After dinner tonight, Allen Lim gave us a lecture; I won't say that I'm summarizing it, but I'm rattling off some things that seemed sort of relevant to me, or likely to be of interest to others.
Main messages, distilled into a teeny-tiny summary for Freds.
- Use the powertap to quantify feel.
- Understand effective training cycles.
- Let the feel guide the training.
(Different messages apply to those beyond this stage.)
Power and SpeedOne definition of power is all the drag forces multiplied by the velocity -- rolling resistance, aero drag, mechanical loss, etc; [Aero drag is itself velocity sensitive, giving us an exponential term -- actually, it's velocity squared, giving us a net velocity cubed term].
Because there is very little difference in power output at elite levels, other tunings become the focus of attention. Aero is very relevant -- the "praying landis" TT position was so much better aero, that the gains it gave in the TT in the Tour of California were pretty much exactly Landis' margin of victory there. Without the goofy position, maybe he doesn't win. Of course, there tactics and race dynamics there too, so it's not quite that simple.
Raw PowerPower input is pedal torque * cadence, and the first main difference between elite cyclists and amateurs is that they push with higher torque. Actually, Lim sort of contradicted himself later. At high power outputs, they also use higher cadence. Where at 200w 75-80 rpm is OK, at 400-500 watts, it needs to get to 110 rpm. There's a physiological reason for this. There's an artery in the leg that gets clamped shut ("occluded") at high torque, stopping blood flow. At low cadences, this shutoff is long enough to have it's own negative effects. At higher cadences, it presents less of a problem. One adaptation seen in some athletes is development of the vessels in the legs that keeps them open at higher levels.
EnergyThe PT computer displays an 'E' reading that is in Kjoules, being mechanical energy as an integration of power over time. There's a conversion to Kcalories of food that goes like this: 1 Kcal = 4.186 Kj, but the body is somewhere between 18-27% efficient, with typical in the 22% range. This is affected by genetics, state of training, and cycling efficiency. 4.186 *22% is close to one, so the conversion is typically cited at 1.086 Kcal per Kj, or about 110% rounded.
One of Landis' genetic jackpots is that he's been measured as 25% efficient in base training, and up to 27% efficient at the start of the '07 tour.
Other factors affecting the conversion rate are cold stress (shivering), whether you are in steady state or doing bursty intervals, cadence, and training level. To burn more calories, do intervals and less steady state with the same energy net: eg, 100w/300w/100w/300w should burn more calories then 200w/200w/200w/200w.
Trying to burn fat while bonked doesn't work -- you need some blood carbs to prime the reaction.
As a side note, Lim doesn't much believe in the "lactate burn" producing performance loss. He thinks it's more an electrolyte imbalance (sodium) causing a localized inability in things transferring through cell walls.
Power MeasurementLim has a picture of a 1905 device that looks close to a modern lab ergonometer, from an
Atwater-Benedict study. "This isn't new with the powertap." What's new is getting it out of the lab and onto the road.
He made a bunch of explanations why heart rate wasn't as good a tool as people once thought, because it doesn't map in the real world as well as it does in the Lab. The published studies show that HR correlates very well to power in the lab, but observation shows that it does not in the real world. It's better to realize that the power is the input variable, and the heart rate is a function of the power (with its own physiological terms, some of which have time state).
Among the physiological terms is "cardiovascular drift", which is where the same power output has different HRs depending on fatigue, and other factors.)
The Popcorn Analogy of Training
This memorable image falls apart a little in the detail, but it's pretty good as a model. You have a certain amount of potential capabilities, imagine them as popcorn in the bag. You apply some training load (heat and time), and you get a certain output (heart rate, popped kernels.) If you apply the right heat at the right time, you get all the kernels popped. If you have too little time or heat, you get unpopped ones. If you go too long, you start overcooked, and eventually get burned and singed kernels, and a broken down overworked athlete.
The great trick in training that that it's hard to tell how many kernels (genetic talent) there are in the bag when you start. With things like power measurement, you can control the training load, and observe the corresponding output HR.
Training LoadMeasures of load are simply (there are more complicated forms):
duration * intensity * frequency * mode
I don't recall that he explained the mode -- interval, steady state, I think, but there weren't example values that seemed to make sense. If steady is 1.0, is interval 1.5?
Intensity can be measured in a number of ways -- as power, or other units. One crude approximation is known as "Foster's index", which takes a reported perceived exertion (RPE) on a 0-10 scale, times minutes. Another measure would be average HRM * minutes, which is called the "training impulse" (TRIMP).
"Periodization" of TrainingAllan admits the term is made up, but it is used in the field. If it were me, I might call it "quantization" in the quanta sense. It's the description of a varying workload over time, with the goal of focusing on different goals -- and explains "peaking" for the first time I've ever really heard in a clear quantitative sense.
At some point I may get his graph, or draw my own, but I'll have to describe it here. On an X-Y chart, we have the X axis as days, and on the Y axis we'll have training load in Watts (kJ?). He drew a curve, but I'd draw histogram bars instead.
On top of this histogram is a two-week moving average of the workload as a curve, and above that is a curve representing the top 2.5% peaks for the same days. (Why 2.5%? not clear; same scale as average suggests watts instead of kJ).
Days where the load is greater than the mean are "overload" days; those below the mean are "recovery" days. The general training model is: overload your training until you cannot maintain the overloaded work; then recover until you can to the overload again. So, a training period will have some number of days of overload, followed by some days of recovery, then the cycle will repeat.
You can overlay a cosine-like (sine starting down) over this, with some phase offset to the right -- say that the overload cycle starts 75% up It isn't really a simple wave, an he didn't explain the function that I recall. I also didn't get the labels clearly, and some are missing in this description. One point in the downward cycle is called "involution", and another on the upward side was "compensation". It did appear that there was a point in the upward curve that counted as the legendary "peak" in the training cycle, late in the recovery phase that one would might want to align with an event date. Or maybe I'm understanding it wrong, and you want the inverse of this curve to define the peak.
There's a missing line in his chart that turns out to be important. Imagine that there is a third plot line that is 2 standard deviations up from the mean line. Lim claims that is the effective limit to your training load peaks -- if you train with peaks harder than that, you will crash, almost immediately.
Finally, the chart was an example of the weekend warrior, who trains hard on the weekend, but little during the week. It showed little effect on the two-week average, and the peaks were way higher than the 2-std deviation rule. This is not effective, Lim says.
For workload increases, Lim recomends no more than a 10% change in one cycle.
What to do when recovering? Well, first stop:
"You don't get strong when you train.
You get strong when you rest."
Sitting on the couch is not bad. Landis is legendary for sleeping all day, and part of the next.
Getting back to the bullets, one simple way to tell when it's time to stop is to see how you feel with light load. If at 100w you feel great, train hard; if you feel crappy, recover instead. Recording this is good for training diary is good, too.
Perception is not
always the same as reality, however. At the 2005 TdF, Landis started feeling crappy and might have abandoned, except that he could see the output and observe that he was actually still putting out the watts in a reasonable amount. He notes there is a critical point 10 days into a stage race where the body starts doing these things, and that this is where things go funny.
The big difference between training loads for Landis in 2005 and 2006 was that he had a lighter load in 2006, and was more relaxed about the process. Floyd describes this as "I drank more beer." In reality, it was limiting the amount of overloading he do.
For example, in 2005, in Spain, Landis went out one day and just did himself in. He went to bed at 9pm, and told Lim to check on him and if he wasn't up in a couple of days, to check to make sure he was still alive. The next day, Landis was asleep in the morning, and Lim went out for a ride. Coming back for lunch, still zonked. At 4 pm, Lim freaked -- maybe Floyd
was dead. What would he do in Spain? Who would he call. He knocked on the door, and heard some noise, followed by a screech:
I'M STILL ALIVE!
Strength/Weight TrainingLim says it can help, but off-season only isn't much good. Perhaps every other day off-season, and every 3 days on-season and no more than 10 days off is probably effective. He doesn't think you need machines, just doing exercises on your own body weight suffice. He thinks the ones that work produce muscle burn making you stop in about two minutes.
You don't want to do it so infrequently you end up with late onset soreness (stiffness), which is the result of damage.
EatingEating on a 2-1/2 hour ride isn't much help, unless its just carbs at the end. For long road races, he advises "real food" for two hours, and the carb stuff towards the end. It's really good to load up on food while riding at the end, and within an hour of concluding.
One aspect to manage is insulin reaction. If you eat too many carbs while not working hard, you'll get an insulin reaction which will move the fuel into the muscle instead of into the bloodstream where you want it. You want this for recovery, but not as gas.
Recovery Riding"There's a lot of vanity in the sport, and the perfect recovery ride plays to that."
You can ride the bike path, looking cool and not sweaty, telling people you're on a cool down.
He advises bringing lunch and eating it on the ride -- a burrito over an hour is good, and gets food into the muscles right away, better than eating sitting down.
After some announcements, we were dismissed. Some people went in search of a bar, I guess. Others went to bed, and some idiots wrote up their notes.
