach to monitoring exercise training. Journal of Strength & Conditioning Research.