Average Power without zeros?

[umd1]

I got in a argument with someone tonight about average power when they were boasting about their 500+W average power for a crit. A quick look at the garmin connect data page showed his number, but downloading into WKO showed as I suspected, that they were not including zeros in the average and the true average was closer to 290, and the NP was about 415.

There was no convincing him that NP (or xPower since he is using GC) is a better metric, he just came back at me with something about duty cycles, dynoing a car engine, and miles per gallon when the engine is off.

So, am I crazy or is he? Is there any meaningul basis for measuring or reporting averages without zeros?

 

[Alex Simmons]

Originally Posted by umd1 .

So, am I crazy or is he? Is there any meaningul basis for measuring or reporting averages without zeros?

He is and no.

Next.

 

[qcwtom]

He's oblivious and has no idea how to use his power meter.

by his logic, you could do a 1000 watt sprint then coast downhill for an hour and have a 1000 watt average for an hour.

 

[umd1]

Originally Posted by Alex Simmons .




He is and no.

Next.

Although I greatly appreciate the validation that I am not crazy, I was hoping for some authoritative information I could point him to. It's obvious to me, but I've run out of my own explanations as to why and he remains unconvinced. Searching the net I found a paper on how xpower and bikescore is calculated but nothing really on why zeros should be included in the average. I guess because it's pretty obvious to everyone else too...

 

[tonyzackery]

Did he win the race or get a high placing? If not, who cares? I know I wouldn't.
Regardless, the guy is pretty strong.
Sounds like he's reporting from an 'up and down' crit course - basically doing a series of short, L6-effort hill repeats where you coast back down after doing a rep. These courses will assuredly skew the power numbers and I don't even find NP to be a reliable measure of the work involved. NP reports high IME for these types of crits.
No use trying to change someone's beliefs.
He's an awesome "bench racer". Great.
He'd be even lamer, in my book, if he's not getting any results...

 

[lanierb]

AP without zeros tells you nothing.

As an aside, those numbers are really high. I see you mention Garmin. Any chance he had "Smart Recording" on? (the default on many Garmins) Because if so there's another problem with the data and all the numbers you are quoting are junk (the 500 and the 290 and the 415) -- most likely way too high. When using power you need to set the Garmin to record every second or it messes up the power data.

 

[hrumpole]

As tempting as it may be to persuade this person, why bother? Take his money. (If this works, I want ten percent

If he's this dense, and prideful, bet him $100 that he can't ride at 500 watts for 10 minutes on a trainer. The trick: this must be done on a turbo trainer--one with a power curve, and the power curve (e.g., avs) has to be over a certain speed for him to win. (Hell, since it's a one-time event, you can calibrate it using the PM). Odds are you can spot him a few miles an hour...

 

[qcwtom]

Some people won't be convinced, not matter how much evidence you put before them.

The best source for this is Coggan's book, Training and Racing with a Power Meter.

 

[Alex Simmons]

Originally Posted by umd1 .
Although I greatly appreciate the validation that I am not crazy, I was hoping for some authoritative information I could point him to. It's obvious to me, but I've run out of my own explanations as to why and he remains unconvinced. Searching the net I found a paper on how xpower and bikescore is calculated but nothing really on why zeros should be included in the average. I guess because it's pretty obvious to everyone else too...

1. NZAP is meaningless because the time spent not pedaling affects the power one can produce when they are pedaling.

2. Because an average is just that - an average of all data (equally weighted - in this sense that weighting is equal time samples).
Otherwise It would be equally valid to remove all power numbers above 400W, so you get a non>400W average power number. Now does that tell your mate any more than a non-zero average power number does?

 

[tonyzackery]

W/o zeros may not have a lot of meaning, but I frequently find it pretty interesting how well it tracks with NP on some rides - even those that involve frequent starting and stopping (around town/in the city). Sure, those rides don't involve a bunch of anaerobic efforts, but "it is what it is".

In OPs case, I like the above suggestion for the bench racing champ to go and see how long he can hold that 500+ watts. I'd bet not more than a few minutes, let alone the duration of a crit...

 

[swampy1970]

There was no convincing him that NP (or xPower since he is using GC) is a better metric, he just came back at me with something about duty cycles, dynoing a car engine, and miles per gallon when the engine is off.
So, am I crazy or is he? Is there any meaningul basis for measuring or reporting averages without zeros?

Huh? Dynoing an engine and miles per gallon when the engine is off? Sounds like Mr Unleaded pencil **** needs some schooling on how dynos work and how they measure "power" (and how complicated it'd be for him to understand all the different types of dynos, their respective merits and pitfalls would be beyond the realms of science and hence far beyond this thread) as well as how to interpret power meter data before he tries to engage his brain and puts his foot in his mouth again. Remind him that his average power was above that produced by Lance when he was going well and would likely serve him pretty well in a 4 minutes and change effort in a world class individual pursuit. I'm guessing even on the pharmaceutical equivalent of AVGas that your oxygen thief of a friend would be hitting anywhere close to those numbers. His ECU needs a retune.

 

[An old Guy]

Originally Posted by umd1 .

I got in a argument with someone tonight about average power when they were boasting about their 500+W average power for a crit.

So, am I crazy or is he? Is there any meaningul basis for measuring or reporting averages without zeros?

Sort of foolish to argue about numbers from a race.

A more important number is the work done (joules). But even the average power without zeros is useful if that is the number you always record. I guess guys that do intervals ignore the zeros between repeats. They even ignore the long warm up and warm down they do. Now they are crazy.

 

[Alex Simmons]

Originally Posted by tonyzackery .

W/o zeros may not have a lot of meaning, but I frequently find it pretty interesting how well it tracks with NP on some rides - even those that involve frequent starting and stopping (around town/in the city). Sure, those rides don't involve a bunch of anaerobic efforts, but "it is what it is".

That's just coincidence, the relationship between them is of course spurious.

 

[tonyzackery]

Originally Posted by Alex Simmons .




That's just coincidence, the relationship between them is of course spurious.

Thanks for the revelation./img/vbsmilies/smilies/wink.gifMy mistake, did I say or imply somewhere that NP and APNZ were related? Apologies if I led anyone astray.

Spurious. Frequent coincidence. Recurring happenstance. Quasi-ubiquitous fluke. Call it what you want, but "it is what it is"...and not at all too infrequently do I see them tracking in step... Just interesting to, and for me...

 

[Woofer]

Kieran is the guy who thinks he's as stronger than a Pro Tour rider because he can put out more watts and doesn't realize the real comparison is w/kg...