Feeds:
Posts
Comments

Archive for November, 2009

Modern Warfare 2: I Played It

And I played it more…mostly in multi-player.  The single-player campaign was, in my opinion, trite and overwrought.  Modern Warfare 1 succeeded for me on the basis of having an elegant story with a few interesting twists and a significant break from standard game narrative meta-rules.  Modern Warfare 2, by contrast, tried to do both too much and too little.  Too many twists were thrown in, they over-used their narrative shocks, and then they ended up, despite all the convolution and drama, with little better than a well-scored Rambo.

Fortunately, that’s all I have to say: Yahtzee says it all, and better than I could.

Read Full Post »

I haven’t really had much economics I’ve wanted to talk about.  Frankly, most of my thinking has been nebulously about some weird policy proscriptions and microeconomics – none of it really firm enough for a full post.  My gaming has consisted of flips between Dragon Age, Mass Effect (yeah, I’m late to the party), and Torchlight.  Borderlands will return to the fray with Dr. Ned soon.

However, Blizzard doesn’t fail to provide me with interesting news to comment on!  No indeed, the folks on the WoW live team have really been cranking it up with 3.3…probably because my prediction of another content patch before Cataclysm is wrong, and they’re being forced to load as many rockets as they can into the 3.3 bazooka.  I won’t be surprised by Cataclysm earlier than holiday season next year, but I’ll be shocked if it arrives before the summer.  Unfortunately, Blizzard has been striving for a year long cycle between expansions.  That didn’t pan out for Wrath, obviously, and so they’ve been trying to spread out the length of the expansion.  A content patch has generally offered up about 4 months of content, on top of the 4 months from the expansion release.  That would mean that 2 patches ought to be sufficient between expansion releases on a yearly cycle.  We’ll be seeing 3, and one of them feeling decidedly…slapdash.

See, 3.2 simply hasn’t felt like a proper patch.  Maybe that’s just me, but it simply has’t felt like a proper content patch.  The raid and 5-man were simply not as content-intensive as, say, Ulduar or Naxx.  While they did present a fully-featured BG, that still doesn’t match up to the level of content we’re seeing in 3.3.  My take is that 3.2 was a tide-me-over once it was realized that Cataclysm was taking too long.  Unfortunately, Cataclysm is taking even longer, and Wrath really doesn’t have any more room for them to extend on for raids…it’s time for the Lich King, so Icecrown needs to last.

So the Blizzard designers have decided to employ EVERY SINGLE MAJOR GATING TECHNIQUE they’ve devised since Sunwell.  Heck, I’m sure the only reason they didn’t put in an AQ style server-event gating the instance as a whole is because they knew that would be blatantly unfair to competitive raiders on low-pop servers.  Also, I’m guessing they’re afraid of the technical madness that went down at the time of the giant AQ invasion.  It’s a shame, though, that they won’t be doing any kind of scripted, repeating server events a la the Scourge Invasion.

Anyway, first, they’ve decided to gate off the wings of Icecrown Citadel, allowing one wing initially, and then opening the other three at intervals after that.  That alone will draw out reaching Arthas for a couple months.  This is on top of the fact that raiders can’t attempt Heroic mode at all until they’ve beaten Arthas in Normal mode.  Assuming that Heroic takes another 2 weeks at minimum to complete, all these hard gates mean the IC won’t be completely beaten for nigh-on 3 months.

That’s not all, though.  They’ve also decided to implement a sort of soft-cap on completion by capping the number of attempts raids may make on the major bosses in IC.  Once the second wing is available, raids will be limited to a paltry 5 attempts on Professor Putricide, the final boss of the wing.  The third wing grants 10 attempts on the two big bosses, and the fourth wing brings the total up to 15 attempts on the 4 hardest bosses in the raid.  Even for the hardcore raiders, this will likely prove to be a severe limitation on their Arthas attempts, though I suspect the best of the best will manage a full clear of Normal within 2 weeks of Arthas become available…

Or I would, but for one final gate.  As time progresses, the number of available attempts will increase.  In addition, zone-wide buffs will eventually start being applied to help raiders, increasing in potency as time passes.  The result is that Blizzard has an unprecendented, built in difficulty nerfing system.  This stands in direct contrast to the prior system of initially releasing dungeons as difficult and then nerfing them down as time passed via incremental patches and hotfixes.  It also gives devs an excuse to release a punshingly difficult Normal mode IC, knowing it will be auto-nerfed as time passes.  This alone means they can make Arthas a punitive fight, pushing back a potential full-clear even further.

I suspect, though, that Arthas will be downed within 2 weeks even still.  I think Normal mode will start rough for the mid-tier guilds, and then be made manageable for virtually everyone via the auto-buff.  It’s Heroic difficulty where things will be, I suspect, very, very nasty, and that should add at least a month or two before we see a full Heroic clear.  That will have extended the race for being “first” a full 4 months, with farming time giving IC at least another 2 months of life, meaning that raiders will be forced to pace out play for far longer than normal.  This serves to give Cataclysm an extra half year before people really start getting antsy for new content.

I’d also like to comment on the whole 20% dodge debuff being applied throughout the zone.  It’s the severity of this buff that really tells me 3.2 wasn’t planned in the itemization.  See, the problem for Blizzard wasn’t hard-mode as such: the jump in item levels from Naxx to hard-mode Ulduar wasn’t that bad, and would have been completely manageable without the “Icecrown Radiance” debuff.  However, the massive gear onslaught in the Trials, plus significant jump between hard-mode Ulduar and Heroic ToC tipped the scales too far.  This isn’t so much a gear inflation issue as a rating inflation issue.  The planned point ratings would reach at the end of the expansion was likely 10-15% lower than it turned out.  They could have dealt with inflated health and armor simply by increasing damage dealt proportionately.  Unfortunately, the rating inflation led to a precipitous increase in dodge and parry.  These two end up leading to highly volatile damage, which would have led to the greatly feared “2-hit instagib” or would have negated many boss swings (basically, the stuff GC has been whining about recently in discussions about “IC Radiance”).

A 20% dodge reduction is enormous, and I assume is Blizzard’s attempt at reducing the rating values rather than effective health (combined armor and health and potentially block).  If that’s the case, that implies that dodge rating + parry was inflated by ~800 points (more if you think they were indirectly targeting parry as well).  That’s a rather sizable difference, easily equal to the amount of those rating a tank would receive upgrading a full tier.  I suspect Blizzard deliberately negated a tier of rating inflation and then increased damage dealt proportionately…essentially removing most of the gains of ToC.

Read Full Post »

Apparently, Activision will never be sure that a horse is dead: they can always rig it with bionics and a robot brain, or infuse it with eldritch energies, binding its decayed flesh to their dark purpose.  So they really will never, ever stop beating it.

Read Full Post »

Hungarian Giant Robot MMORPG

Perpetuum looks like a strange mashup of Battletech robots and dropships with Eve-online.  Some of its bullet points include real-time flora alteration (planting/harvesting/burning in a mighty conflagration that you might see the tears of your enemies glisten in the flickering embers of your strength) and deformable terrain.  The designer in me says those are both a) awesome and b) scary.  Letting your players manipulate the world is like begging for griefing.  Then again, that may be precisely who they’re catering to.  And if the world is big enough, the cost of changing high enough, and the regrowth and change rates sufficient, it may not be an issue.  The engineer in me looks at those and begins shaking about all the possible ways it can go wrong and how amazingly complex those calculations might get, particularly in a “one server for all” environment.

But I really miss giant robot games, so I signed up to test.

Read Full Post »

Think These Work on Arthas?

Read Full Post »

Barry Ritholtz over at the Big Picture is normally a guy I agree with.  He’s smart and obviously has a far better grasp on how equity and asset markets work than I do.  However, today I’m going to quibble with him (and, via proxy, Floyd Norris) over their interpretation of Non-seasonally adjusted Non-Farm Payroll data.  If you hadn’t heard, the seasonally adjusted payroll data from the household survey indicated a drop of 589,000 employed persons.  Obviously, we need to wait to see the revisions of this based on tax data (the BLS reported last month that their preliminary revision for the first quarter of this year indicated an increase in their unemployment total by ~800,000 persons, so they’re not necessarily good at getting these numbers perfect).

Floyd Norris notes that the employment level is substantially better than that.  He looks at job numbers, but I prefer the household survey, because that’s the number we use to arrive at the highly cited unemployment rate.  The number of employed persons increased by 9,000 if you remove seasonal adjustments.  So why is it adjusted downwards by nearly 600,000 for the month?  Because the BLS is attempting to smooth the entire employment curve, so some losses get pulled forward and some get pushed back.  We’ve had several months this year where seasonal adjustments added to the employment figure.  Now, I’m not saying this adjustment is always correct, I’m just noting there’s a reason for it; this month’s data isn’t very good at all, when you take into account seasonal factors.

For instance, while the pickup in college graduate hiring that Ritholtz remarks on is encouraging, and while the unemployment rate, unadjusted, is at “only” 9.5% as Norris notes, we normally expect an increase in hiring in October.  This month’s increase of 9,000 is so small as to be, very likely, within the margin of error.  It might as well be 0.  Historically, October has seen a sizable increase in payroll, with anywhere from 600,000 to 900,000 added to the employment figure in the majority of cases over the past 10 years.  The odd years out?  Recession and immediate-recovery years: 2001, 2002, 2007, 2008, and this past October.  Even October 2008, not exactly a banner month for employment, added an unadjusted 233,000 to the employment figure.  If the best we managed this year is 9,000, that’s troublesome.

Further, while the hiring of college graduates, 25 and up, did jump, this was obviously offset by declines for non-grads (the majority of the US population).  Heck, that even managed to push weekly production wages up for the month.  But that college grad employment offset a decline of around 400k persons 25 and up and 300k persons aged 16-24.  The problematic bit here is that this may have been deferred hiring.  Most years, even the bad years of 2001-2002 exhibit a non-seasonal uptrend in college graduate hiring.  2008 was the first year in a decade exhibiting a systematic decline, and the first year in the last decade where employment for college grads 25 and over ended the year lower than it ended the year before.  2009 seems to be addressing that gap, with companies finally attempting to fill the spots they’d been holding off on for the past year.  In spite of this, wages have remained generally stagnant, with a smaller than expected uptick in this season.

Employment follows some pretty standard patterns over the course of a year, patterns that have been at play for at least 50 years (which is quite a while, given the duration we’ve actually been tracking this sort of thing).  Employment usually jumps in the winter, has a drop mid year, pops again for the fall, before plummeting between December and January.  The reason for the massive adjusted drops in employment at the start of the year were due to the anemic gains in employment, in what ought to have been a seasonally-driven massive increase over months.  It was also the BLS pushing the December->January losses over the next 3 months (we lost nearly 5 million jobs between Dec. and January, NSA, on top of huge cuts between November and December).  As I said earlier, October normally sees a substantial gain in employment.  November usually follows this with minor changes (up 100k at most or down 50k), followed by major declines in December and January.

It’s those final declines which have me concerned.  As things stand, the holiday season is going to be lacking.  My general feeling, though this has no research backing that I know of, is that holiday expenditures are both a social necessity AND a leisure.  They’re one of the few required expenditures with lots of leeway in how much is spent.  As such, in the face of declining aggregate income and sustained debt servicing costs, holiday spending has to decline year over year.  That, in turn, means another inventory cycle where companies are forced to liquidate inventories in the face of weak demand and lay off even more employees, leading to further reductions in aggregate income.  So I’ve got my eye on the final declines of the season, because I’ve been expecting them to take us substantially lower over.  If we see downside surprises in employment numbers, demand will become significantly constrained, increasing deflationary pressures.

Below is the unadjusted employment level since 1998:

NSA_EmploymentLevel

Read Full Post »

I’ve been reading over recent GDP, employment, S&P 500 data, etc., mostly checking for YoY changes, etc.  I was doing this in response to a series of posts by Gevlon regarding his bets on a coming crash in the Western World due to progressive automation of low and medium skill work.  It wasn’t, however, his conclusion I wanted to discuss, but some of the data presented.  For instance, he looks at GDP per worker and notes the smooth increase in this value over the time period he examines: 1989 to present.  It’s notable that this period, excepting the (potentially) ongoing recession, was referred to as the Great Moderation precisely because it exhibited essentially this trait.  Changes in YoY GDP growth were limited, and even the fluctuations brought on by recessions were managed with monetary policy, so they could be managed by companies via staff reductions.  However, the Great Moderation was remarked on precisely because it differed so markedly from prior economic experience.  So I went to the numbers to see what they said, and found some interesting tidbits.

One tidbit I wanted to share doesn’t relate to any the Great Moderation at all (not directly).  Instead, I’d like to talk about nominal debt versus income.  I’ve been hammering, lately, on the impact of debt on balance sheets and economic behavior.  Particularly, I’ve noted the requirement of the consumer to service debts before leisure spending or saving (after taking care of cost of living expenses).  When the ratio of debt to income rises, ceteris paribus, the economic agent has to cut back on other expenses…including positive savings and leisure, and potentially reassessing cost of living.  The debt contract, and thus servicing cost, tend to remain constant.

For an individual person, this isn’t a big deal.  They simply cut back until such time as their income allows for expanding consumption.  For the aggregate consumer, that is, for all the consumers in the economy to do this, though, can be a big deal.  This recognition comes to us thanks to Irving Fisher, after he’d given great thought to figuring out why he misread the Great Depression so dramatically.  In his “The Debt-Deflation Theory of Great Depressions”, he theorizes that a tendency towards disequilibrium in economic systems may arise when deflation and over-indebtedness are both found in an economy.  Of note here is that when a recession occurs, there is often accompanying deflation.  This is due to the drop in aggregate demand, which necessarily accompanies the decline in GDP, leading to lower prices as sellers chase customers.  The recession itself can probably be managed, as long as over-indebtedness is not a feature of the economy.  When over-indebtedness is present at the same time as deflationary forces enter the economy, Fisher proposes this logical sequence follows:

  1. Debt liquidation and distress selling.
  2. Contraction of the money supply as bank loans are paid off.
  3. A fall in the level of asset prices.
  4. A still greater fall in the net worth of businesses, precipitating bankruptcies.
  5. A fall in profits.
  6. A reduction in output, in trade and in employment.
  7. Pessimism and loss of confidence.
  8. Hoarding of money.
  9. A fall in nominal interest rates and a rise in deflation adjusted interest rates.

(As a side note, it’s important to note that he goes to great lengths to say this isn’t necessarily the temporal order of events, merely a logical order.  He even points out a different order things could occur in; the logical order remains consistent and the end result is the same.)

What’s important to take away from this is that “The more debtors pay, the more debtors owe”.  As one can see from the result of the list above, as prices fall (and the value of the dollar grows) the real rate of interest also increases, increasing the cost of debt servicing in real terms and reducing income available for consumption even more, which feeds the cycle.  Eventually, this crashes, and debts are, somehow or another, wiped out.  In the meantime, a lot of capital is destroyed.  I bring all this up so I can present two graphs I made based on personal income data US Dept. of Commerce Bureau of Economic Analysis and the Flow of Funds data from the Fed.

QuarterlyNominalHouseholdDebt

First is nominal household debt, in billions of dollars, since 2003.  As you can see, it was on a pretty steady upward trajectory until the crisis took root, after which US households began attempting to pay down their debt…with some success.  This is nominal debt, not real debt.  Inflation is not accounted for in the above chart.

Next is the above debt values divided by nominal aggregate personal income (income also in billions of dollars).

QuarterlyDebtOverIncome

As you can see here, people were initially successful, reducing debt as a percentage of their income once the recession started kicking in.  However, the recession overcame them.  Despite the consistent reduction in nominal debt, debt as a portion of their income jumped again.  While the portion dropped, it’s still higher than it was at the start of the recession…despite consistent debt reductions.  Hopefully, a recovery takes hold and controlled inflation actually manages to kick into place, because we remain on a precipice as things stand.

The problem facing policy makers is that this debt must be paid down in order to help insure credit returns to being a recession buffer, rather than a depression inducer.  If they allow an uncontrolled deleveraging to occur, though, they risk a depression now.  It’s a fine line to walk.

  1. Debt liquidation and distress selling.
  2. Contraction of the money supply as bank loans are paid off.
  3. A fall in the level of asset prices.
  4. A still greater fall in the net worth of businesses, precipitating bankruptcies.
  5. A fall in profits.
  6. A reduction in output, in trade and in employment.
  7. Pessimism and loss of confidence.
  8. Hoarding of money.
  9. A fall in nominal interest rates and a rise in deflation adjusted interest rates.

Read Full Post »

Older Posts »