Limitations of TSS

[RapDaddyo]

Originally Posted by An old Guy .

Why would I care what 3 professionals have to say about TSS?

Doesn't this apply to everybody and everything?

 

[An old Guy]

Originally Posted by acoggan .


373/330 = 1.13, so if your FTP is indeed 330, you've clearly generated an NP buster. Indeed, even if your FTP is 340, the IF is still >1.05.

On the flip side, though: you don't mention the AP for that hour, but I'm 99.99% certain that it was less 287 W. IOW, even if the NP is an overestimate, it is still closer to the truth than the AP.

Back up that comment. You seem to like to make statements that have no support. And then demand that others provide support for their oppossing opinions. Do what you ask others to do. Support your statement.

----

It appears that you used the following computation:

373/330 = 1.13
(1.13 + .87)/2 = 1
330 * .87 = 287

If the workout was 1 hour at maximun power output, that might be a reasonable result. But the workout was 2 hours. Based on FTP the average power should be less. But the poster did not say he was working at the maximum level he could for the 2 hours. That pushes the average even lower.

 

[wysinwyg]

Originally Posted by An old Guy .


Back up that comment. You seem to like to make statements that have no support. And then demand that others provide support for their oppossing opinions. Do what you ask others to do. Support your statement.

----

It appears that you used the following computation:

373/330 = 1.13
(1.13 + .87)/2 = 1
330 * .87 = 287

If the workout was 1 hour at maximun power output, that might be a reasonable result. But the workout was 2 hours. Based on FTP the average power should be less. But the poster did not say he was working at the maximum level he could for the 2 hours. That pushes the average even lower.

The 373 number was the highest 1 hour NP within the 2 hour workout. The NP for the entire 2 hours was 336.

I can't understand the rest of your post.

 

[acoggan]

Originally Posted by An old Guy .


1) TSS is free to me. I don't use it. I doubt that any professional team pays to use something that is free.

2) I once gave a link to a page on TrainingPeaks.com with the author listed as Andy Coggan that tells why TSS was invented. I believe it was you on this board that said Andy Coggan got a license fee for allowing TSS to be incorporated into produts.

You are a fraud and a liar.

1. I meant that they paid/pay to use WKO+ and the TP website.(They're also paying Hunter to come to Tenerife for a training camp, so that they can pick his brains even more.)

2a) TSS was invented because Hunter was smart enough to realize that coaches and athletes alike could benefit from leveraging all of the data being collected to look at "the big picture".

2b) The only licensing fee I receive is from TrainingPeaks, for use of my various ideas in WKO+. I don't receive anything for their use of my ideas on their website, or by/from anyone else.

 

[acoggan]

Originally Posted by An old Guy .


Back up that comment. You seem to like to make statements that have no support. And then demand that others provide support for their oppossing opinions. Do what you ask others to do. Support your statement.

----

It appears that you used the following computation:

373/330 = 1.13
(1.13 + .87)/2 = 1
330 * .87 = 287

If the workout was 1 hour at maximun power output, that might be a reasonable result. But the workout was 2 hours. Based on FTP the average power should be less. But the poster did not say he was working at the maximum level he could for the 2 hours. That pushes the average even lower.

Here are data for FTP, AP, and NP for 21 individuals who have generated "NP busters" (i.e., IF > 1.05 for ~1 h or longer). As you can see, in every case except one the NP is closer to FTP than is the AP.


FTP Duration ave P ave/FT old NP old IF old VI 240 1.179 188 0.78 252 1.050 1.340 280 1.007 135 0.48 294 1.050 2.178 300 1.028 270 0.90 316 1.053 1.170 315 1.001 272 0.86 334 1.060 1.228 330 1.002 315 0.95 350 1.061 1.111 295 0.998 228 0.77 313 1.061 1.373 295 1.001 232 0.79 316 1.071 1.362 315 1.088 267 0.85 338 1.073 1.266 240 1.006 212 0.88 258 1.075 1.217 262 0.983 247 0.94 283 1.080 1.146 250 1.037 182 0.73 273 1.092 1.500 250 1.003 188 0.75 275 1.100 1.463 250 1.003 186 0.74 276 1.104 1.484 280 1.014 235 0.84 312 1.114 1.328 295 0.398 237 0.80 330 1.119 1.392 310 1.002 283 0.91 347 1.119 1.226 250 1.002 201 0.80 282 1.128 1.403 305 1.041 259 0.85 345 1.131 1.332 310 0.995 261 0.84 351 1.132 1.345 250 1.175 146 0.58 284 1.136 1.945 325 0.975 281 0.86 384 1.182 1.367
As for 1 vs. 2 h, you misread the poster's comments: the 373 W was for the highest single hour.

 

[daveryanwyoming]

Originally Posted by An old Guy .
... I am going to have to ask that you be banned....

Good luck with that....

Let's see, the forum mods can ask whether 'An old Guy' or Andy Coggan has provided more value to the cycling community and these boards.

Gee I wonder where they'll come down on that decision?

 

[alienator]

But it dawns on me that simply telling acoggan that if he wants me to collect data for him he has to pay for it, might in error be considered promoting a product. After all Andy Coggan is not about to pay me or anyone else to disprove his claims.

Why should anyone pay you to collect the data that you so conveniently lack to support your claims? You haven't put up, so maybe it's time the old adage, "Put up, or shut up", be applied to you. As for banning, let's do a quiz: Q: Who does the libeling, acoggan or An old Guy? A: An old Guy. Q: Who continually labels the other a liar, acoggan or An old Guy? A: An old Guy. Q: Who is helpful in the forum, acoggan or An old Guy? A: acoggan Q: Who is the troll, acoggan or An old Guy? A: An old Guy

 

[fluro2au]

This thread was moving along so nicely.

To the mods, seriously, can we have AOG's posts removed when he starts attacking Coggan?

It's becoming quite "OLD"

Paul

 

[Alex Simmons]

Originally Posted by acoggan .


Here are data for FTP, AP, and NP for 21 individuals who have generated "NP busters" (i.e., IF > 1.05 for ~1 h or longer). As you can see, in every case except one the NP is closer to FTP than is the AP.


FTP Duration ave P ave/FT old NP old IF old VI 240 1.179 188 0.78 252 1.050 1.340 280 1.007 135 0.48 294 1.050 2.178 300 1.028 270 0.90 316 1.053 1.170 315 1.001 272 0.86 334 1.060 1.228 330 1.002 315 0.95 350 1.061 1.111 295 0.998 228 0.77 313 1.061 1.373 295 1.001 232 0.79 316 1.071 1.362 315 1.088 267 0.85 338 1.073 1.266 240 1.006 212 0.88 258 1.075 1.217 262 0.983 247 0.94 283 1.080 1.146 250 1.037 182 0.73 273 1.092 1.500 250 1.003 188 0.75 275 1.100 1.463 250 1.003 186 0.74 276 1.104 1.484 280 1.014 235 0.84 312 1.114 1.328 295 0.398 237 0.80 330 1.119 1.392 310 1.002 283 0.91 347 1.119 1.226 250 1.002 201 0.80 282 1.128 1.403 305 1.041 259 0.85 345 1.131 1.332 310 0.995 261 0.84 351 1.132 1.345 250 1.175 146 0.58 284 1.136 1.945 325 0.975 281 0.86 384 1.182 1.367
As for 1 vs. 2 h, you misread the poster's comments: the 373 W was for the highest single hour.

Curious about the red highlighted inclusion given short duration.
and do the first two count, given they are no more than 5% above FTP?

there's something else I'm curious about but another time perhaps

 

[Alex Simmons]

Originally Posted by An old Guy .

You are a fraud.

You are out of line.

 

[acoggan]

Originally Posted by Alex Simmons .


Curious about the red highlighted inclusion given short duration.
and do the first two count, given they are no more than 5% above FTP?

there's something else I'm curious about but another time perhaps

I wondered about that myself when I dredged up the data!

At first I thought it might simply be a typo, but that column contains the formula to convert from minutes to hours, so the value is correct. I therefore think I included it as an NP buster because the individual who provided it had carefully established their power-duration relationship via critical power testing.

As for the 1st two, you could consider the cut-off to be >1.05, or >1.05, so I tossed 'em in there....

 

[RapDaddyo]

And then there's the question of whether the performances are repeatable. If they are repeatable, that can be exploited in a variable power pacing strategy for ITTs.

 

[Alex Simmons]

Originally Posted by acoggan .


I wondered about that myself when I dredged up the data!

At first I thought it might simply be a typo, but that column contains the formula to convert from minutes to hours, so the value is correct. I therefore think I included it as an NP buster because the individual who provided it had carefully established their power-duration relationship via critical power testing.

As for the 1st two, you could consider the cut-off to be >1.05, or >1.05, so I tossed 'em in there....

Visual of the data (minus the 20 something minute effort) showing variance from FTP for average and normalized power for each NP buster:

 

[dkrenik]

Originally Posted by Alex Simmons .


Visual of the data (minus the 20 something minute effort) showing variance from FTP for average and normalized power for each NP buster:

Interesting & thanks for posting. The image looks to validate what AC has been saying for a while: that for highly variable efforts NP comes closer to estimating FTP than AP. How much closer seems to be individualistic. I wonder how this would look comparing AP to xPower?

 

[acoggan]

Originally Posted by dkrenik .

I wonder how this would look comparing AP to xPower?

As I have mentioned before, I tried exponential smoothing when I was developing NP/TSS/the PMC back in the spring of 2003. On average, it doesn't make a significant difference. The same is true re. tweaking the weighting function. Only if you change both, and significantly, does the final value change much.

 

[dkrenik]

Originally Posted by acoggan .


As I have mentioned before, I tried exponential smoothing when I was developing NP/TSS/the PMC back in the spring of 2003. On average, it doesn't make a significant difference. The same is true re. tweaking the weighting function. Only if you change both, and significantly, does the final value change much.

Just for grins I compared xPower to NP on an AC workout I did this past summer. It consisted of ~30+ seconds "on" with ~4-5 minutes rest between efforts. The entire workout lasted 1 hour 12 minutes. My FTP at the time was 250 watts. I've got the following from WKO+ and Golden Cheetah:
AP: 132
NP: 268
VI: 2.04
xPower: 208

I won't pretend to understand the underlying principles in deriving NP and xPower. I'm guessing that as VI gets closer to 1 (how much closer?) the differences between NP and xPower become insignificant.

Is this what you mean by "on average"?

 

[acoggan]

Originally Posted by dkrenik .


Just for grins I compared xPower to NP on an AC workout I did this past summer. It consisted of ~30+ seconds "on" with ~4-5 minutes rest between efforts. The entire workout lasted 1 hour 12 minutes. My FTP at the time was 250 watts. I've got the following from WKO+ and Golden Cheetah:
AP: 132
NP: 268
VI: 2.04
xPower: 208

I won't pretend to understand the underlying principles in deriving NP and xPower. I'm guessing that as VI gets closer to 1 (how much closer?) the differences between NP and xPower become insignificant.

Is this what you mean by "on average"?

No. By "on average" I mean across numerous workouts from multiple individuals.

(BTW, note that differences between WKO+ and Golden Cheetah may arise due to differences in how they handle gaps in the datastream.)