The Rockets have been preparing for this all season -- longer, even -- and here is what they face: a Warriors team that is 24-3 in the playoffs since signing Kevin Durant. Houston has to beat them four times in seven games. Golden State with Durant has lost three times in six playoff series. Last season's Cavaliers might have been one of the greatest single-season teams ever, and certainly were one of the greatest offensive teams ever. They got one game.

But these Rockets were built over the summer, and then tweaked midseason, specifically for this series. They are 50-5 when all three of James Harden, Clint Capela and Chris Paul play. By any measure, they represent the most profound challenge the Warriors have faced in the Durant-Stephen Curry era.

They can match any style: fast, slow, big, small. Hell, they might be able to dictate the style of this series. Some will cast this as a battle for the soul of basketball: isolation showboats and the math geniuses behind them against the protectors of the beautiful game. There is a kernel of truth there. The reality is Golden State has the tools to win both sorts of games. Durant has inoculated them from any stylistic limitations. They have seen every switching scheme in existence, and know how to break them with smart cuts and passes.

Let's dig into some of the major questions and X-factors:

• Houston can surround Harden with four good defenders, and Golden State usually offers one safe-ish place to hide him. The matchups will be fun, even if they switch in a blur on every possession.

We don't even know yet who the Warriors will start: the Death Lineup, with Draymond Green at center, or a more traditional lineup probably featuring Kevon Looney in place of Andre Iguodala. If Golden State starts small, Houston can stash Harden on Iguodala, slot Capela onto Green, Paul onto Curry, Trevor Ariza on Klay Thompson, and the immovable PJ Tucker onto Durant. Against bigger lineups, Houston could match up across positions -- with Harden on Thompson (gulp) -- or slide Harden onto Green, or even Looney.

I would bet (reluctantly) on the Warriors starting Game 1 with Looney. He defended Anthony Davis well, and it just seems like the Steve Kerr move: start with someone who can body up Capela, and keep one ace in the hole. Golden State is also thin on wings -- centers compose almost half of the roster for some reason -- and starting all the best ones deprives the bench of some pop. Smart staggering can mitigate that, but the Warriors are also concerned about overtaxing Iguodala.

• Whether Golden State's centers serve any purpose in this series will be one of its pivot points. Only JaVale McGee -- and basketball gods help any coach who counts on McGee in this round, against this opponent -- can do the Rudy Gobert/Pau Gasol "stand near the rim and raise your arms" thing and bother Harden. None of them can reliably switch onto Paul or Harden. Looney will pull it off one of every three or four times, but he's walking a tightrope. Kerr might try Jordan Bell at some point for added speed.

The Warriors know all of this, and have sometimes gone a third route: trapping Harden with their centers.

That seems like a terrible idea. Harden and Paul will split some traps, or slip passes to their screeners, who can then work 4-on-3 with shooters around them. But ever since David Thorpe suggested it on my podcast Friday, I have come around to it as an interesting tactic -- in small doses: remove the ball from Harden and Paul, and force Houston's screeners to make plays.

Some of those screeners -- Tucker especially -- are pretty good at that. Capela has improved, but not to the point where he scares you catching the ball 20 feet out. And in at least some alignments, the spot-up guys flanking them will be slightly above-average 3-point shooters -- not elite marksmen. Fly at Ariza, Tucker and Luc Mbah a Moute, and they will miss a decent chunk.

• That is one of the most basic swing factors: Whose blah 3-point shooters make more? Houston will leave Iguodala and Green open. Golden State will do the same with the above Rockets, though perhaps not as blatantly.

• Related: I can't wait to see how daring Houston is playing its best shooting lineups. The Rockets are impossible to guard with all three of Paul, Harden, and Eric Gordon on the floor. But Houston sacrifices a little defense to pump up its long-range shooting. Do they even dare play Gerald Green against an offense that will force him to make a half-dozen quick-hitting decisions on every possession? (Ryan Anderson has been excised precisely because he has such a hard time in this matchup. They might dust him off if they get desperate for scoring, but otherwise I'd be surprised to see much of him.) Do they try Ariza at power forward?

They should absolutely dial up the shooting quotient when two of Golden State's stars are on the bench. But against the Warriors' best lineups, Houston might need at least two of Ariza, Mbah a Moute, and Tucker. They might need all three, and they have outscored opponents by a preposterous 25 points per 100 possessions in 222 minutes with that trio, per NBA.com. Interestingly, only 15 of those minutes also featured both Paul and Harden.

That lineup -- Paul, Harden, Ariza, Mbah a Moute, Tucker -- could work, and I'd wager the Rockets use it. Paul and Harden bring enough star power to compensate for the so-so shooting. Still: Any other lineup featuring those three wing stoppers will have trouble scoring against Golden State.

• Back to Golden State's centers: Looney has been the most stable defender among them. The Warriors played around with dropping him back against Harden and Paul, and he managed:

David West will not be able to manage consistently. Houston is going to attack him on every possession he plays. The Warriors could hide their centers on Tucker or Mbah a Moute, and slot Green or Durant onto Capela, but then what is the point of even playing those centers? They offer nothing defending wing players, and it's not as if you are riding out their defensive limitations because of what they bring on the other end.

The Warriors need a ton of Green at center, regardless of whether they start games that way. Those lineups feature four guys who can credibly defend any Houston player, including both Harden and Paul. They are superfast, and the Warriors have some interest in pushing tempo.

• Houston is vulnerable to calculated transition rushes. The Rockets often set up their offense with a shooter in each corner, and Capela under the rim. If Harden drives and misses, the Warriors have a chance to run with four Rockets loitering way below the foul line. That is death. Against most teams, Houston's floor balance doesn't matter. Against the Warriors -- and especially the Warriors with Green at center -- it matters. One of those corner shooters is going to have to loop back toward half court earlier than usual.

• A related battleground: turnovers. Houston has coughed it up on only 9.8 percent of its possessions in the playoffs, a comically low number that would shatter the all-time single-season record. That is one benefit of slowing down and playing isolation basketball: It's hard to turn it over when you don't pass. If that number stays in the same ballpark, the Rockets increase their chances. Ditto if the Warriors suffer too many bouts of the sort of arrogant sloppiness that has occasionally undone them.

• With all of Thompson, Iguodala, Durant and Green on the floor, Golden State can easily hide Curry on someone among the Ariza/Tucker/Mbah a Moute group. He will probably start on Paul, because Curry is a prideful -- and somewhat underrated -- defender, and because Kerr usually gives Curry a chance to guard his own number before pivoting. He will defend Gordon a lot, too. Against some Houston lineups, there won't be a better choice. Gordon has to impact this series with bullying drives on Curry, blow-bys against Green on switches, and pull-up 3s.

But at some point, Curry will slide over to a less-threatening player. Thompson, Iguodala, and Durant can split the Paul/Harden assignments. It will be fascinating to see how Kerr deploys Durant on defense. Thompson has usually drawn the Harden assignment. If Curry is hiding on a wing type, that would leave Durant and Iguodala left as primary defenders on Paul. Guarding Paul over long stretches might be too much for Iguodala. We might see more of Durant on Paul than normal matchups would dictate. (He guarded him at times in the regular season.)

Golden State might even flip those assignments around, and have Durant check Harden while Thompson hounds Paul. Remember: The last time these teams met in the playoffs, Thompson had enough trouble with Harden's battering ram physicality that Kerr eventually assigned Harrison Barnes to Harden. Durant could even defend Capela, leaving Green on Tucker -- and allowing Green to rove as a help defense menace. (Kerr experimented with this in the regular season, too.)

We will see everything. Shaun Livingston will get chances defending both Houston stars. Both teams will switch so much, individual matchups might not even matter. But Curry's will. When things get grimy, Houston will hunt Curry, and have his man screen for Harden. That is one action the Warriors will try not to switch; they don't want Curry on Harden. Harden overpowers him, plows to the rim, and either lays the ball in or finds an open teammate.

The less dangerous the screener, the less dangerous it is to have Curry lunge off of that player, cut off Harden, and then scurry back to his original man. That is one benefit of slotting him onto Ariza or Mbah a Moute. It's still dangerous, though. Everything is dangerous against Houston.

• It will be tempting for Houston to slow the pace, wait for all of Golden State's switching to produce a mismatch, and attack it. Turning the series into a rock fight might even represent its best chance.

But the Rockets have to be careful taking that strategy to an extreme. Harden has a tendency to indulge in dance-offs against wing defenders who don't even constitute any sort of mismatch, and laze into step-back 3s. That has been a good shot for him -- maybe the defining shot of his (likely) MVP season. He will make some.

I doubt he makes enough against Golden State's most capable defenders. Houston will not win this series if Harden falls into the trap of launching too many semi-contested step-backs. It saps Houston's offense of its rhythm. He needs to drive more -- to use his one-on-one game to either get to the rim, or unlock easier shots for teammates.

• This is another reason Capela stands as such an important player. He is by far Houston's best screen-setter. When he is on the floor, Harden is more likely to call for a pick instead of hoarding the offense. Yes, Capela clutters the paint when Harden chooses to drive one-on-one. That's OK. He and Harden have a nice chemistry making plays in tight confines around the basket.

No other Rocket can duplicate Capela's rim-running and shot-blocking. He is their only source of offensive rebounding -- they have rebounded a pitiful 12 percent of their own misses in the playoffs with Capela on the bench -- and scrounging a few second-chance points is always useful against the Warriors' smaller lineups. I'm not sure Houston can win without Capela giving a solid 32-plus minutes in every game. How Capela fares against Curry on switches -- and whether Curry's knee allows for his twitchiest dribbling flurries -- is perhaps the most important bellwether.

• It will be chic to paint this series as a referendum on Paul and Harden. Nonsense. There would be no shame losing to this Golden State team. Paul has already proven his playoff bona fides. Harden has been to the Finals, further than Paul, but he is the one with more legacy points at stake -- especially after his meltdown against San Antonio last season.

The Rockets cannot afford Harden falling back into bad habits and old tendencies. Too many step-backs won't cut it. Houston needs to lean more on those sets where Harden and Paul zoom up from the corners, bolt off a pindown screen, and sprint into a handoff at the top of the arc. No one has defended those.

Harden should have the energy for that, and to defend at his best. That is a lot of why Houston pursued Paul -- to ease Harden's burden, and preserve energy for this moment. Houston can try to hide Harden on Iguodala, Green, Livingston, even Looney -- anyone but Thompson, Curry, and Durant. They can't bank on that, though. Houston's switching and Golden State's turbo pushes mean Harden will be on one of those guys here and there. And in Golden State's system of whirring screens and cuts, everyone is a threat. Every defender has to be on high alert.

If Harden is a weak link, the Warriors will exploit him. If he fights and holds his own, the Rockets have a chance.

• With Durant, the Warriors are equipped to play the slower, mismatch-preying game Houston prefers. They ran the Curry-Durant pick-and-roll more in the last few games against New Orleans, and switching against that usually leaves a smaller defender -- Paul at times -- on Durant. Paul has a knack for burrowing under taller players, and chopping them down. That isn't enough against one of the purest scorers in history.

Houston might be able to slide Paul someplace else, and guard the Curry-Durant combo with two of the Ariza/Tucker/Mbah a Moute trio -- meaning they could switch without losing much size on Durant. But again: Rigging the matchups that way won't be easy.

Houston will have to be very careful shading help toward Durant in the post. Slough away from Green, and he pops up to screen for Curry, Thompson, or some other shooter. With Green's man away from the action, that shooter is wide open if Green's pick connects. The Warriors burned New Orleans with this action over and over. The Rockets have seen it, too:

(Houston committed a defensive 3-second violation just as Durant was about to pass to Curry.)

Capela is aware of it, and knows how to disrupt Durant without straying too far from Green:

• Houston has home court, and the perfect tool set to stress Golden State. I just keep coming back to Golden State's experience, elite shooting, and stylistic versatility. Warriors in 6.