Roanoke race analysis and race-block adjustments
- Athlete: Sidney McGill
- Prepared: 2026-09-28 by Coach Luca
- Races: Virginia's Blue Ridge Go Cross, Roanoke, VA. C1 on Saturday, September 26 (fell off the leader after two laps). C2 on Sunday, September 27 (won, ahead of Manon Bakker).
- Evidence:
- TrainingPeaks 1-second race files for both Roanoke races and both Rochester races.
- All activities from September 18 to 28, and her fitness trend (CTL, ATL, TSB).
- Weather recorded in the daily logs.
- The operator's race observations and Sidney's notes on the coming block.
- Status: internal coaching report. The revision was approved and applied on 2026-09-28 to the calendar files and live TrainingPeaks (see the publish report).
Bottom line
- She did the same amount of work both days. Normalized power (NP) was 223 W on Saturday and 224 W on Sunday; average power was 200 W both days. The difference was how she spent it, not how much she had.
- Both days were a clear step up from Rochester a week earlier:
- Race NP up 5-7%.
- Time above 390 W (150% of FTP) up from 5-7 minutes to about 10.5 minutes.
- A smaller fade over the race.
- This came before the new training block had really started.
- Saturday was a pacing problem, made worse by heat and an unprimed body, not fatigue.
- She was fresh both mornings (TSB about +9).
- She chased the leader at 250 W for the first two minutes and 245 W for lap 1, about 10% above what she could hold, starting from a heart rate of 98 bpm at the gun.
- Laps 2-6 fell to 213-226 W and ran about 20 seconds slower per lap.
- Saturday was 27°C in full sun; Sunday was 26°C and overcast. From lap 2 onward, Saturday cost about 3-4% more heart rate per watt.
- Sunday was a well-paced race. A soft start (189 W for the first two minutes), then 221-233 W laps, with the strongest lap last (233 W) and a higher 1-minute peak than Saturday (354 W vs 314 W). She had efforts left for the finish.
- Sidney's concern about the next five weekends is well founded. The published plan has her racing at TSB -13 to -16 on three October weekends, bottoming at -33 after the October 5-11 block, and starting Nationals at -4. I recommend revising it (details below).
- Race execution fixes are worth as much as training changes:
- A cap on lap-1 pace.
- A staging routine that gets her to the gun with her heart rate already up.
- A heat protocol.
- Never a full rest day on the Thursday before a race weekend.
The weekend in numbers
Freshness is TSB on race morning, from TrainingPeaks. "Laps 2-6" is the average of the per-lap NP for those laps.
Both Roanoke days used the same course: same footprint, 6 laps each, 18.3 km and 18.1 km. Lap times are only comparable within a weekend.
Lap by lap in Roanoke
Sunday's lap 5 is the tactical lap, where neither rider wanted to lead. Laps were split from GPS passes through the finish area.
Why Saturday and Sunday looked like two different riders
In order of evidence:
- Lap-1 pacing.
- On Saturday she went with the leader: 250 W for the first two minutes and 245 W for the lap. That is about 10% above the roughly 220 W she could sustain.
- It came from a cold start (98 bpm at the gun), which makes a hard first lap even more costly.
- The bill came on lap 2: NP down 11%, and every lap after that slower.
- On Sunday she started soft (189 W for two minutes), rode lap 1 at 228 W, and held that level all race. Leading let her set a pace she could hold, and the games on lap 5 gave her a breather before the finale.
- This is the same lap-2 fade seen at Rochester, triggered the same way.
- Heat and sun.
- Saturday's 2:54 PM start was in full sun at about 27°C. Sunday's was under cloud at 26°C.
- Direct sun adds a large heat load that the air temperature doesn't show.
- Her heart-rate cost per watt from lap 2 onward was about 3-4% higher on Saturday.
- She has been in Alberta since late August, so she is not heat-acclimatised.
- "She looked tired" is consistent with heat strain on top of an over-paced first lap.
- No priming before Saturday.
- Thursday was a complete zero (travel; the planned opener, run and core weren't done).
- On Friday's pre-ride she hit 814 W for 5 seconds, but her best minute was 254 W, short of the 260-286 W openers.
- So Saturday's race was her first real race-intensity effort in three days, and it acted as the opener for Sunday.
- Both weekends this season show the same pattern: Sunday beat Saturday at Rochester (214 W vs 208 W) and at Roanoke.
- Cold starts, both days.
- Her heart rate at the gun was 95-98 bpm (116-126 at Rochester), after about 10 minutes between the end of her warm-up and the start.
- That doesn't explain the difference between the days, but it made Saturday's hard first lap more expensive.
What it was not:
- Not accumulated fatigue: TSB was +9 to +10 both mornings.
- Not a lack of fitness: the same total power won on Sunday.
- Not the power meter: both race files recorded normally.
Unknowns that could also contribute (see Questions): sleep on Friday night, Saturday's breakfast and fluids, how long she waited in staging and in the sun, whether she used a bottle, and why Sunday's laps were about 20 seconds faster at similar power. That could be course conditions, drafting behind Bakker, or the cooler air.
What Roanoke says about her fitness
-
She is responding fast. From Rochester to Roanoke:
- NP up 5-7%.
- Surge time up 50-100%.
- Laps 2-6 up about 17-18 W.
- HR cost per watt improved about 7-10%.
This happened in a week that was mostly travel, a pre-ride and racing.
- Her best race of the season came at CTL about 65 and TSB about +9. Rochester, at TSB +20 to +25 after a disorganised lead-in, was flat. Her racing sweet spot looks like TSB +5 to +10: fresh but not over-rested. That supports Sidney's instinct: a little freshening before targets, not a long taper.
-
She is still short of World Cup level:
- Race NP 223-224 W against 241-254 W at her best World Cup rides last January.
- Best race minute 314-354 W against 375-445 W.
The gap is closing, and it is closing through race sharpness as much as through CTL.
Sidney's concerns
Her notes: five very hard weekends coming up; don't overload during that block; Major Taylor could be a light race; a little taper into Nationals; and a worry about arriving in Europe overcooked.
The numbers support her. The published plan's load, projected with the same per-session estimates used below:
Where the risk sits:
- Five race weekends in six weeks.
- Travel every Monday and most Thursdays.
- Hashimoto's, and a recorded history of struggling to recover.
- The parts of the published plan that carry the risk:
- 2:30-3:00 Tuesday rides two days after back-to-back races.
- Endurance tails added to Wednesday's hard sessions.
- Four Sunday top-ups.
What I would keep:
- The key interval sessions. They are what's moving her race numbers.
- The October 5-11 block, trimmed. It's the only week without a race.
Recommended revision
Every key session stays. The cuts come from volume in race weeks, from the October block, and from the days before Nationals, PanAm and Ostrava. Days not listed stay as published.
Runs stay at two a week. Hill run-ups drop to 5-6 reps in the Major Taylor, Nationals and PanAm weeks.
Projected load, published plan vs recommended
Weekly TSS:
Race-morning fitness and freshness:
These are corrected CTL values, with the August duplicate uploads removed; TrainingPeaks currently shows about 3.5 points more. Races are counted at 105 TSS per race day including warm-up, with Major Taylor Day 2 at 80 and Hardwood at 85.
What the revision costs, stated plainly: about 8 CTL points at Ostrava (71 instead of about 80). That is roughly last season's number at the first World Cups (72-75). The differences from last season:
- Fitness direction. Last season it was falling (from 93 after an October spike); this season it is flat to rising.
- Freshness is managed. Race mornings sit at about TSB +5, not a deep hole followed by a collapse.
- She is already racing better at lower fitness. Race power is already 223-224 W at CTL 65. Last December she raced at 204-223 W at CTL about 70.
The rise in fitness moves to Europe.
- December 1-12: the race-free window between Tábor and Glasgow. A roughly 700 TSS/week block there should take CTL to about 76-78 for Glasgow and the December rounds.
- December 29 to January 17: can take it into the low 80s for Benidorm, Hoogerheide and Worlds.
That is the realistic top-10 window. It needs planning before she flies.
Load dial
The revision is a starting point, adjusted by what the next two weekends show.
Race execution changes, starting at Charm City
- Cap lap 1.
- Commit fully for the first 60-90 seconds to win position.
- Then settle so lap 1 finishes no more than about 5% above her target race NP (about 235 W right now).
- If the leader goes harder than that on lap 1, let the gap open and close it on laps 2-3. Sunday proved she can match the leaders when she paces evenly.
- Arrive at the gun warm.
- Stay on the rollers or trainer until call-up.
- Do 2 x 20 sec hard efforts at about 10 minutes and 5 minutes before the start.
- Target a heart rate of 120-130 at the gun, not 95.
- Heat protocol for any race at 22°C or warmer, or in full sun:
- Wait in shade, with a cold towel or ice vest in staging.
- A bottle of ice slurry 30 minutes before the start.
- 500-750 mL of fluid with 500-700 mg of sodium in the two hours before.
- A bottle hand-up from the pit if the commissaires allow it.
- Check the Charm City forecast on Thursday.
- Never a zero day before a race weekend.
- On Thursday travel days, ride at least 20-45 minutes with 3 x 1 min openers, even on a hotel trainer.
- On Friday, hit the pre-ride openers at their targets (3 x 1 min at 260-286 W).
-
Use start targets that fit the course. "Average power for the first 60 seconds" depends too much on course layout: Roanoke's start dropped to coasting within 30 seconds. Use these instead:
- Best 15 seconds at the gun.
- NP for the first 2 minutes.
- Position at the first corner.
I'll update the race event targets when the revision is applied.
- Debrief every race within 24 hours: placing, gap, start row, sleep, breakfast, fluids, conditions, and how the legs felt.
Questions
- What was Saturday's placing and gap, and Sunday's winning margin?
- Saturday: how did she sleep Friday night? What did she eat and drink before the race? How long did she wait in staging, and was it in the sun? Did she take a bottle?
- Was Sunday's course the same, and was it drier or faster? Her laps were about 20 seconds quicker at similar power.
- Did she ride at all on Thursday (September 24)? No file exists.
- Hardwood (November 1): race it controlled for points, race it fully, or skip it to protect PanAm six days later?
- Approval to apply the revision. On a yes, I'll update the calendar files and the matching TrainingPeaks workouts in place (by their IDs), and remove the three Sunday top-ups. Tomorrow's session (Tuesday, September 29) is the first change. If it can't be updated tonight, Sidney should ride the 2:00 version with 2 x 12 min sweet spot.
Data and process notes
- Race files now have clear names ("Go Cross - Race", "- Warm-up", "- Pre-ride") and power recorded on race days. That's a big improvement.
- Warm-ups on the trainer are still recorded without power.
- The September 26 daily log again records a system heartbeat prompt as a "Message from Sidney", the same routing bug as September 23. There is still no direct athlete input for the weekend: no sleep, food or race notes.