I agree with Marius. However, we're missing some key data. So, I'll assume the following:
Fighter #1: Normal speed Full Attack the Evil Mage
Fighter #2: Normal speed Full Attack the Evil Mage
Evil Mage: Snap Disengage (25%), then 50% Move, then 25% Move
After the snap phase, the mage is 10' away.
After the mage's action in the normal phase he's 50' away (assuming Run pace with 50' BMR).
The fighter's actions are no longer possible.
After the Deliberate phase, the mage is 85' away (10' disengage + 50' move + 25' move).
The fighters can now move 50' towards the mage.
Simply put, the fighters foolishly wasted time looking for an optimal opening while the mage scampered.
Now, consider a different declaration:
Fighter #1: Snap speed Pressing Attack the Evil Mage
Fighter #2: Normal speed Pressing Attack the Evil Mage
Evil Mage: Snap Disengage (25%), then 50% Move, then 25% Move
In the snap phase, the mage moves 10'. Then, the first fighter presses 10' (with a -5 penalty for movement on top of the -20 penalty for snap).
In the normal phase, the mage is still in melee, so his movement action is canceled. The second fighter presses 10' (with a -5 penalty for movement). Finally, the mage's third move action is canceled.
At this point, the mage can choose 50% movement. I might be kind and allow that to be treated as a 25% disengage, but I'm all for running away to fight another day (from either party).
In short, a mage might be able to evade a single fighter in pursuit, if the dice are kind. But, he's not going to avoid being cut down by a pair of pursuers.