ב"ה
This is a great strategy session
It is indeed very important to know when to go for the mage and when to go for the threats.
But there is also a thing I usually name "damage economy". Things such as a magebane early on or a rot condition on the enemy mage are both good examples of "damage economy". When you have effects that damage the enemy mage over time, you gain advantage the longer the game goes. So while the damage economy purpose is killing the mage, having a damage economy allows you to go after the threats in order to make the game longer so the damage economy work for a longer time. I also use the term "damage economy" when I count offensive creatures by their damage output.
Just for the example, I would like to discuss the "damage economy" analysis of recent Gencon winning books:
2014) With the straywood aviary book, each bird had a damage output, which increased with the number of Rajan furies. A bird coupled with "rouse the beast" is immediate damage, like an attack spell, but the bird staying to hit again in the next round is damage economy (the more birds you have, the greater your damage output every turn). This would probably qualify as a "rush book", so I would say that the only economy it needs is the damage economy.
2015) With the jinx wizard, the main damage dealers were the mage and the wizard's tower. While the mana economy and the action economy were amazing, and he totally won the tournament because of his action economy, he didn't really have much damage economy except for the wizard's tower. The only game he lost was to the priestess with the 4 guardian angels, which his damage economy just couldn't best. That book relied on enough actions and attack spells (and the spellbind quality of the wizard tower to keep his attack spells) to take down what he needed to take down.
2016) Well, I didn't fully understand this book since I never saw a competitive skeleton necromancer without an armory yet, and I really hope I will understand this book better after arcane duels will publish the Gencon 2016 games, but I will analyze by what I read so far about this book. Like most necromancers, this one bring early the Idol of pestilence, which is one of the best damage economies in the game. Combined with a Deathlock, the Idol of pestilence simply set a clock for the game, so your opponent actually have to come to you (and destroy either the deathlock or the Idol or the necromancer before the Idol and the deathlock kill him). As I understood it, the whole point of the book was bringing damage economy, and then go for "kill the threat" until the enemy mage got enough damage from the damage economy to be finished by the attack spells.
Now, sometimes the most important part of your strategy is knowing when to attempt to "Kill the mage". Against players who are too set in the Idea of "Kill the Threat", I usually bring a zombie necromancer. There are 2 possible outcomes to such a thing, either they learn that sometimes they should just go for the mage or they end up becoming Necromancer players themselves. Trying to "Kill the Threat" against a zombie necromancer is a really bad idea, so those who doesn't become necromancer players themselves just learn that they have to rush them. On the other hand, against a druid it is usually a bad Idea to go for the mage before you kill the tree, even with a rush book (though in this case there are some exceptions).
What's much the same as Magic is evaluating the starting position from archetype, and specifics of your book, and your opponents if you know them. Many books can play different kinds of game, but they are best suited to one most of the time. In essence you are weighing up each of your book's playstyles against each that the opponent can pull off, but mostly it's your 'A' game against their's that matters most, at least at the start. Building a book that can start into different ways is generally a design flaw in my view - but building one that can react to the current situation into different game plans is a strength, and many people confuse the two. MW also has the complexity that you can get any card when you want it so you can 'splash' silver bullets in a reliable way that most MtG decks never can.
Well, I really don't think that "Building a book that can start into different ways is generally a design flaw". Most of my books have different openings against different mages. For example, every non-necromancer non-rush book I build got an alternative rush opening to be played against necromancers. I see absolutely no flaw in having different openings against different mages.