Best way to use a power meter to pace myself up climbs.

Poppa Wheelie

In short, I’m looking for advice on how best to use a power meter to pace myself up climbs.

To help understand my question, I should tell you a little bit about myself first. I’m 53 years old and have had asthma since birth. Whilst my asthma is very well controlled, I do have a low lung capacity for someone of my age, thus I think I go into oxygen debt quite easily. The improvement to my asthma cycling gives to my day to day life cannot be overstated, however I do get frustrated getting half way up a hill and having to stop because I’ve overdone it. Hence I’ve bought myself a power meter to compliment my heart rate monitor and Wahoo Elemnt.

So far the power meter has already shown me that when I perceived I was over exerting myself, I was changing into an easier gear and still producing almost exactly the same wattage as before but at a higher cadence. Consequently that’s why I was failing to summit the climb.

On a turbo trainer I carried out the 8 minute FTP test that’s pre-loaded into the Elemnt and was very surprised that the 8 minute session took an hour to complete!

I thought the result was low based on what I seem to be able to produce when I’m out on the road so I did another turbo trainer session where I did a 5 minute warm-up, 20 minutes trying to hold 220 watts then a 5 minute cool down. Once I’d finished that session, training peaks (free) said I’d established a new FTP of 197 watts. Whilst I’m sure that’s not a ‘proper’ FTP test, the results seems to more or less reflect what I can produce out on the road for an hour.

What I want to know what is the best way to pace myself up a climb. Is it just simply a case of riding in the VO2 max range calculated from the FTP, i.e. up to 120% of my FTP or is there a better way?

Especially as some inclines seem to need more than 120% of my FTP to be able to climb at all.

gethinceri

When you go above ftp you're using energy you won't get back.
Climb in Zone 4 or lower.
VERY simplistic suggestion, there is much more complex advice which will no doubt arrive shortly.

Alex_Simmons/RST

Poppa Wheelie wrote:

On a turbo trainer I carried out the 8 minute FTP test that’s pre-loaded into the Elemnt and was very surprised that the 8 minute session took an hour to complete!

There's something deliciously ironic about that.

Poppa Wheelie wrote:

What I want to know what is the best way to pace myself up a climb. Is it just simply a case of riding in the VO2 max range calculated from the FTP, i.e. up to 120% of my FTP or is there a better way?

Pace climbs based on what you can actually do, not based on some proportion of what you think your FTP is.

If you went too hard and were unable to complete an effort, well you know next time not to go so hard. Intervals are self regulating like that.

How hard one can go depends very much on how long you expect to go hard for. At 120% of FTP one rider might last only a couple of minutes, while others might manage it for several times that duration. Individual capabilities for supra threshold efforts vary considerably.

Poppa Wheelie wrote:

Especially as some inclines seem to need more than 120% of my FTP to be able to climb at all.

If the inclines are too steep for your current gearing, then change your gearing so you can ride at speeds and power that is appropriate for your fitness and the gradients and duration of climbs you expect to ride up. Or ride less steep hills until fitness improves.

sloppyschleckonds

Climbing at Endurance/ Tempo should get you up and over most hills.
There is no shame in running a 30/32/34 rear cassette if it helps prevent you going into the red too early.

Poppa Wheelie

Thanks for all the replies so far. I do have a 32 tooth cassette although I only need it on 10% climbs. If the climb is under 10% I often only need to go as low as second which is 28 tooth.

From recent ride, on a 3 minute climb, I was able to maintain around an indicated 240 watts for most of the climb and provided I kept under 290 watts on the very steepest part (about a minute 30) I was able to summit the hill ok. The overall average incline was 6% but there are parts that must be touching 10%. Average power for that effort was actually 218 watts at 70 rpm, peak cadence was 83 rpm.

From what has been said I appear to be doing much better than I though!

milemuncher1

When you’re climbing, try to fight the urge to grip the bars too tightly, and try to relax your shoulders, you’ll naturally tend to tense up. If your gripping the bars, and tensing up too much, your oxygen demand will be higher than it could be / should be, if you relax a bit. If you’ve got a lower than normal lung capacity, you won’t be able to get as much oxygen into your blood, as someone with a similar VO2 max, with a more normal lung function, and as your leg muscles will be demanding as much oxygen as you can supply them on a climb, you don’t want any other muscles to be adding to it unnecessarily. Work on improving your posture on the climb, to maximise your ventilation, so straighten you back as much as possible, try not to scrunch up your diaphragm. If you use the power meter to try and keep your power in your personal zone 3 for as long as possible, and not let it climb too far into zone 4, if for instance, you’ve got an uphill Corner, you’ll be doing yourself a favour. It’s most efficient to maintain as many contact points with the bike, during a climb ( usually by remaining seated) for as long as possible, then stand during your zone 4 efforts ( for example ), in the corners. Use the power meter to try and reduce the amount of variation in the power, during the climb.

imposter2.0

Milemuncher1 wrote:

When you’re climbing, try to fight the urge to grip the bars too tightly, and try to relax your shoulders, you’ll naturally tend to tense up. If your gripping the bars, and tensing up too much, your oxygen demand will be higher than it could be / should be,

Or alternatively, just ignore this nonsense. Gripping the bars tightly/less tightly is not going to have any meaningful impact on your oxygen usage.

FatTed

Sloppy I cant climb the Col de Galibier at 90% of my so called FTP, maybe at Tempo, My calculated FTP from a 20 min test bears little relationship to what I have actually achieved over an hour on the road.

Poppa Wheelie

Gethinceri wrote:

When you go above ftp you're using energy you won't get back.
Climb in Zone 4 or lower.
VERY simplistic suggestion, there is much more complex advice which will no doubt arrive shortly.

I had disregarded this as you gave so little information for your comment to be of any meaningful use.

However, I've been googling 'using a power meter to climb' and have stumbled across something called w' (w prime). That's what you're referring to and it appears to be what I'm missing.

I'd already downloaded Golden Cheetah but have found it too complex and unintuitive to be of much value to me.

All of a sudden I've got a lot of reading to do on W prime and a reason to learn how to make use of Golden Cheetah.

briantrumpet

Imposter wrote:

Milemuncher1 wrote:

Blah, blah, blah

Or alternatively, just ignore this nonsense. Gripping the bars tightly/less tightly is not going to have any meaningful impact on your oxygen usage.

Stalker

True though.

paul2718

If you have a Garmin then perhaps exploring something like, https://www.trinakan.com/wp/2016/04/02/ ... ct-iq-app/ would be interesting.

It might help you learn how to feel the 'blow up' approaching.

Or not, just an idea.

Poppa Wheelie

paul2718 wrote:

If you have a Garmin then perhaps exploring something like, https://www.trinakan.com/wp/2016/04/02/ ... ct-iq-app/ would be interesting.

It might help you learn how to feel the 'blow up' approaching.

Or not, just an idea.

Thanks but I have a Wahoo Elemnt. W' is the only thing I've found that the Wahoo cannot do. That said neither can a Garmin nativity.

imposter2.0

People have been pacing themselves up climbs long before power meters came along. As Alex suggested earlier, the more you ride, the better understanding you will have of your own limits.

Alex_Simmons/RST

Poppa Wheelie wrote:

Gethinceri wrote:

When you go above ftp you're using energy you won't get back.
Climb in Zone 4 or lower.
VERY simplistic suggestion, there is much more complex advice which will no doubt arrive shortly.

I had disregarded this as you gave so little information for your comment to be of any meaningful use.

However, I've been googling 'using a power meter to climb' and have stumbled across something called w' (w prime). That's what you're referring to and it appears to be what I'm missing.

I'd already downloaded Golden Cheetah but have found it too complex and unintuitive to be of much value to me.

All of a sudden I've got a lot of reading to do on W prime and a reason to learn how to make use of Golden Cheetah.

There is no need to be concerned with W' in this context. Just ride the hill and use the meter as a guide to what to do next time.


But for the sake of explanation, I'll try to keep it simple...

W' is one of two parameters in a simple mathematical model of human energy output capability*. The other parameter is Critical Power (CP).

In this simple model we have two types of energy sources:

- one that is not capacity limited but is rate limited. This is expressed as your Critical Power, which is a (threshold) power output you could theoretically maintain for a long time

- one that's not rate limited but is capacity limited, W', which is like a small rechargeable battery we tap into when the energy demand exceeds the Critical Power.

IOW if your power output exceeds your CP, then you need to draw energy from your W' "battery".

The higher above your CP you go, the faster you drain your W' battery.

Drain the battery completely and you are forced to reduce power output to at most your Critical Power. You can recharge your W' battery but it requires riding at less than your Critical Power, and the easier you go the faster the battery recharges. Stopping altogether will enable the most rapid recharge.

Most people will have a W' somewhere in the range of 5kJ to 30kJ, and there is quite a bit of individual variability in W'. It is trainable but has a sizeable inherited component.

For context, a W' of 15kJ enables one to ride 50W above your CP for a maximum of 15000J / 50W = 300 seconds.
or:
100W above CP for at most 150 seconds
200W above CP for at most 75 seconds

Hopefully you get the idea.


The recharge rate of W' is a little more complex but in essence all you really need to remember is:
- to recharge any depleted W', power output needs to be under CP
- the further under CP you are the faster the recharge rate, and
- the rate of recharge is way way slower than the rate at which we can discharge it. Just like those small rechargeable batteries, we can suck the juice out of them very quickly but it takes ages to recharge them.


The model is useful but its domain of validity ranges from a couple of minutes out to about an hour. In reality we can't sustain CP forever and the model doesn't work so well for very short durations either. It's also sensitive to the inputs used to derive the parameter estimates of CP and W'.


* Originally described in the 1960s by messrs Monod & Scherrer. W' has other names, like Anaerobic Work Capacity (AWC) and in Andy Coggan's Power-Duration model the equivalent indicator is Functional Reserve Capacity (FRC). There are various models and each has their strengths and limitations, and understanding them enables one to determine their application, usefulness and domain of validity.

58585

Alex_Simmons/RST wrote:

For context, a W' of 15kJ enables one to ride 50W above your CP for a maximum of 15000J / 50W = 300 seconds.
or:
100W above CP for at most 150 seconds
200W above CP for at most 75 seconds

I've recently started having a play with best bike split and the topic of optimal pacing is quite interesting. Looking at power data can of course be misleading but it does seem like some people are (naturally?) better at pacing their efforts, and I am not one of those.
Is there any way to test and quantify W' and implement it in pacing strategies for eg 20km time trials with rolling terrain? As far as I can see, best bike split needs an simple user inputted "maximum above CP will you go"

StillGoing

I just ride within myself for 50% to 2/3rds of the way up and if I'm feeling good, put a big effort in for the last bit. Usually ends up passing my ride buddies that went steaming up the hill from the start.

Poppa Wheelie

58585 wrote:

It does seem like some people are (naturally?) better at pacing their efforts, and I am not one of those.

Nor me, that's why I'm asking. Once I got the power meter I was able to see that on a hill, I was changing to an down into an easier gear but still producing the same wattage but at a higher cadence when really I needed to throttle back some more..

58585 wrote:

Is there any way to test and quantify W' and implement it in pacing strategies for eg 20km time trials with rolling terrain? As far as I can see, best bike split needs an simple user inputted "maximum above CP will you go"

Golden Cheetah has a tool that can estimate your CP and W'.

Poppa Wheelie

Alex_Simmons/RST wrote:

There is no need to be concerned with W' in this context. Just ride the hill and use the meter as a guide to what to do next time.


But for the sake of explanation, I'll try to keep it simple...

W' is one of two parameters in a simple mathematical model of human energy output capability*. The other parameter is Critical Power (CP).

In this simple model we have two types of energy sources:

- one that is not capacity limited but is rate limited. This is expressed as your Critical Power, which is a (threshold) power output you could theoretically maintain for a long time

- one that's not rate limited but is capacity limited, W', which is like a small rechargeable battery we tap into when the energy demand exceeds the Critical Power.

IOW if your power output exceeds your CP, then you need to draw energy from your W' "battery".

The higher above your CP you go, the faster you drain your W' battery.

Drain the battery completely and you are forced to reduce power output to at most your Critical Power. You can recharge your W' battery but it requires riding at less than your Critical Power, and the easier you go the faster the battery recharges. Stopping altogether will enable the most rapid recharge.

Most people will have a W' somewhere in the range of 5kJ to 30kJ, and there is quite a bit of individual variability in W'. It is trainable but has a sizeable inherited component.

For context, a W' of 15kJ enables one to ride 50W above your CP for a maximum of 15000J / 50W = 300 seconds.
or:
100W above CP for at most 150 seconds
200W above CP for at most 75 seconds

Hopefully you get the idea.


The recharge rate of W' is a little more complex but in essence all you really need to remember is:
- to recharge any depleted W', power output needs to be under CP
- the further under CP you are the faster the recharge rate, and
- the rate of recharge is way way slower than the rate at which we can discharge it. Just like those small rechargeable batteries, we can suck the juice out of them very quickly but it takes ages to recharge them.


The model is useful but its domain of validity ranges from a couple of minutes out to about an hour. In reality we can't sustain CP forever and the model doesn't work so well for very short durations either. It's also sensitive to the inputs used to derive the parameter estimates of CP and W'.

Thanks for taking the time to post this and I do understand all the 'get out and just ride the thing' comments many people have said. However, I think one of the reasons we all ride a bike when we all own cars is the sense of achievement we get from a good ride.

In my case I absolutely hate failing to summit a hill, it totally destroys any sense of achievement, hence the power meter. Since buying it I've only failed one climb and that was because I was going too far over my FTP. I basically used all my W' and needed to recharge it by stopping.

Whilst W' seems to be the answer to the initial question, I now know that it's rather difficult to quantify and my Wahoo cannot measure it, so I will have to just rely on a mixture of perceived effort and past power data to predict how far over my FTP to push any new climb.

sloppyschleckonds

Poppa Wheelie wrote:

I will have to just rely on a mixture of perceived effort and past power data to predict how far over my FTP to push any new climb.

There is enough advise on this thread. Read through some of the replies again.

Alex_Simmons/RST

58585 wrote:

Alex_Simmons/RST wrote:

For context, a W' of 15kJ enables one to ride 50W above your CP for a maximum of 15000J / 50W = 300 seconds.
or:
100W above CP for at most 150 seconds
200W above CP for at most 75 seconds

I've recently started having a play with best bike split and the topic of optimal pacing is quite interesting. Looking at power data can of course be misleading but it does seem like some people are (naturally?) better at pacing their efforts, and I am not one of those.
Is there any way to test and quantify W' and implement it in pacing strategies for eg 20km time trials with rolling terrain? As far as I can see, best bike split needs an simple user inputted "maximum above CP will you go"

TT pacing is another kettle of discussion fish.

Use of W' is a good thought but an unnecessary complication for longer TTs and probably not so reliable in a day to day sense. Keep in mind over a 20km TT (~30-min) W' is only going to represent a few % of your total energy supply.

It's more useful conceptually for analysing shorter events, e.g. team pursuits and other track events, or say high intensity interval training. That said, the amount of data one has available and can sensibly process during such events is zero to a little so it's effectively a post-hoc analysis tool. You get a some opportunity during a longer TT to view data but even then your attention should be elsewhere.

Pithy Power Proverb: The best predictor of performance is performance itself.

Optimising time trial power output based on your expected Normalized Power capability for the duration, which is what BBS does and something Andy Coggan demonstrated well over a decade ago is a "simpler" approach, in that your mean maximal NP capability already deals with the physiological cost of varying power.

So yes, knowing that for the expected duration you are capable of sustaining X watts is pretty much all you need to know in order to develop a reasonable pacing strategy. If on such a variable course you know you can attain Y Normalized watts, then that's a little more helpful again as a guide.

10 years ago I wrote a paper about it after developing a model for quantifying pacing optimisation based on Andy's ideas. BBS works on the same principal, at least for a part of this model approach.

Mine goes a step further in that it provides not only pre-event strategy and post-event analysis but also assesses variances in environmental conditions en route (by building in the concept of Virtual Elevation by Robert Chung) and quantifies the time improvement possible on that day with an improved pacing strategy relative to best in class.

I analysed TT data from amateurs and pro riders, and examples of excellent, average and poor pacing.

In essence the best pacers followed some basic practices:
- don't start too hard (#1 mistake, sustaining a power well over target capability, only to be forced to slow down after 3-5 minutes)
- a little harder on ascents, perhaps up to 10% over the target average power (but often less than 10%)
- a little less power on descents, perhaps up to 15% less power than target average power (but often less than 15%)
- variance in power due to head/tail wind scenarios should be pretty small, a handful of watts more into wind, a handful less with tailwind

Now in terms of how this actually feels compared with power output, on ascents up to 10% more power actually feels relatively easy, and up to 15% less power on descents feel like you are going really hard. IOW the perception of effort is opposite to the power output relative to target average.

Of course there are going to be variations on a theme depending on the exact nature of the course (length and gradient of climbs, steep descents where speed is so fast you are better to tuck and coast, technical elements if you have a lot of cornering and so on).

But as a starting point, a steady state power pacing strategy that is close to the maximal power you could sustain for the duration is actually a pretty fast strategy, and varying from that a little by applying a bit more power on ascents and a bit less power on descents is faster again. The time differences with such variations are not huge, we are talking seconds not minutes.

Now people already naturally apply more power on climbs and less power on descents and anyone experiencing the data from their power meter for the first time will have their eyes opened about just how much their power varies with even just subtle changes in gradient. The trick is to reduce those differences from the natural tendency by not going too hard on ascents and too easy on descents.

The other one is the opening minutes - using the power meter to keep you in check because it can feel way too easy at this early stage even though you are riding at or just above your target power.

The power meter is helpful in that it can fill the feedback gap caused by the lag in perception of effort when actual effort (power) changes. This is most useful in the opening minutes and at the start of climbs. Otherwise use how you feel on the day to guide you. Power is not the whole story.

cgfw201

My experience is the following
- The more you use it, the more useful it will become
- You might be able to ride at 300w for an hour one day, but for some reason only 250w the next, got to listen to your body ultimately.
- If I've got a long day in the mountains, I'll calculate the number of hours climbing, roughly, and then aim to ride at around 10% less than my FTP for every hour of climbing, so if my FTP is 300W, I'll climb at 270w for 1hr of climbing, 240 for 2h, 210 for 3h etc. This usually leaves me with enough in the tank to spare on the last climb. Served me well for 2 5000m days in the Pyrenees last year.
- Don't waste watts in the valleys if you're in the big mountains. One guy I ride with would biff it around at 350w in the valley roads then get dropped immediately on every climb.

58585

Thanks for a thorough reply!

Alex_Simmons/RST wrote:

TT pacing is another kettle of discussion fish.

This may be a strange question, but have you looked at the psychological effect of having live pacing data?
Anecdotally I think there is a performance benefit when I have a virtual partner/live "segment time" on my head unit and I am up on time. Of course it can backfire as soon as a gap opens.

Alex_Simmons/RST

cgfw201 wrote:

My experience is the following
- The more you use it, the more useful it will become
- You might be able to ride at 300w for an hour one day, but for some reason only 250w the next, got to listen to your body ultimately.
- If I've got a long day in the mountains, I'll calculate the number of hours climbing, roughly, and then aim to ride at around 10% less than my FTP for every hour of climbing, so if my FTP is 300W, I'll climb at 270w for 1hr of climbing, 240 for 2h, 210 for 3h etc. This usually leaves me with enough in the tank to spare on the last climb. Served me well for 2 5000m days in the Pyrenees last year.
- Don't waste watts in the valleys if you're in the big mountains. One guy I ride with would biff it around at 350w in the valley roads then get dropped immediately on every climb.

Interestingly, it was the challenge of developing a pacing strategy for La Marmotte that originally triggered my work on this 10 years ago. Some smart arse said it couldn't be done.

Alex_Simmons/RST

58585 wrote:

Thanks for a thorough reply!

Alex_Simmons/RST wrote:

TT pacing is another kettle of discussion fish.

This may be a strange question, but have you looked at the psychological effect of having live pacing data?
Anecdotally I think there is a performance benefit when I have a virtual partner/live "segment time" on my head unit and I am up on time. Of course it can backfire as soon as a gap opens.

The issue of using simulated competitor feedback has been studied. Modelling for it is problematic for obvious reasons - individual variability.

Here's an example from nearly 20 years ago for simulated 16.1km TT:
http://www.jsams.org/article/S1440-2440(14)00129-7/abstract

And another 2012 study for 4km pursuits:
http://bucks.collections.crest.ac.uk/10129/

No doubt there are others.

Tangentially there was a 2013 review of studies about the influence of feedback deception on pacing and performance. E.g. what happens when you misrepresent distance or time or power feedback to subjects during trials:
https://link.springer.com/article/10.10 ... 013-0094-1

g00se

FatTed wrote:

Sloppy I cant climb the Col de Galibier at 90% of my so called FTP, maybe at Tempo, My calculated FTP from a 20 min test bears little relationship to what I have actually achieved over an hour on the road.

Using 95% of the average power from a 20 minute test to calculate your FTP is just an estimation. I've read that the 95% is for well trained elite athletes and for most people, the percentage can vary wildly. I can't find the reference to it now, but I recall it can be as low as 80%.

Alex_Simmons/RST

g00se wrote:

FatTed wrote:

Sloppy I cant climb the Col de Galibier at 90% of my so called FTP, maybe at Tempo, My calculated FTP from a 20 min test bears little relationship to what I have actually achieved over an hour on the road.

Using 95% of the average power from a 20 minute test to calculate your FTP is just an estimation. I've read that the 95% is for well trained elite athletes and for most people, the percentage can vary wildly. I can't find the reference to it now, but I recall it can be as low as 80%.

FTP would never be as low as 80% of 20-min power.