Bob Rouleau

Paul, I think Dell (LVDELL) is a professor of statistics, so you may not be alone here.

Best,

OldGuy

Wow!! We do testing here in the Navy all the time for JSF, F/A-18,V22, AV-8B, and many others
that come to China Lake to take advantage of our expertise. The Air Force, Army, of course Navy, and Marine Corps, and even the private sector in commercial aviation and we dont have ONE statistician!!
GOD to think the 60 years of testing we do and have done is not valid!!! Our tests, probably like Todds, are designed by engineers, executed by engineers and technicians and analyzed by engineers!! How could we have been so wrong??!!

997gt3north

- as a stats guy i new enough to think there was likely another, thus my guess at 1 or 2

- once you learn how to bs, you always leave yourself room

997gt3north

- i lived with more engineers than most humans will every come across in their entire lives with 5yrs in residence at University of Waterloo
- next to statisticians, they were were the smartest at the school right behind the physics students, the actuaries, the combinatorics guys, the cryptography geniuses, the graph theory space cadets, the pure math guys (very scary people that all ended up working at unnamed places in the government), the architects (all lived in one buidling and slept in labs for 6 years)
- the reason you aren't so wrong is that the stats guys have made very simple tests for you to follow and all you have to do is fill in a few numbers and the answer comes back as to whether the plane is ready to fall out of the sky

OldGuy

Sorry there Northy but we (the engineers) design our own tests and I have been here for 21 years
and there are 3500 scientists and engineers here-All very smart. I dont ever remember a statistician
helping our designs. In fact if they were we would never start the test and always be over budget.
We have always said a statistician is just a mathematician with out a job.

Yargk

Maybe what GT3north meant is that some time ago people came up with the theory behind statistics (according to wikipedia, these are people like Pierre de Fermat, Blaise Pascal, Roger Cotes, Laplace etc.) and this theory is easy to follow in practice. So the error analysis that the engineers do was most likely formulated by a mathematician, even if hundreds of years ago.

RR

Todd,

I know you aren't bragging about your exhaust claims. But what I'm saying is that since your exhaust only saw - as you say - an average of 9hp increase over the stock runs - I think according to your graph in another post, it's only a 2.5% increase which is not enough to compensate for other factors.

Basically it is by chance that you gained 9HP and the chances are very good - unless you tested the car/exhaust to answer those 4 questions I asked - that someone who buys your exhaust won't receive that 9HP gain.

Todd/A.W.E.

RR,

Maybe I am still not being clear enough:

30+ runs.


40+ runs.

After warmup, none. Runs were within 1-2 hp of each other.

After warmup and adaptation, on all of them. Runs were all within 1-2 hp of each other.

The fact that the gains were so slim in relation to what we see from a turbocharged car, meant that we had to be extra careful to ensure the gains were real. We are no strangers to dyno anomalies and skewed data due to incorrect testing protocol. These are *real* gains, and if the testing procedures are followed like I outlined above, customers *will* see the same results.

If not, we have a money back guarantee. Just like we do with all our products.

To date, we have *never* had a client return an exhaust due to it not making claimed power on their car.

I suppose I can understand the skepticism on your end. We are quite aware of the lack of "scientific approach" found in the domestic performance exhaust industry. We've staked our reputation on being the engineering house of performance street exhaust manufacturers, and you'll find posts from us all over the internet calling for more rigorous standards and expectations from clients when it comes to claimed power numbers.

After all, if we are acting as our own regulatory agency in an unregulated market, it only pays off for us if potential clients perceive a value. We want people to demand transparent, documented power testing for their Porsche exhausts. That is where our strong talent and desires lie, and where we want to compete, not in marketing/self promotion...

Yargk

According to this website http://www.isixsigma.com/library/content/c000709a.asp, a sample size of 36 is needed to say that the average hp found by that sample is within 1 hp of the actual average with 95% certainty. So there is MUCH greater than a 95% chance that gains were found and there is a 95% chance that the gains are greater than 7-8 hp. I used a standard deviation of 3 hp, an error of 1, and an alpha of .05.

Philip in AL

Todd, I'm probably not the only one here who believes that your time would be better spent by developing the next GT3 product for us instead of being bogged down in an argument that you simply will not EVER win by getting the other person to change his position. Now go develop a good ECU upgrade for us!

Thanks

Todd/A.W.E.

Hah!

I think I have satisfied my curiosity, and now I know where RR is coming from.

Onto better things!

mooty

LOL, i clearly have much to learn

jsdkms

AWE is the BEST - I am with them all the way. Why are you people arguing over 9 HP - Either buy their system or buy someone elses!

Philip in AL

But wait! I started thinking about this whole dyno mess. If we are going to go so far as to argue over a few hp, shouldn't we also consider the addition to performance due to weight savings. There may indeed be none here; however, one of you math heads can probably calculate this pretty quickly.

How many pounds of weight saved would it take to equal 1 hp? In other words, if the idea is measure an increase in performance, shouldn't we also include a possible gain due to saved weight? I understand that in this particular case, there is probably not enough weight saved to equal 1 hp. Nevertheless, how many pounds would need to be shaved off to give the equivalent of a 1 hp gain? Anyone?

944CS

Using 415 hp and 3, 075 for weight, stock GT3 has 7.4096 lbs/hp. If your GT3 was making 416 hp, it has 7.3918 lbs/hp. Take 7.3918 and multiply by 415 and you get 3,067.

If you lost 8 pounds that is equivalent to gaining 1 horsepower