Define "reacts to each."
Is that Cu for example (well that is what ALL this judgment is based on) - is all that copper wear metal?
YouTube UOA Testing of Amsoil SS vs Penzoil Ultra Platinum on Ram 2500 w/6.4L HEMI
- Thread starter rstcso
- Start date
Is that Cu for example (well that is what ALL this judgment is based on) - is all that copper wear metal?
Clearly the Amsoil produced more wear metals. It's right there in the samples. But several people in this thread have asserted that those findings are bogus and/or not to be trusted.
So if we had 50 samples that showed the exact same wear metals with Amsoil, and 50 that showed lower wear metals with Pennzoil, THEN you would believe the results? Is that what I'm gathering here?
I had one professor who ended class every day with this phrase: "Always be skeptical!" And he had a great point. But I browse through this forum a lot and I think there are many folks who take that idea way too far.
I'm skeptical, and you didn't even answer my question.
The basic notion is you really can't from singular UOA's choose an oil. IOW can you statistically tell me there is a difference, and if so why?
Another question - how much difference does weather, driving style, idle time make in a UOA?
Where is it statistically determined it's from the oil? That's the problem. Observing something is the easy part, ascribing it to one isolated variable is the much harder part.
And there aren't 50 samples of each here, nor is there a controlled environment for data collection. It's the same thing that makes most "real world" data useless when it is properly analyzed. Either way you wouldn't know if 50 is a proper sample size without analysis in such a highly variable environment.
My statistics professors said that and many other things about........numbers
dnewton3
Staff member
OFFRD -
Please read this:
https://bobistheoilguy.com/used-oil-analysis-how-to-decide-what-is-normal/
30 samples is considered an absolute minimum to understand the Stdev of data streams. 50 samples is much better. This is because of the inherently wild inaccuracy of low sample sets when it comes to calculating Stdev. Without knowing what "normal" variation exits, you cannot understand how much or little things will change with the nuances of daily inputs. Further, one needs to understand how to properly interpret data in macro vs micro analysis.
So, to compare/contrast two lubes, you'd need 30 samples of one product, then 30 samples of another product. That's 60 (sixty) UOAs !!!
Perhaps that helps put LSJr's claims in perspective?
Please read this:
https://bobistheoilguy.com/used-oil-analysis-how-to-decide-what-is-normal/
30 samples is considered an absolute minimum to understand the Stdev of data streams. 50 samples is much better. This is because of the inherently wild inaccuracy of low sample sets when it comes to calculating Stdev. Without knowing what "normal" variation exits, you cannot understand how much or little things will change with the nuances of daily inputs. Further, one needs to understand how to properly interpret data in macro vs micro analysis.
So, to compare/contrast two lubes, you'd need 30 samples of one product, then 30 samples of another product. That's 60 (sixty) UOAs !!!
Perhaps that helps put LSJr's claims in perspective?
LSjr also shills for Pennzoil, he also produced the UOAs, so when it comes to skepticism, theres those two facts.
- Joined
- Feb 3, 2025
- Messages
- 67
I’d like to know how they’re getting that PUP to stay in grade, ‘cause it hasn’t done that for me. Also, aren’t these oil coolers known to impart quite a bit of copper to the oil?
I will defend him here. He's very knowledgeable when it comes to data and the proper methods of obtaining that data. However, most of it gets lost on the average Youtube viewer. That stuff makes videos dull and boring, unlikey to be shared, and so on. I don't like that he does it that way, just to be clear, but he's well versed in data if you speak to him in person. Show that you know a bit about that subject, he'll talk your ear off. Sitting in the pits at an NHRA event a couple years ago, we had a 2 hour long conversation about the use of ionic liquids as friction modifiers in engines, transmissions, and differentials. He broke down everything he did, alongside GM engineers, in testing these at ORNL with a ton of graphical data.
What is the cause of the viscosity deviation you’ve seen?
So we’re the problem? We just can’t understand what he’s doing?
No. This video is just beneath the average BITOGer. Take it for what it is. We are not the target audience.
I’m still confused. Are his conclusions valid based on the data he’s using?
Regarding UOA and wear this is what Lake said about his time in NASCAR:
"We could correlate used oil analysis results with actual engine wear. We did both used oil analysis and measured engine wear on the same test engines, and increased measured wear resulted in higher wear metals in used oil analysis. We could change oils and measure the difference in wear either way accurately."
"We could correlate used oil analysis results with actual engine wear. We did both used oil analysis and measured engine wear on the same test engines, and increased measured wear resulted in higher wear metals in used oil analysis. We could change oils and measure the difference in wear either way accurately."
Not really. He's been to HPL's lab, Valvoline's lab, did videos on Mobil 1 etc. and will pretty much talk to anyone about anything related to oil. He's always gotten back to me with any question I had even when he was with Driven, and still does get back to me if I message him.
So your facts aren't really facts at all.
No. I don't condone using singular UOAs as conclusive data, and I hate that he does this. He knows better, we've spoken about this, but in the world of Youtube, there's often a distinction between what's valid and what's likable/relatable. Lab tests like PDSC, TGA, RPVOT, Teost, etc... and live engine tests on a dyno cell are not available to the average Joe. A UOA is, however. Thus... you use what's relatable. He still does videos with these appropriate tests. I'm not defending this approach to Youtube videos. I'm defending the comment of him being ignorant. He most certainly isn't.
The video should be treated as any other Youtube video which is entertainment only.
He also is not a shill for Pennzoil. He's done more videos with Mobil 1 than Pennzoil, as well as videos with Valvoline, Chevron, HPL, Amsoil, etc....
Which may be true for that environment. As frantic as a race may be it is likely much more controlled than the highly chaotic environment of a normal “real world” OCI.
Very valid point. I personally don't think much of small ppm differences in wear metals for your typical consumer UOA.
If it wasn't for LSJ going to Valvoline people would still be questioning whether it was all marketing, which lets not forget 99% of the people did think it was marketing. I personally think that someone that spent a good deal of time in NASCAR and with Driven likely knows a thing or two about oil. I'm sure he's even spent time with formulators being he worked at Driven.
His videos are another story. I'm sure he's trying to create content and also keep it grounded. Lake's videos sit a tad below Lubrication Explained, which is 100% focused on technical. LSJ's are more mass appeal.
His videos are another story. I'm sure he's trying to create content and also keep it grounded. Lake's videos sit a tad below Lubrication Explained, which is 100% focused on technical. LSJ's are more mass appeal.
Similar threads
