How Decisions Are Modeled – A Discussion About Anthem

By | 11/23/2020

Update 2/24/2021: In an update posted on Bioware’s blog, Christian Dailey announced the end of further development for Anthem. The cancelation of Anthem Next (or whatever it was to be called) might make this article appear “wrong”, but what I describe here is how projects are modeled. It just so happens that EA ran their models and decided the opportunity cost to develop Anthem’s rebuild was no longer acceptable.

Opportunity cost for Anthem is too damn high. Thanks ImgFlip. And yes, the rent IS too damn high.

The bullet points that follow are my “final thoughts” on the cancelation of Anthem’s rebuild:

  • Covid-19 happened. Working-from-home isn’t as effective as being in the office, especially when hundreds of people are involved. This slowdown in production creates extra financial pressure since every dollar being spent is less effective. As detailed in the blog post, Bioware is running other projects that are critical to the studio’s success. By choosing to place skilled resources on Anthem, Bioware incurs an opportunity cost for those resources not working on another project. In other words, when Bioware had employees assigned to Anthem’s rebuild, they were effectively reducing their ability to complete other projects. In a world where Covid-19 doesn’t happen, it’s possible that opportunity cost is acceptable and the Anthem rebuild gets the greenlight.
  • Bioware lost two significant studio leaders. At least one of them was a champion of the Anthem rebuild. As with most large organizations, change happens. A change in leadership can often mean priority changes, especially if the new managers/executives don’t “see” things the way that previous leaders did. That change can come in many forms. In the case of Anthem, I believe that Casey Hudson’s exit left it without an executive champion.

The original article, which details how organizations use models to make a decision, starts here:

You may have forgotten about Anthem, but Bioware hasn’t.

In a company blog post, Casey Hudson announced Anthem fell short and would undergo a redesign. This wasn’t too surprising since it was already rumored that Bioware was doing so, but the official announcement made it…official. As you might expect, Reddit went to work with its often deep and nuanced analysis:

Everyone is a business analyst when it comes to Anthem. Everyone.

Joking aside for a moment, there’s probably is a point to be made there: Why is Bioware throwing good money at Anthem? From a completely rational business standpoint, shouldn’t Bioware just put Anthem on server-only support, ramp down development, and save the capital that was marked to support it? Anyone who knows anything about the fallacy of chasing sunk cost knows that it’s time to end Anthem. Right?

Not exactly. Like many things in life, it depends.

Sunk Cost Is Always Sunk Cost

Most of the criticism that I’ve viewed, in regards to Bioware rebuilding Anthem, is something along the lines of: (paraphrasing) “Bioware is chasing sunk cost! Why are they doing that!?” It’s not an unexpected criticism as the sunk cost fallacy appears to make logical sense in this case: Has Bioware fallen into the trap of committing more resources to Anthem, because they already committed a lot of resources to it (the bias of ongoing commitment)? In other words, Bioware should recognize their ongoing commitment bias and walk away, thus leaving Anthem in “maintenance mode” for a few years (eventually shutting it down). The problem with that statement is that it represents the opposite of what a rational actor would do. A rational actor (most multi-billion dollar businesses are semi- rational) analyzes the situation moving forward from the data they have today. Remember, sunk cost is always sunk cost. It doesn’t factor into the equation from this point forward. I know it feels counter-intuitive, as this would seem to be a classical case of loss aversion or ongoing commitment bias, but the reality is that sunk cost isn’t part of the equation moving forward. Now that Anthem has been released, the modeling of Anthem’s profitability starts fresh.

Important Note: As game publishers & developers trend more towards service models such as EA PLAY PRO & EA PLAY, there will be less pressure on individual games to recoup their own initial investment with day one sales. Instead, it is expected those games add value to the service catalog, which in turn generates predictable and steady revenue.

Have finances. Will Model.

This is probably a good time to look at a financial model. For this section of the discussion, and to simplify the model, we are going to ignore related concepts such as opportunity cost. (Opportunity cost is significant and it will come up in most decision-making discussions.)

Update 2/24/2021: We now know that opportunity cost is likely to have sunk the Anthem rebuild, especially in a Covid-19 world.

I am going to focus on what matters most: direct profitability. Now that we understand the initial cost to build Anthem (sunk cost) is out of the equation we can look at Anthem as a profit & loss generator from this point forward. Finally, to keep thing simple, we will look at revenue generation per quarter.

When Bioware made the decision to “rebuild” Anthem, it was likely based on multiple models of outcomes. Traditional business strategy is as follows:

  • Build a financial model
  • Insert different revenue (or loss) scenarios in the model
  • Record the financial result of each model
  • Assign “chance” of each different scenario becoming reality based on market research (e.g. Scenario 1 = 35%, Scenario 2 = 10%, Scenario, 3 = 15%, and so on and so on…)
  • Talk some more. Discuss outcomes, outliers, etc.
  • Make a final decision. Rebuild or Abandon?

In order to help make some sense of this, I mocked up a very simplified scenario model in Google Sheets. (This is not to be taken literally. This is a model. It is for demonstration purposes only. I am not saying this is the literal financial model that Bioware is using.) Using this model, I propose that Bioware invests about thirty million on the rebuild phase (between launch one and launch two) and then runs at a quarterly expense rate of about 5.9 million per quarter.

A simplified financial model that allows the user to enter engagement numbers, expenses, and calculate quarterly revenue. This is internal business modeling, not GAAP. This is not how it would appear on the balance sheet.

The point of demonstrating the model is to show how a company might mock up different profit & loss scenarios based on different conditions. If a company generally finds the conditions to be favorable, there is a good chance they will commit at least minimal resources to the project, in hope that it can generate profit. If a company finds conditions are generally unfavorable, they will likely choose to cancel the endeavor. As I noted above, game services are changing the way that companies calculate revenue and the value they place on each title in the service catalog. In order to fairly judge Anthem as a profit & loss generator, some percent of revenue from EA Play Pro & EA Play has to be attributed back to Anthem. I accounted for this fact in the model (attributed three million per quarter).

The Cost Of Doing Business

The internet social-sphere might hate Electronic Arts, but I’ve always been adamant that EA is not the worst company in America, far from it. The idea that a video game company is somehow “worse” than the biggest carbon emitters or simply, Bank of America, is absolutely absurd. When it comes to judging EA’s actor status, they get a big “shrug” from me. They may have a history of arguably mixed decisions, and they continue to battle pressure from external forces (regarding their monetization practices), but EA is not the worst company the US. While it may pain people to admit it, EA has been walking the line of “good guy” lately, releasing the Command and Conquer source code and producing a Star Wars game without a shred of microtransactions. The bottom line is this: The people working and managing EA understand the cost of doing business. That means producing, creating, and building products that aren’t purely optimized revenue generators. In other words, the people working at EA understand that sometimes they need to be a socially conscious actor in order to safeguard long term customer loyalty.

One of Anthem’s best features was the movement & mobility of the player character. Many other aspects fell flat.

The game industry has changed significantly over the last twenty years, drifting towards a service model in many aspects. Given the connected nature of our technology, a game that “fails to launch” can be rebuilt, re-tooled, or reconfigured to give it a second chance at a cost that is much lower than building a new game. Obviously there is always risk to that strategy, and many games that have a less-than-optimal launch will be put aside in favor of moving on, but Anthem has one advantage that probably influenced EA’s decision to double down: total sales.

Anthem sold well in 2019. It sold so well (assuming that we can trust NPD’s data) it ended up in the top 10 best selling games in the United States that year. As almost any game developer / producer will tell you, it’s a heck of a lot easier to get a person to come back when they already have the game, than it is to get a new person to install / buy it. Regardless, Bioware knows Anthem sold well, and they know trying to recover at least some of their customers (and maybe grab a few new ones) is a calculated risk they are willing to take.

That’s the cost of doing business.

Bioware continues to update their Anthem blog with details and screenshots of where the game is (likely) heading. Header image credit: https://www.ea.com/games/anthem.

Did you enjoy this article? What do you think about the Anthem relaunch? Can Bioware pull it off? Let us know on Twitter!