As far as numbers go, in terms of retention, from what I've studied and what that NPS report would lead you to determine is that USNA grads stay in the Navy longer, but as far as the Marines, their multitude of commissioning sources really causes some stir - but it's not all commissioning source to take into account either. You also have to keep in mind quality spread. My labor prof is showing us this study he is doing and it's really interesting and extensive. In short, he took a look at USMC Commissioning Programs (PLC, OCCs, NROTC, USNA, and prior enlisted who were commissioned in another program, the name slips me now.) He also only looked at ground Marines and not aviation ones. If you look at trends in performance at TBS, the priors were typically higher ranked and did the best at TBS. USNA and NROTC guys fared similarly, and the PLC and OCCs faired the worst. Then, he took a look at the quality spread and came up with trends of dissatisfaction based on the rates at which people leave. In the end, getting what job you want in the USMC really seems to matter - infantry, artillery, and a 3rd but I forget what ground job stayed in the longest. Priors stayed in typically only for their 20 years and almost none past that - so by LTCOL and COL almost all are gone. USNA guys leave the lower ranks at the highest rates, and he found that is typically attributed to to outside job opportunites that present themselves. By the time you get to Colonel and the Flag ranks, USNA and NROTC guys really seem to take over and are by and large the ones leaving at the lowest rates. There's a lot more to it, but it was a very interesting study. So the conundrum of USNA is that its education and connections are very valuable to the civilian world where opportunity presents itself to their graduates more often. However, in the highest ranks, they also appear to advance the most along with their NROTC counterparts and stay in the longest. Does that suggest a "good ol' boy's club" for advancement was my question, and he said he was still trying to figure out a way to evaluate that, but I would tend to think that by and large most of you would agree that's probably not the case...