My final game at Hexcon was of the former variant. I had Caine2 sitting in the open with Blur. He had to be in the open because I needed to dominate the zone (we were playing Outflank) and force my opponent's hand. This was the situation and at this point in time, I was up either 3-0 or 4-0:
My Skorne opponent had Molik Karn behind the left zone and Morghoul + Croak Raiders behind the wall and on the hill. Given this position, just about his only chance to win the game is to kill Caine. Morghoul received Puppet Master but alone his nuke isn't enough so he needs help. That meant the Croak Raiders needed to hit a DEF20 Caine. 5 of them could reach him and what do you know? Double 6's came up for one of them. Luckily for me the rest of the run, which was a long shot even then, failed and I won by cleaning up the zone in the ensuing turn.
I have just finished watching the Warmachine Weekend finals between Haley2 (Jordan Nach) and Lylyth2 (Brian White). I highly recommend the game. If you haven't seen it, click this link right now and stop reading. Spoilers will follow. Ok, so basically Jordan is having an amazing turn where he manages to lock down 2 Ravagores with a Stormwall sitting on Temporal Acceleration and 4 focus. This is the state at the start of his turn (he's just cast spells and moved Haley way back):
So, what Jordan does from here is send the Stormwall up and it nearly kills off both Ravagores. Now, at this point in time Legion basically has nothing left. At least nothing worth attritioning with. At best one of the Ravagores survives and starts hitting the Stormwall. Sitting on Arcane Shield it's highly unlikely to go down from that seeing as it had taken no previous damage. Basically, I believe the only way Brian can get this game is to kill Haley (blue circle). The only way he can do that is an insane run with Lylyth2 (red circle) feating, either shooting her dead herself or a Ravagore getting her.
Now, this is a fairly rare and -very- different situation compared to what we normally consider when we play. It's also actually not too different from my Caine example above. What Brian realizes is that Lylyth does have a shot at this. It's very unlikely to succeed, but it does exist. What she has to do is kill a model, swift hunter up 2", snap fire and kill another, swift hunter again etc. Eventually she is within range of Haley, thanks to her feat, and she has one shot to kill Haley outright. At Pow12 facing Arm14, she needs 17+ on 3d6 to win. Anyone who's played Stryker2 extensively can tell you that although this is a 2% chance on paper, it feels like a lot more when you play.
Brian made his shot and won the tournament. You can tell on the video that poor Jordan can hardly believe his own eyes when the dice come up and Haley dies. It really is insane, but the question I immediately asked myself was: Could he have avoided this? The short answer is, obviously, yes. The thing is that Legion has nothing left. With the Ravagores being damaged a ton by the Stormwall they can't walk out and shoot Haley and even if they could, the 2" extra move she gained from Telekinesis wouldn't matter (with 18" from Lylyth's feat + 6" move). The only thing that Haley has to worry about is that one shot from Lylyth. By simply camping 2 focus instead of TK-ing, the run is mathematically impossible (now, since one Ravagore didn't die there's always the chance that it can be healed up, survive the free strikes and kill her as well, but this chance is the same regardless of what Haley does).
Eliminating risk is a hallmark of great players. Now granted this example was extreme, but then again a 2% chance means that every 50th game you play, you lose. When considering probability, it's important to consider averages, not the actual probability. Your Caine2 assassination run has a 75% chance to succeed? Great, that means it fails one in four times. Always look for ways to improve your odds and take away your opponent's angles.
Sometimes games boil down to plain dice. My last game in yesterday's tournament for example came down to whether 3d6 could meet 11 or not. As pure a coin toss as you can get in this game. When that's not the case though, there's almost always something you can do to decrease, or even eliminate, the long shot risks that keep winning games every now and then. If you have a choice...
Don't Risk It!
Edit: After I wrote this article I participated in a discussion about Brian's actual chances re: killing Haley. He had 2 shots left when he got within range of her (I was unaware that Lylyth2's feat shot could generate snap fire shots) which with 4 fury translates to 1 shot boosted attack, boosted damage, buy another and boost attack. While these odds are still in Haley's favour (even if both shots hit they average 13.5W and they won't autohit + the chain prior to attacking Haley has to work out), these are way better odds than needing to one-shot her. Regardless the principles above still ring true: By camping 2 focus and meat-shielding Haley with Reinholdt, Junior and Squire she is still incredibly hard to kill (not impossible: Lylyth could take out one of the others and get snap fire on Haley but this is yet another long shot due to lack of fury). Many ways to reduce and eliminate risk!