Cyberpunk 2077: A Cautionary Tale for Management

By Julia Hur

Let’s assume there is a deadly pandemic in the world. You cannot eat out, see your friends or family, or pack your bags for short weekend trips. Then, a question arises: How can I still have fun when I am stuck at home? The answer to that question for gamers worldwide was Cyberpunk 2077.  Cyberpunk 2077 is a “next generation” action role-playing video game that every gamer was waiting for - it’s developed by CDProjekt Red, the studio that made The Witcher series, which was sold over 50 million copies and praised as one of the best games in the last generation. Before the pandemic, Cyberpunk 2077 was highly anticipated to say the least. After the pandemic, without any other major releases from industry heavy weights, everyone was dying to play this game - “the game” to rule them all, the game to bring them all and in the darkness bind them. 

Cyberpunk 2077 was scheduled to be released on April 16th, 2020, but it was delayed multiple times, all the way to December 10th, 2020. When the game was finally released, however, one thing was clear. The game was not finished. The company released a game that was simply not done and would need years to completion.  There were so many bugs and errors that made the gameplay sometimes almost impossible. While playing, you can see your own eyeball, have a gun stuck in your head, and see cars suddenly flying off from the road like fighter jets for no reason. Not only gamers shared their favorite glitches that were often most hilarious or most frustrating (or both), but even review sites like IGN joined this glitch game - who can find the funniest glitch ever. “Will there be memes in this game?” When CDProjekt Red was asked this question in 2018, they said on Twitter: “Whole game is going to be a meme.” And it surely did.

Not surprisingly, the game was heavily criticized, the major retail stores offered full refunds, and SONY at some point completely removed the game from their store. The CDProjekt Red management put out an apology video, and the company is subject to class-action lawsuits for concealing or downplaying the technical issues. 

Previous research on goals and motivation emphasizes the importance of goal setting. People who set any kind of goal perform better than those who do not set a goal. Also, people who set a specific, yet ambitious goal perform better than those who do not set a very ambitious goal in many work contexts. But how ambitious can we be? If I barely exercise and suddenly set a goal of completing triathlon next month, wouldn’t it do more harm than good? To set an ambitious yet realistic goal, you need to distinguish between ambition and overconfidence.

Overconfidence is an overly optimistic estimation of our ability to forecast future events. Basically, we think we can predict uncertain outcomes better than we actually can. It is quite common in business world. 80% entrepreneurs predict their chance of success as 70%, while only 33% actually survive after 5 years. Almost all project construction schedules run late by more than 50%. It is not really arrogance or narcissism. It is more about making predictions and having confidence intervals that are too narrow. Just think about the last time you thought you could finish your assignment in one hour and ended up pulling all-nighter. 

Cyberpunk 2077 is a classic example of overconfidence when you take a closer look in their development timeline. The game was first announced in 2012, but with everything else going on in the company, the developers said they did not even start working on the game until 2016. Four years passed without much progress. Then in 2018, after only two years working on the game, the company announced that the game would be released in 2020 and shared a trailer of how this game would look like, which looking back, filled with false promises and plans. 

The developers who interviewed with WSJ said their management kept setting up unrealistic goals and deadlines with “an unwavering belief that with enough hard work and crunch, their games would come together.” However, even with all the hard work by the development team (some said they worked 13 hours a day, 7 days a week), four years were clearly not enough time to finish this kind of game. Other major studios with twice more people and budget spend about 8 years on games like this. As results, as a desperate attempt to meet the deadline, they had to cut out many parts, put some mechanics at the last minute, and give up what they promised in the 2018 trailer including smart NPCs and dialogue options. 

? HOW CAN WE OVERCOME OVERCONFIDENCE?   

There are a couple of remedies for overconfidence. First, you need to 1) generate detailed work plan with small, short-term goals and 2) factor in other responsibilities. Instead of only setting a goal of “finishing the game”, once you start thinking about what exact steps needed to be taken to reach the goal, the estimate tends to be more realistic. List all the short-term goals for the project (e.g., review character designs, cast voice actors, program vending machines) and list all the other responsibilities besides the main project that employees also need to spend time on (e.g., fix bugs in previous games). 

Second, set a confidence multiplier. This kind of “buffer time” approach is quite common in production and construction - whenever an engineer comes up with any estimate, you multiply the estimate with a fixed number. For example, if the engineers estimate the construction of this bridge would take 2 years, multiply 1.3 to that, and make the estimate 2.6 years. If the developers say finishing the driving mechanic of the game would take 2 months, multiply 2, and make the estimate 4 months. The exact number for multipliers should be based on previous projects or common practices in the field. 

Finally, seek out clear feedback. Who would provide most objective and honest opinions about how long it would take to finish this game? Probably not those who really want to release the game in 2020. Probably not those who are not very familiar with day-to-day work tasks. Considering confirmation biases, ask yourself as well - are you collecting feedback only from others who you know would tell you what you want to hear? Are you collecting enough disconfirming evidence? Seek out clear, unambiguous feedback from those who might have different perspectives to better calibrate your estimate.