Incentives in the Starcraft II Beta

As those of you who follow my Twitter feed may already know, I somehow landed a spot in the Starcraft II beta. My best guess is that I got added to a list for leveling a Death Knight to 80 in the Wrath beta, though I suppose it could also have been random chance. I'm actually not the world's biggest Starcraft fan, in part because I'm not all that good at Starcraft, but I resolved to spend some time playing the beta anyway, just in case logging some hours helps my odds of a Cataclysm beta invite.

Revisiting Old Friends

Let's be perfectly clear - this game is basically Starcraft with better graphics, and, given how well that game is STILL selling over a decade later, that's high praise.

The three races are all back, and each has a few new tricks. Terran supply depots can now retract into the ground so that players can use them as barricades. A Protoss upgrade can teleport ground units into any location on the map that has Pylon power. The Zerg have several new ways to spread creep around quickly, without having to establish a new hatchery.

Many units are back and do the same thing that they did before. Others have been dismantled and had their abilities parceled out to other units - for instance, Terran drop ships are now Medi-vac ships, and come with the healing ability formerly found on medics. Overall, the tech trees feel streamlined but not in a simplified/dumbed down sort of way.

Accessibility and Gameplay
As an RTS, Starcraft can't really offer persistent rewards as an incentive to pad the agony of defeat in quite the same way that MMORPG battlegrounds can. Instead, they do their best to make it easier for players to learn to play and find appropriate opponents.

All of the maps come with a "novice" version, which seals off the entrance to player bases with a giant pile of rocks that someone has to blast away before ground units can pass. These things are a speed bump if you've got a decent sized army, but they do a reasonably good job at keeping other players from having half a dozen zerglings in your base inside of three minutes.

Meanwhile, the game has no more secrets. If the other guy DOES flood your base with a dozen hydralisks before the match is 10 minutes old, you get to see precisely how they pulled it off with the new post-match replay. There's a summary of what players built when, and a real-time replay that allows you to see what everyone is doing (including which units the player has selected, how he has them grouped, what their resource count is, etc).

Finally, a new matchmaking system is designed to herd players to the correct rating relative to their opponents. I'm still playing my calibration matches, so I don't know where I'm rated or how effective it will be after it's done, but the plan is clearly to try and arrange even competition. On the downside, you can't really do much of anything else while you're waiting for a match. I guess they're hoping that the sheer volume of players will keep things running at a decent clip, but having to sit and watch it search for several minutes feels painful after having the ability to go do other things in WoW while you wait for your PVP or dungeon queues to pop.


Fortunately, the "estimated time" has been relatively accurate, and never more than three minutes for me so far.

Outlook
In some ways, this is a highly polished product. The graphics look great, the game runs smoothly, and I have yet to encounter any real bugs in the traditional sense of the word (e.g. crashing to desktop). The only thing that is not at all complete is single player (which will not be tested in the beta). You can set up a custom match between yourself and an AI opponent, but the only AI setting that's currently implemented is "build a few buildings, send the player something to kill every few minutes, and generally wait around for them to figure out what all the new buttons do". (The AI will actually say "gg" when you arrive in its base with a fleet of Carriers, only to discover that it decided it was going to build a dozen marines and stop at that.)

The real reason the beta exists to test balance for multi-player. I think there is some work to be done in this area - for instance, the Terrans have a new jet-pack trooper, who can jump straight up sheer cliffs, bypassing your base defenses and landing in the middle of your workers. This is a neat concept, but it seems very difficult to counter, in part because the troops can actually outrun most units in a foot race, even if they don't simply hop up some ledge where your troops can't follow them.

Personally, I'm not the biggest fan of the RTS genre. I'm not all that competitive, so I don't get that excited to know that I just beat someone. Meanwhile, the amount of multi-tasking that you get into when you're trying to manage an economy and an army gets to be more than I want to deal with many evenings. Even so, this revision of Starcraft was good enough to occupy several hours of my time to date, though I have no shortage of other games I could be playing.

In the end, Blizzard has a plan - near instant access to a match against someone who will be a relatively even foe for you. If they can pull it off, as it looks like they will, this game may exceed even its own hype.

Tyranny of the Cash Store Wallet

Between SOE's facebook promotion and the $5 in "free" credit with the expansion purchase ($40), I currently have $17.50 in credit at SOE's Station Cash store. The new expansion just arrived and, as always, I'm short on tradeskill vitality. With just a few clicks of the mouse, I could have a week's worth of vitality instantly, in the form of a potion that nominally costs $10.

Of course, the catch is that I would never pay $10 for that potion. I didn't pay for the credit I have now, but there might eventually be some other item that I will want badly enough to actually pay for. If I save my balance now, that future purchase might be free, or at least deeply discounted. Any yet, somehow simply having that balance in your wallet encourages you to find a way to spend it.

The Adversarial Design of the Item Shop
Cash stores are ultimately designed to get players to spend money. Even so, one wonders if part of the reason why players are so distrustful of the model is because of all of the tricks these stores employ to try and get players to spend more than they planned to.

In all stores, currency must be purchased in increments of $5. If the items you buy aren't priced in such perfect increments, you may need to top off your account again in order to finish spending your current balance. Unlike some other stores out there, SOE uses an easily interpreted exchange rate (1 cent = 1 SC) and does not require a massive bulk purchase to obtain that best rate. Some games use random numbers to make it hard to figure out the real world cost of the store offerings, and require purchases of as much as $80 at once for players to get the best exchange rate. All of these games are hoping that players will fall into the trap of spending faster than they intended once the real money is loaded into a seemingly less real virtual wallet.

From the perspective of a non-subscription game where 90% of players never pay a dime, I suppose that having a player pay $50 and leave in horror when they realize that they burned through it in a month sounds like just what the bookkeeper ordered. In the long term, though, how much more would that player have spent if the game had not burned them? How much future business is lost from that player's bad word of mouth? The risk that cash store developers run is that all of these little tricks have slowly poisoned the well for ALL cash stores, not just the ones that did a bad job on their pricing research.

As the old saying goes, fool me once, shame on you, but fool me twice and the shame is on me. For me personally, the "once" was the time I spent about $150 in three months on an online trading card game. Ever since then, I've always been very suspicious of item shop models, simply because encouraging that kind of spending is actually part of the game design. I wonder how many other players have already used up their item shop mulligan, and regard the genre with similar distrust as a result.

The Upside of Removing Content

In LOTRO, level 65 characters have a very limited selection of dungeons - three level-appropriate small group dungeons, a single full group dungeon, and a single raid. The early story of the EQ2 expansion is whether there was enough content to go around in an expansion that launched with fewer zones than the last expansion that raised that game's level cap. In that context, Ferrel asks why it's worth raising the level cap of an MMORPG. After all, these problems can be directly attributed to the fact that raising the cap "dramatically reduces valid and usable content for players at the max level." Why bother if all the developers aren't even going to offer new spells and abilities?

Variety is the enemy of consensus
One of the key reasons why the WoW automated group finder has been successful is that the system specifically pushes players to agree to run a random dungeon. If there are three players in the queue who insist on one dungeon, a single player who demands a second, and a third player who will only join something else, no one gets a group. The random dungeon finder may compromise on a dungeon that NONE of these players would have chosen on their own, but at least they're doing something instead of sitting around waiting for a queue to pop.

In games that do not have such a system in place yet, variety is the enemy of players ever actually getting to do anything that requires a group. The more options, the harder it will be to broker a compromise group. On top of that, player preferences will rapidly trend towards the dungeon that offers the best reward/time ratio, leaving the rest of the content underused and hard to find a group to complete.

This problem is further exacerbated by the not-so-horizontal progression that is the gear grind. Players have repeatedly demonstrated extremely limited interest in running content that does not offer their characters gear upgrades. So-called casual raiding guilds, whose less strict requirements are more likely to result in players with a wide range of dedication/attendance, are especially likely to get hit with this problem. The leading edge of the guild finishes a dungeon and decides that they don't want to waste progression time by leaving it on the calendar, as the lagging edge of the guild protests that they still need the old dungeon for the gear upgrades. Outside of guilds, this manifests itself in the form of demanding group members who overgear the instance in question to ensure a quick and smooth ride to the finish.

Ask not what your newbie can do for you...
What brings this issue to the top of developers' priority lists is the plight of the newbie. Players are especially disinterested in running content that their characters do not need for complete strangers. New players, unless they were brought directly into a guild by offline friends, are strangers to everyone on the server.

Ask yourself a question: how many hours in the last week have you devoted to running content that does not offer you personally any rewards with one or more complete strangers in your group? If your answer is non-zero, is it typical of your guild? Your server? If you (yes you personally) answered zero, then why should anyone else devote more time?

At the end of the day, telling players to be patient with camping the LFG channels all day in the hopes of finding a group that will not throw them out over their gear is not a viable solution to the problem of getting newbies ready for group content. I've spent entire evenings on such efforts, and it's really not any fun. The two options that developers have tried to date to try and cut down the size of this mountain are literal bribes (as with WoW's dungeon finder) to make raiders want to run the trivial content, or a deliberately induced bottleneck.

If you increase the cap and only have half a dozen dungeons available at the new one, everyone - veteran, alt, and newbie - ends up in those few dungeons, dramatically reducing the time it takes to get consensus on a dungeon run. That, more than anything else, is why games raise the level cap and remove content from circulation, even at times when they do not have enough content. Players will not take care of the newbie unless they are forced to, and that tragedy of the commons would ultimately leave group content to wither and die as players who leave through attrition cannot be replaced.

Automated Holiday Cheer-Finder

WoW.com made an interesting catch regarding WoW's temporary holiday bosses.

Remaking the Rare Holiday Loot Pinata
Currently, players must find a group for each boss the old-fashioned way (they tend to live in under-leveled dungeons, which max-level players cannot queue for in the LFG system). Once you get there, each player in the group, presuming that no one lies (you can verify this by checking your quest log to see how many groupmates also have the appropriate quest), can spawn the boss once for everyone's killing and looting pleasure. This would have caused problems for the random group finder, which does not currently screen out players based on who has or has not done a given daily quest.

With the change, the holiday bosses will be available ONLY via the dungeon finder (can't have a level 80 boss up and running in a level 26 dungeon once you remove the spawn quest). As with the emblems for completing daily dungeons (and soon battlegrounds), players will receive a special reward on their first kill of the boss per day - a bag with a chance to drop the rare loot formerly held by the boss.

Instead of five shots at a five-way roll against your group, you will now have one shot at loot that is guaranteed to be yours. Statistically, it's probably pretty close to a wash, unless you were sneaking in more than five boss kills per day by some method (such as agreeing to tank or heal for a desperate group, with the understanding that you had already used your summons). It also has the side effect of reducing the sheer number of the boss' common loot drops (such as necklaces, trinkets, etc) that enter the world, as most players will only kill and loot the boss for their regular loot once a day - I think I ended up with the complete set of trinkets from Coren Direbrew last brewfest, including the melee ones on my decidedly non-melee mage.

Philosophical Implications
Beyond logistical convenience, the first thing that jumps out at me about this change is that it drastically reduces the penalty for failing the encounter. Under the old system, once you verified that everyone had their summons, it was in your best interest to quietly sit back and let someone else go first. That way, if your group proved incapable of beating the encounter and disbanded immediately, your summon would still be available.

Under the new system, the group can simply try, try again until they get it right. This could be viewed as reducing the already-low difficulty of the holiday bosses, or it could allow Blizzard to crank up the difficulty a bit more without unduly penalizing that hapless first player.

More broadly speaking, there is a clear message that WoW players now expect the convenience of the automated dungeon finder to come standard with as much content as possible. This is understandable, but how will it play with the system's general push towards easy content (to alleviate players' fears of fail pugs)? I'm not convinced that holiday events really are the place for difficult encounters, when they are already limited by both time and the random number generator to begin with, but what about elsewhere in the game?

At the risk of repeating myself, it is going to be very interesting how Blizzard balances the level 85 heroics when Cataclysm arrives.

Item Store Games Are Supposed To Require Item Store Purchases

Syp has a comprehensive roundup of the blog reaction to the Allods Online cash store. To recap in extreme brief, Allods is a pure free-to-play (no box fee, no subscription, just an item store) game that was getting very good buzz until the game's item store prices turned out to be an order of magnitude higher than expected - for example, an absurd $20 for a six-slot upgrade to your bag space.

This incident may or may not do irreparable damage to the game's word of mouth. Publishers can try price tag brinksmanship, but it may not be forgiven, even if prices are later reduced. The bigger story, though, is that there was a giant red flag that should have warned players that something would have to change.

You know your item shop is broken when....
Buried in Keen's recap of the incident is the admission that, as of the patch before the sticker shock and awe, there was no need for players to actually purchase anything from the cash store to complete all of the content in the game. That's a big problem.

Though it's technically possible to build a free to play game in which 99% of the players decline to pay, I'd suggest that this would not be a game worth playing. A game in which 1% pay is a game in which 1% of the players matter. If they want fluffy cosmetic pets, that's where development will head. Hardcore raids? Expect the hardest of the hardcore raids anyone has ever seen. Whatever the specifics, the end result will be a game that will, by design, not cater to the overwhelming majority of its playerbase. That might sound fine if you think you can squeeze yourself into the 1%, but why not just make it a niche subscription game and cut out all the middlemen? (This might, in hindsight, be what happened to Free Realms.)

The most equitable way to implement an item shop is to design it as effectively an hourly fee. You might allow players to "pay" in time (grinding in-game currency to trade to cash-payers for items, or somehow toughing it out without the use of items at some greatly reduced rate), but everyone must pay somehow. Everyone has room to win under this model - the infrequent player gets a top-notch product without having to pay the full monthly fee, while the hardcore player knows that their voice actually does carry more weight.

Where should the price tag lie?
The irony is, if you look beyond the absurd bag upgrade, Allods might not have been as far off as you'd think. Keen estimates that serious raiding/PVP would rack up a $50-75 monthly bill. That's probably double what it should be, but - if the game actually delivers (which its fans seemed to think that it did until this week) - it's not a 10-fold overcharge. If the market standard is that everyone pays $15/month, there should be a niche for a product that comes in the $20-25 range for players in the top tier of content, with service to match. If anything, it's better to put the prices at the highest rate you can imagine and then lower them than establish a price point and have to raise it later.

Whether the damage has already been done is a separate question - this incident is more than enough to kick Allods to the bottom of my list, for "wait and see how this shakes out before wasting any time on it". That's unfortunate, because they might have gotten more right than the market gives them credit for.

The WoW Dungeon Finder And Future Difficulty

Tobold writes about the second demise of Naxxramas. Structurally, nothing about the dungeon has changed or broken since its heyday a year ago, when even non-raiders like myself were pugging it. Its current disuse is purely a failure of the incentive curve.

Easier 5-man content currently gives equal or superior rewards. Because the only real penalty for failure in a MMORPG is loss of time, choosing to do something that offers sub-optimal returns for your time means that you have already failed, even if nothing else goes wrong.

The Leeching System
The dirty little secret of WoW's dungeon finder is that the system is built on trivializing content. Icecrown raiders can easily triple the DPS I was doing a year ago when I cleared all the heroics for the first time, so even a single raider can literally carry two empty slots' worth of DPS through a five man group. Tobold himself titled a post describing the system "Sometimes you just have to leech" just a week ago.

But why would overgeared players want to do this? The answer is that they didn't until Blizzard offered rewards that were absurdly out of line with the difficulty of the content. This simultaneously made success worth the raider's time and nigh-guaranteed success by ensuring that groups would be filled with at least some similarly overgeared players.

Both side of this equation are crucial - without rewards, players won't want to run content, but it doesn't matter how good the rewards are if players believe that the attempt to get them won't be successful. This was the problem that made Blizzard attempt to raise the ante for the unpopular Oculus.

(The same principle holds true for leveling dungeons - players aren't as likely to massively overgear these dungeons if they don't have heirlooms, but character power has increased significantly since 2004 due to more powerful talents, glyphs, etc. The result is that appropriately leveled characters are more powerful than launch-era dungeons were balanced for, especially if you consider that players who voluntarily opt into group content repeatedly when they could solo instead are likely more experienced and competent at running in groups.)

Implications for Cataclysm
Tobold says that the current massive rewards for use of the group system were necessary only to convince players to give the new system a try. I am less convinced. There was a non-trivial period of time (at a minimum, the entire 3.2 era) in which it was very hard to get groups for 5-man content, even though that content was likely to be easily cleared, because none of the rewards were relevant to current tier raiders.

Raiders are generally among the most active group players in WoW, and I am not convinced that the current level of demand for the dungeon system is sustainable if Blizzard fails to provide motivation for raiders to participate. The random dungeon rewards in Cataclysm will certainly start at least a tier below the top raid rewards at launch, but I would be very surprised to see them stay there beyond the expansion's first content patch (or, at most, two patches in). The bigger concern is that, as I mentioned above, there's a second side to the effort-reward equation. It does not matter how good the potential rewards are if players are not confident that they will receive them.

The real test of the automatic group finder will be seeing how it tolerates the early months of the expansion, when few players will be overgeared, making the content relatively more difficult. The dungeon finder may have expanded the demographics of players interested in group content, but many of those players now equate dungeons with quick, easy, and highly rewarding content. Blizzard will have a lot of unhappy players on their hands if the competent players that make the current ease of the dungeon system possible decide that only pre-made groups are viable due to the increased relative difficulty of the content.

Future Proofing Expansion Experience Curves

EQ2 got its latest expansion on Tuesday, which meant that, inevitably, it got its first characters to reach the new level cap (90) on Tuesday.

This isn't especially unique to EQ2, even if the time involved was unusually short (reportedly as low as three hours after the servers came up) due to some of the game's quirks. First off, there are seventy five slots in player quest logs, not counting collection quests. This allowed players to "bank" an unusually large amount of exp to redeem immediately on launch day. On top of that, there are a variety of experience bonuses (recruit a friend, paid exp potions, and unpaid veteran reward exp potions) that are fair game even on expansion launch day. All of these added up to some very quick leveling.

Personally, I picked up nearly half a level in an extended launch night session. Judging from past experience, I will probably hit level 90 in about a month and be "done" with the solo quests, tradeskills, and reputations for Lyriana in around two months (plus or minus depending on workload and how beefy LOTRO's Volume III patch ends up being). That would put this expansion approximately on par with the time I've spent on other recent AAA expansions (WoW's Wrath, and LOTRO's Moria expansion), before working on alts or serious "endgame" content.

How long should an expansion run?
Though a single month or two may seem short, even in the context of an expansion every year, it's worth remembering that this figure represents a month just to get from the old cap to the new one. If you're joining the game from scratch, as I did about a year ago, you've got far more time ahead of you before you get to the level ranges where most of the players are.

Where you could make an argument is for a slightly tougher experience curve at launch and reduced requirements as the expansion ages (especially if the next EQ2 expansion does not increase the level cap - the game has only been raising the cap every other expansion for a few years now). However, what purpose would that really serve?

Having the steepest experience requirements on launch day, which is typically when there will be the least content available, runs the risk of leaving players without enough to do. Indeed, SOE had to spend a significant amount of time this patch adding content to the 60's because that level range relied too heavily on group content that players can no longer count on being able to complete. Moreover, if you ultimately do opt to speed the leveling curve at a later date, your latecomers may actually end up at a disadvantage in terms of less time to obtain gear and cash before their next level.

At the end of the day, experience levels are an arbitrary number. What matters is whether the expansion provides players, of whatever level, with enough to do.