Wednesday, May 26, 2010

Zoncolan power calculations revised

Basso on ZoncolanBasso climbing Zoncolan (CyclingNews)

Last post I proposed an estimate of the power used by Basso and Evans while climbing the mighty Zoncolan stage of the Giro d'Italia.

However, no sooner did I publish my estimates than CyclingNews published claimed numbers from Basso's SRM. That Basso had an SRM was surprising to me: this stage is one of the rare days in pro racing that riders use a 110 mm BCD "compact" crank. In the past, they've not made the investment (or their sponsors haven't made that investment) of a separate power-enabled 110 mm spider. But 110 mm BCD power meters are gaining increased acceptance as people realize that "compact" gears are suitable for racers, too, in addition to the substantial power meter market of performance-oriented cyclists who don't race. So they probably want to promote their compact option, and a epic stage like this one attains substantial visibility.

In any case, the numbers: Basso averaged 395 watts for the final 10.1 km, for a claimed 5.68 W/kg, implying a body mass of 69.5 kg. I'd calculated an average power of 6.08 W/kg. Ouch!

Let's check assumptions. I assumed a drivetrain efficiency of 0.97, which is considered typical for a derailleur system in good working condition. However, drivetrain efficiency is generally better at higher powers, and the pros ride at exceptionally high power, so I'll increase this to 0.98.

Another assumpption: I assumed a coefficient of rolling resistance of 0.5%. Assuming the road was in excellent condition (as is usual for grand tours) and the tubular tires used by the riders are relatively low rolling resistance, this is probably too high. I'm willing to reduce this to 0.3%.

What about bike mass? I assumed equipment + bike was 1/8 the total weight of the rider. Assuming Basso is near the 69.5 kg implied by that article, that's 8.4 kg. Well, silly UCI rules limit the bike mass to no less than 6.8 kg. So that leaves only 1.6 kg for helmet, clothes, contents of pockets, water bottle, etc. So there's no room to reduce this estimate.

Then there's wind resistance. I'd assumed CdA = 0.3 m² / 50 kg body mass. For Basso, this would yield 0.42 m². Evans was standing a lot during the climb, Basso less so. I'll reduce that 10%, to 0.3 m² / 55 kg, yielding 0.38 m² for Basso. Lower than this seems unreasonable. An excellent time trial position is around 0.23 m². When climbing steep hills riders tend to be relatively upright, far from an optimized time trial position, on bikes designed for handling rather than for aerodynamics.

Then there's drafting. Basso didn't pull the whole climb, certainly not at the bottom, and Evans even took a few pulls before he was dropped. I'll assume that Basso drafted half the distance, and while drafting, wind resistance power was reduced 30%. This reduces wind resistance power by 15%.

So I ran these numbers: the result was 5.76 W/kg, of which 0.24 W/kg was from wind resistance, 0.35 W/kg from rolling resistance, 4.52 W/kg from climbing. The disagreement is down to 0.08 W/kg, assuming zero wind, and neglecting the pushes Basso reportedly received from enthusiastic tifosi.

But what about that body mass? To resolve this difference I'd need to assume Basso weighed 68.6 kg, an error of 900 grams from his claimed weight. Dehydration during the stage could easily account for that.

So with these numbers, "theory" and "experiment" seem to agree rather well. Using the same assumptions, Evans averaged 5.56 W/kg for the stage. Barely good enough to win a Low-Key Hillclimb, let alone set a record up Old La Honda. But at the end of a gruelling stage, in the midst of a gruelling stage race, you'd expect some signs of fatigue to be showing.

In any case, it's clear that those who've accused Basso or Evans of doping in the Giro are unfounded. These power numbers don't come even close to some of the numbers we've seen in the past, including from Basso himself. See, for example, the Science of Sport blog post on the subject.

Monday, May 24, 2010

running the numbers from the Zoncolan (2010 Giro d'Italia stage 15)

Yesterday's Giro stage finished on the epic Zoncolan climb, a truly brutal slope which I've vowed to climb myself some day. If numbers alone can induce pain, these numbers surely qualify:

Zoncolan profileZoncolan profile: ouch!!!

I watched the last 50 km on Justin.tv (linked via CyclingFans): absolutely riveting stuff. Sure, L'Alpe d'Huez has its character and history, but the mighty Zoncolan is in a separate category of "beyond". With the pathological gearing used by riders on the stage (for example, 34-52 front with 11-29 back), riders were at least able to maintain a decent cadence, but that couldn't hide the brutality of the slopes. This hill is truly every rider for himself.

Given this, it's interesting to see what the numbers say about what transpired on the mountain. I took a few time checks from the main group (ignoring the break) starting at the 10 km to go point. The climb "officially" begins 100 meters sooner. But it didn't appear to me there was much of a grade, if any, in these first 100 meters. I estimate the pack could cover that in around 8 seconds, time which should be considered for comparing times reported for the climb.

From this pack, it was initially a war of attrition as one rider after another fell off the pace. Finally, it was just Ivan Basso, Cadel Evans, and Scarponi. Scarponi was the first to fall back, while Evans bravely struggled on, fighting to avoid showing Basso any weakness. But finally Evans, too, was gapped off, leaving Basso to solo in for his greatest victory since his two year suspension from the Puerto blood bank affair. Evans regrouped a bit, regained Basso's pace temporarily, but then lost more ground approaching the finish. Scarponi almost recaught Evans, but not quite, finishing third.

Here's my data from Evans and Basso:

Basso:
kmsecsgapmetersgradekphnet kphVAMnet VAMW/kg
000530      
512170113212.0414.7914.79178117816.10
717830144115.4512.7214.13196518396.56
922310164510.2016.0714.52163917995.74
102437017308.5017.4814.77148517735.36

Evans:
kmsecsgapmetersgradekphnet kphVAMnet VAMW/kg
000530      
512170113212.0414.7914.79178117816.10
7181734144115.4512.0013.87185418056.18
9229261164510.2015.1614.14154617515.38
1025167917308.5016.0714.31136617174.87


The key numbers are probably "VAM", the rate of altitude gained between time checks ("net VAM" is the rate from the 10 km point), and "W/kg", which is my estimate of the power delivered by each rider to his pedals. To estimate this, I assume extraneous mass (bike, clothes, etc) was body mass / 8. I then assume CdA = 0.3 m² / 50 kg body mass. Further, CRR = 0.5%, mass density = 1.1 kg/m³, and drivetrain efficiency = 97% . I assume uniform grade between "time checks" (this affects only wind resistance, which is small and uncertain anyway). Daniel Coyle says in Lance Armstrong's War that a rider needs to be able to sustain 6.7 W/kg to win the Tour. During the EPO era, this has certainly been typically the case: power estimates during sustained climbs tend to be close to this. For example, Alex Simmons estimated Contador delivered 7.03 W/kg climbing Verbier during last year's Tour de France.

So first it's clear neither of these guys were climbing to that standard, which is a nice rebuttal to those who claim Basso must be back on "the program". Indeed, Gilberto Simoni, who's clearly not on the program this year, was able to climb the same hill with a VAM of 1850 m/hr in 2007, for which I'd calculated a power-to-mass ratio of 6.28 W/kg. Even that power is below the Contador/Armstrong standard of "excellence".

But these numbers also are telling tactically. It appears Evans made a mistake trying to follow Basso's unsustainable pace from 5 km to 3 km remaining. Evans would have been better off riding his own pace, then managing or even perhaps closing the gap which Basso was able to open. You can see Basso faded toward the finish; Evans, however, faded even more. Evans should have practiced the patience which won him Fleche-Wallone, where he let Contador have a gap on the final climb of Mur de Huy, only to close that gap then pass Contador for the victory. Perhaps Evans overestimated his strength.

Sunday, May 23, 2010

Powertap Torque Test: Post-Service

Saris sent me back my powertap wheel after I'd sent it to them for torque-tube test. Hats off to Saris: they did it w/o charge and returned it to me quickly. But had they actually done anything?

Unfortunately I didn't follow up their quick return with a quick re-test. Honestly, I found it liberating to not have power numbers telling me how lame I was every single ride. It was fun to just ride. I didn't need the power meter to tell me if I was going hard or not.

But eventually I couldn't ignore it any longer. So time for the torque test, again.

My procedure was slightly modified, after what I'd seen a friend do with his. To optimize the crank orientation (the weight needs to hang perpendicular to the crank or the torque estimate will be off), I grabbed the rear wheel and rotated in, forward and backward, until I saw the Powertap torque reach a maximum value. But here I deviated from my friend's approach: instead of using this number, I grabbed the brakes, let things settle out, then used that number. So my data taken here, like data previously, were done with the brakes holding the wheel. My reasoning was I wanted to make sure the wheel was stationary, and the brakes could do this better than my hand holding the rear rim.

Okay, cut to the chase:
Powertap retestRetested powertap, compared with previous test of my wheel and a reference ("C") wheel.

So it appears as if nothing has changed.

To test whether either wheel significantly deviates from the optimal curve, I did a t-test:
========== My wheel ===============
pre-service
Correlation coefficient = 0.110
Regression coefficient (SLOPE) = 0.008
Standard error of coefficient = 0.026
t - value for coefficient = 0.294
Regression constant (INTERCEPT) = -9.222
Standard error of constant = 2.829
t - value for constant = -3.260
post-service
Correlation coefficient = -0.061
Regression coefficient (SLOPE) = -0.004
Standard error of coefficient = 0.025
t - value for coefficient = -0.173
Regression constant (INTERCEPT) = -6.143
Standard error of constant = 2.769
t - value for constant = -2.219

====== Reference (C) wheel =========
Correlation coefficient = 0.300
Regression coefficient (SLOPE) = 0.011
Standard error of coefficient = 0.013
t - value for coefficient = 0.833
Regression constant (INTERCEPT) = -0.259
Standard error of constant = 1.596
t - value for constant = -0.162
The t-value is a test for significance: if it's less than 1, the
result is generally consistent with random variation combined with the
actual value being zero.

So on the reference wheel, the t-value for both the offset and the deviation of slope from 1 have low t-values. This is what I'd expect, assuming all torque is transferred to the hub from the crank. And Andrew has already shown this tends to be the case by directly applying torque to the cassette.

On the test wheel, the t-score for the offset is well in excess of 1 (the slope t-score is small, however). This is consistent with there being a fixed torque error in the hub. The two tests measured different offsets, but the separation between measured values is comparable to the standard error of either estimate, so there is no
statistical proof anything changed with an the offset. However, the probability of a wheel without an offset producing a similar set of results is very small.

So the conclusion? Saris didn't change anything. But beyond that I don't know. Next up: I need to get the wheel to someone with a spider-based power meter. A comparison with that, especially on a relatively steep climb, ridden with a smooth cadence at low speed, will be telling.

Anyone want a deal on a used Powertap?

Saturday, May 22, 2010

Creative Labeling

After Skyline to the Sea, amidst all of the Cosco food provided by PCTR which generally falls into the "hydrogenated fat + corn syrup" category which I prefer to not put into my body, especially when trying to repair the considerable damage from such a hard effort (a recipe is only as good as its ingredients, and my recipe for me is what I eat), I found something which for me saved the day: small protein-fortified balls by Clif called Shot Roks.

Obviously they were sponsor products, not part of the Cosco run. They were mostly sitting untouched: honestly, they looked fairly threatening. Little hard balls. But after I tentatively tried one, I had another, and another, then two more, then... good stuff at the end of a very depleting day, and the protein was just what I craved after the long run.

So what's in Shot Roks? Here's the listed ingredients:

ingredients
INGREDIENTS: Clif Protein BlendTM (Whey Protein Concentrate, Milk Protein Isolate, Whey Protein Isolate), Organic Evaporated Cane Juice Crystals, Organic Evaporated Cane Juice Syrup, Glycerine, Inulin Syrup (Chicory Root Extract), Organic Peanut Butter, Oat Flour, Natural Flavors, Organic Tapioca Maltodextrin, Salt, Organic Corn Starch, Confectioners Glaze, Soy Lecithin, Colored with Caramel Color and Beta Carotene, Carnauba Wax.

Now, the curious thing here is #1 in the list is "protein blend". Ingredients lists must be provided in the order from highest to lowest fraction by mass. So obviously there's more "protein blend" than anything else, right?

Well, not really. After "protein blend" is a series of carbohydrate sources: "Organic Evaporated Cane Juice Crystals" and "Organic Evaporated Cane Juice Syrup". Cane juice is cane juice, right? Then glycerine (or "glycerol"), an alcohol which is counted on ingredients lists (somewhat controversially) as a carbohydrate (controversial because it's metabolized differently than sugars, and in fact commonly causes gastrointestinal issues for many people). Then after that, inulin syrup, another sugar source.

Curious, isn't it, that the protein sources are all grouped together in a "blend", but there's no "carbohydrate blend".

The clue is in the nutritional information. The protein sources are essentially all protein, the carbohydrate sources essentially all carbohydrate. So the ratio of carbohydrates to protein listed there doesn't lie: 38 grams of carbohydrate versus 20 grams of protein. So despite the labels, the protein sources clearly are not the most common component: it's more carbohydrate than protein.

But that's not fair, really. Included in the carbohydrate total are non-digestible fiber (3 grams) and 8 grams, including the glycerine, of "other carbohydrates". That leaves 27 grams of sugar.

So a fairer listing of the ingredients would be something like the following:

ingredients
INGREDIENTS: Clif Sugar BlendTM (Organic Evaporated Cane Juice Crystals, Organic Evaporated Cane Juice Syrup, Inulin Syrup (Chicory Root Extract), Organic Tapioca Maltodextrin, Confectioners Glaze), Clif Protein BlendTM (Whey Protein Concentrate, Milk Protein Isolate, Whey Protein Isolate), Glycerine, Organic Peanut Butter, Oat Flour, Natural Flavors, Salt, Organic Corn Starch, Soy Lecithin, Colored with Caramel Color and Beta Carotene, Carnauba Wax.

The sugars are front and center, followed by the proteins, followed by the alcohols, followed by the "real food" (peanuts and oats) and glue (ingredients included for their mechanical properties, like wax and color).

So after a trail marathon, there's probably nothing wrong with a nice dose of sugar, as long as it's from a less refined source such as evaporated cane juice (as opposed to bleached white sugar, or high-fructose corn syrup). Just don't get the idea these are primarily protein, despite the deceptive ingredients list.

Wednesday, May 19, 2010

Tour of California Stage 3 spectating

I rode over Hwy 84, first misty, but then a steady rain at the summit, near Skylonda (but technically Woodside). 84 is normally too heavily trafficked to be popular with local cyclists, except as a descent where good riders can outpace most cars. It's the easiest route up to Skyline in the area: Old La Honda Road, Kings Mountain Road, and Page Mill Road (in order of increasing difficulty) all much-preferred. TOC tried to get Page and OLH, but residential resistance meant they had to go with 84. In any case, I rode 84 today in the spirit of the race.

Fortunately, the skies cleared as I descended to the west, a long gradual downhill into the town of La Honda. after passing through, a left onto Pescadero at the "5 km to KOM" sign took me into the redwoods.

A short rise, a bit of flat, then a right over the bridge marking the climb proper of Haskins Hill. Here it was chilly, the roads a bit damp, but thankfully no rain. This climb is the highlight of the annual Pescadero Road Race each June, so I think about that race every time I climb it. Really, really nice.

Tunitas summitThe pack crests Tunitas on way to Haskins Hill (Franz Kelsch)

At the top I waited for the riders. It was a bit longer than I'd anticipated: the pace was obviously restrained. But eventually, they came amidst the usual motorcade: first the break, then a gap of around three minutes, then the peloton taking up the full width of the road. At least the break seemed determined: the pack was at a conversational pace. Armstrong, in fact, was riding next to the Radioshack car, conversing with someone inside, his teammates mostly at the front of the pack chatting together.

After they passed, I was starting to shiver from the chill, so I headed back down the way I'd come. As I descended a lone Spidertech rider was climbing the hill, well behind the broom. I wonder if he finished.

Going back to my office I climbed Alpine Road, descending Page Mill. Both absolutely spectacular roads with amazing views. It's too bad the race couldn't have passed that way. Still, I'm grateful it was able to even get the route it got: cycling events always face resistance on the roads surrounding Woodside, CA.

I'd always known the real action would begin on Bonny Doon. And this turned out to be even more true than I'd predicted: not even a chase group. Just the leaders moving along, the pack moving along, and lots of cars. I didn't even get any photos out of it: I'd brought a good camera, but had forgotten to insert the memory card. So was it a mistake to watch from Haskins Hill, rather than ride down to Davenport for the final, critical climb? I don't think so. I saw them on my home turf, in a sense. These are roads I've ridden many, many times, roads I know very well. That's really cool. I felt like I was sharing something with the international pro peloton. Letting them in on the secret which is the spectacular beauty of the roads in the open space and parkland of the otherwise crowded and sprawled San Francisco Bay area.

Monday, May 10, 2010

Berkeley Hills Road Race: Working Reg

Yesterday was the first time I've ever worked registration at a sanctioned race. Berkeley Hills is a classic Nor-Cal course: San Pablo Dam Road, Castro Ranch Road, Alhambra Valley Road, and Bear Creek Road provide almost idylling riding in the hills east of Berkeley. The route is highlighted by the "Three Bears", three short but fairly steep climbs along Bear Creek Road. From the south, first Mama Bear (where the race once finished), then two bumps of which the second is generally considered to be "Baby Bear", then a descent over the bridge @ Happy Valley Road to the final climb, Papa Bear, at the top of which the race now finishes. For laps preceding the last, following Papa there's a fast, straight descent to another bridge and a short but steep climb to San Pablo Dam Road. Then it's back around the loop and another encounter with the Three Bears. The race is a war of attrition: survive a seemingly endless series of short hill intervals until the final, slow-motion sprint to the line. There's no faking your fitness here.

2010 Berkeley Hills RRMike Nachtwey photo from the road

Each of the past two years I raced E3 here. In 2008, I found myself off the front following the neutralized climb out of the start area. Very strange the riders didn't even seem sufficiently motivated to stay with the lead motorcycle's modest pace up the climb. I decided to take the gift and run with it. But the fun lasted only a lap and a half, and then I was caught and passed, my reserves spent. I forced myself to continue and get a finishing place, despite the temptation to pack it in.

In 2009, with less fitness I decided to ride smarter. Despite my lack of preparation due to a series of colds and allegy problems in the winter and spring, I was surprised at how well I was able to hold the pace. Until, that is, on the second lap when a rider went down in front of me and I crashed. Rather than continue on with bruised ribs and a potentially damaged bike, I spent the rest of my day in the nearby feed zone, handing bottles to teammates.

This year, with my focus on running in preparation for Skyline to the Sea, I was even further behind in my cycling-specific fitness. My club was co-sponsoring the race with the Berkeley Cycling Club, and so my responsibility, unless I was in top form, was to support the race rather than participate. So I volunteered for registration.

Due to the toxic influence of urban sprawl, the Berkeley Hills Road Race is forced to start bright and early @ 7:30 am. That means registration volunteers show up dark & early by 5:30 am.... After awakening at 3:50 am, I first ran to meet my carpool, then we drove to pick up another volunteer, then onward to the race start.

It was fun, and went a lot smoother than I'd anticipated. A few race-day snafus: a lack of pins and a lack of rider release forms had to be at least partially rectified with an emergency early-Sunday run into Orinda. But my group, 35+ 1-2-3, was fantastic. These guys all knew the drill: every single one showed up with license out and most even had their own pins sparing our unfortunately limited supply. Even though other categories tended to arrive earlier, registration of my master's group went almost without a hitch. Well, almost: I accidently swapped the numbers of two riders with the same last name. But a quick visit to the laptop guy took care of that, swapping their numbers in the database.

Afterwards, I finally got to see Don Becker's built-up Ruegamer Überlight, a true work of weight weenie genius. The bike is simply amazing, making my own racing bike (weenied out in its climbing wheels) look like a boat anchor. I couldn't help but notice his Aerolite pedals had a wonderfully hollow bore: they'd be perfect candidates for the Metrigear Vector.

Don BeckerDon Becker with his Ruegamer Überlight, at 2009 San Bruno Hillclimb

My duties done by 9am, I set off for home, stopping along the way to watch the riders pass. Intermittent rain eliminated any envy I might have had for them. But eventually I moved on for good, running up Wildcat Canyon to Shasta past Grizzly Peak then eventually down to Euclid, Hearst, UCB, and Berkeley BART. From the Civic Center station to home brought my net running for the day to 11.5 miles. For me, that's solid, and combined with the early wake-up, I was tired.

But it will be fun to be on the racer side of the reg table again. Hopefully at the Sierra Road KOM Hillclimb (if it happens) in conjunction with the Tour of California Stage 4 starting in San Jose. I just need my body to give me a hint it's ready for a hard climbing effort. I'm not sure the long run is going to help towards that, but at least it was fun.

Wednesday, May 5, 2010

height vs mass for sub-27 minute 10 km times

Chris Solinsky
Chris Solinsky

In comments to the Science of Sports blog post on Chris Solinsky's remarkable sub-27 minute 10 km time, data on the height and mass of those who've managed to break that barrier were listed. It's interesting to see how the numbers scale:

height vs mass for sub-27 min 10 km runningheight vs mass for sub-27 min 10 km running times

While constant BMI certainly describes the upper bound of these data, at the lower margin, the trend is closer to a constant ratio of mass to height. It was proposed a factor may be the ability to dissipate heat, which becomes more difficult for larger runners than smaller runners at a fixed BMI.

For fun, I also plotted my numbers, and while I'll never run close to a 27 minute 10 km (40 minutes would make me happy), my dimensions certainly fit in nicely with this super-elite crowd:

height vs mass for sub-27 min 10 km running + DJCI added myself to the plot

Tuesday, May 4, 2010

Giro d'Amsterdam begins Saturday

It's amazing to me all three of the Grand Tours within a 12-month window begin in Holland. I'd love to go to Holland, especially Amsterdam, and take in the refreshingly bike-friendly culture in which the need to slow down cars to protect public safety is well established (in stark contrast to the "drivers own the road" mentality in good-old-friendly U-S-A). But from a racing perspective, the convoluted roads and complex network of traffic circles make racing there quite a challenge. Challenge is good when it's a test of skill, but with the Russian Roulette of a peloton surviving an almost endless series of traffic circles in the final kilometers of a win-at-all-costs pro bike race, not so good unless you're into random blood-letting.

So here's the profiles for the first three stages of the Giro d'Italia, from CyclingNews, each stating in Amsterdam:





Heck, I can't ride from home to the local market (in San Francisco) without doing more climbing than two of these stages, and a ride across this compact city is more climbing than all three combined.

Saturday, May 1, 2010

quick review of California state propositions

This blog is mostly about bicycles, and I ride my bike to the polls in California several times per year, so anything having to do with elections also qualifies.... California's primary election is 8 June this year: less than 6 weeks away. A quick scan of statewide ballot measures:
  • Proposition 13: Definitely no. Property owners should pay a tax proportional to the value of the property. Seismic integrity is an important aspect of value. Additionally, putting loopholes in the tax code allows owners to classify upgrades as "seismic retrofitting" even if they contribute to other aspects of a building's value.
  • Proposition 14: Definitely yes. The present system of party-based primaries is polarizing and inertial. I think both parties benefit from a broader base of input. For example, fiscally conservative yet socially progressive voters can contribute to less dominance of the wacko fringe in the Republican party, while the Democratic candidates will be less smug in their support base.
  • Proposition 15: Based on who's supporting this one, I'm leaning positive. But public financing of elections is a complex matter. I'm not committing, but if I had to vote now, it would be yes.
  • Proposition 16: Absolutely no. Requiring 2/3 majority for anything short of constitutional ammendment is a farce. Give democracy a chance.
  • Proposition 17: No way: I couldn't possibly be more against this one. Don't punish those who go car-free and are therefore able to unplug from the insurance grid.