Let us consider two characters, Scylla and Pierre. We're going to ignore gifts for the moment and just concentrate on stats, which break down as follows:
The Value of Knowing Your Opponent
Our system is based on stats representing style and approach, rather than the specifics of the "how". (For instance, the classic Amber DRPG leaves open the question of whether a fight is Warfare, Strength, or Endurance. In a conflict here, you never have to deal with that question.) We leave the "how" to color posing. In a conflict, both players get to pick their approach. They don't know which approach the other person picked until after the comparison is resolved. At that point, the stat they picked influences the poses in the scene (which ought to be consistent with the approach chosen) and the resolution.
Because the stat distribution generally reflects a character's approach to life, not just his combat orientation, it is often possible to guess what stats a character is strong in, by observing his IC actions. For instance, most characters in heavy armor with heavy weapons favor Force, while swashbucklers and courtiers tend to favor Grace, and studious characters usually emphasize Wits.
This IC knowledge, whether gathered through previous conflicts or through observation, can certainly play into your IC thinking about how to approach a conflict comparison, as well as the play of the scene in general. If you are fighting a duel, for instance, you probably want to set the duel's terms to your opponent's disadvantage, and pick your stat in the comparison based on how you think he'll approach the conflict.
In this first fight, Scylla and Pierre aren't previously acquainted, and fight a duel. Logically, in such a situation, each character takes the approach that they know they're best at. So:
Scylla opts to go with Grace because it's her highest stat, a 9.
Pierre uses the same logic, and also chooses Grace, but for him it is a 10.
Because both characters picked the same stat, it is added to itself, so it is Grace + Grace vs. Grace + Grace.
Scylla's Grace + Grace (9 + 9) is 18
Pierre's Grace + Grace (10 + 10) is 20
Modulo any further modifiers, such as gifts or a bad turn of the randomizer, Pierre has a small advantage on Scylla. If he gets lucky, it may be enough for a consequence
The next time they meet, Scylla knows that Pierre is a graceful fighter, and she figures Pierre will choose Grace again. Obviously his Grace is pretty high, and it worked for him last time. Since she knows she probably loses on a pure Grace fight, she decides to fight smart (literally) and chooses Wits (an 8). It's not her best stat, but if Pierre does call Grace, she'll be using her best stat anyway, so she goes for second best.
Pierre does exactly as she predicts, and calls Grace (10) again.
This means the conflict will be Grace + Wits vs. Grace + Wits
Scylla's Grace + Wits (9 + 8) is 17
Pierre's Grace + Wits (10 + 9) is 19
Scylla bet wrong, and played into Pierre's highest set of two stats. Once again, she faces a narrow defeat.
Now Scylla knows that Pierre is very strong in Grace and Wits, so he'll probably call one or the other — most likely Grace, since he seems to dig that. Thankfully, Scylla's sheet is pretty well rounded, so she dips into that depth and calls for Resolve (7).
Pierre could call Wits, but there's no scenario he can see where that helps him, so he calls Grace (10) again.
This means the fight will be Grace + Resolve vs Grace + Resolve.
Scylla's Grace + Resolve (9 + 7) is 16
Pierre's Grace + Resolve (10 + 1) is 11
Pierre's spike has turned around and bit him on the ass, and Scylla has a pretty major advantage against him. Worse, he knows he's in pretty deep trouble for any future fights, since she knows not to throw Grace or Wits.
Scylla's going to call Resolve again, since she has no incentive not to. She's pretty sure it wins if Pierre throws Grace or Wits. Force is the wildcard, but she's betting that if he calls Force, hers is respectable enough to maintain her advantage, so Resolve (7) it is.
Pierre has an interesting call to make. He can guess that she's likely to call Resolve, and he's going to get murdered on it. If he throws strong (Grace or Wits) then he'll lose the same way he did last time, but it will at least be minimizing the failure. He could risk Force on the hopes that her Force is even worse than his. She hasn't used it yet, so it's probably low, but the odds that it's lower than his, at least substantially, are slim. So, knowing how Grace turns out, he decides to go for Wits (9) hoping that his margin there is a little broader.
This means the fight will be Wits + Resolve vs. Wits + Resolve
Scylla's Wits + Resolve (8 + 7) is 15
Pierre's Wit's + Resolve (9 + 1) is 10
Pierre's gamble did not pay off, and he pretty much gets the tar beat out of him, though not as badly as if he had called Force - that would have been 12 vs 4, and would have sucked even worse. At this point, Scylla's got his number despite his earlier advantage, and he knows he doesn't want to cross blades with her again until he's put some points into stats, or picks up some appropriate gifts.
Scylla has spent 7 more points on stats (54 total) than Pierre (47). Those 7 points could buy Pierre's force up to a 6, which would not change anything that's happened, but it would mean that if he threw Force it would at least be 12 vs 7, which is a bit less embarrassing, though it might be more cost effective to split them between his Force and Resolve. Though as a math-inclined reader pointed out, just dumping them all in his Resolve probably best of all.