Quote:
Yes, but if an oil tends to do well under all of these conditions I believe it safe to say that it stands out. PP has done this which is why I like it.
If PP has done significantly better than another oil over several UOAs under the same conditions in the same engine, it might be concluded that PP stands out above the other oil in that engine.
What seems to typically happen here is a driver does a couple of UOAs each on one or two oils in their engine and conclude which is best, often with very little difference in the UOA results. Or, they compare their UOAs with other UOAs done by other drivers on other engines under different conditions and conclude that Oil A is better than Oil B, again with very little difference between the UOA results.
When I look at the data here, I see the vast majority of UOAs with low metals, regardless of engine and oil used. On those relatively few occasions where UOA metals are high, it’s usually due to engine break-in or a problem with the engine. I don’t recall any case where the consensus was that the metals were high because you used the wrong brand oil. (There might be a few exceptions where someone tried an oil quite different than that recommended by the manufacturer and had high metals in their UOA).
Therefore, if one considers all the BITOG UOA data on healthy, broken-in engines, one might conclude that it doesn’t matter what kind of oil is used; there are no oil standouts for the vast majority of engines using typical OCIs. Such a conclusion has just as much validity as the conclusion that Oil A is better than Oil B in my typical engine based on the UOA data at BITOG. This is because the UOAs here contain too many variables to achieve accurate conclusions.
However, given the size of the BITOG membership, we could probably get some reasonably valid data by running a more controlled test. If there were, say, 30+ members who had healthy broken-in engines of the same family and were willing to participate, there might be a chance at finding a preferred oil (if such exists) for that engine family. It would probably take at least 6 months, but it could be done.
In the case of 30 participants, simply take three oils, divide them equally among the members and have each one run at least one (much better to have at least 2) UOAs under controlled conditions. The controlled conditions would need to be things like a new air cleaner, same oil filter, a pre-flush run of the oil to be tested, and perhaps a few other things (like the OCI, and perhaps fuel grade and possibly brand). This would generate much more valid data than the essentially random data that is here now.