For the decklist as i ran it at that tournament, I didn't have much I felt like bringing in, so there were only so many cards I was going to take out.  Had I more cards for the mirror, I may have decided differently, but at the time my logic was thus:

Regarding vault: as far as I'm concerned, this doesn't stick around after you cast it, but is a colorless dark rit unless you've got a key.  I'd often prefer to skip a draw to potentially tutor up the win than pay 4 on my upkeep to untap my artifact.
Regarding crypt: my impression of this card is that it's useful for a quick ramping in the early game.  If you don't have any immediate use for the mana, so you hold on to it, or if there a bunch of turns after you cast it, it gets pretty bad.

  Regarding the tutors: the deck has three game winning combos: tinker + counterspells, tez + counterspells, key + vault + counterspells.  The tutors help you put those combos together, and once you do that, the game is over, and it doesn't matter how many cards you spent to do it.  How does this jive with treating the matchup as a regular control mirror?  Because the more mana you have, the more counters you can cast in a turn, and the more likely you are to win a counter war.  Having more mana also lets you not tap out when you cast your scroll, tfks and the like, so you're more likely to cast them rather than just sitting back on untapped blue mana, and your opponent is more likely to let them resolve in the face of your untapped blue mana.
  Additionally, tutors can act as pseudo duresses.  If you don't know what's in your opponent's hand, and they cast a tutor to your one counter in hand, you're put in a tough position.  If they already have a game winner in hand, they could be getting the force to push it through, and you lose if the tutor resolves.  If they already have a counter in hand, they can get the game winner, and you lose if the tutor resolves.  If you think they already have a counter or a game winning card, you're pretty much forced to counter the tutor and bluff further counters in hand.  Essentially, the tutor just duressed your counter.  Being able to put opponents in that situation, especially if you win when the tutor resolves, is good enough to spend a card if it resolves.
  Lastly, having the tutors makes for strong gifts without losing your good cards.

[Disclaimer: this post is not meant to be offensive, but to defend myself and my play.]
My name is spelled Smutko.
I'd ask: what would have made the semi's epic in a way that the quarters were not?  They were both Tez mirrors, in which the small mistakes were the ones that decided the match.  It also took a while.  it even came to coin flips in the end.  i thought it was okay on the epic front.

As for your play attitude / hunger, I think you assumed that my not having played vintage before would make me a bye for you.  I think it made you play sloppily, as you didn't top aggressively enough, and vamped for a draw spell for no reason.  It then put you on tilt when you realized things weren't going right.  I'd put it down less to hunger and more to (a) getting more experience playing more than 7 rounds in a day, which few vintage players have probably ever done in their lives, and (b) being able to get out of your own head and play the way you think you should regardless of round, opponent, or anything else.

Game 1: I mull to 6 and my hand requires that I go for turn 2 tinker, with FoW backup.  The hand has no long game potential, but I'm on the play, which means to beat it, you have to have double FoW, or mana drain mana turn 1 + drain + FoW.  Unfortunately, you had the latter (keep this in mind for later in the match when your luck went the other way).  I was then out of cards, and lost.
sideboarding: out - mana crypt, mana vault.  in - control magic, extirpate.

of note: if the game lasts longer than 6 turns (3 per player), it becomes a control mirror in which vintage experience is pretty mediocre, and knowing your way around magical cards in general, control mirrors in particular, is what will win for you.  I mulliganed aggressively for hands that would do that - survive an early attempt at comboing, and then be good for the long game, i.e. lots of mana, play an attrition war.  Mana crypt and mana vault are in no way useful for this strategy, and if you expect the game to go long (and you should), you should side them out.  If you keep them in, and then lose to mana crypt damage because (surprise!) the game went long, you can't blame luck.

Game 2: This was the game where my inexperience with vintage shined through.  I mana drain a thirst in order to deprive you of cards and cast tez on my turn with counter backup.  You cast tinker, and I incorrectly assumed you'd be getting DC.  If your tinker-brobot was a DC I would have won on my turn.  Instead, it was a titan, and you nuked my lands, so i couldn't cast tez.  But, i still had business in hand, and when i truthed the guy back to your hand, you then (a) discarded it to thirst, and then (b) got a do-nothing with your vamp.  oops.  You let me get back in the game, so I did, and tinkered for the win.

Game 3:  I didn't take match notes or anything, but my memory of the game is this:  We both kept double duress, lands, e truth hands.  mine had a third duress, yours was short on mana and had two drains.  you played mana crypt early, and not for any sort of big effect (top twice for no reason often?).  the game went long, and near the end, we both had depleted hands (due to all the duressing), but you had a top.  You topped into and played the titan with me at 15 (1 fetch, 2 thoughtseize).  i had 3 turns before i'd die to titan beats.  You were at 8, i believe, from the crypt.  I had fetch to get a black source (titan took mine) and a vamp in hand.  you had 0 cards, or maybe 1 but i knew it was irrelevant, i'm not sure, but a top and a fetch.  I did the math:  you had drawn 2 drains, 1 FoW, leaving you with 5 counters left in your deck, minimum (you could have negates, or all sort of shenanigans).  After that many turns, you had ~40 cards in your deck.  top + fetch = 6 looks.  6 * 5/40 = .25 chance what i tutor for resolves, and i've taken a turn off my life to do it.  If i do nothing, and wait for crypt, i have to have it hit two turns in a row = .5 * .5 = .25.  Same chances, but if i sit, i can still topdeck on my next turn.  I played to my outs.

lastly, as for losing with your opponent at 1, i don't think that means you got particularly close.  tez isn't a deck that wins by incrementallly affecting your opponent's life total.  my life total is nearly irrelevant, except for how many turns it'll take your tinkerbot to kill.
Was it unlikely?  sure.  [must be nice.] yeah, it fucking is.  i dont think that means you "played better than me," and I didn't see much of the 'good' you'd so willingly trade for luck.  If you're serious about getting past the quarters next time you top 8, don't look to blame your loss on coin flips on the mana crypt that shouldn't have been in your deck, much less in play for that long.  I'd recommend playing some non-vintage formats.  Seems to help Owen.

Hello.
My name is Peter Smutko.
I play in/around Chicago, mostly t2 and ext.  I'd mostly consider myself a PTQ player.  I played in my first vintage tournament yesterday, and split the finals with T(F)K, and finally learned to tell mana crypt and mana vault apart in my head.