Proof of OE/Gintani dyno manipulation

[pumper206]

Originally Posted by spdu4ea

3 things...

1st: I'm not affiliated with powerchip, ESS, AA, GIAC, etc. While I'm not free of all bias (if anything, it used to be towards Gintani), I think I'm about as objective as you're likely to find.


2nd: As Evolve mentioned, Jeremy @ Gintani's impressive list of graphs still left out some manipulatable factors. Personally I'd just like to see the correction factors for each run -- graphed if possible (since their own screenshots displayed a certain fondness for photo editing software).


3rd: I apologize for not being as clear as I could have been in my initial post. I see now that many people looked at this graph:



...and thought "Hmm... That is unusual but it doesn't prove anything."



To those people: The probability of this "unusual" distribution occurring naturally is 0.003% (0.00003). In a world with few absolutes, I'll call this proof:

Your normal bell curve:


Where the tuners fall on that curve:

Wow...Statistics/Probability. I haven't seen that since college!! I think that's some of the best proof I've seen on the forum to date. Great effort spdu4ea.

[mastek]

Originally Posted by Kyoshi71

Seriously, don't you get at all embarassed when you need to post such BS? You have any understanding of the word "conjecture"?

If that is BS to you ...then you are far from any help I can type.
And if you happen to have any 'proof' of any points your making or the sad attempt to debunk mine - feel free to post them - otherwise your insults and comments such as "makes no sense" and "wrong again" and "BS" are the very definition of "conjecture"
Do you have any understanding of the word "hipocracy"?

And yes, i am a little embarrassed, ...that i engaged you.

[FStop7]

To me this is the most interesting thing posted throughout this entire debate, and I don't mean just this thread. Thanks for posting it.

[Singletrack]

Originally Posted by mastek

If that is BS to you ...then you are far from any help I can type.
And if you happen to have any 'proof' of any points your making or the sad attempt to debunk mine - feel free to post them - otherwise your insults and comments such as "makes no sense" and "wrong again" and "BS" are the very definition of "conjecture"
Do you have any understanding of the word "hipocracy"?

And yes, i am a little embarrassed, ...that i engaged you.

For the record, I don't agree with any of your statements either. I also don't see how they are the least bit relevant to this discussion. I'm sure there are people out there just like you are describing. But to say that they are in the majority is pretty absurd. No one needs to debunk that, because it isn't a fact; it's conjecture. Just like Kyoshi said.

[jmflukeiii]

Originally Posted by FStop7

To me this is the most interesting thing posted throughout this entire debate, and I don't mean just this thread. Thanks for posting it.

Yes this is nice and good work to the person putting it together. However, this is a very small sample size and doesn't necessarily fall under the idealized normal distribution. And honestly, if someone put the time in, they could build a software that addresses each single revolution per minute more and adjust the fuel supply, ignition, etc. to create a more consistent gain throughout the powerband. I'm not arguing for either side, I really don't care, but I don't think that a bell curve with n=4 is necessarily a good comparison. Whats to say that the results should actually be a normal curve? Maybe there is a limited extreme of variation with AA, ESS, GIAC falling in that extreme (where a majority of tunes would fall), the curve extends far along out to the right where you find OE - this would still mean they had a low probability of producing the consistent returns they did, but it could make it more likely than placing this in a normal distribution.

[J08M3]

^--- I agree with some points. However it does show 3 almost the same, and 1 quite a bit different. It could be that OE knows something none of the other tuners know. One way that can be proven to be true would be by the dyno challenge PC has proposed. Until then, you are correct it doesn't prove anything about how well the OE tune is. But it does prove OE's tune is quite a bit different than the other 3, which by the graph are all fairly similar. That graph in addition to the dyno graphs from OE are combining to show an odd set of circumstances with the OE tune. There's also all the info posted in this thread about how to cheat dyno numbers and what the results may look like if cheated.

[jmflukeiii]

Originally Posted by J08M3

^--- I agree with some points. However it does show 3 almost the same, and 1 quite a bit different. It could be that OE knows something none of the other tuners know. One way that can be proven to be true would be by the dyno challenge PC has proposed. Until then, you are correct it doesn't prove anything about how well the OE tune is. But it does prove OE's tune is quite a bit different than the other 3, which by the graph are all fairly similar.

Very true. Something is obviously much different, be it bending the truth or an extremely detailed tune; either way, graphs won't prove it. Honestly, it is actually very rare to find normal distribution present in anything; it is really just a statistical "theory", if you will, (much like classic supply and demand in economics having very little to do with actual asset pricing in reality).

If they did what I proposed they could do in my previous post (addressing each RPM on its own throughout the entire powerband), software would probably be very very expensive (which it is not).

[J08M3]

I added a couple more sentences while you were quoting me

"That graph in addition to the dyno graphs from OE are combining to show an odd set of circumstances with the OE tune. There's also all the info posted in this thread about how to cheat dyno numbers and what the results may look like if cheated."

[jmflukeiii]

Originally Posted by J08M3

I added a couple more sentences while you were quoting me

"That graph in addition to the dyno graphs from OE are combining to show an odd set of circumstances with the OE tune. There's also all the info posted in this thread about how to cheat dyno numbers and what the results may look like if cheated."

I understand and am in no way arguing for OE. I really don't know much about this situation; I just like statistics and much of what I do is based around it so I thought I'd just chime in because I like graphs

[spdu4ea]

Could you explain this graph:

Sure. Each post-tune dyno graph has variations in the power gained at various rpms points over the baseline run. If a tuner had exactly the average amount of variation, they would be at 0 on the graph. GIAC has almost average variation, ESS has a little less than GIAC. OE on dynojet and AA both have slightly more variation than average. OE on Gintani's dyno had much, much less variation.



What normal distribution looks like... 95.4% of the data should fall within +- 2 standard deviations of the mean. 99.7% should fall within +- 3 standard deviations, and 99.99% should fall within 4, and 99.9999% should fall within 5. Gintani fell outside of 5 standard deviations. (6 is the statistical limit -- anything outside of that is essentially impossible rather than just really really really really unlikely).


Potential criticism:

1) Sample size is too small...

Response: The calculation of standard deviation factors in sample size (small samples will have a larger standard deviation than larger samples of the same distribution).


2) This assumes a normal distribution

Response: It does assume a normal distribution... However, a normal distribution seems to hold when I add additional tuners to the existing graph:



or if I use them to calculate a new standard deviation:




3) There is a dynojet bias to the data

Response: Possible, but as you can see above Evolve's DD data fit the normal distribution -- as did GIAC's load-bearing mustang dyno. However, both of those graphs also included an aftermarket air filter and Evolve's before/after was done on 2 separate cars/days so they are imperfect comparisons (I would of course welcome additional data)

[jmflukeiii]

Good response. The other manufacturers do fall into this normal distribution quite nicely.

[J08M3]

I think Gintani should respond to the graph above. I would love to hear an explanation of why their tune/dyno run is no where close to what other tuners are getting.

[jmflukeiii]

Originally Posted by J08M3

I think Gintani should respond to the graph above. I would love to hear an explanation of why their tune/dyno run is no where close to what other tuners are getting.

Because they are just so goooooooooood will be the response. But as I mentioned, their tune should be $10,000 if they put in the effort required to create what they claim they have created.

[Sales@Gintani.com]

To Mike & Mike's dyno expert: Can you please show all dynos comparing "tractive effort" (measured, not calculated values) for all. (Tractive effort is what a dyno actually records/measures to the wheels. Other variables can be calclated) This is just nonsesse colating data from different dyno brands that read/interpret data in different ways. Tractive effort is measured to the wheel. Please show that for real transparency.
What you have successfully done is let all consumers know that "all dynos" can be manipulated with a little effort (this does not say much for those who practice in this). Why not dissect DynoJet or Mustang next??? Same issues will always exist.
Please be my guest and ask for all details when using my dyno. I have no problem in displaying that the weather station is connected each time for every customer. With the weather station connected and "turned on", the variables that come into question (from other vendors) cannot be manipulated. THE END.
BTW: Powerchip does not own, ever owned a dyno nor operated one. Other posts exist where Powerchip refute dyno testing, and do not believe dynoing is a real measure of their products performance. Please feel free to search.

I remember a "dyno shootout" which I gladly wanted to participate in. This was cancelled due to other tuners backing out. I am down for this to go ahead on an independant dyno anytime. Any one want to jump in the ring for real?

Jeremy

[e46e92love]

I get one thing from threads like this: exhibit A for why I give Dinan the extra money.

Cheers,
e46e92

[Kyoshi71]

Originally Posted by e46e92love

I get one thing from threads like this: exhibit A for why I give Dinan the extra money.

Cheers,
e46e92

+1000
If this isn't the one thing that we can conclude from all of this, I dont know what is?

[J08M3]

Originally Posted by Sales@Gintani.com

To Mike & Mike's dyno expert: Can you please show all dynos comparing "tractive effort" (measured, not calculated values) for all. (Tractive effort is what a dyno actually records/measures to the wheels. Other variables can be calclated) This is just nonsesse colating data from different dyno brands that read/interpret data in different ways. Tractive effort is measured to the wheel. Please show that for real transparency.
What you have successfully done is let all consumers know that "all dynos" can be manipulated with a little effort (this does not say much for those who practice in this). Why not dissect DynoJet or Mustang next??? Same issues will always exist.
Please be my guest and ask for all details when using my dyno. I have no problem in displaying that the weather station is connected each time for every customer. With the weather station connected and "turned on", the variables that come into question (from other vendors) cannot be manipulated. THE END.
BTW: Powerchip does not own, ever owned a dyno nor operated one. Other posts exist where Powerchip refute dyno testing, and do not believe dynoing is a real measure of their products performance. Please feel free to search.

I remember a "dyno shootout" which I gladly wanted to participate in. This was cancelled due to other tuners backing out. I am down for this to go ahead on an independant dyno anytime. Any one want to jump in the ring for real?

Jeremy

But the question of this thread is not about Powerchip. It's about your dyno, which again you have not answered. No one in this thread is claiming PC has a good/better product. They are questioning your dyno

[duk]

Somehow that wasn't the response we were all waiting for.

[Aries932]

I am pretty sure I am correct with regards to defending Gintani and to get solid evidence if Gintani is up for it I would like to offer up a challenge.

Gintani is out of Van Nuys from the looks of things and so am I. So if Gintani or anyone with a Gintani Supercharger is up for it there is a HK MotorSports Dyno maybe a mile down the road from you on Oxnard in Van Nuys that we could meet up at and prove that you are actually making these gains.

I will even pay half the cost of the dyno which is $95 for 3 runs. I will then take the dyno information and post it on this thread.

Seems like the best way to solve this whole stupid arguement. What do you think?

[808MGuy]

Originally Posted by Aries932

I am pretty sure I am correct with regards to defending Gintani and to get solid evidence if Gintani is up for it I would like to offer up a challenge.

Gintani is out of Van Nuys from the looks of things and so am I. So if Gintani or anyone with a Gintani Supercharger is up for it there is a HK MotorSports Dyno maybe a mile down the road from you on Oxnard in Van Nuys that we could meet up at and prove that you are actually making these gains.

I will even pay half the cost of the dyno which is $95 for 3 runs. I will then take the dyno information and post it on this thread.

Seems like the best way to solve this whole stupid arguement. What do you think?

What does the supercharger kit have to do with this thread? That will only add another variable to the argument and cloud the issue even more.

[crummer]

Originally Posted by spdu4ea

Response: It does assume a normal distribution... However, a normal distribution seems to hold when I add additional tuners to the existing graph:



or if I use them to calculate a new standard deviation:

Day four, page seven, we have these two "gems" of information. Building on the original data the OP has now included several other tuners. (Hats off to tuners / members who may have contributed) Analysis from these other sources and the documentation from the OP begins to paint a telling picture of how abnormal the OE@Gintani dyno graph is.

Hats off to the OP for all of the work, I am not sure how many hours you have in this analysis, but your efforts to date have elevated the stature of the community.

Originally Posted by Sales@Gintani.com

To Mike & Mike's dyno expert: Can you please show all dynos comparing "tractive effort" (measured, not calculated values) for all. (Tractive effort is what a dyno actually records/measures to the wheels. Other variables can be calclated) This is just nonsesse colating data from different dyno brands that read/interpret data in different ways. Tractive effort is measured to the wheel. Please show that for real transparency.

What you have successfully done is let all consumers know that "all dynos" can be manipulated with a little effort (this does not say much for those who practice in this). Why not dissect DynoJet or Mustang next??? Same issues will always exist.

Sorry, but I can't help but feel you are a half step behind the analysis here. Assuming the OP is not manipulating data behind the scenes or eliminating random oddities of data he is capturing, the analysis is showing that across 5 or so vendors, the deviation is within naturally occuring limits. If anything, through multiple data sets the OP is determining what acceptable limits are (tractive effort or not) and educating the consumer on how verify their dyno runs are not manipulated.

Originally Posted by Sales@Gintani.com

BTW: Powerchip does not own, ever owned a dyno nor operated one. Other posts exist where Powerchip refute dyno testing, and do not believe dynoing is a real measure of their products performance. Please feel free to search.

Not relevant, don't care. Hopefully they consult dynomometer experts when they do decide to measure their software. Could be a smart business move as their primary business model is to make performance software not operate a dynomometer.

Originally Posted by Sales@Gintani.com

I remember a "dyno shootout" which I gladly wanted to participate in. This was cancelled due to other tuners backing out. I am down for this to go ahead on an independant dyno anytime. Any one want to jump in the ring for real?

Jeremy

That would be a fun time and allot of work , but for now I think the community would settle for a 3rd party assesment of your software to verify your own dyno results.

[ianbiz]

this is good..