Just to add to the interesting-ness of the analysis, from the Velonews story on the stage:
“Leipheimer rode a time trial bike and was 17 seconds ahead at the midpoint time check. Vande Velde, on a road bike with clip-on aero bars, made up 16 and a half of those seconds in the second half.”
.
Just to add to the interesting-ness of the analysis, from the Velonews story on the stage:
“Leipheimer rode a time trial bike and was 17 seconds ahead at the midpoint time check. Vande Velde, on a road bike with clip-on aero bars, made up 16 and a half of those seconds in the second half.”
Yeah…I’m sure they both paced the same too :-/
BTW, by my figuring, if there was only the weight difference (i.e. exactly the same drag for both setups), I’m getting that the extra 0.9 kg should only have been ~7-10s slower overall by itself…or, the entire difference that JV claimed there was between the 2 setups *including *the drag deltas.
Like I said, hmmmm…
Jordan - agreed, ‘best’ is too broad
But a table/article on Zipp’s view on what tyres are most aero for each wheelset would be great
The idea of having different width tyres for front and rear might also seem as innovative (to some) as the idea of the ‘606’ wheelset
And to the guys rambling on about the R5 vs P4 - get your own thread !
Just to add to the interesting-ness of the analysis, from the Velonews story on the stage:
“Leipheimer rode a time trial bike and was 17 seconds ahead at the midpoint time check. Vande Velde, on a road bike with clip-on aero bars, made up 16 and a half of those seconds in the second half.”
Yeah…I’m sure they both paced the same too :-/
BTW, by my figuring, if there was only the weight difference (i.e. exactly the same drag for both setups), I’m getting that the extra 0.9 kg should only have been ~7-10s slower overall by itself…or, the entire difference that JV claimed there was between the 2 setups *including *the drag deltas.
Like I said, hmmmm…
I also get ~10s difference caused by the weight (and not taking into account the worse aerodynamics of the R5ca) on the last 8 km - and JV actually said 20s presumedly including everything…
I also get ~10s difference caused by the weight (and not taking into account the worse aerodynamics of the R5ca) on the last 8 km - and JV actually said 20s presumedly including everything…
Yep…almost all of the 10s difference between the weights (assuming equal aerodynamics) was in the second half of the course (I had only ~1s of difference in the 1st half).
Thanks for the confirmation though…it’s not like I haven’t ever “stuffed” an analysis like this before 
Good luck on getting the promoters to change the results ;-).
Any way you look at it LL was favored over CV. Given that G-C did an analysis and choose weight over aero for the frame and the actual results seem to bear out their analysis, I trust there numbers more. Of course on race doesn’t necessarily mean a lot, but it would be interesting to see G-C math.
Styrrell
Good luck on getting the promoters to change the results ;-).
Any way you look at it LL was favored over CV. Given that G-C did an analysis and choose weight over aero for the frame and the actual results seem to bear out their analysis, I trust there numbers more. Of course on race doesn’t necessarily mean a lot, but it would be interesting to see G-C math.
There’s nothing stopping you from doing your own analysis and seeing for yourself…it ain’t rocket surgery
But yeah…I’d be interested in seeing that GC analysis as well, because right now what I’m seeing so far that tweet by JV doesn’t pass the “smell test”.
I’m suspecting that the biggest factor in that equipment decision was what Christian thought would be fastest
Self conscious roadies who don’t want to too fredly.
You called, sir?
Doesn’t the tweet roughly match the Col De tipping point presentation. IIRC that gave 200 grams being equal to 8%. The tweet is more or less saying that 600 grams is equal to 3%. Those smell roughly the same in GA.
Also are you saying for the last 10 years Christian has underperformed vs LL? After all LL has been the better rider and supposedly had a 40 watt frame advantage.
Styrrell
Doesn’t the tweet roughly match the Col De tipping point presentation. IIRC that gave 200 grams being equal to 8%. The tweet is more or less saying that 600 grams is equal to 3%. Those smell roughly the same in GA.
No, it doesn’t match considering that the drag difference between a P4 and an R5ca is MUCH greater than the drag difference assumed in that presentation.
But, don’t believe me, just do the math yourself. You’ll either see what I’m saying or figure out where I made a mistake. Either way, I’d be happy to see the effort.
Also are you saying for the last 10 years Christian has underperformed vs LL? After all LL has been the better rider and supposedly had a 40 watt frame advantage.
You act as if VdV has never been a decent climber or TT’er…besides there’s no telling their relative states of fitness at that event. If this is really what you’re pinning your opinion on, good luck with that.
“You act as if VdV has never been a decent climber or TT’er…besides there’s no telling their relative states of fitness at that event. If this is really what you’re pinning your opinion on, good luck with that.”
No read what i wrote originally. Decent TT and Climber doesn’t equate to not as good as LL. And LL relative state of fitness was good enough to win the overall at the race.
Its bot that I doubt your math, its that I doubt your madel, but I haven’t seen it yet so I can’t say for certain. My guess is you made the classic mistakes of
A) Assuming their is only one acceleration on the whole course (the beginning). Check out a few powerfiles. Even good TTers vary speed more than that.
B) Assuming that RR doesn’t go up with weight or that its so negliable it doesn’t count.
C) Assuming that the only way weight hurts is due to gravity. Fatigue due to rocking the bike and other minor things don’t ever count.
D) Assume the micro accelerations are so small, even when lugging up a climb that they don’t count.
Admittedly its not a lot to hang my hat on, but whats your explanation for how CVV gave up over 40 watts of aero, and virtually tied an on form LL. Even without the 40 watt handicap, he wouldn’t be expected to be that close.
Styrrell
“You act as if VdV has never been a decent climber or TT’er…besides there’s no telling their relative states of fitness at that event. If this is really what you’re pinning your opinion on, good luck with that.”
No read what i wrote originally. Decent TT and Climber doesn’t equate to not as good as LL. And LL relative state of fitness was good enough to win the overall at the race.
And so was VdV’s, based on his performances throughout the week…so, once again, your point is?? Without power files from both, historical and for the event, you really are just guessing at who “should have” beaten who…
Its bot that I doubt your math, its that I doubt your madel, but I haven’t seen it yet so I can’t say for certain. My guess is you made the classic mistakes of
“Classic mistakes”? Ummm…OK…
The equation of motion of a cyclist is pretty easy to find, or even just come up with from first principles. The model I’m using seems to be good enough even for VE testing…and I’m sure it’s perfectly fine for determining the major effects in this case.
A) Assuming their is only one acceleration on the whole course (the beginning). Check out a few powerfiles. Even good TTers vary speed more than that.
You haven’t done the math on this one either, huh? Let me give you a hint…cyclists, even in an all out sprint accelerate VERY slowly (relatively speaking) and the mass difference being discussed here (i.e. ~1% of the total mass) is exceedingly small. F=ma, but multiplying a small m by a small a get’s you one really small F.
Just because a model doesn’t explicitly account for every minor term, it doesn’t mean that it’s not accurate. Many times inconsequential effects are left out because…well, they’re inconsequential.
B) Assuming that RR doesn’t go up with weight or that its so negliable it doesn’t count.
BZZZT. Wrong answer. Power to overcome rolling resistance = Crr * m * g * V. I assumed a conservative .0050 for the look I took.
C) Assuming that the only way weight hurts is due to gravity. Fatigue due to rocking the bike and other minor things don’t ever count.
Really? You’re going to try to account for that? Puh-leaze…show me it’s even an effect on performance first. Seriously.
D) Assume the micro accelerations are so small, even when lugging up a climb that they don’t count.
Now I know you haven’t done the math. Ask jackmott to fill you in on the micro-acceleration red herring.
Admittedly its not a lot to hang my hat on, but whats your explanation for how CVV gave up over 40 watts of aero, and virtually tied an on form LL. Even without the 40 watt handicap, he wouldn’t be expected to be that close.
Who said 40W of aero? Taking the JV tweet at face value, for the difference in aero of the setups to account for the P4 being 10s faster over the first half of the course, I’m figuring only about 15-22W (depending on the tailwind) for that to happen. The problem with that drag difference, even as low as it is (considering it is comparing a “squarish” road frame vs. a “superbike” TT frame) is that if that’s the case, then the TT bike is in no way 20s slower over the second half of the course…not even half that amount…and with no tailwind, the P4 is actually faster on the 2nd half as well. Personally, I think the aero difference was greater, but then the claim of 20s slower over the 2nd half of the course REALLY doesn’t make sense.
Why can LL be “on form” but VdV can’t be??
All I know is that they were within 1s up the hill…and a simple analysis tends to indicate that, all other things being equal, VdV’s frame choice may have cost him the win despite the weight difference. Show me logically and numerically where that analysis is wrong.
Heck, just show me ANY reasonable model for that course that matches the info in the JV tweet…in other words, what CdA difference would account for the 10s faster/20s slower over the 2 halves of the Vail course given the 0.9 kg difference? Even if there was NO drag difference between the setups, there still wouldn’t have been a 20s difference over the 2nd half of the course from just the weight difference alone.
In other words, attach the 0.9 kg to the R5ca setup and VdV would have only been only 6 to 9 seconds slower over the 2nd half of the course, depending on the tailwind. The only way to match the info in that tweet is for the CdA of the P4 frame to somehow change at the halfway point to have higher drag than an R5ca. Sorry…not buyin’ that…
Doesn’t the tweet roughly match the Col De tipping point presentation. IIRC that gave 200 grams being equal to 8%. The tweet is more or less saying that 600 grams is equal to 3%. Those smell roughly the same in GA.
No, it doesn’t match considering that the drag difference between a P4 and an R5ca is MUCH greater than the drag difference assumed in that presentation.
But, don’t believe me, just do the math yourself. You’ll either see what I’m saying or figure out where I made a mistake. Either way, I’d be happy to see the effort.
Also are you saying for the last 10 years Christian has underperformed vs LL? After all LL has been the better rider and supposedly had a 40 watt frame advantage.
You act as if VdV has never been a decent climber or TT’er…besides there’s no telling their relative states of fitness at that event. If this is really what you’re pinning your opinion on, good luck with that.
You are cracking me up. You seriously think that you have a better idea of what bike to select than the director of a Protour team? Especially a Protour team that is sponsored by Cervelo and undoubtedly is advised about by the engineers at Cervelo about the performance characteristics of the team bikes.
The part that really takes the cake is that the actual splits almost exactly match what the supposed performance advantages would be for the two sections of the course for a heavier TT bike vs. a lighter setup and you still think they selected the wrong bike.
You were saying it might have had something to do with pacing? Do you honestly think that LL or CV don’t know how to properly pace a TT and somehow the splits reflected poor pacing rather than the impact of the bikes?
“You act as if VdV has never been a decent climber or TT’er…besides there’s no telling their relative states of fitness at that event. If this is really what you’re pinning your opinion on, good luck with that.”
No read what i wrote originally. Decent TT and Climber doesn’t equate to not as good as LL. And LL relative state of fitness was good enough to win the overall at the race.
And so was VdV’s, based on his performances throughout the week…so, once again, your point is?? Without power files from both, historical and for the event, you really are just guessing at who “should have” beaten who…
Historically its pretty cut and dried. CVV is a fine rider LL is better.
Its bot that I doubt your math, its that I doubt your madel, but I haven’t seen it yet so I can’t say for certain. My guess is you made the classic mistakes of
“Classic mistakes”? Ummm…OK…
The equation of motion of a cyclist is pretty easy to find, or even just come up with from first principles. The model I’m using seems to be good enough even for VE testing…and I’m sure it’s perfectly fine for determining the major effects in this case.
So you’re saying that if you knew the set up both were riding before the race you would’ve preduicted a tie`
A) Assuming their is only one acceleration on the whole course (the beginning). Check out a few powerfiles. Even good TTers vary speed more than that.
You haven’t done the math on this one either, huh? Let me give you a hint…cyclists, even in an all out sprint accelerate VERY slowly (relatively speaking) and the mass difference being discussed here (i.e. ~1% of the total mass) is exceedingly small. F=ma, but multiplying a small m by a small a get’s you one really small F.
Just because a model doesn’t explicitly account for every minor term, it doesn’t mean that it’s not accurate. Many times inconsequential effects are left out because…well, they’re inconsequential.
Yes but my contention is that their are many more of the small accelerations than you think. You basing your calcs on your (and pretty much my) acceleration rate, not a pro tour rider and my contention is that their are many many more than just 1. Even a small diffedrnce adds up if you add enough of them.
B) Assuming that RR doesn’t go up with weight or that its so negliable it doesn’t count.
BZZZT. Wrong answer. Power to overcome rolling resistance = Crr * m * g * V. I assumed a conservative .0050 for the look I took.
Ok I’ll give you that one, why don’t you just cut and paste the whole model you used?
C) Assuming that the only way weight hurts is due to gravity. Fatigue due to rocking the bike and other minor things don’t ever count.
Really? You’re going to try to account for that? Puh-leaze…show me it’s even an effect on performance first. Seriously.
Like I said classic mistake. you have to balance a bike you rock them back and forth, you move them around curves, etc The mistake isn’t not accounting for them its thinking that if you don’t or can’t account for them that they don’t exist…
D) Assume the micro accelerations are so small, even when lugging up a climb that they don’t count.
Now I know you haven’t done the math. Ask jackmott to fill you in on the micro-acceleration red herring.
Ah yes that classic math that shows heavy wheels go up hill faster because if you have lighter wheels you have to add weight to the frame and pedal in a non steady manner. So just how much heavier of a wheel should’ve LL used to really have put the hammer down? My God.
**
Admittedly its not a lot to hang my hat on, but whats your explanation for how CVV gave up over 40 watts of aero, and virtually tied an on form LL. Even without the 40 watt handicap, he wouldn’t be expected to be that close.
Who said 40W of aero?
Cervelo, when they came out with the S5, it wasn’t exactly this situation, but reasonably close S5 vs Trek SC. 3% vs level.
Taking the JV tweet at face value, for the difference in aero of the setups to account for the P4 being 10s faster over the first half of the course, I’m figuring only about 15-22W (depending on the tailwind) for that to happen.
I thought you didn’t believe his math now your using it???
The problem with that drag difference, even as low as it is (considering it is comparing a “squarish” road frame vs. a “superbike” TT frame) is that if that’s the case, then the TT bike is in no way 20s slower over the second half of the course…not even half that amount…and with no tailwind, the P4 is actually faster on the 2nd half as well.
*Opps now he’s wrong again *
Personally, I think the aero difference was greater,
Now your back to agreeing with Cervelo
but then the claim of 20s slower over the 2nd half of the course REALLY doesn’t make sense.
And they are wrong again
Why can LL be “on form” but VdV can’t be?? ''Where did I say that? They both can and I suspect that the first 2 riders on any Protour TT are on form. Just that LL would be expected to beat CVV straight up most days given equal equipment. Spot Levi 20-40 watts or whatever you think it is and VdV wasn’t on form, he would’ve had to have had a once in a lifetime performance.
All I know is that they were within 1s up the hill…and a simple analysis tends to indicate that, all other things being equal, VdV’s frame choice may have cost him the win despite the weight difference. Show me logically and numerically where that analysis is wrong.
If G-C releases their math use that. Not too long ago you were arguing that their math for the S5 was pretty rock solid, did they get stupid in the last month or two?
Heck, just show me ANY reasonable model for that course that matches the info in the JV tweet…in other words, what CdA difference would account for the 10s faster/20s slower over the 2 halves of the Vail course given the 0.9 kg difference? Even if there was NO drag difference between the setups, there still wouldn’t have been a 20s difference over the 2nd half of the course from just the weight difference alone.
Again my point is we dont race computer models. I could care less about the model, real life works pretty well in this case.
In other words, attach the 0.9 kg to the R5ca setup and VdV would have only been only 6 to 9 seconds slower over the 2nd half of the course, depending on the tailwind. The only way to match the info in that tweet is for the CdA of the P4 frame to somehow change at the halfway point to have higher drag than an R5ca. Sorry…not buyin’ that…
Like we started on this , good luck getting them to change the results.
Styrrell
Just to add to the interesting-ness of the analysis, from the Velonews story on the stage:
“Leipheimer rode a time trial bike and was 17 seconds ahead at the midpoint time check. Vande Velde, on a road bike with clip-on aero bars, made up 16 and a half of those seconds in the second half.”
Yeah…I’m sure they both paced the same too :-/
BTW, by my figuring, if there was only the weight difference (i.e. exactly the same drag for both setups), I’m getting that the extra 0.9 kg should only have been ~7-10s slower overall by itself…or, the entire difference that JV claimed there was between the 2 setups *including *the drag deltas.
Like I said, hmmmm…
Looking at his tt build, I truly wonder if it weighed 6.8Kg or is really lighter. I wonder if there are ways to get around this, especially since many riders now swap to different bikes during climbing stages, within the actual stages after starting with other bikes.
You are cracking me up. You seriously think that you have a better idea of what bike to select than the director of a Protour team? Especially a Protour team that is sponsored by Cervelo and undoubtedly is advised about by the engineers at Cervelo about the performance characteristics of the team bikes.
Show me the math 
Oh wait…perhaps JV was talking about a different sort of “calculus”…hmmm…interesting thought.
I’ve done the simple physics and the JV tweet doesn’t pass the “smell test”…what are you bringing to the table besides a second-hand appeal to authority?
The part that really takes the cake is that the actual splits almost exactly match what the supposed performance advantages would be for the two sections of the course for a heavier TT bike vs. a lighter setup and you still think they selected the wrong bike.
Except…it doesn’t match. Like I described, if one adjusts the CdA down for the P4 setup such that it gains 10s over the first 8k of the course (with the given mass difference) then it doesn’t “give up” 20s in the second half of the course. Even if there was NO AERO DIFFERENCE at all between the 2 setups, then the heavier setup doesn’t even give up 10s over the second half of the course. The only way for that to happen is if the CdA of the P4 got suddenly worse (or the R5ca suddenly better) halfway up the course. Yeah…I don’t think the P4 has an “airbrake” option :-/
It’s simple math. Do it yourself (or google up a spreadsheet). That’s a much more productive approach than just attacking and ridiculing me for having done so myself.
You were saying it might have had something to do with pacing? Do you honestly think that LL or CV don’t know how to properly pace a TT and somehow the splits reflected poor pacing rather than the impact of the bikes?
That was in response to the relative splits of the riders, NOT the predicted splits of the 2 setups…there’s a difference.
"Except…it doesn’t match. Like I described, if one adjusts the CdA down for the P4 setup such that it gains 10s over the first 8k of the course (with the given mass difference) then it doesn’t “give up” 20s in the second half of the course. Even if there was NO AERO DIFFERENCE at all between the 2 setups, then the heavier setup doesn’t even give up even 10s over the second half of the course. The only way for that to happen is if the CdA of the P4 got suddenly worse (or the R5ca suddenly better) halfway up the course. Yeah…I don’t think the P4 has an “airbrake” option :-/
It’s simple math. Do it yourself (or google up a spreadsheet). That’s a much more productive approach than just attacking and ridiculing me for having done so myself. "
Tom, simply put you’re saying I’ve got a calculation and Cervelo’s made a mistake. I’m not doubting that the above is what you;re calculation shows.
All I’m saying is that Cervelo is a pretty smart group. They published the Col De Tipping point, you once said that it wasn’t correct. Damon Rinard recently confirmed that yes, it was. Now they choose equipment based on some calculations, which looks like a good choice from the results, and again you think the calculation was wrong, and the results were due to something else. Thats fine, my arguement is that Cervelo’s model seems better than your model. I haven’t seen either of them, so there isn’t much more I can reasonably say.
Styrrell
Tom, simply put you’re saying I’ve got a calculation and Cervelo’s made a mistake. I’m not doubting that the above is what you;re calculation shows.
No…that’s not what I said. I said that a simple look at the physics of the situation says that there’s something amiss with the JV tweet. It doesn’t match what’s physically possible.
I’ve also asked others to do the calculation themselves to either confirm that or point out where I made a mistake. MTM’s quick look at just changing the mass for a given system (i.e. NOT changing drag) basically confirms what I’m saying.
All I’m saying is that Cervelo is a pretty smart group. They published the Col De Tipping point, you once said that it wasn’t correct. Damon Rinard recently confirmed that yes, it was.
Again, you’re misconstruing things. The conclusions of the “Col” presentation have always been correct, and I’ve never said otherwise. It’s just that the one slide which showed the CdA difference at first showed a value of “.009 Cw” (you can search webarchive if you’d like to see) and then later (actually after I asked what those units of “Cw” were supposed to mean) was changed to .0045 m^2…but the results weren’t changed. Folks were taking the .0045 m^2 number and extrapolating what that meant as power savings between the 2 bikes…and then saying that Cervelo was claiming different drag values between the bikes over time. However, as Damon confirmed, and you are well aware of since it was pointed out to you previously, that .0045 value was incorrectly changed. The original value, and the value used for the conclusions of the study, was .009 m^2.
Now they choose equipment based on some calculations, which looks like a good choice from the results, and again you think the calculation was wrong, and the results were due to something else. Thats fine, my arguement is that Cervelo’s model seems better than your model. I haven’t seen either of them, so there isn’t much more I can reasonably say.
Since I think it would be valuable for you to learn some of this stuff yourself, I’ll point you to a spreadsheet you can play around with yourself:
http://whitemountainwheels.com/SpeedPower.html
I think you should be saavy enough to be able to scare up the course profile on your own
Get back to us after you play around with that for awhile. That spreadsheet incorporates the same basic model I use.
BTW, how can you make a judgement that “Cervelo’s model seems better” when as I pointed out above (and MTM confirmed) even if you merely added 0.9kg to the R5ca setup (i.e. there was NO aero difference) one STILL wouldn’t lose 20s over the last 8k of that course? Heck, it wouldn’t even be 10s slower. That doesn’t sound too accurate to me. Of course, that’s assuming that the contents of the JV tweet were actually the results of any Cervelo modeling. I haven’t seen the actual results of any Cervelo modeling either…only what was reported in the JV tweet, which as I’ve pointed out ad nauseum, doesn’t pass the “smell test”…