Thorzain had better be as ready and prepared as humanly possible. His build orders and reactions should be so well practised that his hands flow across the keyboard without thought. But it's the mental, not the physical that will be decisive for him. MC has played in front of a roaring, live audience. He has defeated the most venerate names in Korea and strutted confidently to his prize. His mind is sharp and hard with no room left for nerves and distractions. I am worried that he will push Thorzain to his mental limit - attack, fake, juke and harrass, constantly, in each game until Thorzain cracks under the pressure. As Arrian attests in A Finale in Five, a best of five can be a rattling, crushing experience. Against MC, an unshakable veteran with the eyes of a hawk, it only takes one slip for everything to come tumbling down.

Prediction: The way Thorzain played Tyler was conservative but also predictable. As they have never been worth so much, I am not sure how much scrutiny Thorzain's builds have taken. If MC bothers to, he might find more than one hole. But if he underestimates Thorzain, if he enters these games unprepared, he will be hard-pressed to win by experience alone. That said, I do think that MC will be familiar with Thorzain's style, and that he will have the answer to it that Adelscott and Tyler did not. Thorzain is a great player but MC is on another level.

Liquibet Stat: MC: 6249 Thorzain: 1871

Kas and Adelscott are like two strong, stubborn bulls about to butt heads. To generalize, they both play a safe style with only a pinch of harassment. Macro, tweak, and react is how they win games. Because they play so similarly, the winner will be the one with the most refinement. Novelty and gimmicks will have no role in these matches. I expect long macro games with the winner being decided by who edges out the tightest macro and safest defense. If the games go as I expect, it will be a wearing experience for the players. In long, drawn-out games one misjudged cycle of units can be the difference between life and death. Everything counts – each detail under your control has to be refined, every action be with a purpose.

Adelscott defeated two top Terrans to make it to the quarterfinals. His series against Qxc went the distance and showed that he has the mettle of a champion. Kas, however, is a much different opponent. He is not overly aggressive, and is more than content to win by outmanoeuvring his opponent in both macro and micro.

Prediction: Adelscott won more than one of his games against MVP and Qxc by having a clutch defence and, like an avalanche, slowly building his advantage. I do not think Kas will be as eager to finish the fight as Adelscott's past opponents. From what I have seen, he is not prone to rash decision making and has terrifying macro. Kas will take it but it won't be easy.

Liquibet Stat: Kas: 4288 Adelscott: 3770

NaNiwa may have qualified for the TSL as a controversial figure, but those days are somewhat behind him. Instead, he has been largely surpassed in that category by the player he faces in the Round of 8. Since steaming through the third TLOpen as an unknown, CrunCher has steadily added to his tournament resumé, and is no longer the wild card he once was. In his TSL run so far, he has defeated two fan favorite zergs in Mondragon and IdrA in forum-rage inducing fashion. To many aggrieved fans of his victims, CrunCher has come to represent everything that is wrong with protoss. Sweet revenge would come in the form of a PvP loss here.

But perhaps it is a NaNiwa who deserves that distinction. While CrunCher is the consummate protoss troll, mixing cannon rushes and six gates with the turtle'n'max strategy, NaNiwa has long relied on good old fashioned brute force to win his games. Of course, like all top players he can mix it up, and surely will in this series, but if NaNiwa's serial destruction of his MLG Dallas opponents are any indication, the Swedish protoss loves nothing more than a quick and brutal stomping.

Prediction: PvP is a weird match-up to call. The player with the advantage is the player who is most comfortable with the many permutations and reads that are required. It'll be fun to see the phoenix/void ray dynamic we've seen on rare occasions. But most likely this PvP will be won in the trenches with three and four gate micro. On the basis of his MLG Dallas run, the advantage in that category should go to NaNiwa, 3-1.

Liquibet Stat: Naniwa: 7676 Cruncher: 772

BoxeR was widely predicted to be the most likely Korean to fall out of the TSL in the Round of 32, and the same was true in the Round of 16. But the Emperor of Starcraft does not relinquish his throne so easily, and edged two difficult opponents to make this Round of8 appearance. Perhaps it shouldn't be treated as a surprise; this is Boxer after all, but up against yet another strong player, the creeping feeling that his luck will turn is back.

That is what HasuObs is hoping. The second ex-WC3 protoss of BoxeR's campaign, HasuObs has proven himself with maybe the most difficult TSL run so far, defeating HuK and MorroW with just a game to spare. Indeed both of these players have advanced by the thinnest of margins each time. That experience will help both in another long series, which this promises to be. Both players should also be comfortable with the external factors related to the TSL by now and these should be less of an issue than they may have been previously.

Prediction: I've bet on BoxeR's opponents twice now, and I've gotten burned. I'm not going for the trifecta. I think Boxer should have more comfort with the playing conditions than he did previously, and I think he can defeat HasuObs, especially with the help of the TvP practice he put in for his Code A match-up. HasuObs is a strong player, as evidenced by his TSL match-list, but I think the emperor has had a charmed run through this TSL that will continue at least one more round.

Liquibet Stat: Boxer: 6388 HasuObs: 2048