The riders who get faster every season aren't braver. They're more systematic. They use telemetry the same way an engineer uses a wind tunnel: change one variable, measure, decide, move on.
This is the 5-step method that consistently drops seconds for amateur club racers. It works whether you're chasing a 10-second drop or your last tenth. If you're new to reading the data itself, start with our beginner's guide to motorcycle telemetry — the rest of this post assumes you can read a speed trace.
Key takeaways
- Don't try to get faster everywhere — find the single biggest time loss and target that.
- Use lap overlays to find the gap, not your perception of how the lap felt.
- Make one change per session so you know which change moved the trace.
- Test whether you've actually improved by comparing the same corner across sessions, not just lap times (which depend on traffic, tyre temp, and conditions).
- Most riders find 0.5–1.5 seconds of "free time" in the first 2 sessions of doing this properly.
Step 1 — Set a clean baseline lap
You can't improve what you can't measure. The first session of any track day should be a baseline — not a hot lap, not a "rusty rider" lap. A clean, consistent rep of the lap as you currently ride it.
What "clean" looks like:
- 5–8 laps with no traffic, no overtakes, no big mistakes.
- Tyres up to temperature.
- Brain calibrated to the circuit.
When you upload to a tool like ApexIngest, pick the 3rd-fastest lap as your reference, not the fastest. The fastest is often a fluke — third-fastest is repeatable. (The lap overlay feature makes this straightforward — pick any two laps to compare.)
If you've been to the track before, also pull your best lap from your last visit. That's the lap to beat.
Step 2 — Find the biggest delta
Open your lap overlay view. You should see two traces — your reference lap and a target (either your previous best, your mate's lap, or a target time you've set).
Look at the time delta trace. Where does the gap grow most?
Sector 1: +0.05s ← small loss
Sector 2: +0.12s ← small loss
Sector 3: +0.71s ← here. focus here.
Sector 4: -0.08s ← actually faster
Lap delta: +0.80s
In this example, you don't have a "lap problem." You have a Sector 3 problem. Working on Sectors 1 and 4 is wasted effort.
Zoom into Sector 3. Which corner specifically? In ApexIngest, the corner coaching panel will tell you in plain English: "Lost 0.4s in T7 — throttle pickup 0.6s late."
Step 3 — Form a single hypothesis
This is where most amateurs go wrong. They look at the data, see five things they could improve, and try to fix them all on the next session. Result: nothing changes consistently and they can't tell what worked.
Instead, form one specific hypothesis:
- ❌ "I need to ride T7 better."
- ✅ "If I get back to 30% throttle 5 metres earlier in T7, I'll save 0.3 seconds."
Look at your throttle trace through T7. Where exactly are you opening the throttle? Where could you realistically be opening it? The hypothesis has to be:
- Specific — you can describe it in one sentence.
- Measurable — the data will show whether it happened.
- Achievable in one session — not "I'll lean 5° more" (that's a year of riding).
Step 4 — Test the hypothesis on one session
Go out for one session focused only on the change. Don't try to also fix your braking into T2. Don't try to be smoother in T11. Just T7 throttle pickup, every lap, for the whole session.
Tips for testing:
- Pick a physical reference: a kerb edge, a marshal post, a paint mark. "I'll be on the throttle by the second blue kerb stripe."
- Don't worry if your lap times are slower during the session — you're learning a new pattern. Speed comes when the new pattern is automatic.
- Stay aware of the rest of the lap. Don't crash trying to fix T7.
Step 5 — Compare and decide
Back in the paddock, upload the new session and overlay the new "T7-focused" lap against your baseline.
Three possible outcomes:
1. The trace moved. You're faster through T7. Bank it. Move on to the next biggest delta in your data.
2. The trace didn't move. Same throttle pickup as before. You didn't actually execute the change. Either the hypothesis was wrong (the kerb you picked is too early) or you reverted to the old habit under pressure. Re-set and go again.
3. The trace moved but you're slower overall. The change worked locally but unbalanced something else — usually you carried more speed but ran wide and lost the next corner. Adjust the hypothesis: "I'll get to throttle 5m earlier AND tighten my line by half a metre."
The wrong response in case 2 or 3 is "this method doesn't work." The right response is "my hypothesis was wrong, here's the next one."
A worked example: 1.2 seconds in two sessions
Real example from a rider on a Yamaha R6, Donington Park National, summer 2025:
Baseline: 1:21.4
| Corner | Time loss vs target |
|---|---|
| Redgate | +0.2s |
| Craner Curves | +0.1s |
| Old Hairpin | +0.5s |
| Coppice | +0.3s |
| Goddards | +0.3s |
Session 1 hypothesis: "I'm braking too deep into Old Hairpin and rolling onto throttle 0.4s late." Change: Brake 10m earlier, get back to throttle by the apex paint mark. Result: Old Hairpin -0.4s. New time: 1:21.0.
Session 2 hypothesis: "Goddards exit — I'm waiting for the bike to be fully upright before pinning it." Change: Trust the rear, pin it from 30° lean. Result: Goddards -0.3s, but Coppice gained +0.1s (carried more speed in). New time: 1:20.6.
Two sessions, two hypotheses, two trace-confirmed wins. 0.8s. The remaining 0.4s came from a third session focused on Redgate entry.
The rider didn't get braver. He got specific.
What slows this method down
- Looking at the lap as a whole instead of finding the biggest single delta.
- Trying to fix everything at once. Your brain has one slot for "thing I'm working on this session." Use it.
- Comparing against laps with traffic or wet patches. Always overlay against clean reference laps.
- Trusting feel over data. If the trace says you're slower, you're slower. Feel lies. Data doesn't.
- Skipping the "decide" step. If you don't explicitly bank or reject each change, the method becomes wandering practice.
FAQ
How many sessions does this take to see results? Most riders see 0.5–1.5 seconds of improvement in the first 2 sessions of using a structured telemetry method, because there's almost always one or two corners where they're leaving obvious time on the table.
Do I need a coach to do this? Not strictly. ApexIngest's automated corner coaching describes the biggest issues in plain English. But a coach can help validate your hypotheses faster — especially for technique changes that are hard to self-diagnose.
What if I don't have a faster reference lap to compare against? Use your theoretical best lap — the sum of your best sectors, even if you didn't put them all together in one lap. Most telemetry tools (including ApexIngest) calculate this automatically. It's usually 0.3–0.8s faster than your actual best, and it's a realistic target.
Should I work on slow corners or fast corners first? Wherever the biggest time delta is. There's a myth that slow corners are where you find time — actually, fast sweepers can carry huge time losses because you're committing for so long. Always follow the data, not the theory.
How do I know if a change is real or just lap-to-lap variation? A real change shows up consistently across 3+ laps. A one-lap improvement is variance. ApexIngest highlights consistency by showing the spread of all laps through a corner, not just the fastest.
Try the method on your next session. Upload your existing data free — speed traces, corner coaching, and lap overlays included. Or learn the basics in our beginner's telemetry guide.