Road to Sub 4 — Vikram's Marathon Journey

Zone	Pace range	Feel
Easy	6:20–7:00/km	Fully conversational · 80% of all running
Marathon	5:40–5:50/km	Controlled · can speak in sentences
Threshold	5:10–5:30/km	Comfortably hard · short phrases only
Interval	4:55–5:10/km	Hard · 800–1200m reps
Repetition	4:30–4:50/km	Fast · Parkrun PB lives here

Strength log — Arnold's Pump Club

How to log a session

1. Finish your Pump Club workout
2. Take a screenshot of the exercise list screen
3. Drop the screenshot in a new chat with Claude and say "log my Pump Club session"
4. Claude parses the workout, classifies muscle groups, adds it to your calendar, and flags any conflicts with your run schedule

No API integration needed — the screenshot is enough. Claude reads exercise names, sets, reps, phase number, and date automatically.

Strength ↔ run conflict guide

Arnold's Pump Club doesn't know your run schedule. Use this as your personal conflict guide — when to do the full session, when to modify, when to skip leg work entirely.

Sunday — long run day skip all leg work

No strength on long run day. Full energy goes to the run. If you want to do something, upper body only — arms, shoulders, chest. Nothing below the waist.

Monday — day after long run modify leg work

Go by feel — if legs are heavy from the long run, modify or skip what feels wrong. You correctly bailed on Lunges on Apr 27. Trust that instinct.

Tuesday — easy run day full session OK

Easy run + full strength is fine — just do the run first, strength after. Or split them (run morning, strength evening). Don't do heavy leg strength before a run.

Wednesday — strength day full session OK

Wednesday is your dedicated strength day — do the full Pump Club session. If life forces a run onto Wednesday (flex day), do it in the morning and strength in the evening, or pick one.

Thursday — easy run day full session OK

Same as Tuesday — run first, strength after. If Pump Club hits legs hard on Thursday, consider doing upper body only on Friday's strength session to avoid carrying fatigue into Saturday's Parkrun.

Friday — strength only (no run) light legs only

Rest day from running but strength is fine. Keep it lighter than Monday/Wednesday — legs should feel fresh for Saturday Parkrun. If Pump Club serves heavy lunges or squats, do upper body only.

Saturday — Parkrun day no strength

Parkrun only. No strength before — legs need to be fresh. If you want to do something post-Parkrun, upper body only, very light.

Recent sessions — Pump Club log

Sessions parsed from screenshots. Add new ones by dropping a screenshot in chat.

Workout #13 · Phase 3 Apr 29 · 5:16pm

Legs + upper

BW Rear-Foot Elevated Split Squat 3×10 · BW Reverse Lunge 3×15 · BW Lunge 3×10 · BW IYT Raises 3×25 · Pushup 3×12/12/8

Thu Apr 30 run was easy 8.2km @ 6:34/km HR 125 — correctly dialled back. Good call.

Workout #12 · Phase 3 Apr 27 · 5:50pm

Lower body + core

Hip Raise 4×20 · Superman W 4×12 · Wall Sit 4×~41s · Bodyweight Lunge (bailed — day after long run ✓) · Crunch 3×20 · Chair Dips 2×20/15

Impact on next run: Hip Raises + Superman W are fine. Lunges correctly skipped. Thursday easy run unaffected.

Drop a Pump Club screenshot in chat → Claude parses it → session appears here

Race predictor — TCS Toronto Oct 18, 2026

Calculated predictions — from your real data

Ordered by recency and reliability — the Working Estimate is auto-recomputed on every upload and is the one to trust. The rest are kept below as reference points, roughly in order of how much weight they should carry.

Working estimate — auto-computed, built off your current fitness

3:43:13–4:02:38

Two-stage model, deliberately NOT anchored to last year's marathon or the April time trial — fitness changes too much over months for an old race to set today's baseline. (1) Base: your current best 5K, extrapolated using a marathon-distance fade rate that's typical for recreational runners in general (not derived from your specific race history). (2) That base is then adjusted by how your last ~2 months of volume, longest run, and aerobic HR efficiency compare to the same window a year ago — the actual "body of work" signal.
Wider range than before — without a recent race to calibrate against, this carries real uncertainty. Treat the trend below as more trustworthy than this specific number.

Current-fitness base: 4:16:03 · Recent-work adjustment: +9.0%

Working estimate over time

3:29 faster since 2026-05-25 · 13 data points

● Actual upload ○ Reconstructed from historical data

2025 Marathon — historical result only

Not used to anchor the current estimate

~4:17:52

You ran 42.6 km @ 6:02/km · HR 154 on Oct 19, 2025.
Per your own read, and correctly so — fitness changes too much over 10 months for this to set today's baseline. Kept here as a fact of record, not as a prediction input. The legacy speed-scaling model derived from this race (4:16:31 (legacy, from Oct 2025 race)) is shown for reference only and doesn't drive the Working Estimate above.

Around the Bay extrapolation

Historical note · also not a current-fitness anchor

3:55–4:05

You ran 30.2 km @ 5:35/km · HR 151 on Apr 12, 2026.
Same logic as the card above — this is 4+ months old, and fitness moves too much in that window for a single April effort to represent where you are now. Kept for the record; the Working Estimate above is built from your current best 5K and your last ~2 months of training instead.

Garmin VO2max — trailing 60-day average

Live · auto-updated on every upload

47.1

Garmin estimates VO2max on every run, but it swings 5–7 points day to day (device/algorithm noise, not real fitness change) — a single reading isn't meaningful. This is a 30-run rolling average over the last 60 days, which smooths that out. Up 2.6 points vs. the same 60-day window last year (44.4). Shown as supporting context on aerobic capacity trend, not a second competing time prediction — the Working Estimate above is still the number to trust.

Riegel formula — from best 5K

Reference only · speed ceiling

3:31:19

Formula: marathon = 5K × (42.195/5)^1.06 · Based on current best 5K PB, auto-updated on every upload.
⚠ Your 2025 marathon at 4:17 confirms this generic formula significantly overpredicts for you. Shown as a speed ceiling only — not a target.

Historical data — from Runalyze (2025–2026)

Parsed from 178 activities across 16 months. This is the actual training history the predictions are grounded in.

TCS Toronto Marathon 2025 — debut

Oct 19, 2025 · Your baseline · the time to beat

4:17

Distance

42.6 km

Avg pace

6:02/km

Avg HR

154 bpm

Goal 2026

sub-4:00

17-minute PB target. You ran 6:02/km for 42.6km. Goal is 5:40/km. That pace felt comfortable at 30km (ATB, Apr 2026) — the question is whether it holds over the full distance with peak training.

Training load — year-to-date (Jan 1–Aug 19)

Metric

Total km
20km+ long runs
Longest
Best 5K

2025

km
runs
2 km
44/km

2026 ↑

km
runs
2 km
24/km

+29% volume · +200% more 20km+ long runs · 19s/km faster on best 5K · longest run 21km → 30km. Auto-computed year-to-date through 2026-08-23.

2025 long run staircase — the completed template

2026 long run staircase — same scale, so far

Easy builds · 20km+ · 30km+ · Race (42km) · Both charts use the same km-to-height scale, so bar heights compare directly. 2026 has already logged 24 long runs ≥15km vs 2025's 18 across its whole build (including race day) — more frequent long-run exposure this cycle, reaching 30km back in April rather than only in the final month. Peak in the current plan is recalibrated to 32km, matching 2025's actual peak.

Training signals — directional only, not percentages

Based on 2025 full-year data (128 runs, 1,168 km) and 2026 Garmin history. No invented numbers.

Speed — strong ✓

Best Parkrun: 22:02 (2026-04-04). Last 9 Parkruns (90 days): 24:22, 24:52, 23:11, 24:55, 27:21, 24:11, 23:05, 23:55, 23:10.

Marathon experience — confirmed ✓

4:17:52 on 2025-10-19 (Toronto - TCS Toronto Waterfront Marathon). 42.6km @ 6:03/km.

Endurance — matching last year, not exceeding it

2025 long run peak: 32 km (Sep 28). 2026 plan was originally going to push to 36 km, but recalibrated Aug 19 off actual training volume — peak comes back to the same 32 km. The improvement needs to come from pacing and efficiency, not extra distance.

Weekly volume ahead of last year at this point

Same point last year: 23km, then 9km. This year, same window: 37km, then 39km — ahead of where you were a year ago. 2025 peak was 50km/week (2025-10-13). Auto-computed from upload processed 8/23/2026.

Consistency — strong ✓

94 runs in 2025 through this date (794km), then 113 runs in 2026 (1021km). 2 real gap weeks this year (2026-02-16, 2026-07-27), otherwise consistent.

Aerobic efficiency — plateaued in 2026, not concerning

HR at ~6–6:30/km pace by quarter — 2025: Q1: ~138, Q2: ~139, Q3: ~139, Q4: ~143 bpm. 2026: Q1: ~137, Q2: ~134, Q3: ~137 bpm. Auto-computed from actual runs in that pace band each quarter — no resting HR in this data source.

Next calibration point: The Jul 12 half marathon time trial was skipped — a 27km long run (Jul 19, 6:13/km, HR 134) filled that role instead. The plan itself was recalibrated Aug 19 off actual training volume (peak long run lowered from 36km to 32km). Continued weekly Garmin data through the peak phase will keep sharpening the estimate from ±10 min to ±5 min.

AI coaching analysis

Enter your latest race times and Garmin stats for a detailed coaching assessment. The AI will use your actual data — not guesswork.

Recent 5K (mm:ss)

Recent 10K (mm:ss)

Half marathon (h:mm:ss)

Longest recent run (km)

Additional context

Vikram's roadto sub 4:00

Vikram's road
to sub 4:00