The progress

Seven months, three sports, no editing.

Real intervals.icu history, February through August 2026 — nothing logged before Feb 6. Every number below is pulled straight from the account, including the ones that don't tell a flattering story.

~2,954 kmcovered under his own power
2,134 kmon the bike outdoors
730 kmrun
89 kmswum

Plus roughly 251 hours of total training time. Distance/hours are honest estimates from the same monthly data as the charts below — the "~" is doing real work, not decoration.

Running — the clear win
6:21/km ↓ 5.8% faster
at the same heart rate (131 vs 132 bpm)

Easy-run pace went from 6:44 to 6:21/km, Feb → Jul, with heart rate essentially unchanged — that combination is a real aerobic improvement, not just a harder effort.

Aerobic efficiency (speed per heartbeat) on easy/moderate runs. Higher is better. May (n=3) and August (n=2) are thin months — hover for exact values.

MonthRunsKmPaceAvg HR
Feb967.56:41/km134
Mar14123.36:07/km140
Apr1091.96:11/km137
May14104.75:59/km147
Jun12138.56:32/km136
Jul14141.46:11/km138
Aug*662.96:22/km136

*Aug partial (through the 14th). Table is all runs; the chart above and headline number use the easy/moderate subset only, which isolates fitness from single hard efforts.

Swimming — real, modest, thin
2:12/100m ↓ 3% faster
heart rate down 11% (133 → 118 bpm)

Pace and heart rate both moved the right way, Feb → Aug — a real improvement, just on a smaller sample than running.

Average pace per 100m, pool swims. Lower is better. 4-8 swims/month.

MonthSwimsKmPace/100mAvg HR
Feb610.82:17133
Mar815.72:17129
Apr715.42:17130
May510.62:17129
Jun612.92:09132
Jul715.22:14124
Aug*48.82:12119

*Aug partial. 43 pool swims total, 89.2 km — volume stayed flat all year, only pace and HR moved.

Riding — two tracks, not one shrug

Power only started getting recorded in July, so a power trend genuinely doesn't exist — but distance, time and heart rate cover the whole year, and they tell a real story once outdoor and indoor are looked at separately, the way this project always tracks them.

Outdoor · efficiency
+16%speed per heartbeat
Feb–Apr avg vs. May–Aug avg

Speed (km/h) per bpm. Confounded: heart rate rose alongside speed (124→134 bpm), so part of this is harder rides or hotter weather, not a clean fitness gain.

Indoor · session length
57 → 84min+47% longer
at flat-to-lower heart rate

Avg trainer session length, Mar–Jul vs Feb. This is the cleaner signal — no speed/terrain confound indoors.

MonthOutdoor ridesAvg speedIndoor ridesAvg session
Feb723.3 km/h158 min
Mar320.8 km/h993 min
Apr321.1 km/h884 min
May1029.3 km/h679 min
Jun327.6 km/h1082 min
Jul1025.2 km/h282 min
Aug*228.0 km/h355 min

*Aug partial and thin on both tracks — 2 outdoor, 3 indoor rides. 2,134 km / 88.2 h outdoor, 53.1 h indoor, 141.3 h combined, two known GPS/mislabel artifacts excluded (below).

What got excluded, not averaged in: a GPS-corrupted "105-minute" ride that actually covered 321 meters, a mislabeled FTP-test attempt with no usable power, and 3 trainer sessions mislabeled as outdoor rides (kept in ride counts, excluded from speed since they have no real distance). The 259W outdoor FTP quoted elsewhere on this site is a model estimate, not a completed test — benchmarks.yaml still shows an empty test log as of today.