5w-30 compared by LSJr - NAPA, Amsoil, RP, ST

rijndael said:
I agree. This isn't some industry white paper or something in a scientific joural. It's a video for general consumption on YouTube.
Click to expand...
And this is where I sit with it, masses of people not understanding what it actually means and spinning their narratives like wildfire. It’s destructive because it’s fake news essentially when one doesn’t understand the limitations of the testing and their own ability to interpret what it may infer. Lake himself has limitations to that, but because he states his take for entertainment purposes in part it is indirectly throwing shade at other oils. Where is the disclaimer stating the nuances the audience should be aware of?
 
OVERKILL said:
The issue is the use of UOA's and their accuracy/repeatability. As @dnewton3 said in his comments earlier in the thread:

- these tests ignore the statistical variability regarding "normality" (though he does acknowledge the existence of variation, he does nothing to account for it)
- I'm not a fan of his "total wear metals" method; I don't believe adding data values for separate elements is a good way to understand "wear"
- singular UOAs are NOT by any stretch a proper way to compare/contrast one lube to another; small sample sets are rife with variability which cannot be accurately predicted without decent quantity of data (30 samples min)


If I have a Chronograph with 5% variability and I fire one round of each: Hornady Whitetail, Federal Blue Box, Winchester Western, and the results stack like this:
- Hornady: 2,732fps
- Federal: 2,756fps
- Winchester: 2,714fps

And conclude that the Federal is the fastest. This is the same test, with the same parameters.

Would you describe these results as conclusive?

In fact, technically, the Chronograph test is more direct than a UOA, because you are measuring bullet speed directly, while with a UOA, "wear" is being inferred from parts per million contamination measurement of a lubricant with a tool that's blind to contaminants with a size above ~5 microns.

The UOA approach is more akin to doing a tire wear comparison, but determining wear by measuring the amount of rubber and other tire compound constituents collected on a 4x4' section of the track.
Click to expand...
I am picking up what you are putting down. That is what i meant about "as conclusive as could be i na 30 minute video"

As to your analogy of FPS measurements, with the the particular lot # on the box, the gun, etc, the federal would have to be concluded as the fastest......but that really does not tell the whole story does it? What really matters is the group size print on the target. It is really the combination of the ammo and gun and all its variables......just like an engine. Just becuase one gun shoots hooter loads better, does not mean the exact same gun model with do the same thing....I have played that game, and something tells me you have too.

So yes, is the video conclusive to all engines in all applications, no. I also dont see much of a difference in 5ppm of wear anyway. Kind of like other things in science, whereas we take a snapshot of today and conclude that it should always be this way sort of thing.
 
Yeah, where is the trend with multiple UOAs being on a single lubricant to account for noise or variation? 3 at a minimum is usually the standard here.

A single dyno run isn’t telling what the results would be over an OCI for a daily driver not being tracked. Even two of the same series engine may show differing results ; even if there were trends established on this one engine of that series.

We already know racing oils aren’t suited for street use long term, for example. What about use in context?
 
Glenda W. said:
IMO that would be the beginning of an interesting test. A problem though is it’s still only using one(wear) of many things an oil needs to do well to be considered great.
Click to expand...
That’s fair, if he combined it with some of the other tests he’s done in past videos like the sheer stability test and the oxidation resistance test he'd present a much better case for saying one oil is better than another.

And you’d probably see a lot of results where theres not a clear winner but oils that do some things better and others worse like his comparison of Mobil 1 0w-40 oils. Euro FS sheered but had better oxidation resistance than ESP x3 and GT C40 which had greater sheer stability.
 
ATex7239 said:
Hot Take: a small business owner on YT is not going to have the resources and budget of XOM, SOPUS, etc. It is what it is.
Click to expand...

or some guy like me who doesn't have more than a couple hours a day of free time and down time, so YT provides information to those who seek it, right or wrong , it serves a purpose. Just like the women who clean windows with see-through dresses, lol, it is out there and easily available.
 
I like Lake, and Napa synthetic seems like perfectly good oil to me. I bought a case some time back, bought for a great price on sale. My Sentra slowly worked through the case, running on annual changes, and the car never burned any oil, so I'm calling it GOOD! The Nissan 1.8 engine only held 3 quarts of oil, kind of a flaw to my way of thinking, so the fact that it didn't burn oil was a very good thing!
 
ATex7239 said:
Technically, no. But also kinda yeah, if you set out to find out which one’s faster with the equipment and resources you have at your disposal. Does your example or this LSJR video stand up to rigorous peer review, no. But does it present information in an easily digestible format for mass consumption and provide some entertainment value, I think so.
Click to expand...

burbguy82 said:
I am picking up what you are putting down. That is what i meant about "as conclusive as could be i na 30 minute video"

As to your analogy of FPS measurements, with the the particular lot # on the box, the gun, etc, the federal would have to be concluded as the fastest......but that really does not tell the whole story does it? What really matters is the group size print on the target. It is really the combination of the ammo and gun and all its variables......just like an engine. Just becuase one gun shoots hooter loads better, does not mean the exact same gun model with do the same thing....I have played that game, and something tells me you have too.

So yes, is the video conclusive to all engines in all applications, no. I also dont see much of a difference in 5ppm of wear anyway. Kind of like other things in science, whereas we take a snapshot of today and conclude that it should always be this way sort of thing.
Click to expand...

Guys, THIS is where I was going with this:
ltslimjim said:
Yeah, where is the trend with multiple UOAs being on a single lubricant to account for noise or variation? 3 at a minimum is usually the standard here.
Click to expand...

And this is the point @dnewton3 was making as well. We are discussing a sample size of 1, with a test that has known variability. Dave would have liked to have seen 30 tests, I would have liked to see at least 5, preferably 10, of each, then comparing the averages. This, while still not perfect, is far closer to giving us numbers that can be compared, at least in the context of what a UOA is able to tell us for comparative purposes.

Multiple tests of the same lubricant to account for run-to-run variation in both the equipment and the test don't make the video any longer. Sure, they mean there's considerably more work that he has to do, but it then makes the information massively more valuable because it's at least approaching statistically valid.

Going to the gun analogy, if we fire off a box of each, cycling through each as we go, not doing 20 of one, then 20 of another...etc and then average the results, it's far more statistically valid in the same way.
 
I think he approached this with the general assumption that most folks, in their mind, wouldn't even bother to compare the big dogs, Amsoil, etc. to the cheap Warren or rebranded Valvoline. Granted this wasn't anywhere near enough to compare to the detailed tests for oil standards, but to compare the time and resources necessary to do what he did, to shooting off a few boxes of shells, well thats not realistic, is it?
 
srfdude said:
I think he approached this with the general assumption that most folks, in their mind, wouldn't even bother to compare the big dogs, Amsoil, etc. to the cheap Warren or rebranded Valvoline. Granted this wasn't anywhere near enough to compare to the detailed tests for oil standards, but to compare the time and resources necessary to do what he did, to shooting off a few boxes of shells, well thats not realistic, is it?
Click to expand...
I think you missed the point.
 
Last edited:
Glenda W. said:
I think you missed the analogy.
Click to expand...
Not at all. Granted the results of the ammo test is more rigorous, and probably more accurate. That obviously wasn't the case with the oil even though he was using 10's of thousand dollars of test equipment, to say nothing of the time involved. To me it suggested that the multiple specs re: the ad packs were surprisingly right on. The base oil probably is probably where differences would be evident, but how long would that take to come up with a result?
 
OVERKILL said:
Guys, THIS is where I was going with this:


And this is the point @dnewton3 was making as well. We are discussing a sample size of 1, with a test that has known variability. Dave would have liked to have seen 30 tests, I would have liked to see at least 5, preferably 10, of each, then comparing the averages. This, while still not perfect, is far closer to giving us numbers that can be compared, at least in the context of what a UOA is able to tell us for comparative purposes.

Multiple tests of the same lubricant to account for run-to-run variation in both the equipment and the test don't make the video any longer. Sure, they mean there's considerably more work that he has to do, but it then makes the information massively more valuable because it's at least approaching statistically valid.

Going to the gun analogy, if we fire off a box of each, cycling through each as we go, not doing 20 of one, then 20 of another...etc and then average the results, it's far more statistically valid in the same way.
Click to expand...
That is what I was saying with lot number on the box. Shooting 3 out of 20 in the box would not be an accurate test.....to your point.
 
OVERKILL said:
Guys, THIS is where I was going with this:


And this is the point @dnewton3 was making as well. We are discussing a sample size of 1, with a test that has known variability. Dave would have liked to have seen 30 tests, I would have liked to see at least 5, preferably 10, of each, then comparing the averages. This, while still not perfect, is far closer to giving us numbers that can be compared, at least in the context of what a UOA is able to tell us for comparative purposes.

Multiple tests of the same lubricant to account for run-to-run variation in both the equipment and the test don't make the video any longer. Sure, they mean there's considerably more work that he has to do, but it then makes the information massively more valuable because it's at least approaching statistically valid.

Going to the gun analogy, if we fire off a box of each, cycling through each as we go, not doing 20 of one, then 20 of another...etc and then average the results, it's far more statistically valid in the same way.
Click to expand...
Without a proper mathematical analysis of any test like this I wouldn’t buy it either for or against.
 
You must log in or register to reply here.

Similar threads

J
Castrol 5W-40, QS 5W-40, QS UD 10W-30, Amsoil SS 10W-30 Volatility Results - Dec. 2024
Replies
1
Views
2K
edyvw
goblin
Valvoline Restore and Protect - 2017 Hyundai 3.3 V6
2
Replies
29
Views
4K
tired
J
Valvoline Restore and Protect 5W-20 Volatility Results - Dec 2024
2
Replies
34
Views
6K
JAG
Don_M
MIA since 2016 ... I'm back - with a (subjective) observation
2
Replies
25
Views
3K
SimpleSyrup
S
2019 Toyota Corolla SE 65,000 Miles
Replies
10
Views
1K
Snowcrash
Back
Top Bottom