6 - Felt B2
2 - Cervelo P5
1 - Cervelo P5-X

3,4,5 - no idea

---

Make Inside Out Sports your next online tri shop! http://www.insideoutsports.com/

I'm guessing number 6 is Ventum.

It's interesting how asymmetric number 6 is compared to all others.

---

Genetics load the gun, lifestyle pulls the trigger.

Disc brakes?

Cool laser charts.

---

--Chris

My guess is you posted the results in your post:

Premier Tactical
Ventum One
P5-X
P5-6
Andean
B Series

---

--Chris

What is going to be your y-axis on the charts?

---

blog

Honest guess, or just your heart's desire?

Honest guess

---

Make Inside Out Sports your next online tri shop! http://www.insideoutsports.com/

We'll obviously find out soon, but I'm betting that at best you have 1 and 2 reversed. I won't be surprised if you're more wrong than that.

I know the answer!!!

---

Brian Stover USAT LII
Accelerate3 Coaching
Insta

If I'm wrong, my P5-X opinion remains unchanged.

---

Make Inside Out Sports your next online tri shop! http://www.insideoutsports.com/

Good question that I thought a lot about. Grams of drag at 30 mph. There are pluses and minuses to this, but ultimately this needs to be a report that most people understand, and I think those units better lend themselves to comprehension than CdA or CdX

I think that's what Cervelo did

Can you tell us the range of drag from 1-6 or do we have to wait? IOW did bike 1 really blow away the others or does it just look like that?

---

--Chris

I will reverse your 1 and 2 but leave the rest the same.

Ventum One for the win.
Premier Tactical
P5-X
P5-6
Andean
B Series

---

http://www.sfuelsgolonger.com

Hope we get uncertainties with that data, point estimates alone aren't very informative

purely for entertainment purposes, I'm hoping #1 and #2 are the two disc brake bikes.

---

--------------------------------------------
TEAM F3 Undurance

You'll have to wait. Otherwise I would have left the y axis. One of the big question is, of course, how close these bikes actually are, aerodynamically speaking

I'm not sure what you mean. You're hoping for uncertainty? Or you want uncertainty to be quantified and disclosed?

Boo!

---

http://bikeblather.blogspot.com/

In the meantime, we'll just amuse ourselves by inventing our own labels for the axes.

I'm guessing it's median daily UV exposure vs. transverse frame cross-sectional area.

---

Eliot
blog thing - strava thing

I'm not insinuating any foul play with this ....but I'd also like to see an accounting of the $$ collected and where they were spent..

Transparency is a good thing.

---

"Good genes are not a requirement, just the obsession to beat ones brains out daily"...the Griz

I am guessing:

Ventum
Andean
p5x
p5
premier
felt

---

_________________________________
Fit Endurance Coaching - Head Coach|Facebook
USAT L1 Coach | BikeFit Certified Fitter | Contributing Writer - Triathlete Magazine | ROKA

The tunnel time alone was $3,600. The Cervelo P5-X stem quill and seat post were $600.

I don't think I need to go into the rest of the expenses. But they were material.

Still good for the Premier Tactical as it's the cheapest of the bunch, excluding a used Felt.

---

--Chris

What's interesting is that there isn't really a bike that performs well at low yaw without also performing well at high yaw. With your testing protocol, fast is fast.

I don't half want bike 1 to be the Felt. I suspect it will be the Ventum.

Edit.. I wrote 6 instead of 1.

So you put a lot of thought into it and came up with the wrong answer?

I'm guessing the only one I'd buy is a Premier Tactical cuz it's the nicest looking.

Yes, I'm thinking the big ass seat tube of the ventum is causing drag on that side of the graph, thus my prediction... :)

I'll pick the P5 as the fastest.

But why would that be asymmetric?

1. Andean
2. Ventum
3. P5X
4. P5
5. Premier
6. Felt

---

2024: Bevoman, Galveston, Alcatraz, Marble Falls, Santa Cruz

Dunno, that's just my guess!

Wrong answer for you? Yes.

Grams of drag at 30 mph is used by nearly every manufacturer that does or has done comparison testing. Regular people are reading this report, not just armchair aero quants. Like my sister, who did her first half Ironman last year and rides an alloy Novara.

I have no problem publishing the CdA values outside of the report, but for the purpose of the report, it needs to be accessible to a broad audience

What would you want on the y axis? CdA? I've thought about this a bunch of times and I can't think of a significant benefit over the other. What am I missing?

---

blog

HOW ARE WE SUPPOSED TO GUESS WHEN THE AXES AREN'T LABELED?!?!?!

---

------------------------------------------------------------
Any run that doesn't include pooping in someone's front yard is a win.

Yes. Because it's the one unit that doesn't require additional test information to be useful.

"Grams of drag" is not only a misuse of units (drag is measured in force, not mass), but it requires additional information (such as test air speed and air density)...and then most consumers of the data will STILL need to "translate" it into some other meaningful unit or speed for context. Yes, Kiley said "at 30 mph", but many times I've seen that key info not be related upon repetition/other distribution, plus it adds to the confusion of "but I don't ride that fast", etc. Edit: And then there's the problem of manufacturers deciding to test at lower tunnel speeds (ostensibly to be more "realistic", but really to show low gram values) and then not being overly diligent about pointing out that speed difference. It just all adds to confusion.

Yeah "grams of drag" has been the most common units used in reporting this type of data, but that's no reason for continuing its use if there's a better and more clear alternative. All of the above is why Specialized has switched to showing their drag charts in CdA.

Or...you can just report the CdA and allow the user to directly "translate" it into their conditions of interest. Plug in your air speed and density of interest and go. Least steps and chance for misunderstanding.

---

http://bikeblather.blogspot.com/

Don't you hate it when people provide well thought out, logical answers to questions on the internet?

Thats all great Tom. I'm sure Kiley will give you exactly that and you and a few others here will do all the data masturbation you want.

But this is going in Triathlete mag. Not everyone has dentists for parents. They're not going to be plugging anything into anything.

---

"One Line Robert"

The asymmetrical result is odd in this test. Would only have expected that on a bike with one chainstay or a MTB style single fork ("Lefty") or something else large and asymmetrical. Disk brakes can be seen in the data and the chainrings but that's pretty small.

So...tell me, how are they going to be using "grams of drag at 30mph"?

How will they put it into context for themselves? Explain to me how that's somehow more useful??

---

http://bikeblather.blogspot.com/

For example, in Velonews today it was reported that the new Cervelo R5 has "a claimed savings of 44 grams of aerodynamic drag over the previous R5 frame"

That's what I'm talking about...no mention of test speed or air density. No context. WTF does that even mean??

---

http://bikeblather.blogspot.com/

Hello kileyay and All,

Please refresh my memory ...... was the yaw sweep 15 degrees each side?

---

Cheers, Neal

+1 mph Faster

It means nothing to you. We're well aware. I also don't disagree with you, btw.

But why is it so hard for you and others to be self aware enough to understand that it doesnt matter to most end consumers? Give them a graph or number with some sort of rank order that provides context to the relative improvements over each bike, and thats enough (right or wrong - thats not the argument here). Do you think 90% of this sport is talking about air density and speed as it relates to this type of testing? If so, I'd say it'd be in your best interest to spend less time on ST.

---

"One Line Robert"

New album cover for the upcoming 35th anniversary release of Pink Floyd's "Dark Side of the Moon" ?

---

"They're made of latex, not nitroglycerin"

Tom,

I don't know what the audience of this report is, nor do I know the breakdown in quant aero sophistication of the 120+ people who contributed to the GoFundMe.

What I do know is this:

• Grams of drag is easier to observe on a chart; easier to convert on a rough basis into "watts", which is the term I'm betting *most* people understand; and consistent with nearly every other study of this type ever done
• The journalist I spoke with from Triathlete magazine could not conceptualize anything other than minutes over an Ironman, and basically said "la la la la" when I tried to talk through the caveats
• Before undertaking this study, I could not have answered the question stevej asked, and I sure as shit couldn't tell you the equation for CdA
• Specialized publishes the vast majority of their 'studies' on YouTube in seconds over 40k

Here's the context given in the rough draft of the report:



Feel free to help me out with something better

And don't worry, there is a very quick segue into an explanation on why time-distance assessments require much more intense and *individual* quantification. And then we go ahead and step you through that.

10 degrees

And yet, invariably when reports like this are produced there's a cavalcade of folks asking "what does this mean to my split time?"

Sounds like there might be better ways of contextualizing the information, no?

---

http://bikeblather.blogspot.com/

i know you don't actually think you are the ''norm'' (who does on this site anyway), but you are DEFINITELY not the norm. NARPs want data that tells them how much faster they will be over a distance in the easiest, most direct way possible.

...and "grams of drag" does that how?

---

http://bikeblather.blogspot.com/

I know what you mean, but...

NARP = Non-Athletic Regular Person

That person doesn't care about speed or distance, except maybe the distance from their car to their goddamn barcalounger, where they can sit for over six hours each evening and let the cable TV wash over them

well, it translates (for most people) into seconds/k quite easily.

yea, so everyone slower than me. right?

I think that's what he meant. (This is just a random graph I grabbed off the internet to show what eisforurgent meant).

You're nothing if not consistent, Tom. But I like the way Mark Cote handled this question in 2008, which is to simply give Tom A. the data he wants separately from most everyone else

You're joking right? The entire triathlon industry is kept afloat by NARPs.

---

"One Line Robert"

Any more easily than .01 m^2 CdA ~= 1s/km ~= 10W (at a wide range of race speeds)? I'd say "no".

And that's assuming you're sure at what test speed the "grams of drag" was measured of course. And if they don't specifically say it (such as on Cervelo's R5 website today), then you can't really be sure, can you? :-/

Here's the thing...most people have to be reminded of the "rules of thumb" for conversion anyway, so it doesn't really save anything to report in "grams of drag". The advantage of reporting in CdA is that IF someone DID want to do a more detailed analysis, they're ready to go. No other assumptions about the data are necessary.

Let me ask you this...what units of drag data do you think are used as inputs into prediction programs such as "Best Bike Split" and the like?

---

http://bikeblather.blogspot.com/

Christ, Tom. Most people who will see this in Triathlete Mag cant even interpret what ^ is or means. Let alone that entire stretch of numbers and other special characters. You obviously don't get it. And then you launch into all this other shit - that again a NARP or a Fred just doesnt care about. And cant interpret anyway.

You're just pissed the industry wont bend to your will on how to report on such things. Much like your personal war against disc brakes. Its sort of odd really.

FURTHER - Kiley already said he'll give you what you want. Which is more than the bike industry has done to stop delivering disc brake bikes. What more do you want?

---

"One Line Robert"

Not going to put error bars on the drag charts that compare all the bikes, just because it gets too busy and you can't actually see anything meaningful. That chart in the OP is hard enough to see as it is.

But there will be a lot of and maybe even an excessive amount of information on variability

Ummm...you might want to search around some more for where he and Chris Yu agreed that data should be reported in CdA.

If you'd like, I can show you the "standard" output spreadsheet from the Win Tunnel. Wanna guess what units it's in? ;-)

---

http://bikeblather.blogspot.com/

What's actually "odd" is that you aren't answering my questions...and instead just attacking me. Ummm...OK.

---

http://bikeblather.blogspot.com/

I'll answer.



Yes, absolutely.

In many cases here, I am discussing very small differences in drag. 0.01+ comparisons are in the minority. So then we start getting into 0.00x. What's easier to understand, 30 grams or 0.003?

This is the choice for the body of the report and in my opinion it's the right one for the audience. All of the main charts I can easily publish in CdA -- whether that's with the published report or after the fact I can't say. But you will get it eventually, I promise.

Isn't it chainrings, chain, cassette, front and rear derailleurs, and the assymetrical dishing of the rear wheel to accommodate freehub body? Perhaps one bike does a poorer job of normalizing that difference, somehow. Or a "better" job of amplifying it. Or yeah, an assymetry in the frame design itself.

Eye catching on the graph. And interesting how the Andean is the only bike which attempts to fair the chainrings, perhaps they fucked it up and accidentally directed loads of air onto the drive chain or something weird.

Negative x axis on a yaw sweep the drive side, right Kiley?

Nowt better than a bit of uninformed speculation! :-)

Actually now I'm looking closer at the P5x it almost forms a pocket around that area. Maybe it's the p5x!

Decimal points aren't hard to move. You want it in terms of .001 m^2 instead of .01? OK, that's ~0.1s/km. .003? OK, ~0.3s/km.

30 grams of drag? (at 30mph, right? Standard atmosphere? Just checking...)

Let's see...I know that ~0.1 lbs of drag (or, 45.5 "grams") measured at 30mph tunnel speed is ~equivalent to 0.5s/km of time difference. 30/45.5 = ~2/3....2/3 of 0.5s/km is ~0.33s/km...yeah, that's SO much easier :-/

---

http://bikeblather.blogspot.com/

Cervelo was 1,460

The kink in #6 tells me it's a disc bike. I'm going to go with a somewhat unpopular view:

6. P5-X
5. Diamondback Andean
4.
3.
2. P5
1. Premier Tactical

Edit: I'd like to add that there's a chance that the Ventum is #6. I've been told by two people that the Ventum "isn't great" at yaw but I can't think of why it would do well on the drive side and poorly on the NDS.

Can't edit for some reason. I mixed up DS and NDS. Also the rear wheel could contribute to the asymmetry.

I think it would be amazing if Premier came in from nowhere and not only undercut the price of all of the other bikes (they did already) but also beat them aerodynamically as well. Personally I think it's the best looking of the bunch as well.

---

--Chris

I'm thinking along the same lines. The asymmetry in 6 suggests it's a bike with disc brakes and the p5x has the worst integration with brakes that appear to have been an afterthought. Then, Kiley has already suggested he's going to sell his Andean after this season and I can't help but think that the test results here may be a factor. So I'm leaning towards 6 and 5 being P5X and Andean.

I think there have been hints of the Premier Tactical being surprisingly good - there's also a fairly consistent presence of Dan on this forum which makes me think he's very excited about this. Kiley has also been very positive about the Tactical... so it must be scoring well. The P5 has always been considered to be on of the best... so I would expect it still is. So I'd agree with the top 2 as well.

It would definitely be cool to see the cheapest bike be the fastest and the most expensive bike be the slowest. That's going to create some animated discussion if that would be the case.

OMG only on ST would someone argue that "grams of drag" is easy for the masses to understand.

So the rule of thumb is valid if you are riding at 30 mph. How does one get that rule of thumb converted to say 24 or 25 mph?

What most people seem to be doing when analyzing these type of things is that they get to xx amount of watt savings. Let's say its 10 watts saved (at 30 mph) but they ride at 23-24 mph. Then they guess and say its more like 8 watts at their riding speed.

I'm with Tom on this one but I understand why it's being published in grams of drag. Sign me up on the list for someone who would like to see CdA data.

---

blog

I'm saying that I hope that the measurement uncertainty around each data point is included. Otherwise, we can't make statements around statistical differences/significance from just visualizations of the point-estimate data from each bike.

Who's doing the statistical analysis? I'm sure they'll take this into account, pretty stats-101 stuff

And yet most people still will not buy it.

I guess I have to concede round 1...

Are you planning on having the report "peer-reviewed" in some sort of way before publishing it? This would add credibility by having someone (or people) outside of the scope of the test scrutinize it before publication.

Hello kilejay and All,

You pays your money and takes your chances ..... (apologies Mark Twain)

Selections 1 and 2 look very much like a typical P-5 plot ..... A Gambrel roof shape ...

"Gambrel roofs are an elegant, Old World design harkening back to Colonial times and older. The roof design allows for more living space and head room within a compact space. The framing doesn’t have the glaring flaws of other styles but there are air sealing and insulation considerations." [empahsis added]

Sounds like (and looks like) a Cervélo to me ......

---

Cheers, Neal

+1 mph Faster

The more sophisticated analytics are done by a guy who goes as cyclenutnz on the forums. He runs a company called velogicfit.com that makes software to help bike fitters. After the Best Bike Split modeling down by Ryan Cooper, I don't know that there is anyone in the world who has done more time-course modeling for cycling...maybe there is, but I haven't ever seen their stuff. Anyways, cyclenutnz has done all the time-course modeling to convert the CdA data into time savings/gain on a given course, for a given rider with x/y/z characteristics, on a particular day, on a given bike...wow, that's a lot of parameters. No wonder this stuff is hard for people to conceptualize.

To answer your question more directly, yeah cyclenutnz calc'd the 95% CI levels using the baseline vs. control for both bike only and rider-on tests.

In addition, I've put together some 'first day of stats-101', more 'back of the envelope'-like ways of assessing the variability which serve no purpose other than to make this report accessible to a broader range of people.

Absolutely. There will be more than one heavy hitter with analytic cycling chops reviewing it prior to publication.

But that is for the purpose of making it better...more accurate or more refined or better written or with fewer errors. I don't really think that needs to be an advertising point because I think and hope that the report will speak for itself.

Your title for this threw made me spew my beer!

For the general purposes of what you're trying to accomplish and getting info out to the general public, nah your data doesn't really need to be peer reviewed or noted as such. Essentially if you're simply testing bikes to see which is fastest for you under your set conditions, the results will be what it is.

But for "insiders" or frankly anyone who care for the protocols, result validation, process validation; peer reviewed data is the only acceptable data regardless of the acccomplishments/background/expertise of the testers and or said completed data regardless of how expansive the data report turns out to be.

Long post to say it would be of greater value to adversite that your data has been reviewed then to hang your hat on how great a report you think it is. ETA: and it can be a great report, I would just assume to acknowledge that this data has been reviewed would be of great value with an test and the "results"/discussion that is absolutely to follow.

---

Brooks Doughtie, M.S.
Exercise Physiology
-USAT Level II

Look, I'm having several people go through it. At this time I don't feel comfortable saying who. None of them have 'signed up' to be totally behind this study or how we executed it (some weren't there)...I just asked a few smart, experienced people to read it and help and offer any thoughts on how to make it better. We didn't know there was some standard you had to meet or some level of peer social proof you had to display for your testing to be legitimate.

So help me out...what's the standard? How does this study earn legitimacy in the eyes of people who 'care about protocols' if the ability and experience and expertise of the people involved is not enough?

Thank you for the update!
This more exciting than the original moon landing! - I haven't slept or showered in anticipation of the results!

It's great your getting all the data reviewed by knowledgeable people. Promote that much more then the completeness of your report. That improves the credibility of the study.

---

Brooks Doughtie, M.S.
Exercise Physiology
-USAT Level II

I'll also guess 60 grams of drag between #1 and #6 at zero yaw and close to 100 grams at 15(?) degrees of yaw (I never read the original thread so I don't know the protocol).

How many seconds does that save me in an imperial metric 2/3 half Ironman?

Wine.

---

--Chris

I am going to guess just by the colors chosen that 6 is the Felt (white for the baseline) and the Cervelo's are dark and light blue.

I think what Tom and others are saying is that your study borders on a "scientific experiment." Which I, being an engineer myself, would agree with. But that's great! I think that's where your heart lay in the first place; trying to standardize a testing environment, removing as many variables as possible etc. etc. This is essentially what any kind of scientific experiment (on the physics/engineering/aerodynamics side anyway) tries to do.

I think you have inadvertently arrived at this standard you refer to and if you so wish, could complete the "process" and call it an official "peer-reviewed scientific study" and even publish it in "aerodynamics weekly" (or whatever journal willing to publish it).


In the "scientific community" of which you could now call yourself a member of, legitimacy is brought about by transparent reporting of your experiment, and asking appropriate "peers" to review your work prior to publishing. Sometimes when you submit your "paper" to a scientific journal, they will have a board/pool of reviewers and get a number of them to review it. Have you found any similar published articles online? Who reviewed those?

I'm my opinion you have no obligation to undergo any of the above process. I think you've been extremely forthcoming with your intentions at the outset, transparent throughout your testing process and steadfast in sticking to your guns regarding the likely audience of this report and it's level of detail.

However, as an engineer I will always say "YAY Science!!" or as Jessie in Breaking Bad said, "Science BITCH!"

All this talk about protocol, etc...

While this MOP'er won't be basing any purchase decisions on the results - 1 minute here or there doesn't matter to me in a 70.3 or Ironman - I'm willing to bet the results will be more accurate than any testing ever published by Bikeradar or any other tri-related magazine. Looking forward to the results (and yes, I donated a small amount). :)

I also agree that 99% of the folks outside of this thread that see the results, will just want to see a basic graph showing the differences. Nothing more complicated.

I think whatever unit of measurement you used on the graph, if that could then be stated as "= x seconds over 40km" then people (in the main) will be very satisfied.

---

29 years and counting

Thanks for your work on this.


Everyone seems to want to pick this test and you apart about how they would have done it different/better etc. but i haven't seen anyone else step up to the challenge.

So thanks (even if you are an @** sometimes)

Also, I'm betting on the Premier

Do the manufacturers know the results yet or do they have to wait like the rest of us?

---

--Chris

Are those real results, or the cover of the Dark Side of the Moon album?

---

I'm closer to the feathered end of the spear than the point.

It's the red bike. The red bike is always fastest.

---

***

I agree with this. I'm in the market for a new bike and if I'd be looking at the charts, I'd want to know which one is the fastest, and do I like the price and how it looks and then go from there. As to how much faster it truly is will depend on too many other factors to get an accurate number.

I'm also with Tom on this one. It should be published the same as anyone else in the industry would. Have another chart for the people who don't care about that kind of stuff.

---

Make Inside Out Sports your next online tri shop! http://www.insideoutsports.com/

So just like Cervelo and Trek. Which is what Kiley plans to do.

Did you even read the thread?

---

"One Line Robert"

I read some of it last night. Next time I'll be sure to get the cliff notes from you.

---

Make Inside Out Sports your next online tri shop! http://www.insideoutsports.com/

I don't spoon feed.

---

"One Line Robert"

I don't really care.

---

Make Inside Out Sports your next online tri shop! http://www.insideoutsports.com/

You CLEARLY do care. You won't stop posting and participating in threads. You do realize that the ENTIRE industry publishes in grams and not cda values, right? You "don't care" so much that it's become alarming how much you do care.

Do you just follow me around to comment on my posts?

---

Make Inside Out Sports your next online tri shop! http://www.insideoutsports.com/

So, can someone just tell me if this test confirms that I can go back to buying my bike based on price, color and style?

Oh, ha. Thank you for explaining this to me; I didn't realize that's what Brooks was saying.

So yeah, let me clarify: this isn't for the Global Institute of Aeronautical Study Quarterly Journal...or whatever. It's not for academia. This is for regular people. In my opinion, it's the same audience that is reading Trek's or Cervelo's 'White Papers', except it's done by people who don't have a dog in the fight and with greater transparency. I think we can draw a comparison to Flo's tire study, except that this is about bikes.

Personally, I'll be basing my next purchase on the results of this test.

I can't afford to waste time training to overcome the incremental grams of drag/CdA delta between these bikes; running my dental practice and commenting on ST posts about aero gains takes up too much of my free time. The only way I can get any faster is by buying the fastest bike. I just hope the results are out so that I can test ride my new whip before IM LP!

I just "calls 'em as I sees 'em" ;-)

Yeah, some of the problems with using "grams of drag" as a reporting value are that most either don't ALSO include the explanations of "when measured at 30 mph" for the grams value, and also "when traveling at a wide range of race speeds" for the roughly equivalent watts and secs/km translations...or, if they are included, people don't see them or understand them, or just simply forget or ignore them...which invariably leads to the question like Stevej made. I've seen it happen too many times to count.

Not to mention that the original ROT used 0.1 lbs as the drag force value, and using 50g as the equivalent introduces an additional ~10% error (since 0.1 lbs is actually 45.5g)

I understand the desire to "simplify" the results for a wider audience, but after observing this a bunch, I don't think grams of drag is the best way to do that. It adds more confusion and uncertainty than it takes away.

IMHO, Specialized is doing it the most "right"...publish the drag chart in CdA (perhaps include the simple explanation that 0.01 of CdA ~= 10W ~= 1s/km) and then report an overall, yaw- weighted predicted time savings over 40K. Boom. Done.

---

http://bikeblather.blogspot.com/

I'm sorry you wasted the time to create a parody account. Let's try and guess which forum member this might be?

---

Make Inside Out Sports your next online tri shop! http://www.insideoutsports.com/

I'm with you on this. And as SBRcoffee said in post #91, most triathletes who will look at the results just want to see a comparative chart showing which bike has less drag. They don't really care about the numbers beyond perhaps wanting to convert them into seconds per 40km/56m/112m, and that can be done just as, or more easily using CdA.

The Flo tire study gave both charts, grams and CdA. Both charts look identical, so if a relative layman is just looking to see which bike/tire has less drag visually on the chart, they don't really care what the y axis scale is. They just want the data presented as a visual. Given that, you might as well make the unit CdA imo.

Specialized is doing this the right way? If I recall correctly, in their most recent report, they left the units off the axis entirely, and didn't even tell us what the grid lines represented so we could judge on a relative basis. I might as well just leave it blank like I did in the OP and then put a big arrow facing downwards that says "FASTER". Because that's what Specialized does.

I recognize that the aero benefits of helmets and clothing are dependent on the user and how well they fit. Assuming same setup, are the aero benefots of these bikes rider dependent at all?

There is no aero benefit from one bike or another that's going to justify riding a bike that doesn't fit you, if that's what you're asking. The bike only represents about 20% of the drag in the entire system, so it makes much more sense to optimize the aerodynamics (and comfort and sustainability and all that) of your fit and position on the bike than it does for the bike itself.

All that said, these bikes all fit me and fit me well in a standard configuration. I'm pretty much at the bottom of the stack envelope for all of them, and we didn't have to do anything crazy to them to make them work. We used 5mm of spacers here and there and moved the cups 1cm or so fore and aft to bring things into alignment, but that's about it. Those who have done and do serious comparative bike testing tend to set up their test equipment in lower stack configurations with a rider or mannequin that fits well on the bike. There is plenty we can pick on Cervelo for in the P5-X testing, but appropriate size/fit of the equipment to their mannequin is not one of them.

Are the results from these kinds of tests individual? Yes. Interactions occur everywhere between bike and rider. There will always be some differences aerodynamically. But, I think if you execute this kind of testing in a world-class tunnel like A2 or LSWT (with proven repeatability) on a slate of bikes that fit your rider or mannequin and appropriately normalize all the variables you can, then the results you get will apply quite broadly.

For the rest of us who are engineering challenged. I'm assuming 6 has the most drag and therefore is the slowest while 1 is the fastest?

That's right. The y axis is drag. So higher is slower along that axis, while lower is faster. 1 is the fastest bike here.

Should have put the caveat this assumes the bike fits properly. Thanks for the explanation. I'm on a P3C and looking to upgrade in next six months so this will be very helpful.

Since, for the specific test conditions, CdA and grams of drag are directly proportional a simple solution would be to put 2 y-axes so the reader can read the chart in his/her preferred way.

---

Ale Martinez
www.amtriathlon.com

Hello Ale Martinez and All,

Too easy .................. and spoils lots of back and forth posting ..............

---

Cheers, Neal

+1 mph Faster

This teasing and subsequent inevitable slowtwich reality TV are just too sad. I don't even care about these results anymore. Publish them or don't.

the flow study is good for most people and if i remember correctly it includes ,watts , cda graphs and timesaving table in one report
so there is something for everybody ;-)
but i guess its either cda or time saving as somebody that dosnt understand cda likely dosnt understand watts either and just wants to see how much time they save.

Thanks for doing this, and putting up with the haters.

That said, I'm sure that Specialized is going to release their Shiv140.6X in a couple of months and it will be the "fastest bike ever" "designed in the WinTunnel" and beat whatever you put at the bottom of the graph, using their own test protocols.

Then the debate will continue.

---

'It never gets easier, you just get crazier.'

6 is so asymetric that I'm guessing it's the Ventum. Which is a pity, because it's not a bare frame, it's a bike with 1.4l incorporated (but no other integration yet - why they don't split the box for repair kit and solids is beyond me, but might relate to patents). I hope I'm wrong.

1 is probably the P5-6, in which case I hope nobody reads this because I'd like to get a second-hand one and they haven't reached affordability yet. That bike is like the Porsche 911 of triathlon.

---

'It never gets easier, you just get crazier.'

Showboater!

YES!!!

CdA is the only sensible way to present drag data. Fine, if you think that people won't understand it, also present the data in another form. Or 10 different forms. But CdA is the only one that is free of numerous caveats and fudge factors.

I think the public has mostly been *made* ignorant of CdA by it's lack of use.

BTW, I also believe the "time savings" is the best value to report in addition to CdA, because it is less sensitive to assumptions than watts or grams.

Am I missing something about the Ventum? I've seen the bike in person at least a dozen times now and can't recall any asymmetries. Also, supposedly Z frames do well at low yaw and less well at high yaw.

The chainstays are quite different on either side. I'm not aware of anything on the others that would cause that degree of asymetry.

---

'It never gets easier, you just get crazier.'

Disc brakes

And even then that 24-25mph is the pointy end of the field. I would reckon that back 2/3rds are averaging 20mph or less.

---

_________________________________
Fit Endurance Coaching - Head Coach|Facebook
USAT L1 Coach | BikeFit Certified Fitter | Contributing Writer - Triathlete Magazine | ROKA

I don't understand why it can't be published in both cda and grams and whatever else?
Is the internet running out of space? Publish in whatever you want with a hyperlink to whatever Tom wants.

The TEST was in 30 mph. The conversion to a range of race speeds, including yours, is valid by the rule of thumb:

10 grams of drag measured at 30 mph ~= 1 watt at common race speeds

100 grams "" ~= 10 watts ""

How can I communicate this in a way that there is not confusion?

Actually, that's the value that requires more assumptions than any other. And some of those assumptions -- like yaw weights, which Tom mentions offhand as if it's a simple thing -- are quite complex and involved and dependent on other variables that are specific to a given course and rider and day.

There's really no downside to also reporting in CdA. And I already said I would do this in some way shape or form.

And time-course figures are core to the report as well. You will know how much time these differences imply for a variety of courses and days for riders of different ability levels.

But the way the report is written there are certain inferences I cannot make and comparisons I cannot draw (to other research and studies) without putting units in grams of drag. Where possible, I can also report in CdA.

This is never going to be all things to all people. But my hope is that most people, especially those who funded it, can interpret the results from the way they are presented in a way that is meaningful to them

And even then that 24-25mph is the pointy end of the field. I would reckon that back 2/3rds are averaging 20mph or less. //

And your point? I think we can all figure out that if the test is at 30mph and we ride less, then our watt savings will also be less. They have to pick some speed to do these at, so 30 is a good round number that highlights differences a lot more effectively than 20 would. If you did all the tests at 20mph you would likely be within the margin of error most of the time and the tests would be less valid.


I would rather know that I was saving 10 watts going downhill slightly from a test than say 2 watts doing 14mph up a grade. One I could be fairly confident in my savings, the other would be a toss up...

I believe 1 is P5-6 as well ... the graph -10 to +10 looks very similar to the Cervelo White Paper.

I hope that

a) Premier Tactical is the other bike in 1 or 2 ... just because I like the looks and the way the owner communicates

b) The Felt B2 is not No 6 ... this would put an outdated mid-range frame on par with many of the latest "superbikes" - although I am not sure Kiley would use the Andean if it is basically just as aero as his old B2 (assuming the Andean is not No 1 or No 2).

My point was that it would be interesting for once to see scaled findings. Say watts saved at 25/20/15mph

---

_________________________________
Fit Endurance Coaching - Head Coach|Facebook
USAT L1 Coach | BikeFit Certified Fitter | Contributing Writer - Triathlete Magazine | ROKA

Power to overcome aero drag is proportional to V^3. Very simple calculation.

Forget this BS of "watts saved" entirely. It depends on too many unique variables.Triathletes are supposed to be a smart bunch on average, and plugging numbers into an online calculator does not require a high IQ. Plus, anybody who understands "watts" or has a PM should have some inkling of their CdA. If it's .25 for instance, and the aero test shows a .01 difference between one frame and another, then that is 4% of your total aero drag.

If you aren't using yaw weights for other metrics then you don't need them for "time savings" either. The beauty of "time savings" is that it won't change very much for a slow rider vs a fast one, compared to watts or grams which will be 8x higher for a 30mph rider vs 15mph.

But CdA is the best thing to use.

So when do we get to see results? :))

Read further. No, it's NOT only valid at 30mph. In fact, the ROT is estimated for "typical race speeds" (i.e. ~18-19mph on up to mid-to-upper 20s) based on drag force data TAKEN at 30mph. THAT is the part which is commonly left out and/or mis-understood...and you're at least the 2nd person in this thread alone to have done so. I rest my case ;-)

---

http://bikeblather.blogspot.com/

It can be. I like the suggestion of publishing the drag charts with something like CdA on the left vertical margin, and measured "grams of drag" on the right...but, I think I would include a large "measured at 30mph tunnel speed" on the grams label.

---

http://bikeblather.blogspot.com/

Don't use grams as your basis for the reference, use CdA.

Think of it this way: CdA is the actual value of interest being measured by the force balance in a wind tunnel. Otherwise there would be no need to monitor the tunnel speed or the air density. All of those values are necessary to put the output in context. But, once you have the CdA, it's relatively easy to apply that info to a wide range of conditions.

You might have to start out with a quick explanation of what a drag coefficient is and the concept of "drag area", but in the end it will be worth it. I think most who will be interested in this report will already have somewhat of an idea already...and for those who don't, you'll be giving them some additional insight they wouldn't get otherwise, and a tool to use in the future.

---

http://bikeblather.blogspot.com/

As Ron mentioned, scaling doesn't matter much when we get down to time differences.
As an example - using IM Cairns model (with yesterdays weather conditions) and data from Felt IA release, we get the following time differences between the IA and a Shiv

Time difference in m.ss, IA faster in every case.

Adjusted weather factors how conditions change for the slower riders. Unadjusted gives everyone the same weather.
The wind conditions make the table above slightly unusual - normally the numbers for each split are more consistent.

If you want to understand power scaling per speed, I included this handy table in a post a couple of years back:


So a 10w difference for a 4.30 rider (40kph) scales down to 3 and a bit watts for a 6.30 (27kph) rider. But as you see from the first table - the time differences don't scale the same way.

Without giving anything away about Kileys results, you are going to see the same consistency of time advantage across different rider speeds.

Just a suggestion.... you may want to include some math or an example/evidence that shows this. "Race speeds" is still kinda vague IMO. I know the math would be over the head of almost everyone and putting it in the actual report will probably lose most readers. This statement has been put out there several times but I have yet to see someone show the math publicly that proves it.

I've done the math myself and it works out mostly. But for those that are OCD like me, I will say the 100 grams ~= 10 watts is just a generalization. There are those that want more than a blanket statement but I know that's not the norm.

---

blog

Put the math in it and if the reader doesn't want to follow who cares. The readers should strive to be educated and math formulas can help.

---

Make Inside Out Sports your next online tri shop! http://www.insideoutsports.com/

Without reported error, your data is about as useful as an editorial. I know that not that many people understand this...I have pulled out my hair for the past 20 years trying to explain things like this to marketing people who don't understand what science actually is. If you truly want to contribute something, then you need to present the data for people to interpret; otherwise you are just presenting what you have said in the past is just a whole load of unbelievable marketing drivel. You will have become what in the past you claimed to have taken great offense at, and spurred you to take other people's money to do this testing.


The MEANINGFUL part of the data IS the data; not YOUR interpretation of the data (which includes the way that you present it).


Stephen J

---

I believe my local reality has been violated.
____________________________________________
Happiness = Results / (Expectations)^2

I agree. I've been watching this very intently. I understand not putting error bars on the graph to keep it clean. But will the SD (or SE) be provided at least in a table or similar? If the data is messy with massive error then whether the differences are meaningful comes into question.

You guys are complaining about something you haven't seen yet. Kiley talked about all this stuff up front. Wait and see what you get and then complain if you think you didn't get your money's worth (I assume you donated?).

Since I've seen the nearly finished paper, there is reporting of margin of error. It is not being put on the chart simply because it would be messy. There is also plenty of math and time comparisons in there. It is not going to make the most hardcore engineers, and likely many ST'ers, perfectly happy, but I don't think we ever set out to do that. The testing had limited scale that was based on funding and logistics. I personally feel like the objectives were met in the instance. And, like most studies, more questions were also brought to the fore.

---

---

Heath Dotson
HD Coaching:Website |Twitter: 140 Characters or Less|Facebook:Follow us on Facebook

Not complaining, just commenting. I fully agreed there is a rationale for omitting error on the graph. Zero issues there.

Come on guys, lets see the results!

You can thank Tom A et al for the delay. It's going to take me a bit of time to redo everything in CdA. I guess I should have run this by you all first rather than studying every other study of this type ever and following suit

Did any of those guys contribute to the gofundme?

My totally "sucked-from-beer-bottle" guess is that #3 is the Ventum.

Cervelo P5-6 and Tactical #1 or #2.

What did I win?

Schwinn Varsity with Scott DH bars & a Bikestream hydration system?

One note is the kink on the NDS of #3 and #5 looks an awful lot like the kink in the P5 and P5x data from Cervelo's own testing. So, actually, scrap that last prediction. My final answer:

1. Premier Tactical
2. Ventum
3. P5
4. Diamondback Andean
5. P5x
6. Felt

If that ends up being the case it would be a tough choice between the Premier Tactical and the Ventum. Hard to argue with the utility of the Ventum's integrated hydration but the Premier Tactical is also one heck of a value and a well thought out "total package".

Shit you have been around as long as I have

I have a pair of Scott DH bars hanging around w/original elbow pads in good shape.
I remember the bike stream as well.

---

Brian Stover USAT LII
Accelerate3 Coaching
Insta

I can't speak for Tom, but yes I contributed ($100).

---

blog

Ventum is ugly though. Same for the P5X and other like bikes. YMMV.

---

--Chris

As my Dad used to say, "There's always time to do it right."

You're welcome ;-)

---

http://bikeblather.blogspot.com/

Yup...and one of the first ones too, IIRC...

---

http://bikeblather.blogspot.com/

And my dad says, "shit or get off the pot".

Happy father's day!

---

/kj

http://kjmcawesome.tumblr.com/

I don't find the Ventum's integrated hydration appealing. If they came out with an option to mount a standard bottle there plus have some integrated storage I'd be more interested. But they why bother paying a massive premium when you can have the Tactical?

How far away are the results?

The appeal for me with the Ventum's storage isn't that they have integrated hydration - lots of bikes have that - it is the AMOUNT of integrated hydration. I race full distance with Infinit and need two bottles to fuel me for the whole bike leg. I also pick up water from the aid stations which goes BTA. So I already have to put an aero bottle on the frame or behind my arse for fuel, and unless I pick up from special needs I have to do both.

I like it to be simple. With the Ventum I could put the whole lot in the integrated bottle and put water in my BTA cage. Job done, with little or no aero penalty.

Not far. I'm trying to turn around an updated draft by the weekend, and then as mentioned several people are editing / reviewing it. Assuming they turn it around somewhat quickly and don't have major changes I am hoping for the middle of next week.

I can understand that. I race IM with a concentrated nutrition drink but I prefer it to be cool or at least not hot. I pack a bottle in a small disposable cooler with an ice pack to swap at special needs. The "stop" has never cost me more than 10-15sec and sometimes I want something else from my bag anyway.

Is there a front page article being written to go along with the results or will the results just be dumped into a new thread?

Does Monday count as midweek?

---

--Chris

There will be a new thread. I don't know whether Dan will cover it. I'm about to send him a draft, and I guess he'll make a decision then.


I said midweek next week, didn't I? If I didn't, that's what I meant.

You did say next week, but you said it last week, which I would think is this week. But you mean what is now next week I guess.

---

--Chris

Just keep doing what you're doing. Thumbs up emoji or something like that.

I agree. I'm just excited.

---

--Chris

Hello kileyay and All,

Thanks for a nicely run aero project ....

Ta - Da!!

http://forum.slowtwitch.com/...ost=6331699#p6331699

Yes .... Cervélo .... signature plot.

---

Cheers, Neal

+1 mph Faster

Where are the actual results?

Why not put them somewhere where they can be found?

There has been an article on Slowtwitch homepage back in the day

https://drive.google.com/file/d/0B8RrY_TRtezdc2pYSHJ1cVllUjQ/view

The full results were clearly released using the Scientology model.

---

29 years and counting