It's not particularly difficult, but there are a lot of tricks and quirks to it. Probably best to start with the base catalogue for the faction you're adapting and work from there. Take a look at how it's been done and learn from that.
basically a measure of how consistently a faction performed. 0 deviation = they all scored the same, high deviation = many of them scored far from their faction average
so for factions that have very few lists, it's not very helpful. but for factions with quite a few, it could be seen as an indicator of how much the faction's performance varies depending on circumstances
for example, if you compare knights to tau, they each had 27 lists, knights have a higher average and have less deviation. so they no only scored higher more often, they also scored well more consistently
well i cant find any rhyme or reason to how roster faction is determined. some people put the faction of their biggest detachment (the ITC method), some people put the keyword that all of the units in the army share (the rulebook method), some people put the faction of their warlord (the 7e method), and some people apparently pick it at random(?)
Yeah okay, still new units right? Along with the Primaris. Far more than most armies got up until now.
I guess I can understand the frustration it might cause, everyone wants their army to be a special snowflake I guess. But it should be clear by now that most specialsauce will come from unique stratagems, relics and warlord traits, not much else.
I for one stopped playing 40k during 2nd ed. and came back to it tail end of 7th. I didn't like the 7th Eldar codex much, and I don't like the 8th either. They're undoubtedly fine codices to play the game with, but I miss the overabundance of choice that I had in fitting out my units with equipment during 2nd. Won't hear me moan about that too much though, the game has changed so I guess I have to accept that or seek my fun elsewhere.