Every fighter dreads getting ill in camp. You've spent weeks building fitness, dialling in your intervals, sharpening your timing — and then your body decides to fight something other than an opponent. The instinct is to push through. The data says otherwise.
Here's how Fight Coach AI handled a real illness that hit 10 days before the WKKC English Championships.
The Warning Signs Were in the Data
Before I even felt ill, the numbers were shifting. My last bag session on February 5th looked off compared to the weeks before it:
| Metric | Typical Session | Feb 5 Session |
|---|---|---|
| Max HR | 188–192 | 180 |
| Time in Zone 5 (185+) | 8–12% | 0% |
| Interval recovery HR | 125–140 | 106–111 |
I couldn't find Zone 5 at all. At the time I thought it was just a flat day — we all have them. But looking back, the recovery HRs were unusually low (106–111). My heart wasn't being challenged enough to need real recovery. That's not a sign of fitness — it's a sign the body was already diverting resources.
Then the sleep data confirmed it.
The Sleep Collapse
Over the next few nights, my recovery metrics fell apart:
| Night | Deep Sleep | HRV | Resting HR | Score |
|---|---|---|---|---|
| Feb 6 | 57m | 69 | 48 | 83 |
| Feb 7 | 15m | 67 | 50 | 71 |
| Feb 8 | 5m | 57 | 52 | 61 |
| Feb 9 | 50m | 56 | 52 | 75 |
| Feb 10 | 57m | 45 | 53 | 37 |
Deep sleep — when the body physically repairs itself — crashed to 5 minutes on Feb 8. HRV dropped from its usual 65–77 range down to 45. Resting heart rate climbed 9bpm above my healthy baseline. Garmin flagged the Feb 10 sleep as "long but not restorative" — nearly 8 hours in bed, 113 minutes awake, stress score rated POOR.
My body was fighting a cold and losing the recovery battle.
What the AI Coach Said
This is where having persistent, data-aware coaching matters. Fight Coach AI had been tracking my metrics daily for weeks. It knew my baselines. It knew I had a competition on February 22nd. And it knew the numbers didn't lie.
On February 14th — eight days out from competition and eight days into illness — the daily coaching briefing was direct:
"You're 8 days out from WKU and just cleared an 8-day illness layoff. The good news: your sleep recovered well (last night scored 85, HRV 59, RHR back to 50). The concern: you've lost a week of peaking time, and your fitness window is now very tight."
It didn't panic. It didn't tell me to train through it. It looked at the data and gave me a realistic assessment: the fitness from 7 weeks of training was still there, but my body needed to finish fighting the cold before it could fight anything else.
The next day, when the numbers had stabilised but not fully recovered, the coaching adjusted:
"Do a short, moderate reactivation session — 30–40 min total, 4–5 x 90s fight intervals at Z4 (180–184bpm) with full recovery between. Focus on getting consistent max HRs and clean recovery HRs (back to 130–135). If you feel flat or HRs won't climb, pull back — one poor session won't hurt; re-injury will."
Zone 4, not Zone 5. Recovery between intervals, not chasing peak HR. The goal was to test the engine, not redline it.
The Recovery Curve
Once I stopped training and focused on rest, the data started turning:
| Night | HRV | Resting HR | Score | Direction |
|---|---|---|---|---|
| Feb 10 | 45 | 53 | 37 | Bottom |
| Feb 11 | 55 | 51 | 77 | Recovering |
| Feb 12 | 53 | 52 | 79 | Holding |
| Feb 13 | 59 | 50 | 85 | Climbing |
| Feb 14 | 53 | 52 | 79 | Plateau |
The HRV bounced from 45 to 59 over three days once I stopped loading the body with training. Resting HR dropped back toward baseline. Sleep scores climbed from 37 to 85.
But the recovery wasn't linear — it oscillated. HRV bounced between 53 and 59 for several days. That's the immune system still working. Fight Coach AI tracked this and held off recommending hard training until the trend was clearly upward, not just bouncing.
Why This Matters
A human coach checking in once a week would have seen me on February 5th, said "good session" (it looked fine on the surface), and scheduled the next hard week. I'd have trained through the early illness, extended it, and shown up to competition either still ill or freshly recovered with zero sharpness.
A generic training app would have served me the next workout in the programme regardless of what my body was doing.
What Fight Coach AI did instead:
- Detected the decline before I felt it — the Feb 5 session data showed reduced output that I dismissed as a flat day
- Correlated training data with sleep metrics — connecting the poor session to deteriorating HRV and deep sleep
- Recommended complete rest — not active recovery, not light sessions, actual rest
- Monitored recovery daily — tracking HRV, RHR, sleep quality, and SpO2 to determine when the body was genuinely clearing the infection
- Planned a graduated return — Zone 4 reactivation before Zone 5 sharpening, with clear criteria for when to progress or pull back
The result: instead of 10 days of declining, illness-extended training, I got 9 days of actual recovery followed by a structured return that preserved the fitness I'd built across 7 weeks of camp.
The Bigger Picture
Getting ill in camp isn't unusual. What's unusual is having a coach that watches your recovery data overnight, notices the 4bpm resting HR creep before you notice the sore throat, and adjusts your entire training plan within hours.
This is what data-driven coaching actually means. Not dashboards. Not charts you glance at and ignore. A system that understands your baselines, detects when something's wrong, and makes real coaching decisions based on what the numbers are telling it.
Seven days to the WKKC English Championships. The engine's still there. Now it's about arriving healthy, sharp, and ready — not overtrained and hoping for the best.