It's killing me but..........

AshesGlory said:

No argument from me. But can the VP strategy be reliably quantified to make it useful beyond the conceptual phase?

VPP strategies invoke two important questions: (1) Why? (2) How? The attached document, extracted from a larger document on the subject of power management, may shed some light on the first question. As to the second question, there are many ways to do it, ranging from experience-driven guesswork to a computer-derived ride script for a specific course under forecast conditions (e.g., wind).😎

RapDaddyo said:

VPP strategies invoke two important questions: (1) Why? (2) How? The attached document, extracted from a larger document on the subject of power management, may shed some light on the first question. As to the second question, there are many ways to do it, ranging from experience-driven guesswork to a computer-derived ride script for a specific course under forecast conditions (e.g., wind).😎

RD, very interesting read. First thing that struck me is that the effect of 10% more power (250W -> 275W) on the flats and on a climb.
In the case of a 5% climb this results in a speed increase of 0.42 m/s, and at 0% grade it is 0.38 m/s which is a very small difference (0.04m/s), and I suspect it would be even smaller with a more realistic TT frontal area (0.6 m^2 was used, 0.35m^2 is more realistic)
EDIT just ran the numbers thru analytic cycling for frontal area = 0.35, all other parameters same as in your test.
250W 5% -> 5.22 m/s
275W 5% -> 5.67 m/s
=> 25W buys you 0.45 m/s, 8.6%

250W 0% -> 12.48 m/s
275W 0% -> 12.93 m/s
=> 25W buys you 0.45 m/s, 3.6%

The % increases in speed are pretty much the same as for a frontal area of 0.6. In fact the numbers 5% grade number increased slightly, hmmmm.

AshesGlory said:

EDIT just ran the numbers thru analytic cycling for frontal area = 0.35, all other parameters same as in your test.
250W 5% -> 5.22 m/s
275W 5% -> 5.67 m/s
=> 25W buys you 0.45 m/s

250W 0% -> 12.48 m/s
275W 0% -> 12.93 m/s
=> 25W buys you 0.45 m/s

You're missing the point. What matters is not the absolute difference in speed, but rather the percentage difference relative to the baseline. If you work out some hypothetical courses, speeds and times you will see why.

RapDaddyo said:

You're missing the point. What matters is not the absolute difference in speed, but rather the percentage difference relative to the baseline. If you work out some hypothetical courses, speeds and times you will see why.

No, I got that. I thought frontal area improvement would decrease the difference, but it increased the difference.
What simulator did you use to get the headwind/tailwind comparison?

AshesGlory said:

No, I got that. I thought frontal area improvement would decrease the difference, but it increased the difference.

I haven't tested to the limits of frontal area (low to high), but I'm not sure there is any realistic frontal area that would negate the advantage of increasing power on an upgrade and decreasing power on the flat or a downgrade. Of course, it is asymmetrical when constrained by a target NP. But, even with this constraint the benefits (distance covered in time x) of increasing power on an upgrade more than offset the cost (distance covered in time x) of decreasing power on the flat or downgrade.

AshesGlory said:

What simulator did you use to get the headwind/tailwind comparison?

The effects of wind are very much a function of the crosswind performance of specific wheels. The wind tunnel data I use are from the Texas A&M and UC-San Diego low speed wind tunnels.

RapDaddyo said:

The effects of wind are very much a function of the crosswind performance of specific wheels. The wind tunnel data I use are from the Texas A&M and UC-San Diego low speed wind tunnels.

Sorry, badly phrased query. I meant what analysis tool did you use so you could conclude
A similar relationship holds for wind. At a constant power (e.g., 250W), bike speed is slow against a headwind and fast with a tailwind. Again, an incremental increase in power results in a large proportional increase in bike speed against a headwind and almost no increase in bike speed with a tailwind.

AshesGlory said:

Sorry, badly phrased query. I meant what analysis tool did you use so you could conclude
A similar relationship holds for wind. At a constant power (e.g., 250W), bike speed is slow against a headwind and fast with a tailwind. Again, an incremental increase in power results in a large proportional increase in bike speed against a headwind and almost no increase in bike speed with a tailwind.

The tool is my own, but you can test different assumptions at Analytic Cycling. The benefits of variable power pacing are more pronounced with changes in grade, but still non-trivial with changes in wind direction. The reason I had to build my own tool is that I am not interested in the analysis of a single data point, but rather optimization of power management for a course as represented by a GPS tracking file. Wind is an overlay, currently assuming constant velocity and direction for the duration of a course. Maybe a later version will include the ability to forecast changes in conditions part way through a course as might be of value in multi-hour TTs (e.g., full-length triathlons). It basically always holds that if you apply a X% increase in power at a slower bike speed whether due to grade or wind you will get a greater percentage increase in bike speed than if you apply the same X% increase in power at a faster bike speed, approaching X% on the upside and 0% on the downside. Although I must say it's fun to play with extremes such as a 5% upgrade with a 25 knot tailwind versus a 5% downgrade with a 25 knot headwind.

RapDaddyo said:

The tool is my own, but you can test different assumptions at Analytic Cycling. The benefits of variable power pacing are more pronounced with changes in grade, but still non-trivial with changes in wind direction. The reason I had to build my own tool is that I am not interested in the analysis of a single data point, but rather optimization of power management for a course as represented by a GPS tracking file. Wind is an overlay, currently assuming constant velocity and direction for the duration of a course. Maybe a later version will include the ability to forecast changes in conditions part way through a course as might be of value in multi-hour TTs (e.g., full-length triathlons). It basically always holds that if you apply a X% increase in power at a slower bike speed whether due to grade or wind you will get a greater percentage increase in bike speed than if you apply the same X% increase in power at a faster bike speed, approaching X% on the upside and 0% on the downside. Although I must say it's fun to play with extremes such as a 5% upgrade with a 25 knot tailwind versus a 5% downgrade with a 25 knot headwind.

The headwind/tailwind conculsion is not intuitivley obvious to me, but maybe I am just being a bit dim.I can see that as speed of the bike is a proxy for speed at which the force is applied so agree that at lower speeds extra power will provide a greater % increase in speed. But in the headwind/tailwind example the opposing force is proportional to the square of the effective headwind, which is higher in the headwind case, so its not intuitively obvious to me how much of this offsets the lower bike speed.
Just wondered if you have the equivalent of "table 1" for the headwind/tailwind scenario.

Today's prologue at the Dauphine was a text book course for VPP. Basically first half uphill, second half downhill.
The winner had also had the best split at the top of the hill by some 4 secs. The 2nd place had about 17th place at the same split then had the best downhill time of the day.
It was also interesting to see the riders practicing "variable frontal area". On the uphill their frontal area was higher than the downhill - many of them were on the base bar. This enabled them to have a more efficient body position on the slower uphill at the expence of aerodynamics but allowing them to produce more power for the same oxygen cost due to a more efficient position.

AshesGlory said:

The headwind/tailwind conculsion is not intuitivley obvious to me, but maybe I am just being a bit dim.I can see that as speed of the bike is a proxy for speed at which the force is applied so agree that at lower speeds extra power will provide a greater % increase in speed. But in the headwind/tailwind example the opposing force is proportional to the square of the effective headwind, which is higher in the headwind case, so its not intuitively obvious to me how much of this offsets the lower bike speed.
Just wondered if you have the equivalent of "table 1" for the headwind/tailwind scenario.

Well, I've done a lot of work with both grade and wind. I didn't quickly find a table exactly like Table 1, but here's a chart I made up for a set of assumptions about a flat 40K ITT course. I did the analysis with winds of 5, 7.5 and 10 m/s. The time gains of a VPP strategy were obviously greatest with the strongest wind at about 45 seconds for a rider with an FTP of 250W. The fastest VPP strategy is to push the power on the upwind leg as much as possible while retaining the NP=250W constraint. BTW, this chart assumes 7.5 m/s wind.

AshesGlory said:

Today's prologue at the Dauphine was a text book course for VPP. Basically first half uphill, second half downhill.
The winner had also had the best split at the top of the hill by some 4 secs. The 2nd place had about 17th place at the same split then had the best downhill time of the day.
It was also interesting to see the riders practicing "variable frontal area". On the uphill their frontal area was higher than the downhill - many of them were on the base bar. This enabled them to have a more efficient body position on the slower uphill at the expence of aerodynamics but allowing them to produce more power for the same oxygen cost due to a more efficient position.

I put a link to the video of the Dauphine prologue in the Aussie, Thread about Nothing, Thread. The constant look of determination and pain, plus his constant out of the saddle pushing the pace on all the rises and corner exists suggest that nothing that held back at any point of the course. A majestic ride... one that was apparently done for training. :eek:

Contador wasn't on the basebar much on the uphills, he wasn't even on the saddle much on the main climb - prefering his out of the saddle, give it some greif, style. A lot of people have gone on about the pedalling efficiency of Armstrong in the mountains - aka high rpms blah, blah, blah. Sure, inbetween the Cols or on easy bits he might have but given that all of his as$-whooping epic rides, apart from the Alpe Dhuez time trial, involved big gears and out of the saddle work, it makes me wonder if these 'experts' actually watched the Tour. There are plenty of youtube clips on this - German TV clips tend to be longer, often including the entire footage of the entire final climb.

RapDaddyo said:

Well, I've done a lot of work with both grade and wind. I didn't quickly find a table exactly like Table 1, but here's a chart I made up for a set of assumptions about a flat 40K ITT course. I did the analysis with winds of 5, 7.5 and 10 m/s. The time gains of a VPP strategy were obviously greatest with the strongest wind at about 45 seconds for a rider with an FTP of 250W. The fastest VPP strategy is to push the power on the upwind leg as much as possible while retaining the NP=250W constraint. BTW, this chart assumes 7.5 m/s wind.

Just a quick question following a looksie of the pdf graph.

If a rider has an FTP of 250, how is he/she supposed to hold an extra 10% (~275watts) for around 35 minutes for that uphill and/or headwind section?

Typically, what I can hold for 35 minutes is very similar to what I can hold for 1 hour. 10% extra is a lot of watts. I'd argue that if you can pull that extra 10% above what you believe is your threshold for 35 to 40 minutes that your threshold isn't what you believe it is (ie it's actually higher than what you thought). That's 35 minutes of L5 in one chunk or 35 minutes of 'not going to happen'.

I'd like to see anyone pull an extra 7+% watts out of the bag for that amount of time over what really is their threshold. A rider with an FTP of 250 isn't going much north of 260watts for any length of time and certainly not 30+ minutes. Even if you do pull that cat out of the hat then you've set yourself up for 25+ minutes of fail if the outbound leg was the hard part.

Even if the event in question was a 10 mile (16km) TT then maybe, just and I mean 'just' as in you have one of 'those special days' where the moon, mars and uranus align and you find cycling nirvana, maybe there'd be a possibility of riding for half the headwind/uphill leg with an average that far above threshold but you better know in advance exactly what's going on for the rest of the course - wind coming out from side roads, small hills, traffic... when you're that far in debt, the little details become ominously large ride/goal failing events.

swampy1970 said:

Just a quick question following a looksie of the pdf graph. If a rider has an FTP of 250, how is he/she supposed to hold an extra 10% (~275watts) for around 35 minutes for that uphill and/or headwind section?

First, the chart is not intended to represent what is possible to execute, but rather to illustrate the implications of alternate strategies constrained only by NP=250W. If you have followed my comments on VPP, you will have noticed that I have stated that a VPP plan is both course and rider specific. In my approach, I use the rider's entire MMP curve as a constraint.

Look, I don't care that you don't believe in variable power pacing. You are an advocate of constant power pacing. Fine. I'm not going to try and persuade you of the merits of VPP. But, I don't appreciate the cheap shots as in your post above about "experts."

RapDaddyo said:

First, the chart is not intended to represent what is possible to execute, but rather to illustrate the implications of alternate strategies constrained only by NP=250W. If you have followed my comments on VPP, you will have noticed that I have stated that a VPP plan is both course and rider specific. In my approach, I use the rider's entire MMP curve as a constraint.

Look, I don't care that you don't believe in variable power pacing. You are an advocate of constant power pacing. Fine. I'm not going to try and persuade you of the merits of VPP. But, I don't appreciate the cheap shots as in your post above about "experts."

Hey, I'm all ears for anything that would help to go faster... and has results to back up the math. Given that one of the best time trial riders ever to grace the planet teamed with one of the better exercise physiologists and the guru that lead to the rise of the all dominating "British Cycling track team" couldn't get the results they believed were theorectically possible, despite the resources at hand, gives one a reason be at least be sceptical. I'd class those guys as some of the 'experts'

Quoted post said:

The time gains of a VPP strategy were obviously greatest with the strongest wind at about 45 seconds for a rider with an FTP of 250W. The fastest VPP strategy is to push the power on the upwind leg as much as possible while retaining the NP=250W constraint. BTW, this chart assumes 7.5 m/s wind.

Why model something that's not attainable? True, it's always interesting to see the "what ifs" but when using the impossible to present an arguement, especially when the impossible is showing compelling data to make one believe that this strategy is great, then that's when things go pair shaped.

Asking a rider to go that far (10%) over threshold isn't possible in a time trial of that length, unless the rider got his threshold wrong. If that was the case then the rider in question would also do better if they rode at their true threshold.

Care to model the effects of riding 10% over threshold into a headwind for 30 minutes? At which point does the rider go bang Hiroshima style? I'd say less than half way through. Repeat that for a mere 7% over and you'd still be pressed to make it to the half hour mark and still be able to give it full gas, ie close to FTP.

swampy1970 said:

Why model something that's not attainable? True, it's always interesting to see the "what ifs" but when using the impossible to present an arguement, especially when the impossible is showing compelling data to make one believe that this strategy is great, then that's when things go pair shaped.

The chart you are referring to was not intended to advocate a specific pacing strategy. That would be possible only by knowing a rider's full MMP curve. Rather, the chart simply shows the time gains from various upwind/downwind power management scenarios that meet one criterion: NP=250W. But, take note that all of the strategies are faster than constant power.

swampy1970 said:

Care to model the effects of riding 10% over threshold into a headwind for 30 minutes?

No, but you can model such scenarios to your heart's content at Analytic Cycling.

It's worth noting that a flat out/back course with nothing but headwind/tailwind resistance changes offer minimal (but not zero) advantages for variable power pacing, due to the small differences between power targets in the two segments. A more common course with some grade changes increases the opportunity for time gains.

Interesting debate on variable pacing strategies. Personally, I feel some mental benefit to varying pace. It tends to break up the session in discrete blocks. Grinding along at a specific steady pace usually ends up in an effort by me during the last half to just keep the average from falling watt by watt; and I have ended up with higher averages in rolling hill training that just hammering away at a steady pace along a level river bike path. I tend to just look at the averages on the cpu at the end of the ride rather than looking at NP on the downloaded data, so maybe NP evens it out anyway. .

What is also true is the development of what I call the dieseling effect of a steady diet of SST 20 minute intervals for an extended period. I find that I can diesel along in that SST range almost infinitely, but when a group or pack brings it up into match burning territory for brief spurts, I have problems responding and then settling back when the pace returns back into the zone. My first fast group ride of the season was a real eye opener to me. A few weeks of over/unders and shorter more intense intervals seems to have solved the problem

Kopride,

Just like you use hills to hammer away, use other geographical points to keep your interest on flatter rides. If you ride a loop multiple times just reference the ride by laps left or even just ride the ride corner to corner if things are getting desperate.

Sounds like you need to add a turbo to that diesel engine of yours. Moving your 20 minute effort to, or just slightly above, threshold should pep things up nicely. Chances are when you do this for the first time you'll either "run out of legs and die a thousand deaths" and not finish the last one or you'll suffer like you never thought possible... but if you get to that stage where you can do a few of those in a session then its hugely beneficial. When I was racing I found sessions like this better for those hard one minute gap closing efforts than shorter intervals.

The comeback is proving to be hard. I'm still experimenting with the CompuTrainer, and am not sure if I have the gearing right. In the past I used to ignore speed and cadence, use the lowest gear and grind my way through the different power levels. Now I'm on the big ring and 4/3 cogs from the smallest rear cog - doing between 18 and 22 mph at a steady cadence of around 88. Having said this, I usually do the last 4/5 minutes upping the cadence to 95-100 and the speed up to 28mph.

At this moment in time I can't imagine staying with the Nagoya racing team as I did a while back (that's me at the back in the yellow jersey). I'm really confused as to what my real power output is. Riding the CT at a lower cadence and lower gears in the past seemed to be pretty close to the wattage I could generate in the gym. Oh well, battle on.😕

swampy1970 said:

Kopride,

Sounds like you need to add a turbo to that diesel engine of yours. Moving your 20 minute effort to, or just slightly above, threshold should pep things up nicely. Chances are when you do this for the first time you'll either "run out of legs and die a thousand deaths" and not finish the last one or you'll suffer like you never thought possible... but if you get to that stage where you can do a few of those in a session then its hugely beneficial. When I was racing I found sessions like this better for those hard one minute gap closing efforts than shorter intervals.

Again, interesting strategy to add the turbo, but on fast group rides, the brief efforts in my hilly parts mean I have to be able to sustain 400-500 watts for up to a minute, with max watts around 700; and there is some sustained climbing where I need to be well above 300 for 3-6 minute stretches followed by tempo riding at 220-250 for sustained periods. There are some occasional coasting stretches, and drafting as well so the average does come down butit is hammer hard up the hill, coast briefly, cruise a flat, and then hammer another roller. And the hills in SE PA are not Rocky Mt long, but they are relentlessly rolling, almost 7000 ft of total climbing in 90 minutes . I need to test to get a solid number, but I am guesstimating FTP at about 280 (based upon some 20 minute intervals). The real problem is when I have to throw on those afterburners in the 400-500 range, I'm just not settling back down. So assuming that I can creep that FTP up 10 watts in a few months, which is, despite the miracle gains routinely boasted on the internet, would be real progress given that the low hanging newbie gains are long past. I'm not sure that a FTP of 290 would be the big difference, as much as learning to get used to those brief stints and learning to settle. Hey, it might be that these young club racers are just faster--but when I look at the average watts for the ride, it is well below my hour TT average, usually in the range of 215-220 max. I just think that I should be hanging with these guys, or at least hanging on when they push hard, given my FTP. I would think an FTP of 250 would be enough, and my estimate of 280 means I should be doing this ride with gas left in the tank.