I was thinking "5 minutes after the first rules were published and people realized the game had dice."
On topic of the OP:
The game may not be adversarial between the various players, but the game contains more than a few adversarial elements. Since the game contains loss conditions, it automatically has the reverse. I "win" when my character doesn't die. I 'win" more when my character manages to advance -- using any metric of advancement -- whether by increasing level, gaining coin, or getting to write down that the local baron owes me a favour.
Further, since the game is a rather simplistic model for a large and complex reality, it invariably has bits that work really well - they reasonably reflect effectiveness and probability of success -- and bits that work less well -- that are not so reasonable. It behooves players who want to succeed to drift toward those rules that provide the greatest chance of success.
As for the provision of advice: it can go too far, certainly. But is the character played by the player supposed to be an expert i.e. would the character be reasonably expected to see the value in a the recommended tactic? Is the player offering advice more skilled? Will the character be closer to its conception if the advice is followed? Did the player receiving advice miss something in the verbal description or forget something that happened 6 weeks ago that the advising player caught or remembers?