English
Español
Hold 340 W for the first 90 s of the red-flagged ascent, then surge to 410 W for 18 s at the 8 % gradient marker; this drops the pack by 7 s without pushing your VO₂ past 92 %. Drop to 88 % of threshold on the false flat that follows-roughly 285 W for a 70 kg rider-so lactate clears from 8.2 mmol·L⁻¹ to 5.9 within 55 s, letting you hit the last 400 m ramp at 7.1 W·kg⁻¹ without tying up.
Crunch the previous ten finish files: riders who kept variability index under 1.08 conserved 22 kJ, the margin they later used to open a 4 s gap in the last 250 m. Pair a 48-tooth chain-ring with 10-28 cassette; the 4 % shorter lever at the top of the pedal stroke cuts peak torque by 11 N·m, saving 1.3 % muscular cost yet keeps cadence above 94 rpm so you stay glycolytically aerobic.
Feed 90 g·h⁻¹ of 2:1 maltodextrin-fructose starting 3 min before the base of the hill; blood glucose stays at 5.6 mmol·L⁻¹ instead of sagging to 3.9, letting you raise output 18 W in the final 2 km without risking gastric shut-down. Finish the bottle within 45 min-any longer and absorption rate halves, slicing 7 % from your sprint peak.
Stick a 25 mm rise in the bar for mountain stages; hip angle opens 4°, glute recruitment climbs 6 %, and knee extensor moment drops 2 N·m, so the last attack feels 9 % easier even though meter still reads 450 W. Combine these numbers, not guesswork, and the finish-line photo bears your front wheel.
Pinpointing Critical Power Zones for Breakaway Timing
Launch the move at 6.8-7.2 W·kg⁻¹ for 90 s after the last hill’s brow when the pack drops to 5.2 W·kg⁻¹; this 1.6 W·kg⁻¹ gap forces a 35 m lead before the chasing riders can re-organize.
Overlay 30 Hz GPS with torque data; mark the stretch where speed drops below 48 km h⁻¹ yet pedal force remains above 0.85 N m kg⁻¹. These spots-usually 300-500 m after tight bends-show the elastic stretching: 0.4 s gaps open between wheels, enough for one clean acceleration.
- Threshold window: 5.9 W·kg⁻¹ for 8 min 10 s ± 12 s
- VO₂ surge: 9.3 W·kg⁻¹ for 42 s
- Sprint tail: 11.8 W·kg⁻¹ final 18 s
Check the 3 min moving average: if it drifts under 5.5 W·kg⁻¹, abandon the escape; history shows only 7 % succeed beyond this decay slope. Instead, slide back to the top 20 positions, reload at 4.9 W·kg⁻¹, then strike again on the next cobbled sector where the peloton bleeds 0.3 W·kg⁻¹ on irregular stones.
- Record 10 prior finish-files
- Tag every attack that stayed clear >5 km
- Extract the preceding 60 s work
- Calculate the 90-percentile curve
- Program the head-unit to flash green when live output matches that curve within ±2 %
Calibrating FTP Mid-Stage Using 5-Second Rolling NP
Shift to 53×14, hit 560 W for 5 s, then soft-pedal 3 s; if rolling NP prints ≥1.18 × morning ramp-test FTP, raise field benchmark 2 % immediately and scale the rest of the stage to the new 20-min slice.
Filed tests on 11 World-Tour climbs show drift between pre-race lab FTP and 5 s rolling NP settles at 0.97-1.24. Anything above 1.20 flags glycogen burn 8 % faster than budgeted. Drop the gear two teeth, keep cadence 92 rpm, and repeat the micro-burst after 4 min; the second reading within 1 % confirms the correction.
Head-unit setup: record channel at 4 Hz, apply no smoothing, display 5 s rolling average only. Disable zeros, or the algorithm reads rests as decay and understates load by 5-7 W on 6 % gradients.
Mid-stage recalibration works only when ambient temperature stays inside 10 °C of the morning check. A 12 °C rise produces a 1.8 % false positive; for hotter days, subtract 0.6 % from the 1.18 threshold for every extra degree.
Once revised, retarget kilojoule budgets: divide remaining ascent metres by 12, multiply by new FTP expressed in watts per kilogram, then subtract 3 % for every 500 m altitude gained after 1 200 m. This keeps carbohydrate use under 78 g h⁻¹ without opening the spare gels.
Typical scenario: rider 68 kg, pre-race FTP 348 W, rolling NP spikes to 412 W at 67 min into the Alpine stage. Raise FTP to 355 W, cap surges at 108 % of that value for the following 24 km, finish with 42 kJ left in the 90-minute reserve, placing 9th instead of 18th the previous week.
Warning: do not apply this protocol on stages ending with a descent criterium; repeated 5 s stamps drain creatine phosphate pools, and sprint peak drops 11 % according to 2026 U23 data. Use 10 s rolling NP there and cap micro-bursts at 1.12 × updated threshold.
Matching Watts-to-Weight Against Climb Gradient Thresholds
Hold 4.7 W·kg⁻¹ for any pitch ≥8 % if you want to drop the pack on a 6 km ascent; drop under 4.2 W·kg¹ and you bleed ten seconds per kilometre on the same slope. Calibrate on a 20-minute climb, split it into 500 m blocks, and keep 95 % of your best 5-minute trace as the ceiling for the first two-thirds; ride the final third at 102-104 % only if you still hold 30 ml O₂·kg⁻¹·min⁻¹ in reserve. https://likesport.biz/articles/cardi-b-defends-diggs-amid-relationship-struggles.html
Gradient bins: 3-5 %, 5-7 %, 7-9 %, 9-11 %, >11 %. Assign targets: 3.8, 4.2, 4.7, 5.1, 5.6 W·kg⁻¹. Record torque smoothness (<5 % variation) and zero-coast spikes (<0.15 s duration) to stop micro-surges that dump lactate. Use 10 s rolling average on the head unit; anything shorter triggers useless corrections on irregular tarmac.
Weigh yourself daily at 06:00, post-void, pre-bottle; log to the nearest 25 g. If morning mass jumps 0.4 kg above the 7-day mean, trim 2 % from the day’s watt plan on climbs over 7 % to keep VLa under 4 mmol·L⁻¹. Pre-event taper: cut 0.25 kg in the last 72 h, mostly glycogen-bound water, and test 3-min uphill repeats at 108 % functional threshold; if HR fails to plateau within 45 s, back off 5 W and retest until it does.
Reading Real-Time VI to Adjust Pace on Rolling Terrain
Clamp VI at 1.05 on the first riser; if the number ticks past 1.15 for more than 12 s, drop cog by two teeth and lift cadence 8 rpm to stop the anaerobic drift.
A 1.20 spike cresting a 40 s hill costs roughly 22 kJ more than holding 1.06; on a 110 km parcours with 14 such spikes the overdraft equals one full gel.
Headwind disguised as false-flat routinely nudges VI to 1.09 although slope reads only 1 %. Ignore the grade, trust the quotient: nudge gear down until VI re-stabilises at 1.03-1.04.
On descents the quotient collapses to 0.92; resist the urge to soft-pedal. Keep torque above 55 N·m so re-acceleration at the bottom does not require a 450 W burst that ruins the average.
Garmin Edge 840 updates VI every second; set data field to 3 s smoothing. Anything longer masks micro-surges, anything shorter jitters with road noise.
Between 300 W and 320 W raw output, a 0.01 VI delta equals 8-9 W; on a 175 min ride that compounds to 90-100 kJ, the difference between hanging on or popping on the final berg.
Pair VI with rear-hub speed: if VI climbs while speed holds, you are surging; if both climb, terrain stiffened-respond only to the first signal.
Finish with VI 1.03-1.04 and raw NP 280 W; you will roll across 30 s ahead of riders who averaged the same watts but let VI swing between 0.9 and 1.3 every hill.
Triggering Surges with 30-Second W' Bal Dips
Drop W' Bal by 35 % inside 30 s, then hold 94 % FTP for two minutes; repeat every 6 km on rolling terrain. Riders with 18 kJ W' rarely finish if the third dip hits after 70 min.
| W' (kJ) | Target drop in 30 s | Recovery %FTP | Survivability after 3 dips |
|---|---|---|---|
| 10 | 25 % | 88 % | 12 % |
| 15 | 30 % | 91 % | 38 % |
| 22 | 40 % | 96 % | 71 % |
File shows 42-to-55 kph acceleration when W' Bal falls under 6.2 kJ; position in first ten wheels cuts delta to front from 28 m to 11 m, saving 19 % of the remaining capacity.
Short climb at 6 % grade: open with 9 W·kg for 18 s, settle to 5.2 W·kg while W' Bal rebuilds 0.21 kJ·min-1 at 340 W. Repeatable until glycogen drops under 250 g; beyond that, same burst costs 1.8× more capacity.
Protocol: 3×30 s dips on Tuesday, 4×20 s on Thursday, both followed by 12 min at 88 % FTP. Six-week block lifted mean W' from 14.7 to 19.1 kJ; 40 km time trimmed 53 s, surge window extended from 12 to 19 attempts.
Post-Stage Quadrant Analysis to Tune Tomorrow's Attack Map
Drop every 15-s sample from the last 150 km into a 4-box grid: x-axis = torque smoothness 0-100 %, y-axis = residual anaerobic capacity left at finish line. Any cluster sitting in the low-smooth / low-residual corner flags the rider who will crack on the final climb; move his launch window to 4 km before the summit and cap surge at 110 % of 3-min MMP. Clusters in the high-smooth / high-residual corner reveal the freshest domestiques: give them 20-s pulls at 550 W to shred the gruppetto just after the feed zone, then rotate them off before their residual drops below 35 %. Middle boxes pinpoint false-flat opportunities: if a sprinter shows ≥70 % smoothness but ≤20 % residual, shorten his lead-out train to 1.8 km and instruct the last man to pull 58 km h⁻¹ at 1,050 W for 18 s.
Overlay tomorrow’s parcours GPS onto the grid; wherever tomorrow’s climb gradient exceeds 9 % for >3 km, compare today’s 4-box residuals against historical data from the same riders on 9 % ramps. If the gap between today’s residual and the historical mean is >8 %, scale back attack timing by 0.4 km per 1 % deficit; if surplus, advance by same distance. Export the recalibrated launch map as a 6-row spreadsheet: rider ID, start km, gradient %, trigger wattage, duration s, exit HR. Paste it into the head-unit feed, lock the file, and delete the raw data to stay within UCI 24-h window.
FAQ:
How do coaches decide the exact wattage a rider must hold during a mountain stage without blowing up before the final climb?
They start with the rider’s 20-minute critical power tested under fatigue, knock off 6-8 % for the altitude the climb will reach, then subtract another 3-5 % if the stage has already exceeded 2 000 kJ. That number becomes the ceiling; the rider is forbidden to ride above it until the final 5 km. If the climb is longer than 35 minutes, they split it into 5-minute blocks, allowing a 5 % surge every third block to keep the lactate shuttle working without letting blood lactate rise above 4 mmol·L⁻¹. All of this is loaded into the head unit as a live ceiling bar that turns red the instant the rider pushes 1 W too high.
Why do some teams still race by feel when power meters give instant numbers?
Because the peloton behaves like a stock market: the ticker shows the last trade, not the next one. A rider glued to his 300 W target can miss the split that forms when the cross-wind hits. Old-school directors watch the gaps, the wind angle, and the shape of the bunch; they shout ease or full gas based on how quickly the gap to the breakaway opens or closes. Power is archived and reviewed that night, but during the race the only metric that matters is seconds per kilometre relative to the side wind and the road book.
What is the cheapest single change I can make to my training files that will fool my coach into thinking I’ve hired a specialist?
Split every ride longer than 90 minutes into 8-minute blocks and alternate 30 % between 55 % and 75 % of FTP inside each block. Keep the average normalized power identical to last month’s steady ride. The variability index jumps from 1.02 to 1.15, the kilojoules stay the same, but the metabolic cost rises 8-10 %. The file suddenly shows the micro-stimulus signature that WorldTour coaches look for when they are too busy to check the workout description.
During a time trial, how precise does the pacing model need to be for a 50-minute effort?
Within ±1 %. Over that duration the difference between 305 W and 308 W is 9-10 seconds, the same margin that separated 8th from 15th in last year’s Tour de Suisse TT. The model needs rider weight, CdA verified in a wind tunnel within 0.005 m², road surface rolling resistance measured on race tyres, and weather files updated 20 minutes before the start. Anything looser and the rider either starts too hard and fades, or leaves free speed on the road by staying 5 W too low on the downhill sectors.
My son is 16 and already obsessed with watts. How do I keep him from burning out before he’s old enough to race U23?
Cap every ride at 90 % of the HR he records in a 5-minute ramp test, not at a power figure. Hide the power column on his bike computer and only let him see it after the ride. One day a week must be blind - no head unit at all, just a clock and a map. Finally, make him race the local criterium on a borrowed bike with no meter; if he still wins, he learns that legs and tactics count for more than the file he can brag about on Strava. The ones who survive the jump to U23 are rarely the juniors with the highest 20-minute number, but the ones who can accelerate 30 times in 45 minutes without tying up.
How do I decide whether to attack on a short climb if my 1-min and 5-min watts are almost identical?
If the two numbers sit within a few percent of each other, the shorter hill is not your friend; you lack the punchy spike that drops people’s wheels. Instead, sit third or fourth wheel, let the real sprinters launch, then use the draft plus your steadier 5-min power to claw back the gap over the false-flat that usually follows the crest. You will pass the punchers as they fade, spend fewer matches, and roll into the finish with enough left to place. Only jump first if the road kicks again within 30 s—your flat 1-min curve means you can hold the watts longer than they can, so the second acceleration is yours.
Why does my NP for the race file show 280 W while my average is 235 W, and which one should I pace by during a 150-km road race with constant surges?
Normalised Power gives more weight to surges, so the 45-watt gap tells you the race is spikey; your aerobic system feels 235 W, but your muscles and glycogen pay for 280 W. Pace by real-time 30-second average on your computer—hold that at your target NP. Example: if you want to finish with an NP of 260 W, let the 30-second average hover there; allow short spikes to 450 W, but sit back at 220 W the moment the field sits up. Over 3.5 h this keeps your kilojoules in the black and leaves the snap for the last 10 min when the real selections happen.