First Time MMO PC Builder, Part 2 (Initial Build)

Last weekend I sat down in a room with a box of parts - carefully selected after much deliberation - and embarked to turn them into a computer.  Two key lessons right off the bat:

1. The internet is your friend.  I had a ton of questions that I was able to answer through guides - this step by step guide on Tom's forums was a key reference - and asking Google questions.  I don't know that I could have pulled this all off without that knowledge available (on a computer that wasn't lying in pieces on the desk).
2. Allow time.  My goal was to install the bare minimum parts needed to see if the ones I had installed were working correctly.  This still took me most of the weekend.  More to the point, as a first timer there were definitely some things I ended up ordering, and it was just lower stress to know that finishing the project up this weekend was part of the plan.  

Anyway, here are my experiences approximately step-by-step.

Cables Everywhere
The first thing I did was open up the boxes to find and read (yes, read) manuals, examine cables and connectors, etc.  I was struck by how many cables there were.  My case alone came with 4 fans, each of which has a three pin power connector and a speed adjustment switch.  Throw in motherboard hookups for the front-of-case USB and audio plugins and you're talking 10 loose wires in the empty case before I even started putting anything in there. 

Some of the surplus cables that I have no planned use for, along with two of the fan hookups, stick out the side of the case as I wrap up for the night.  Many of these will have to live in the cable management compartment, above.

I can immediately see why so many people talk about "cable management" as a major part of building.  My case comes with a little nook, a bit under two inches wide and maybe half of the case's height, where apparently I can literally stuff the tentacle-like mass of excess cables writing out of the end of my power supply.  Speaking of which, I immediately had an unfortunate surprise in that the power supply had to be installed fan-side up because there is no venting in the floor of my case.  It would have been easier to stash the surplus cables if it had been the other way around.

In my shopping list post, Xaxziminrax II suggested that power supply manufacturers add excessive cables so they can then charge us to not include so many in more expensive "modular" units.  On one hand, I can see how it could be cheaper to have one model that everyone can live with, rather than separate models for people who actually want eight hard drives (the number of SATA power plugs on my power supply) and no PCIE plugs for graphics because they're running a server with intergrated or no graphics, versus people who want one hard drive and multiple GPU's.  That said, my power supply may have been excessive - the two universally required connectors for the mother board are joined by two for graphics cards, eight all-purpose four-pin MOLEX connections, the eight SATA connectors, and a pair of floppy disk connections - yes, not one but TWO of something I haven't seen on a computer in a decade - for good measure.

The swarm of cables threaten to overrun the hard drives.

Ironically, all these cables and I still didn't necessarily have as many as I needed or the right shapes and sizes.

• The fans came with three pin connectors designed to be connected to motherboards, but A) the motherboard only has two chassis fan connectors compared to the four fans in my case and B) this means allowing the motherboard to determine whether and when to run the fans, and I had some trouble figuring out how to get it to do that properly.  A cheap solution to this problem is to buy some adapter plugs to convert the three pin inputs to the molex connectors that I have so many of.  A slightly more expensive but functional solution, which I opted for, involved ordering a fan controller (which will allow me to adjust the fan speeds from outside the case, because I'm definitely not opening the side to get at the controls on the fans).  
• The motherboard came with a pair of SATA cables for connecting hard drives to the motherboard.  I have a pair of hard drives, so this should have been good, except for a minor problem; the cables have right angle connectors (you can see one, with the white tip on the black cable, pictured above).  The case has a floor mounting for an SSD, which would allow me to leave two of the three slots behind the lower case fan un-obstructed by hard drives.  This is also a cheap fix (under $4, shipped), but one I didn't think ahead to order.
• As long as I'm ordering cheap cable-related items, I also snagged a sack of cable ties.  I got about 10 of these with various components, but I expect to use a lot more than that in the process of securing all of the fan cables.  On the plus side, this only cost $5, and I needed some cable ties to deal with my poorly organized home theater setup in any case.  

Putting it together
I showed my wife my progress at several stages along the way, and she remarked that in some ways it was like a very fancy Lego set. This is an exaggeration, but assembling a computer is nowhere near as complicated as I expected.  The only tool I needed was a screwdriver.  When you deduct all the time I spent looking up answers to questions like "can you plug a three pin fan connector into a four pin plug on the motherboard?" (answer: yes, the fourth pin is optional, and there's a plastic tab to force you to pick the correct three pins), the process was also remarkably fast.

1. Install power supply.  This could go later, but the power supply is heavy enough that I'd rather lock it into place early.  I'd be even more nervous about waiting if my case had a top-mounting, because then I'd be juggling a large heavy object over an assembled motherboard.  Also, for people working in primarily plastic cases, this may be helpful in reducing the risk of static shock by adding a metal object - my case is almost all metal, so I was probably pretty safe even before strapping on the wristband. 
2. Place hard drives in bays.  Ironically, I ended up undoing this work when I figured out that the cables weren't going to work the way I had it set up, and I'll need to re-do it again this weekend to get it back the way I wanted it.
3. Get out the motherboard, screw in "stand-offs" to the case for future installation, install input/output bay shield (more on this in a few steps).
4. Install CPU on motherboard.  This was the step that scared me the most, because it's theoretically possible to damage the components.  That said, it's apparently gotten a lot easier.  My CPU did not have pins on the bottom, instead a mostly flat surface that sits in the flat compartment with tabs that physically prevent you from doing it wrong.  All you have to do is remove the cap, set the chip down, and pull the lever that used to hold the cap back into place to secure the chip. 
5. Install CPU cooler.  My cooler selection was common enough that I was fortunate to have pictures on the forums.  This was helpful, because the instructions aren't that clear if you don't already know what you're doing.  There's also a step where you need to apply thermal paste - hopefully the amount I added was correct, but I'm not opening it up to mess with it unless I encounter problems down the line.
   
   

   

   Note how close the CPU cooler sits to the top RAM slot.  Fortunately, my motherboard did not require that slot if you are only using two DIMM's, and I should not need to upgrade to 4 anytime soon since I started with 8 GB of RAM.  I could see why people suggest installing the RAM first if their situation was different, and it's possible that I would need to replace the CPU cooler with something smaller if I do want to try this in the future.

6. Install RAM.  This is easy enough, chips are notched so that they can't be installed backwards.  The only thing that's surprising about this process is that it actually does take some amount of force - not excessive but it definitely don't just slip in like a USB plug or anything.  
7. Now that the motherboard is loaded with the things that do not have cables attached to them, mount it on the standoff's.  Here's where I had a bit of a struggle with the input/output plug shield.  This thin piece of metal is apparently used to make sure the jacks on the back of the machine are touching a piece of metal that is touching the metal case, to ensure that any static charge that happens to be on my USB plug doesn't end up on the motherboard.  The catch is that some need to be bent out of the way to avoid blocking ports, while some need to be NON-BENT to avoid blocking the installation process.  This part actually took me a non-trivial amount of time.
   

   

   Here is the exterior of the I/O shield, you can kind of see the little metal tabs going into the case.

8. Start connecting cables!  The motherboard gets two inputs from the power supply.  The hard drives get one power cable and one data cable to the motherboard.  The case-front USB, audio, and even the power button/light all have their own spots on the motherboard (the manual: it is your friend).  Fans have spots.  And so-on.  

Once I got this far, I had reached my minimum goal of being ready to test functionality.  Having onboard integrated graphics means not having to plug in a graphics card and wondering whether that is why the whole system didn't turn on at the moment of truth.  I didn't connect the data hard drive because it won't be needed, and obviously I didn't even have enough cables to hook up all of the (arguably excessive) fans for my build.  Because I'm using an external DVD drive, that was one more thing that could be deferred. 

I double checked all of the connections, hooked up an old monitor, double checked again, plugged the thing in and pushed the power button.  Never before has the absurd "no keyboard detected, press F1 on non-existent keyboard to enter setup" message been so welcome - if the fact that I hadn't plugged in the keyboard yet was the biggest problem (and the machine was well enough to recognize this), I was in good shape. 

The guts so far.

Ironically, I got further than I'd originally hoped.  The machine accepted my wireless keyboard/mouse and the external DVD drive without any issues, and was easily set to prioritize the DVD drive over the empty hard drive.  I was then able to go on and install windows onto the SSD without issues - and let me note that the machine boots nice and quick off of the SSD.  I hit a minor snag because Windows didn't have a driver for the ethernet port on the motherboard, but fortunately the DVD that came with the motherboard did, and I was then able to go online for my antivirus and Windows Update needs.

There are a number of tasks left for next weekend.  I will need to install a graphics card, a wireless network card (the finished machine won't be sitting on the desk with the router), move and install the secondary hard drive, install the fan controller I ordered and connect all the fans.  Then will come the more involved tasks of actually securing all the cables in a way that does not leave loose cables to block airflow or get caught in moving parts.  After that comes stress testing and overclocking (nothing too agressive, but on paper I should be able to bump the stock CPU speed by several hundred MHz based on the hardware I have).

More on this project - and eventually the fun part of actually taking it for a spin in games - next week.

Rise of the Raid Finder

MMO-Champion has been doing some Armory scraping and produced some statistics on WoW's new raid finder.  Out of a sample of 4.4 million level 85 characters who have been active since patch 4.3 in December, over 30% have completed the Deathwing Raid using the new automated raid finder on easy mode. 

As Chaud notes, it's hard to get to the proportion of players who have completed the raid.  The percentage of players who have cleared the zone can be no higher than the percentage of characters (i.e. if no one had ever cleared the zone on an alt, which is unlikely).  We can almost certainly say that 70+% of players who have a level 85 character that has been active since 4.3 have yet to clear the raid finder. 

Likewise, one imagines that many traditional raiding guilds are taking their people through LFR for gear and practice, which would mean that not all of these folks are new to raiding. A part of the long-term test of this system will be how many players beat the raid finder once to see the grand finale and then never return. 

Even so, the numbers have to be viewed as a success, as they will only increase over the remaining months until the new expansion.  The dungeon finder effect also has a way of amplifying itself - as more people beat the content, groups will gain experience and be better able to triumph even with newbies on board.  For better or worse, this feature is assuredly here to stay. 

First Time MMO PC Builder, Part 1 (Shopping)

I often lay out New Year's resolutions that either get done immediately or not at all, and this year's plan to build my first computer ended up in the immediate camp.  My dissatisfaction with my current laptop, coupled with current prices was enough to push me off the fence early - there's always something over the horizon or some future discount you can snag, and I've definitely stuck it out too long on previous machines because I wanted to wait for just one more thing to come out.

The Past and the Plan
I'm never built a computer from parts before, but I've done various sub-portions of this task - installing RAM, graphics cards, wireless cards, and hard drives, re-partitioning and installing Ubuntu on an old laptop that is now reborn as a websurfing machine, etc.  Basically, what I hadn't done before was pick out my own components and assemble a machine. 

I'm hoping to get 3 years out of this machine as my primary gaming platform and then have it retire to work as a media server or whatnot.  My approximate budget was $800 for hardware not including the graphics card - I have an old card I plan to use until I see whether this whole endeavor catches fire and sinks into the swamp, and I knew going in that this was something easy for me to swap out later.  Either I will buy a long-term solution or a medium term one that I plan to replace (and possibly carry over into the computer after this one).  This is a machine I plan to live with and upgrade, so I'm not as concerned with minimizing expenses on everything that doesn't directly inflate my benchmarks, especially for stuff that would make my life more difficult as a first-time builder.  I decided early on not to mess with any design with either dual graphic cards or water cooling for this reason. 

I also made a deliberate decision to break the project up into two weekends.  I was certain that there would be something I would need to order, that I would run out of time, etc etc.  My goal for week 1 was to install the bare minimum parts I need to confirm that things are working, leaving add-ons, final cable-routing, and actual setup to a second weekend. 

The Parts
As with system builder custom, I will break out separately mail-in rebate values, since these are intentionally designed to be difficult and time-consuming to actually collect. 

• CPU: Intel "Sandy Bridge" Core i5 2500K - $220
  Based on reviews and benchmarks, this appears to be the sweet-spot on price performance.  It's a solid chip that can overclock well, if I get that far.  The guys at Tom's tried to put the comparable priced AMD chip in their $1200 build last month and came out much worse off, despite a major investment in graphics. 
  
• Motherboard: Intel Z68-based ASUS P8Z68-V LX: $125 ($10 mail in)
  The Z68 chipset supports the integrated graphics capabilities of the Sandy Bridge processors.  This can theoretically save power (no need to fire up a graphics card for basic Windows functions), provides extra monitor slots if I try to install more monitors than my graphics card has outputs, and also means that I have a backup option for troubleshooting in the event of a malfunctioning graphics card (which happened to me on my last machine). 
  
  It was actually remarkably difficult to get a comparison of what the differences in the various P8Z68 models - the board's fan club has a comparison chart.  The -V boards effectively do not have a second graphic card slot - which fit with my plan in any case - but you quickly get down into the weeds on all kinds of numbers like USB 3.0 slots etc, which snowball around to hit whether or not your case actually has extra USB 3.0 plugs to plug those into, etc etc.  In the end, I made a judgement call that this board fit my budget and was good enough for my needs. 
  
• SSD (system+program drive): Corsair Force GT 120 GB: $179.99 ($30 mail in)
  This relatively splurge immediately pushed me to the high end of the budget.  Pretty much everyone agrees that an SSD system drive is a great upgrade.  In this approach, you spend about $80 on a 60 GB drive that will take Windows, basic programs, and maybe one MMO (though WoW's 28 GB may be pushing this). 
  
  As someone who plays a bunch of different MMO's, I really wanted to have room to share the SSD love with more than the one game, but I couldn't justify the cost.  Then this drive went on sale for $150 with the rebate.  Even then it's still a bit more of my budget than I technically "should" be spending on storage, but this buys me the space to have my top 3-5 games of the moment on the SSD.   
  
• Data Drive: 500 GB Western Digital Caviar Blue - $80
  Gamers tend to covet the Caviar Black model, which has gotten very costly in the wake of flooding in Thailand.  Still, I didn't view going without a data drive as an option; even with the larger SSD, there are still going to be 100 GB of game clients I use once a month that are going to have to go somewhere.  This was a price that I could afford, and it's still very good from a performance standpoint. 
   
• RAM: 8 GB (2x4GB) Corsair XMS 3 DDR3 1600 - $48 ($10 mail in)
  Lots of people just go with the cheapest 4 GB DDR3 1600 kit they can find and call it a day, and I was initially inclined to go along with this.  The upgrade to 8 GB turned out to be under $20.  I like being able to run in windowed mode with a browser sometimes, this upgrade was cheap, and reportedly sometimes an improvement. 
   
• Power Supply: Corsair TX-650 V2: $90 ($20 mail in)
  The power supply is one of those components that system builder challenges tend to skimp on, while experienced builders caution that this is a highly important part that can fail catastrophically and greatly affect performance in the mean while.  Throw in a touch of uncertainty about how much power my as-yet-unselected final graphics card would require and I was stuck on this one for a while.  Then a well-reviewed model with enough juice to fill my needs went on sale.  One thing you don't get in this price range is "modular" cables - it's apparently cheaper for them to build one model that leaves dozens of excess cables hanging out in the case than to make the cables fully detachable.
  
• Optical Drive: External HP 24x DVD burner 1270e - $30
  This is an unconventional choice - most people go with an internal drive that transfers data through a faster SATA port on the motherboard, rather than a slower USB port.  This difference for me is that I don't get my software on souvenir coasters anymore - my current laptop doesn't have an optical drive and I've done fine without for almost a year and a half.  I also don't use the computer for watching DVD's when I could be watching them on the TV instead. 
  
  Bottom line, I need this drive only when I go to re-install windows, or perhaps an MMO recent enough for the client disks to still be worth using (e.g. SWTOR if I get it in the near future).  In exchange for a slight drop in speed, I have one fewer thing to install internally, and, more importantly, I have a drive that can be shared amongst the present (and possibly future) computers in the Armadillo household that do not have working optical drives (my Alienware, my wife's Macbook Air, my old Dell desktop whose drive broke several years ago) for the price of one internal drive.  Your mileage may vary but this was an easy choice for me. 
  
• Case: Antec Three Hundred Illusion - $70
  This choice baffled me for a disproportionate amount of time precisely because it is relatively hard to measure objectively.  You can get a case for $30-50, but it's likely to be hard to work with and require additional purchases.  You can get a case for $150, but that really doesn't make sense in the budget range I was aiming at.  In the middle are compromises, but it's hard to tell which of these compromises really matter.  There's also so much personal preference involved that every case has someone who swears it was a pleasure to work with and someone who hated it. 
  
  In the end, I decided to go by popularity.  I liked the look of this case, and I liked what I read about it in the September edition of Tom's $2000 build.  Unlike the entry level version of the Antec Three Hundred, the Illusion comes with a pair of intake fans.  I actually had to go back and make a slight switch to my planned motherboard at this stage - I'd been planning on paying slightly more for the P8Z68-V LE, which has headers to support two front USB 3.0 ports, but this case only had USB 2.0 ports.  It's actually remarkably hard to get an answer to a seemingly simple question like "does your case have USB 3.0 ports", but it seems like the answer in this price range is generally "no". 
   
• CPU Cooler: Cooler Master Hyper 212 Plus: $29
  I chose this based on overwhelming popularity, and the fact that it was paired with the case I finally selected in the linked build at Tom's.
  
• Anti Static Wrist Band: $5
  Waste of money?  Perhaps.  Do I look silly clipping my wrist to the metal case?  Perhaps.  If this is snake oil, at least it's a bit of snake oil that only cost me $5 and that I can reuse for all my future computer building/modification needs. 
  
• Monitor, Keyboard, Mouse, Wireless network card, and Starter Graphics Card: Hand me down's from past systems ($0)

Grand Total: $875, before $70 in mail in rebates

There are a few miscellaneous cable purchases yet to be added to the total, so realistically I'm slightly over where I was aiming, primarily because I jumped on an SSD deal that was a bit of a reach for my budget. 

Overall, so far I am happy with all of my choices, though some gave me headaches during installation (tomorrow's topic, this one is already long).  I will say that I was definitely struck by how difficult it is to get all the answers about what types of plugs your components have/need, etc.  I've also probably made a number of mistakes that some of you may be laughing at me for as you read this. 

Even so, I'm glad I embarked on this project.  I definitely feel like I know more about computers in general and what makes mine in particular tick.  Maybe knowing exactly how to reinstall and even replace components at will can eventually save me some money. All that aside, it was actually kind of fun.  If the end result is a solid mid-end computer - which, incidentally, will be far more than I would have gotten for the money through any other approach, I'll be happy. 

Keep Flipping Hits SWTOR

According to the latest episode of the Darth Hater podcast, keep flipping has come to SWTOR's world PVP planet, Ilum. 

The gang is playing on a server where they believe the Sith outnumber the Republic by 5 to 1 (probably in part because the Sith side on that server is home to the guys who spent 3 years recording 100+ episodes before the game even came out).  The Republic has no chance to take objectives in a fair fight, which means that the Sith has no opportunity to claim the incentive rewards for taking the objectives back. 

As a result, when a lone Republic player shows up, the majority of the Sith do not attack that character, and berate those on their faction who do so.  This creates the opportunity for one of the hosts to grief his own faction ... by doing what the Sith are supposed to do and defending their territory from the Republic.

Here's a tip for Bioware from the PVD history office: The last time a new release MMO by the Mythic half of Bioware-Mythic had a keep flipping problem, they responded by blaming WoW for letting players who would do such a thing into the genre.  Hopefully you guys have a better approach?

P.S. People who reach the level cap in the first month can always expect some rough edges, but the Darth Hater crew seems unusually downcast/concerned by the state of the endgame.  Raiders are reporting the content easy but buggy and threatening to leave.  While I think it's possible that retaining players in the traditional MMO endgame is just less important to the game's strategy than having them play through all eight carefully crafted storylines, I hope this isn't Bioware's best effort.

How Much Can A Class Change?

Ferrel points out a fundamental issue with Cleric DPS in Rift: any DPS soul can currently pick up a set of abilities that causes a percentage of their DPS to be turned into passive healing for their entire group.  This puts a ceiling on how good Cleric DPS can be for obvious reasons - no one would want a pure DPS if clerics dealt as much damage AND provided free healing - and Trion has apparently recently attempted to add a new stance that increases DPS and reduces healing to help work around the issue.

Dr. Ferrel's prescription: Nerf the "broken" Justicar soul (which provides the passive cross-role healing), to clear the way for "unique" fixes to increase the DPS of the four DPS souls (rather than the non-unique buff), so that once the dust has settled Trion can remedy a lack of love for the three healers (in particular the Purifier, which Ferrel claims "needs a redesign" due to a "terrible" signature mechanic/"gimmick"). 

Personally, I haven't spent enough time with Rift to conclude whether any - or all eight - of the Cleric souls are in need of such dramatic changes.  That said, the discussion raises a design question that goes beyond the current state of one class in one game - how much is it fair to change the way classes play and feel in a live MMO? 

Changing the game
My limited experience suggests that Ferrel is absolutely right about the constraints the passive healing builds (called *-icar because you're tacking some Justicar onto whatever soul you're actually playing) place on the design for the rest of the class, or even the entire game.  One of my comments from the game's launch was that the sheer amount of passive healing generated by solo DPS builds - including Clerics - appeared to be harming the challenge of the non-instanced group content.  (There was a nerf along these lines in the game's first major patch, but Ferrel's post would suggest that the problem persists.)  Meanwhile, the value of free-form customization is greatly diminished when every Cleric build that will ever solo or DPS must spend most of its off-tree points in a specific soul. 

That said, my character is a Cleric because I actually like the current playstyle.  The types of changes Ferrel proposes - such as moving key abilities deeper into the tree - would have a huge effect on the leveling experience for all clerics.  Moreover, the philosophy behind the adjustments would dramatically alter the way the class feels - from a slow DPS'er with zero downtime due to a constantly regenerating health bar to hopefully a higher DPS class that is more dependent on active healing to survive. 

My opinion as someone who hasn't been a fulltime player of Rift since its first month doesn't really matter all that much, but how many others who are still in game chose the current Cleric class because they preferred its slower pacing?  If there are significant numbers who would be dissatisfied with the change even if it results in a mathematically superior (better exp/hour or whatever) class that loses the current feel, how do their needs balance against the arguably more severe consequences being suffered by players who would trade the survivability for DPS?  (To what extent did the current abundance of disgruntled Cleric DPS directly result from the ease of leveling the current Justicar splash builds?) 

Parting Caution
Two parting caveats to this discussion:

1. In my experience with class balance as an MMO player since 2004, this is not the first time I've seen players of a specific class argue that one mechanic of their class should be nerfed to clear the way for future buffs.  The nerfs that even players of that class agree are needed generally happen.  The compensatory buffs don't always materialize. 
2. Speaking more generally, I can think of one big example of a company which believes that classes can and should be radically redesigned every time the team feels that just one more revamp will solve the problems.  I know of relatively few players who are entirely satisified with this aspect of World of Warcraft, even when they agree the the problems with the status quo are legitimate and the proposed changes are objectively superior. 

My oddly famous blogroll widget

My evenings have been strangely packed so far this year - in part a consequence of having been out of town for a week over the holidays and therefore having a bunch of deferred errands to run.  As a result, the blog has been relatively lacking in content, so I suppose it was inevitable that the most interesting part of PVD this week has been... the "most recent post" sorting feature on my blogroll widget. 

First, Psychochild's year-end wrap-up named me as the number three traffic source to his site.  It didn't occur to me that I personally represent a significant portion of this traffic, because I often use my blogroll as an improvised RSS reader, until Indy commented that he also uses my blogroll for the same purpose.

Then Wilhelm started playing with post tags on his archives, and my blogroll apparently concluded that his blog had last updated in 2006.

If anyone else is working on a post about my blogroll widget to add to the coincidental January 5th publicity blitz, I suppose there are a few more hours left in the day?  (Actual content incoming, really.) 

Three Predictions for 2012

Here are a few predictions on the state of MMO's in the coming year.  Ironically, though I hadn't planned it this way, the three topics I came up with address (albeit in a different order) Wilhelm's top three questions for 2012. 

Go Big, Go Small, Go Free, or Go Home
The subscription MMO isn't dead, but there are basically two very specific circumstances under which it can work:

1. Have a nine-figure budget like the reported $100 million in venture capital that founded Trion or the even larger figures that EA is rumored to have spent on SWTOR.  Ever notice how the three corporations able to foot this type of bill - Blizzard, Trion, and EA-Bioware - are the ones who are still touting the subscription model?
2. Serve an un-filled niche, such as sandbox PVP (see Eve, Darkfall, or perhaps the forthcoming Dominus) or old-school group MMO (see Vanguard, lots of room for a newcomer in this genre).  The big-budget one-size-fits-all MMO that includes solo, group, raid, and PVP has to make compromises  to fit all these activities under one roof.  This makes it possible for a more specialized game to offer something that the big guys cannot.  However, as CCP found out last year, this also means that your entire company lives or dies by its ability to continue to keep one segment of the market satisfied.

Note what's not on that list.  The best licensed IP's out there don't guarantee you $15 a month - see DCUO and LOTRO.  Neither does implementation of a specific feature in what's otherwise a one-size-fits-all MMO (e.g. RVR in Warhammer and PvPvE in Aion).  Even the huge budget is no guarantee of success - probably the most remarkable thing about Rift's progress is how much discipline the team has shown in implementing only what they can actually accomplish and accomplish well.

The bottom line is that if you have yet another fantasy MMO, you're not solidly in one of the two categories above, and your business plan depends on collecting a $15 monthly subscription - FFXIV and Tera come to mind, along with Copernicus if they're not thinking F2P - you are in for a rough time in today's crowded market.  Of course, you're also in for a tough time in the crowded free market, but at least the bar is lower to get potential customers to actually try your product.   

SWTOR will have high churn... and high revenue
Both sides of the discussion on SWTOR's longterm prospects tend to assume that the game will be a failure if there is a mass exodus by the 90 day mark.  Ironically, there has never been another MMO so well-positioned to handle a high rate of churn. 

Yes, the game has guilds and PVP and dungeons/raids, all the traditional MMO trappings that tend to do poorly with high churn.  As long as Bioware was spending whatever ungodly amount they spent on this game, there was no reason NOT to support these playstyles and collect the associated revenue.  However, the core thing that has everyone raving is the Bioware story.  With the past Bioware games, the customer pays once for the box, and maybe once more if the expansions/DLC are worth purchasing, no matter how long it takes the player to complete the game or how many times they replay it.  With the monthly fee, EA gets paid every month for every playthrough and replay, regardless of how little or how much content Bioware actually adds to the game in future patches.   

With such a focus on a highly replayable single player story, SWTOR doesn't need half a million year-round subscribers.  They can get the same effect with 1.5 million players who pay 4 months out of the year when new content is added - or when players choose to replay the old stuff.  I don't see how Bioware can lose here - which is probably why they got so much of EA's money to spend in the first place. 

Mists of Pandaria Will Ship This Summer, Or Heads Will Roll
Many intelligent people are predicting that Pandaria won't ship until Q4, and there is strong basis for making this call - Blizzard is not known for shipping its products on time.  This round, however, I think the stakes are higher. 

Blizzard spent 2011 losing subscribers by the millions - to Rift, or wherever else - and SWTOR will not help this situation.  No amount of spin about how the lost players are in the less lucrative Asian markets, or how players have returned to WoW after the launches of past competitors, can change the reality that Blizzard will continue to lose customers and money until something changes.  A scenario in which the content that was available in early December 2011 is still the only content available in early October 2012 is unacceptable. 

My guess is that we will see the paid closed beta phase of Pandaria (courtesy of the annual pass) kick off in May-June, with an aim for an early Q3 release.  Delaying this product further is not like delaying Starcraft II or Diablo III, which do not have monthly fees - every month means more subscribers lost from the current live WoW service.  I'm prepared to believe that Blizzard might let the expansion slip anyway, but I think that there will - and frankly should - be consequences if this occurs. 

What do you all think will happen in 2012?