www.fram.com/media/1070/fluidfilterrating.pdf
"End-users should also be wary of numbers games – they can also be played with B ratios. B ratios higher than 75 indicate little additional improvement in filtering efficiency. Furthermore, the test procedure to develop the B ratio is valid only for B value up to 75. For B ratios of 75 and higher, there are not enough particles in downstream liquid samples to make counting them statistically significant. Another numbers game has to do with illegitimate comparisons. Even though multi-pass data does lend itself to comparison of filters from different manufacturers, it is still necessary for the end-user to be very alert to various ways in which information can be presented. The end-user should be certain that valid comparisons are made, and that these are supported by recognized and accepted test methods."
"End-users should also be wary of numbers games – they can also be played with B ratios. B ratios higher than 75 indicate little additional improvement in filtering efficiency. Furthermore, the test procedure to develop the B ratio is valid only for B value up to 75. For B ratios of 75 and higher, there are not enough particles in downstream liquid samples to make counting them statistically significant. Another numbers game has to do with illegitimate comparisons. Even though multi-pass data does lend itself to comparison of filters from different manufacturers, it is still necessary for the end-user to be very alert to various ways in which information can be presented. The end-user should be certain that valid comparisons are made, and that these are supported by recognized and accepted test methods."