Marvel Heroes & Free-to-Play Games

An Inside Look at Marvel Heroes, from a Past Employee

My Game Industry Debut

If you’ve read my introduction post, you’ll hear me talk about bringing my club in for an informational office tour, which then led to internship, part time, and full-time employment. This was with a company called Gazillion, based out of the Bay Area. At the time I joined, they had just launched Marvel Heroes 2015, which was a massive update to fix an initial launch one year prior that was poorly received on a number of levels. After the reduction-in-force (aka RIF or layoffs) that ended my employment along with ~1/3 of the company at the time, they went on to launch another update a year later — Marvel Heroes 2016 — and eventually porting the game to consoles while re-branding the game to Marvel Heroes Omega.

Marvel Heroes very recently shut down, and it was a very messy situation to say the least. The Marvel Heroes story is a super interesting one, and I’d like to dig into some of the reasons why I think the game couldn’t pull itself out of a sinking hole, putting aside some serious documented issues happening within the office, ranging from unfair work expectations to misconduct allegations towards the CEO. Let me also be clear that the CEO with allegations was a board member at the time that I was working there, not David Brevik (see below). Dave is an amazing person and very easy to work with (notice I didn’t say work FOR).

I’ll let you all do the separate readings regarding the office allegations, and focus on specific challenges the game faced that were ultimately too difficult to overcome.

Let’s Do a Crash Course, Shall We?

I can’t assume everyone knows how games operate, what terms & genres are, and other stuff. So I’ll go over some of the stuff relevant to Marvel Heroes and give some context so this is entirely more interesting, because it actually is super in-depth and fun to talk about when you understand some simple concepts & background.

The Action RPG Genre (ARPG)

The development of Marvel Heroes and the beginning of Gazillion was largely fueled by this man right here:

For the gamers out there, you will probably know who he is. For those who don’t, his name is David Brevik. He is an icon in the industry, mainly because he co-founded Blizzard North way back in the day — which eventually turned into Blizzard Entertainment. If you don’t know what Blizzard is, or if the name strikes a chord, it’s probably because they made worldwide news when they launched World of Warcraft (WoW) back in 2008 — a game that is still the #1 MMORPG to this day with millions of players online that all pay subscription fees on a monthly or yearly basis. At its peak, WoW had 12 MILLION subscribers, each paying a regular amount to keep their access to the online service active. This was in 2010, and since the 2008 debut they’ve launched multiple expansions over the years. Here’s a look at all of them, with the smaller icons being updates across the various expansions that introduced major, top-tier raids that involves a level of game design intricacy that is difficult to even explain, much less complete at the highest difficulties:

But Brevik wasn’t really at Blizzard for WoW’s debut, although he has said he was there when they had many talks about it. He was most involved with two games that literally defined a genre that we know as the ARPG genre. Those games are Diablo and Diablo II. They were unlike anything anyone had seen, and it’s experiences like this that invoked actual fear and excitement that people hadn’t felt before in this format. When you ask people about ARPGs, and specifically what hooked them in Diablo and Diablo II, they’ll often mention a couple of key unique features: randomized loot, procedural dungeons, hack & slash action, and the staple isometric camera.

Randomized Loot was an entirely new concept and, for the same reasons that people love playing slot machines — the prospect of getting something amazing that is also completely unexpected — it was a no-brainer why people were hooked when an ultra-rare item dropped from some monster you killed. They even gave you distinct audio ques so you knew when the good stuff dropped. It’s textbook classical conditioning. You hear this sound and you know you’ve gotten something special.

Randomized loot is integral to the genre, not just because it gave that exciting feel of the possibility you’d find hidden treasures that made you more powerful, but because it helped drive incentive to keep playing where things can otherwise get quite repetitive. More on that later, as I’m convinced it played a part in dulling the Marvel Heroes experience.

Procedural Dungeons were another amazing feature of ARPGs. While the Towns were static, every time you entered a new dungeon instance, it was completely different. I could play the game through millions of times and never have the same dungeon, and it prevents replayability getting washed out over a period of time. If I run a dungeon that always looks the same, I will very quickly map the most efficient route and behavior becomes automatic rather than reactive. Automatic behavior as a driver of the main game experience is a deal-breaker for most players. Why play the game for multiple hours when it’s dull and not entertaining?

Hack & Slash action is a huge driver. It made you feel powerful. It was badass. It got super intense at times. You’d be slashing through hordes of monsters and having to choose your positioning carefully, watch your health, and manage your resources, like mana & stamina for powers and running, respectively. It was also completely unexpected sometimes to walk into a room to find 50 monsters in a tiny space that immediately turn on you. This action combat drove excitement and heart-pounding moments that kept people coming back.

The Isometric Camera refers to the viewpoint the player has. It’s a top-down, angled third-person perspective. Your character moves in a 3D space, but the camera stays put. Marvel Heroes specifically had some cases where this had to be adjusted, particularly in their raids since they aimed to also be an MMO. Here’s some perspective shots for an example:

Free to Play (F2P) Games, Microtransactions (MTX), and Games as a Service (GaaS)

Marvel Heroes was a F2P game. Anyone with the hardware to run it could download the game and play instantly. Of course, Gazillion is a business with employees, investors, and overhead that they need to pay. They therefore make their money by selling digital in-game content to their players in various forms.

Leading up to launch, it’s more or less standard practice to offer “Founders” or “Early Access” packs for varying prices that can grant early play privileges to players along with other exclusive in-game content. After launch, different packs & bundles can be sold for numerous occasions or remaining on the store for the life of the game (aka evergreen content).

Microtransactions & Lootboxes

Microtransaction refers to purchases that happen within an in-game store. There are different types of items you can buy, but they can largely be grouped into two categories: durable and consumable items. Durable items are things you can buy once and own forever: costumes/skins, mounts, or characters/heroes. Consumable items are as you might guess — you use them once and they give you some sort of benefit and disappear. These can include boosts that last for a period of time (experience, luck/drop chance, or some other unique benefit), items that help expedite in-game mechanics like crafting, and “lootcrates”, or random-chance boxes that have varying levels of opportunity to get different kinds of items. The most desirable, of course, being a low chance to receive. From a business perspective, though, the strategy heavily leans on what your store weight leans more heavily on — durable or consumable. I’ll delve into this at a more granular level later, but it’s a large factor into why I think Marvel Heroes couldn’t recover.

You might have recently heard about lootboxes, and it wouldn’t be the first time they’ve received scrutiny. The big recent headline came from EA/DICE’s Star Wars Battlefront II. It’s a long read, so I’ll give you the TL;DR. Essentially, the time it took for players to rack up enough credits to unlock heroes like Luke Skywalker would be around 2 days of actual playtime — for a single character. The workaround would be to buy random loot boxes, and people were enraged that the progression system heavily pointed towards a suggestion to buy lootboxes. But the Battlefront community has an even larger point to be upset about a system that encourages purchases, especially ones that have a random reward — the game wasn’t free to play, you had to pay at least $60 to buy it.

More common criticism of lootboxes comes from the random element of the rewards, and the lack of regulation to disclose the chance to receive each item in them. Players are often upset that they pay varying amounts of money and still don’t get what they want, and question if they even can get that one special item. The only saving grace is that they usually see someone else running around with it.

Gambling?

So let’s address the elephant in the room: how is this not considered “gambling”? That is actually a great question that I wouldn’t be able to provide a confident answer if I hadn’t gone to work at my next company after Gazillion — GSN Games. GSN Games (yes THAT GSN as in Game Show Network the TV company) made mobile social casino games. Our competitors were Caesar’s, Zynga, and others. The apps had varying types of games ranging from video bingo to slot machines. They made MASSIVE amounts of money selling “tokens” that are then used to play the games. What boggled my mind was that while you could buy, play, spin & win, you could never actually cash out the winnings for money. Why the hell would people pay money to gamble when they couldn’t recoup their winnings?

GSN was also confused by this, and they did extensive research as to why by asking their actual customers what their incentive was to play. It turns out, and this is where the older-skewing demographic of 50–60 something year-old women came into play, they just liked playing something and winning from time to time. It was a pass-time. They didn’t care so much that they couldn’t cash out, they just liked winning.

And it’s that single point on not being able to cash out that rules it out of the “gambling” definition. This allows them to operate in all states without massive legal headaches and state exclusions.

In order to be “gambling” it has to have three elements: you must wager something of value, you must have a game of chance where you can win or lose that thing of value, and you must be able to take that value out in cash.

The frustrations are still understandable. Someone could get the best thing first try and someone could spend hundreds of dollars and still not get what they want. Although not technically gambling by a legal standard, the draw to lootboxes is identical to the draw in the excitement a potential jackpot on the slots presents. Counter Strike: Global Offensive’s weapon crates (which yield varying community-produced skins to make your weapons look cooler) is almost a literal slot machine. It’s just missing the handle to pull, which is instead a button. Not a big difference. See for yourself:

Separating emotions isn’t really relevant, because the drive to buy in is still the same. If anything, it feels more like a slap in the face because you can only use that item or thing in-game and that value is useless otherwise. Except Valve is walking a super thin line on the edge of the grey zone and the black zone. They have a community market, where users can sell their skins — some actually ranging over $1,000 for the rarest — to other users. This value is gained by either selling skins to other people or depositing actual cash to your account. You can’t actually cash out that value, but you can certainly use it to buy games on their portal. So if anything, Valve is at most risk of being targeted for potential gambling, since you are still technically able to obtain something external. But Valve is a massive company and they have the cash to throw when it comes to stomping out litigation fires, so barring some massive worldwide collaboration suit from the biggest countries, I don’t see it being reversed soon.

So Why F2P?

You’d be right to ask, it seems like there are so many edgy controversies. For one, whether players like to admit it or not, the idea that they can play a game for free will never get old. It may cause big challenges with player retention, a topic I’ll elaborate on later, but if there’s no barrier to entry outside of them downloading the game, they will play. And as long as there are games to play, there will be variety.

That’s the second portion, the business element. From a business perspective, F2P games offer lots of hedged risk and a more stable business model. In pay to play games, if you’re not a Blizzard, Ubisoft, or EA, developing a game for two years is a big bet to make that you’ll recoup your investment and have profit when you launch the title at a fixed price. Your timeline is to get funding, develop, hit all your timelines for development & marketing/press activities, and launch. Whether or not future updates (often called Expansion Packs or Downloadable Content/DLC) happen is dependent on if there is confidence that it’s worth the extra investment from profits or additional investor funding.

With F2P games, you take much of that out of the equation. While your game costs nothing, you essentially can offer exclusive content to players — which often has a similar price point to the full pay-to-play games, if content is enticing enough — and then continue to offer paid content in-game. Pay-to-play (P2P) game players have a fixed lifetime value (LTV). That value is whatever the price of the game is and whatever other expansions you release, barring some in-game shop that offers something like realm transfers or faction changes like WoW. In F2P, the reality is that ~87–93% of your entire user base, EVER, will not pay you a single dime. But that 7–13% will drive your entire business. They will have LTV ranging from tens of dollars to, and I’m not exaggerating here, tens of THOUSANDS of dollars. Since this model ensures a more steady stream of revenue, it’s less risky because you can set out a content plan over years and provide assurance that the business can meet forecasted goals, and if not adjustments to remain on track are easy to pivot towards.

Games as a Service (GaaS)

This concept was put into the spotlight by Blizzard with World of Warcraft. They were not the first to put out a GaaS model, but it drew from the idea that deviated from the traditional “Let’s develop a game, release it, and let people play it. That’s it.” Think any game you played on early consoles before the internet was a thing viable to use. Mario, Ocarina of Time, etc. Those games were not a service.

Most games these days are, actually. It encompasses a company’s ongoing commitment to provide an online experience, and that requires constant service and support. While there are many games that succeed today that aren’t GaaS, a vast majority that are topping the charts on sites like Twitch.tv (a livestreaming platform, bought by Amazon for ~$1 Billion) are GaaS. It’s almost an industry standard, but it’s not fair to say that’s an accurate statement yet. Nonetheless, most developers are driving towards GaaS models, as it generally means a longer business timeline and game lifecycle.

Quick Recap

I went over some basic ideas, topics, and terminology to make sure everyone is on the same page. I apologize if this was repetitive for some of you, and my hope is that a consistent linking of my gaming-related posts will help to guide those who need some context & background to the right place so I don’t have to consistently have an entire post dedicated to it. Nonetheless, I touched on the general background of Marvel Heroes & Gazillion, the ARPG genre & its foundations, F2P games & microtransactions, and the games-as-a-service model.

On to the “juice” of the matter :)

Monetization Strategy & F2P Challenges

In Part 1, I outlined how F2P games hedge lots of risk versus P2P games when it comes to a business timeline. This, of course, does not come with a giant bag of risks and challenges either. As with any content people pay for, your players need to actually see value in it. Even though most people will never pay you anything during a F2P game’s life cycle, it doesn’t mean they don’t acknowledge that value.

One major challenge is making sure you offer MTXs that don’t embrace Pay-to-Win (P2W) mechanics. P2W is a term the community associates with something someone can buy to give them an advantage in the game, thereby rewarding and enhancing the game experience for those who have the cash to spend. Keeping in mind the stats I shared about a vast majority of people in F2P games that will never spend any money ever, this alienates 87–93% of your user base instantly. And, with most games being GaaS oriented — or always online, thereby leaning on the strength of a large, active set of players at all times to make the experience fun — P2W models will cripple games. Games that pursue such models actually have data that more closely mimics pay-to-play games, since they generally boom & bust — relying on a massive surge in revenue initially, some big spenders to keep them afloat for a little while, and then a shutdown of operations.

If a successful F2P game is to expect a healthy life cycle and a happy user base, item offerings need to make players feel like they’ve spent money on something worthwhile while not alienating the rest of the players who didn’t spend money (or feel like they are pressured/obligated to spend in order to have fun, since that barrier was removed when they were told the game was free to play). I touched briefly last time on in-game items being broadly separated into two categories: durable & consumable. As you might guess, players love durable items because they can buy them once and own it forever. Consumables have mixed reactions depending on what they are and what benefit they provide.

You might start to see a bit of a trade off here when it comes to operating a successful business and keeping players happy. The store balance of item type needs to be in order while supporting the business timeline. What % of store items are durable vs. consumable determines your business strategy and how you allocate company resources to development of those features. Ideally, you’d want a store that is more or less balanced between durable and consumable items, although durable items often end up taking up more share. As such, if you have a store that is mostly durable items, your job is to develop and release that content on a cadence that will allow your business to continue, since once people buy this item they own it forever.

This was one MASSIVE problem with Marvel Heroes’ model. I’m going off of a fan wiki page, since the Marvel Heroes websites & forums have since been shut down, but they have the store vendor information including quantities of each type listed. In the link, there’s the cost in “G” which was their premium currency. Essentially, 100G = ~$1 USD, and store items ran from 50G — ~1,800G ($18). While I don’t have exact numbers, the overall spread is still spot on. I’ll list each out, mark it as durable or consumable, and assign it a total % share of the item offerings in the store:

• Costumes (Durable) — 69%
• Heroes (Durable) — 12%
• Pets (Durable) — 3%
• Team-Ups (Durable) — 8%
• Fortune Cards (Consumable) — 4%
• Ingredients (Consumable) →1%
• Misc (Consumable) — 2%
• Consumables (i.e. Boosts, etc) — 1%

If you’re looking at the above %s and are scratching your head, don’t feel strange. The store was 92% durable items. Here’s what that looks like at a granular level, per above:

Here’s what that looks like at a high-level view of durable vs. consumable:

I know it might seem dumb to show you a 92/8 pie chart, but sometimes visualization puts things into deeper perspective. This means, if you are to run a successful business, you NEED to churn out new Costumes & Heroes. Pets & Team-Ups (or followers/companions) were cool — and actually most serious players had all of them because each had different bonuses that applied better in different builds — but they weren’t the primary money drivers. Costumes (visual changes to your hero, which were always derived from a past Marvel story line/comic) and Heroes (actual different playable characters, with their own set of powers, skill trees, and voice over (VO) lines, were the primary drivers. I’ll take a moment to discuss Heroes specifically, because Costumes encompassed a similar production path but with less depth, although still very challenging in their own right.

Marvel Heroes was different than many ARPGs in that it deviated from the standard model of having you choose at character creation what type you wanted to be. Typically, you’d select from some generic classes like Warrior/Barbarian (Heavy Melee), Rouge/Scoundrel (Light, Agile Melee), Wizard/Conjurer (Magic Caster), Witch Doctor/Summoner (Debuff/Creature Summoner), or Paladin/Healer (Buff/Healing Caster/Melee). While each Hero in Marvel Heroes generally encompassed one of those types, they were still vastly different from each hero and many were hybrids of the main archetypes outlined above.

Marvel Heroes was clear about their commitment to release one hero a month along with other content, and that was — in my opinion — one of the largest pitfalls for the game. At the turn of the year, in an attempt to retain players and realize revenue early on (and thereby foregoing much in the future), Marvel Heroes would sell “Advance Packs”. These would essentially give buyers access to every hero to be released during the year, with unique costumes (also TBR), and other goodies like unique boosts & currency. Since players were buying into unreleased (and sometimes unannounced) content, this helped keep them invested and anticipating what was to come. The obvious problem here is that once you take money from players for something like this, you are locking yourself into a timeline that supports that sold content, and it severely restricts your ability to shift resources to develop other key game content down the line.

Creating all-new heroes was no small feat, either. Think of the level of depth it takes to create something like the Barbarian in Diablo III. There’s character VO, modeling, animations, visual effects (VFX), active & passive skills, item development & tier progression (level 5 items, level 20 items, max level items — each fitting into the primary stats that benefit that hero type, strength and vitality specifically in this instance). Those are conceptualized, built, and then everything is tested before it ships (goes live). That’s for ONE character, and while the 58 Heroes all drew from the base archetypes of other games, they still needed their own development of things I just mentioned. Now think about how that had to be delivered on a monthly cadence, with each Hero being completely different in every way, all the while fitting into the Marvel universe. That is one HELL of a task, and it’s nothing short of incredible the team was able to keep that up for a few years.

Brand Partnerships: A Blessing & A Curse

There’s one small bit I want to touch on that you might have glazed over at the end of that last paragraph. Each Hero had to fit into the Marvel universe. This is another thing I’m convinced made it too difficult for Marvel Heroes to climb out of its sinking hole. Undoubtedly, aligning and partnering with a brand like Marvel has incredible prospects and is an enormous draw for a potentially massive audience. People LOVE Marvel movies, the comics, and playing as their favorite Heroes. Who else would pass up an opportunity to kick butt as Iron Man, Thor, The Hulk, Spiderman, or even Dr. Doom? I sure didn’t and many others didn’t either.

But there’s a reason why Marvel is the successful brand that it is. They explore other media outlets, and allow the contractual release of licensed content. However, Marvel’s entire business is propped up on their ability to make sure that every bit of content released is aligned and consistent with a spiderweb of lore that has arguably transcended a level of detail that can be feasibly mastered by anyone. Nonetheless, they make the rules and understandably make great efforts to protect their brand, which means any licensed content must be severely vetted before going public.

This included Heroes, Costumes, Team-Ups, Pets, any new game story content or modes, or basically anything that went into the game. Outside of the game, all forward-facing assets (Facebook images, Twitter images, advertising images, etc), social media posts (yes, text), official statements — all had to be approved by Marvel before going out. Each of these had varying ranges of approval times, ranging from a couple days to 2–3 weeks. If it sounds crazy, it’s because it is. But with a brand as huge as Marvel, you HAVE to do this. Please don’t mistake me here, I am not knocking Marvel. Their brand was stamped all over our game and everything we did. That’s just making sure your brand is being represented properly, and it’s why they’re successful they way they are today.

It did make business absolutely tough on our end, and that’s one challenge to consider if you’re ever considering entering into a licensing agreement with a major brand. I didn’t even mention that for each Hero being created for a monthly release that each voice over line, each animation, each visual effect, each description on every skill (including the skills themselves), and the names for every item that they could possibly wear — all had to be approved by Marvel and done so that the Hero could be released on time.

That’s also leaving out Costumes, which were also heavily vetted. I remember seeing one Costume had a rejection with a note that gave a specific color shade to use (which I couldn’t actually tell the difference, but we never argued and just forwarded to the designers) and a separate note that said “Crotch bulge size is too big, please reduce size and resubmit”. Yes, crotch size. It was a bit of a laugh at the time, though.

So as a small studio who has committed to one Hero a month, Costume(s), in-game events, content aligned with movie releases like The Avengers: Age of Ultron, and producing other new game content, you can see how this can quickly turn to utter insanity. There were numerous times when the Hero or Costume we’d had down to final approvals — scheduled to launch the following week — was rejected and needed to be resubmitted for a turnaround time that would put us past when we’d planned on releasing. From a communications standpoint, it’s not much of a challenge when your players are informed and not given hard release dates for said content. From a business standpoint (going back to my point about being able to project a better long-term timeline for KPI like revenue) that makes it a much bigger challenge.

The best way I could describe our marketing, being one of the members of that tiny team, is reactive marketing. We always had a plan, but it so frequently got disrupted due to rejections for things so minor (again, see my point about it being necessary from Marvel’s point) that we were seemingly always coming up with things to put on sale and gather assets last-minute (also needing rush approval) to avoid a massive money gap. Hero rejected? It won’t go live, what do we do? Costume(s) rejected? What other costumes can we put on sale?

Also, we haven’t even nodded at the fact that people generally play small groups of Heroes, mostly of the same type or those that they are fans of. If the new Hero/Costume coming out doesn’t appeal to those who don’t already own it (which was a large share of active users who bought Advance Packs, generally members of the Golden Cohort — which I’ll clarify shortly), they won’t buy it. The sole driver of the business assumed the new content would sell, but not everyone wants to play Jean Grey in her new costume (even though she was seriously badass in the game) or Star Lord following the first Guardians of the Galaxy movie in theaters.

We eventually found ourselves more frequently running premium currency deals, which typically do very well. But larger frequency in those types of sales devalues your #1 base driver of your revenue, and eventually sales fatigue happens. Sales fatigue basically means sales happen so often they mean less and less to your players, and they start anticipating them more, thereby spending less on a regular basis and waiting for the next one to come up. That is a recipe for death in a F2P game, especially if it’s concerning your premium currency. The business model starts shifting from a focus on new content (which was the original plan & driver) to propping up the business on sales, which generally alienates your Golden Cohort users, and the death spiral begins.

What is the “Golden Cohort”

Ok, so I’ve now mentioned this term a few times. What is it? The Golden Cohort is your segment of users who have been with you since the beginning, and they drive a vast majority of your business for the life of the game. They are, by far, the most engaged people. This is another blessing/curse scenario. If you are to look at lifetime value (total amount spent, ever) of users in your game, organized by the month they were acquired and spanning over the life of the game, it might look something like this (this is purely me punching numbers into a spreadsheet, and is NOT actual data):

The X-Axis here is months after launch, with negative numbers representing the months leading to the launch month, denoted by “0”. That revenue initially can be realized due to pre-sale content packs, which often gains access to early sessions like closed/open beta testing in addition to the promised exclusive content. As you can see, ~45% of your TOTAL revenue for the entire life of the game comes from people who play/pay in the three months leading up to the launch, at launch, and the following month. The spikes you see down the line are major content updates where you might expect lots of user acquisition efforts and press coverage, bringing more new players & buyers into the game. Again, this is a very, very rough estimate to get a picture across to help illustrate my point. In reality, the bumps become less significant and it tails off much more significantly, so your Golden Cohort might actually be ~60–75% of total lifetime revenue.

What’s good about this is it’s pretty easy to discover what your users want, because they are the ones sticking around and understand the game the most. What’s bad about this is that your “vocal minority”, while they might provide solid feedback and positive input, are now the “vocal majority”. You must constantly cater to them, since they drive your business — lose one of them and it’s not just “another user gone”. Once the GC users start dropping off, it’s a dire situation.

Furthermore, in catering to that small group, you’re catering to a group of very experienced players and often don’t have the time to invest into what will make the game a better experience for new players which will help the game grow and extend the lifetime of the game, including improving the overall experience for all players since there’s a larger player base. Why is it that new users don’t retain and drop off so quickly? What’s a sticking point? Is the game too hard at first? Are there too many systems that they are overwhelmed by and can’t get acclimated to? Are there too many Heroes to choose from? Do they use their one free Hero unlock on someone and find out they don’t like that Hero?

These are all questions that help great games stand the test of time and prevent users from coming in, churning out, and never coming back again. You have one opportunity to capture a player. It’s highly unlikely that they come back once they’ve decided the game was undesirable, so it’s best to get it right the first time. But if you have no resources to dedicate to optimizing this because you’re stuck on a pre-determined timeline catered to those who you’ve made a promise to, you simply won’t be able to. Over time, that experience inevitably becomes dull for even your most invested players. While efforts to expand the experience to other player bases might normally be welcomed, thinly spread resources can easily shift focus from your core group of players and they’ll lose interest entirely.

Which is exactly what eventually happened with Marvel Heroes.