Glossary.

The terms you'll meet in Kimbia. Each one starts with what it tells you about your session; the arithmetic comes after, for whoever wants to dig. Training yourself is something you learn.
VDOT
One number that places your running level, and from which all your training paces follow. The higher, the faster. Daniels & Gilbert (1979).
In detail
A running fitness index computed from your race times (Jack Daniels' method). The higher, the faster. Kimbia uses it to derive all your training paces. It's the scientific cousin of MAS.
Where this comes from: Daniels & Gilbert
MAS (VMA)
The word your club uses for roughly the same thing as VDOT. Kimbia shows both so you're never lost in translation.
In detail
Maximal Aerobic Speed - the speed at which your oxygen uptake plateaus. It's the everyday term in French running clubs. Kimbia shows the VDOT ↔ MAS equivalence so everyone finds their bearings.
Where this comes from: Daniels & Gilbert
Split (segment)
A chunk of a session: a kilometre, a rep, a track lap. It's the brick Kimbia analyses - the whole per-lap detail is built on it.
In detail
A portion of an activity: a kilometre, an interval rep, a track lap. Kimbia keeps it exactly as your watch cut it, and calls it a "lap" throughout the app - it is the row in the per-lap detail.
Power (watts)
Intensity measured in watts, no detour. Unlike pace or HR it reacts instantly, and it doesn't lie in the wind or on a climb.
In detail
Effort measured in watts by a power meter (crank, pedals, or watch/pod for running). Unlike speed or HR it reacts instantly and doesn't depend on wind, gradient, or cardiac fatigue: it is the most objective measure of intensity. Kimbia draws the curve whenever the activity carries watts - including the ones a service estimated from pace, which no sensor ever measured.
Where this comes from: Allen & Coggan
FTP
Your reference in watts: what you can hold for about an hour. Without it raw watts mean nothing - it's what turns them into zones of your own. Allen & Coggan (2006).
In detail
Functional Threshold Power - the highest average power you can hold for about an hour. It's the reference that turns raw watts into personal training zones (like max HR for heart-rate zones). Estimate it with a 20-min test (≈ 95% of the average) or a ramp test.
Where this comes from: Allen & Coggan
W/kg (watts per kilo)
Your watts divided by your kilos. It's what decides on a climb, and the only honest way to compare yourself with someone of another build.
In detail
Power relative to your body weight: your watts divided by your kilos. It's what matters on climbs and for comparing athletes of different sizes (400 W weighs more on 60 kg than on 90 kg). Kimbia computes it as soon as you enter a weight and an FTP.
Where this comes from: Allen & Coggan
Smoothing (τ)
The slider that decides whether you're looking at every sensor twitch or the overall shape of your session. Raise it for the trend, lower it for the detail.
In detail
The chart's “Smooth” slider removes GPS and sensor noise so the trend stands out. τ is the filter's half-window: each point is combined with its τ neighbours on each side (median then a light mean, to keep transitions sharp). The equivalent in seconds shows next to the slider - it depends on the activity's duration. τ=0 shows the raw data.
Drift (HR)
How hard your body found it to hold the same effort as the session went on. The higher the drift, the more holding your pace cost you.
In detail
The rise in your heart rate at constant effort: average HR over the last 30 % of the analysis window against the first 30 %. The window keeps only the work laps, stitched end to end, and starts at the steady-state onset - the head of an effort measures the cardiac ramp, not the athlete, and recoveries would drag everything down. It rises with heat, dehydration and fatigue. Under 5 % the aerobic execution is stable; past 10 % the session was very hard, very hot, or misclassified. In the per-lap detail, the "Drift" column applies the same idea inside each lap.
Pace drift
Did you slow down along the way? Read it as a pair with HR drift: slowing with a calm heart, or holding pace and paying for it in beats, are two different sessions.
In detail
The twin of HR drift, in pace: your pace over the last 30 % against the first 30 %, as a percentage (positive = you slowed down). Same bounds, same blocks, same window - the two read as a pair. In the per-lap detail, the column applies the same idea inside each lap: it says whether you held or blew up on a rep, where the lap average hides the damage.
Lap role
The label saying what each lap was there for: warm-up, work, recovery… It's what lets Kimbia analyse the real work only. You can correct it.
In detail
The label Kimbia assigns to each lap from its intensity and place in the session: warm-up, ramp-up, work, recovery, ramp-down, cool-down, and "paused" when the watch was stopped. It lets Kimbia isolate the work block - what surrounds it counts in no analysis. You can correct it by hand.
RPE (perceived exertion)
One number from 1 to 10, given right after the session: what did it cost you? Subjective - and that is its strength: twenty years of studies back it, and it captures what no sensor sees. Borg (1982), Foster (2001).
In detail
The Borg CR10 scale: 1 very easy, 10 maximal. Foster showed that a single rating given within half an hour of the effort, multiplied by duration, makes a training load as reliable as heart-rate methods - no sensor required. The rating absorbs what heart rate misses: heat, sleep, stress, soreness. On Kimbia, the RPE prices the session (RPE × minutes) and the feeling colours the curve with your morale; the journal marks sessions still waiting for their rating with a red dot.
Where this comes from: Borg · Foster
Training load
What a session cost you, as one number, so your week can add up. Kimbia keeps two thermometers, one on the heart, one on how it felt, because neither is enough alone. Edwards (1993), Foster (2001).
In detail
What a session cost you, as one number. Kimbia keeps two, because neither is enough alone. The first comes from the heart: minutes spent in each heart-rate zone, weighted Z1×1 through Z5×5, plus a Z0 at ×0.3 below the first threshold - the Effort shown on every session. The second comes from you: your RPE rating times how long it lasted - the Session RPE. That day's load is the sum of the sessions you did, on whichever thermometer is drawn. An hour on the bike counts as 1/3.5 of a running hour: the constant was measured three ways on our own data and all three land between 3.4 and 4.7. On the Load chart, Effort draws in black; the Session RPE line carries one extra signal - its colour is your morale, green through good stretches, crimson through rough ones.
Where this comes from: Edwards · Borg · Foster
Effort or Session RPE
Your load is read on two thermometers: Effort, read by your heart, and Session RPE, read by you. When Effort runs higher than the RPE, the session cost your cardio what your head never noticed.
In detail
The two thermometers share no unit, and no common order of magnitude either. To read them together each is scaled to its own average over the range shown: ×1 is your normal, ×1.3 a third busier. Change the range and the reference is recomputed. What matters is not the level but the gap: when the Effort curve runs above the Session-RPE one, your sessions cost your cardio more than they cost your perception; the other way round means you were taking badly a load your heart found ordinary. Averaging the two would erase exactly that disagreement.
Fitness (42 days)
What your body has absorbed over the last six weeks. It moves slowly, and that's what makes it trustworthy: one big outing doesn't send it flying. Banister & Calvert (1975).
In detail
A rolling average of your daily training load over roughly six weeks, weighted so recent days count more. It moves slowly, which is the point: it tells you what your body has actually absorbed, not what you did yesterday. Read it against your own best - the highest it reached in the last twelve months - rather than against anybody else's number.
Where this comes from: Banister & Calvert · Allen & Coggan

Where this comes from

The published work Kimbia leans on. The constants themselves are recomputed on your own data rather than taken as given.

  1. Banister & Calvert 1975 · 1980

    The impulse-response model: from a daily load, two rolling averages - one slow, one fast.

    The paper +

    A systems model of training for athletic performance. Australian Journal of Sports Medicine, 1975;7:57-61. — Planning for future performance: implications for long term training. Canadian Journal of Applied Sport Sciences, 1980;5(3):170-176.

  2. Edwards 1993

    Zone-weighted heart-rate load, 1 through 5. That is each session's Effort.

    The paper +

    Edwards S. The Heart Rate Monitor Book, 1993.

  3. Borg 1982

    The scale your RPE rating rests on.

    The paper +

    Borg GAV. Psychophysical bases of perceived exertion. Medicine & Science in Sports & Exercise, 1982;14(5):377-381.

  4. Foster 1998 · 2001

    The RPE rating times minutes makes a load, with no sensor at all.

    The paper +

    Monitoring training in athletes with reference to overtraining syndrome. Medicine & Science in Sports & Exercise, 1998;30(7):1164-1168. — Foster C, Florhaug JA, Franklin J, et al. A new approach to monitoring exercise training. Journal of Strength and Conditioning Research, 2001;15(1):109-115.

  5. Allen & Coggan 2006

    The 42- and 7-day constants, since become the convention.

    The paper +

    Allen H, Coggan A. Training and Racing with a Power Meter. VeloPress, 2006.

  6. Minetti et al. 2002

    The energy cost of a slope, measured out to extreme gradients. It's what brings your pace back to the flat.

    The paper +

    Minetti AE, Moia C, Roi GS, Susta D, Ferretti G. Energy cost of walking and running at extreme uphill and downhill slopes. Journal of Applied Physiology, 2002;93(3):1039-1046.

  7. Karvonen et al. 1957

    Intensity read inside the heart-rate reserve, from rest to max - the one form that compares across athletes.

    The paper +

    Karvonen MJ, Kentala E, Mustala O. The effects of training on heart rate: a longitudinal study. Annales Medicinae Experimentalis et Biologiae Fenniae, 1957;35(3):307-315.

  8. Friel 2009

    Decoupling: the efficiency of the first half against the second, over the same window.

    The paper +

    Friel J. The Triathlete's Training Bible, 3rd edition. VeloPress, 2009.

  9. Daniels & Gilbert 1979 · 2013

    Race times turned into a single index, from which every training pace follows.

    The paper +

    Daniels J, Gilbert J. Oxygen Power: Performance Tables for Distance Runners, 1979. — Daniels J. Daniels' Running Formula, 3rd edition. Human Kinetics, 2013.

  10. Hellard et al. 2006

    The model's limits in real monitoring - why we draw no form index.

    The paper +

    Hellard P, Avalos M, Lacoste L, Barale F, Chatard JC, Millet GP. Assessing the limitations of the Banister model in monitoring training. Journal of Sports Sciences, 2006;24(5):509-520.