Shared publicly  - 
 

EA is not a person
We tend to anthropomorphize companies. We accuse companies like EA of repeating mistakes and of failing to listen to our outpouring of anger or grief.

The reality is that companies cannot learn.  There is no conscious entity representing the company as a whole.  The aggregegate behavior we ascribe personality and humanity is emergent from a million tiny interactions between processes, data, culture and individuals. 

The only real piece of a company that can learn is the individuals inside the group directly dealing with a specific problem. 

Large companies destroy individual learning
Assume that a team or individual in a large company makes a big mistake.  Some of the engineers and designers see the data and figure out a solution.  They may or may not fix the immediate issue. 

If the company was a person, we would expect a person who makes a mistake to be conscious of the issue and diligently avoid the same problem again in the future. 

However, what happens next in a big company?  Often, in large companies, those newly educated people go separate ways. The stink of failure prompts less involved management to push the most knowledgeable team members either out of the company or onto unrelated projects. Politics, status, shame, churn, etc all prevent lessons from spreading. So the institutional learning becomes heavily fragmented.  Conscious understanding of the problem is lost in the muddle. 

The next group that tackles a similar problem may have only the slightest hands on understanding.  They go through a birthing process of formation, fighting for understand and then forming new social norms.   If teams have personality, then this new team has both radically different memories, skills and goals than the team that came before them.  _Imagine waking up next to your spouse each morning and they were a totally different person with only fuzzy secondhand memories of who you might be._

The human brain lacks the tools to understand large groups
In smaller groups, a single educated person can have a massive impact.  They can speak up and let the group know about their past experiences and the group can use tools like consensus and respecting expertise to shift directions. 

A consumer is an ape in khakis.  They imagine that all groups act as either individuals or small groups informed by expert individuals.  It is a cheap, effective mental model of society that has served us well for a hundred thousand years. You can empathize, put yourself in their place, with a small group and predict with reasonable certainty how it might react in the future. 

EA or any giant organization is not a Person. Nor is it a small group. Empathy fails.  Modeling the dynamics using simple human models fails.  And people get upset. 

When an individual doesn't do what we think they should do, we feel a breach of trust. In return, we may try to salvage this relationship by asking for explanations or requesting a change of behavior.  When there is no obvious response, we escalate our communication in volume.  We may try to ostracize or shun the individual. 

We do the same thing with companies.  It doesn't work. These are the wrong tools for understand massively emergent hive dynamics. The people we are talking to are some PR firm or talking head.  They are in a different organization with different rules and constraints than the people involved in whatever specific issue may have caused a breach of trust. There exists no relationship.  There can exist no honest communication.  

The best companies create an illusionary facade of such expected human properties, yet it only goes so far.  Once the group size is large enough, it simply isn't physically and socially possible for a company's behavior to be understood in terms of primitive models of empathy and trust. 

Any consumer's faith in the continuity and humanity of a brand or large company is no more justifiable than faith in the Spaghetti Monster (or similar variants). 

Put your faith in people, not companies 
The good news? People are indeed learning when horrible things happen.  However, it is typically individual learning or occasionally, small group learning.  Failure is a wonderful teacher as long as it doesn't kill you. 

These individuals are the single best chance any consumer has of avoiding those mistakes again in the future. 

Instead of directing ever increasing volumes bile at the false target of a company try an alternate approach:

- Learn who the individuals are working on a problem.  Separate the person from the company. 
- Accept that they made mistakes and will be better for them. 
- Engage them in a conversation.  Figure out what you can add to their knowledge. 
- Support them when they do something similar, but better.  Even if it isn't in the same original company or even in the same IP or brand that you learned to love. 
- Avoid screaming.  If you wouldn't say it to your trusted teammate, don't say it to a developer.  They are human too.  As humans, we tend to tune out weirdly angry people when engaged in polite conversation. 

This is may seem like a high bar for interactions between creators and players.  Yet it occurs every day between smaller indie developers and their passionate fans.  Just because big companies are structurally inhuman doesn't mean that game development as a whole needs to be that way. 

All the best, 
Danc. 

PS: My heart goes out to human beings working on SimCity.  Independent on the brand promise, independent of the company they are in, independent of the launch issues, the design problems they are trying to solve are fascinating, cutting edge stuff.  

They woke up next to you with a fresh new personality complete with different memories, skills and goals. And you immediately started yelling at them. :-)
57
25
Zack Hiwiller's profile photoSteve Kunzer's profile photoWilliam Payne's profile photoBrian Green's profile photo
21 comments
 
There is truth in what you say, but... companies are run by intelligent people, and intelligent people should put in place processes (and a culture) that prevent major mistakes happening, and minor mistakes ever making it out into the public domain. Really smart people put into place mechanisms such as lessons learned logs, they make sure learning feedback from previous projects are proliferated and fed into subsequent projects. Of course a company isn't an individual but it's reasonable to attribute blame to a company (knowing that you're actually blaming the individuals responsible for leading that company, and any culture they are responsible for fostering). If the executives who lead a company are not responsible, then who is? Few other people have sufficient power, or accountability to influence the myriad of factors that lead to such serious errors.  By all means don't blame the developers, or the host of people with no ability to change the big things. Avoiding hostility towards the people working on the problem is a good thing, but don't let "the company" (i.e. the executives) off the hook - they are directly responsible, and are the only people who can fix the major problems.
 
It is SO EASY to say "they should put in place processes to prevent that from happening".

Tell ya what. Try it. Find out for yourself how incredibly difficult it is to take a 5000+ person organization and "put in place a process" that will produce the expected result. I've spent decades doing it, and I can tell you that such statements tend to belie an over-simplified view of people and process.

Dan said it great: humans lack the cognitive tools to predict and work within groups in an instinctive way. The only way large organizations generally improve is through repeated iterative failures - but with each step, we tend to crucify the people involved with those failures.

Blaming "executives" may be a convenient mental short-cut, but it turns out to not reflect reality well. Finding someone to blame for a problem (that in reality has a broadly dispersed responsibility) does not help solve the problem. In fact, it furthers the myth of personification that Dan is decrying.

If you want to truly understand large group behavior, get comfortable with the inhuman paradoxes of scale.
 
Okay +Jason VandenBerghe calm down. I've spent a long time designing and putting in place processes to prevent stupid decisions being made, I've spent decades managing projects and programmes, learning lessons, documenting those lessons, providing feedback and proliferating those lessons. Of course these things depend upon human beings, and "executives" are human beings too.

But "executives" are responsible for the culture of an organisation and for the kind of decisions that cause the fiasco EA are suffering here. Oddly enough I've also managed a large scale programme for EA games too, and I'm pretty well aware of how the culture of the company works. 

Saying "it's just group behaviour" and "nobody can predict it" provides excuses for people whose job it is to put processes and culture in place to deal with the result of things driven by the culture they foster.

I'm absolutely not saying it's easy. But I absolutely am saying it IS the responsibility of those "executives" and they must be held accountable for it too. If not, too many people suffer the consequences.
 
+Steve Kunzer +Jason VandenBerghe In my experience, you're both right.  Executives are in charge of the culture, but it's much easier to establish a corporate culture than to change it.

I've seen this repeatedly in MMO games, where if you just let a culture grow without strict guidance, it tends to not be very friendly.  Look at the culture of any competitive online game and you'll see what I mean.  Trying to "clean it up" just leads to the feeling of fighting a hydra, ban one racism-spewing idiot and some mother-insulting moron springs up in his (or sometimes her) place.

It is possible to change.  League of Legends spent a lot of time and effort to clean up their community, and the early reports were that it was working.  By rewarding good behavior with recognition, the incidences of rude behavior have decreased.  But, make no mistake, it was a concerted effort and imposed from the top-down.

It's entirely possible to create a culture that respects mistakes instead of shunning them, that institutionally values analysis and sharing of information, and that essentially learns from the past.  But, it's incredibly hard to maintain that as the organization grows large and damn near impossible to do so after a culture of mistake-shaming and willful ignorance has already set in.
 
Having strong leadership in a company gives you the ability to learn. As long as the company has trust in the CEO and his team, they should be the ones that learn and pass the knowledge, as employees come and go. I guess this is the reason that high ranking officers in large companies get very high salaries - you don't want them go and lose the accumulated knowledge.

What is a strange approach, is human investment in a product. People tend to associate themselves with certain objects (both material and immaterial), can't let go and the blame the company, if their needs are unsatisfied. If you don't like some product, simply switch to something else, instead of trying to force company to change. Voting with the wallet is the best strategy ever invented.
 
+Steve Kunzer Thanks for engaging. :) I appreciate your willingness to examine the problem from both ends. I don't calm down, not really, not ever, but thanks anyway. ;)

So, I never said "its just group behavior" and I certainly never said "no one can predict it". Those are (common, boring, oft-repeated) polarizations of what I said.

To be clear: the fact that the complexity of group dynamics make the true outcome of change to those dynamics nearly impossible to guess with certainty ahead of time does not mean that there aren't reasonable ways of dealing with that problem. There are, and I'm sure you have used them. So, we agree. However, this problem (as you pointed out) is full of more grey area in the middle than binary poles on each side, so the truth is more like "it's damned hard to control a process of large group change, so bring your best people, and be ready to have as many failures as successes".

That said, it's still a cop-out to blame the leaders. Are they to blame? Sure! Should they be held accountable? Sure! Aaaaaaand so should everyone else who contributed to whatever toxicity we're discussing. I've spent about 90% of my time in this business cleaning up other people's 'challenges', and I can tell you that the source is never one person. And only very rarely is it one strata of people.

You're right that culture starts at the top, but in no way does it stop there. Generally, playing the blame game is a nice way to survive in large-scale environments, but it's rarely a correct assessment (with notable exceptions, several of which I have experienced first-hand), and is fundamentally a way of avoiding the hard, frustrating, and often embarrassing work of connecting and collaborating across discipline and salary-level lines.
 
In some companies it's common to write a postmortem after a crisis and publish it widely so that others can learn from the issue. This doesn't guarantee learning since not everyone reads it and memories fade, but it does mean that the knowledge is spread pretty widely.
 
+Jason VandenBerghe As you just said - don't polarise my views either. I'm not saying "it's all the leaders fault", and I hate blame cultures. However, corporate cultures start at the top. Of course they are contributed to by all levels of an organisation, but they start at the top. It's not easy to change a corporate culture, almost impossible, but you have to deal with the situation as it is, not how you'd like it to be.  I'm not here talking about the detail of the things they have to do, I'm talking about the bottom line of accountability. The bottom line (and there has to be a bottom line, like it or not) is that someone made the decisions that rippled down into the fiasco that happened, and someone helped nurture the corporate culture that allowed the fiasco to happen. Blame? Well, yes and no. First, get the problem solved (executive decision), then address the accountability. You cannot say on one hand "it's a cop out to blame the leaders" and then accept that they do have accountability. Take a sports analogy for example: the players in the team may be responsible for performing (or not) out on the pitch, but the manager picks the team, decides the tactics, chooses the substitutes, and directs the team to adjust depending on what happens in play. Ultimately the manager takes the flak and is sacked (typically) when the team lose games. The whole team has responsibility, but sole accountability lies with the manager. So I'm not saying "oh let's all blame the management of EA", just as you're not saying "don't blame anybody" (at least I don't think you're saying that). First, fix the problem, then go back to those who are accountable and hold them accountable. As much as they're accountable for the end result (winning) they're also responsible for connecting and communicating across all lines within the business.
 
+Steve Kunzer Sounds like we're in violent agreement, with slightly different opinions on where to begin, but precisely the same opinion on where to finish. Excellent! <3

I would return to Dan's point and express that all of this, for me, supports the point that personifying EA and accusing 'them' of not learning, etc, misses the actuality of the situation by a wide margin. Which is not to say that there should not or could not be lessons learned and responsibility distributed, etc, but I still heartily support the original post, in the context of our chat. In particular, I think the point about how the people who learn the lessons from mistakes generally go their separate ways afterwards bears remembering.
 
I was at EA when they put out a super-complicated sports title that completed bombed. Everything that could go wrong on that project went wrong and it showed in the final product. It cost EA millions. I very enthusiastically went to the internal postmortem for that product because I wanted to see if in it they publicly (at least, public within the teams at the studio) could say "Hey, look at these terrible, terrible things we did that snowballed out of control. Let's figure out how to not do them ever again" and then the studio would level up and not have some of those endemic problems.

Unfortunately, what happened was that the various leads of the team went on stage and blamed external licensors, contractors, "unforeseeable contingencies" and basically everything else that wasn't their own practices.

Five years or so later, someone decided that that particular franchise would be a good one to reboot. Many of the internal stakeholders were the same, they had been on different projects in the interim, but the studio was largely filled with the same managers and heads of pods. The peons changed, but the knights were (mostly) the same.

They largely made the same mistakes, surprise, surprise and the game bombed again. What did they blame in their postmortem? The market not being ready, Unforeseen technical challenges" and so forth. Management essentially bought off on it. Since there were so many external scapegoats and it is difficult to crow about your leads on one hand and then punish them when they knowingly do wrong on the other, it's easier to just decide that the external is really the problem and your internal people, processes and technology are peachy keen.

So yes, the pieces do change, so it doesn't make a lot of sense to compare the first team to the team that tackled the reboot. True in many cases. But even if the people stay the same, if there isn't any personal incentive for forcing people to own up to reality, as messy and convoluted as it can be to identify, then there can never be growth on either an individual or collective level.
 
+Zack Hiwiller This is the main thing that actually stops the lessons learned from spreading within a company, so that members can learn from the mistakes that other groups have made - everyone's busy pointing fingers and covering their behind.  Excuse-making is much, much harder to effectively do within small groups.
 
+Ricardo J. Méndez wrote, "Excuse-making is much, much harder to effectively do within small groups."

Not in my experience.  As soon as you get 2 people, politics and finger-pointing can start.  It's just more likely that someone will have the courage to go against groupthink to take a long, hard look at internal reasons in a smaller group. But by no means is this easy to do even in a small group.  Our culture generally doesn't accept failure with grace, and admitting fault is nearly a deadly sin in some cases (especially if the legal department is involved).
 
+Brian Green "As soon as you get 2 people, politics and finger-pointing can start."  

Sure, people can make excuses at any company or team size, but I've never seen it happen as effectively within a small group.

I am not sure if it is for the reason you mentioned, because the groupthink has less reinforcement, or because mistakes have much more immediate repercussions on both the company's and your own immediate future; it has always been apparent to me that the emperor's nakedness gets called out easier the smaller the group is.
 
 Dysfunctional learning and group size
As +Brian Green says, politics and protection of acquired hierarchical power are not limited to large groups. However these issues are massively exacerbated as group size increases.  When there are weak bonds between contributors and slow feedback cycles, both of which are common as group sizes increase, small things like fits of hubris or self delusion can easily take hold. 

Accountability
This is a fascinating word with so much baggage.  For me it brings to mind a certain unquestioned hierarchical moral philosophy that permeates American corporate culture. It has connotations of punishment instead of rehabilitation.  I personally find it mildly repulsive and actively poisonous to the creation of a strong creative, learning culture. 

Once a company becomes large enough, there's an implicit assumption by many that it should organize in a regimented authoritarian militaristic fashion.  We get these odd little fantasy ideas of a good company:
- Strong leader who stops the buck and makes hard decisions.  
- Loyal soldiers that fall on their swords when they fail to protect the queen.    

Instead I like to talk about responsibility, capabilities and learning.
- What can we do? What is a blunt assessment of our capabilities?
- What can we do better than last time?  
- What can we do better than anyone else?
- How do we keep the team agile and able to adapt to new information?
- How do we give individuals the power to make awesome things?
- How do we change the product to become the best thing that this team can build?
- How do we remove middlemen between the makers and the players to achieve deeper understanding and learning? 
- How do we shrink and improve feedback cycles? 

More openess to improvement through experimentation and failure is an essential aspect of how we do our job. It pays more dividends than pushing for 'accountability'.  'Accountability' is what you reach for when everything else has failed. 

Something I find useful when thinking about accountability: 
- Are you in a high power or lower power situation?

If you feel responsible and empowered to do something, there is rarely a need for accountability.  You make the change happen. 

If you feel like you are a cog and you can't improve things unless someone else takes the lead, your desire for accountability is really a sublimated PC call help or the ability to personally effect change.  

To me, 'accountability' becomes one of many corporate code words for a certain poisonous culture of hierarchy and dominance.  Often this isn't actually fixable.  You can tweak the worst aspects of it with various reorgs, training and consultations.  Maybe you remove a particularly cancerous manager. Or sooth the emotions of a shell shocked team.  But the basic flawed assumptions about how humans work are too ingrained to change.  There is an implicit judgement of humans as dumb brutes in need of Old Testament order. 

My solution has always been: Get the hell out.  We live in a golden era of entrepreneurship.  Find a band of like minded folks that isn't polluted and build a team that works.  

(+Zack Hiwiller I'm guessing that even if the 'leaders' in the little army group you talk of had been 'accountable' and were demoted, reeducated or discharged, the group would have gotten another militaristic empire builder and the same scenario would repeat again. )
 
+Ricardo J. Méndez Sadly, I have personally seen it happen in smaller groups just as easily as I've seen it in a larger group.  It's driven the desire to avoid the stench of failure when going to a new project, or people just trying to protect their own fragile egos from self-examination.  It's not healthy and it doesn't solve anything.  In fact, in every case it's lead to the dissolution of the team I've been part of.

In larger groups, I think lack of self-examination is largely caused by groupthink, or perhaps more precisely social cueing.  If you're sitting in a post mortem meeting for an unsuccessful project and the manager is only blaming external forces and nobody calls him or her on that bullshit, most people are going to sit and say nothing.  They don't want to be perceived as a person who "rocks the boat", even if everyone else is thinking the same thing.  If other people understand that the culture shuns the "stench of failure", then accepting that failure as something you should have changed can be a "career limiting move".  Even if identifying the problem, fixing it, and recording that for the future would be the ideal way to handle the issue to make sure it doesn't happen again.

As I said above, larger groups are harder to change due to inertia.  So, yeah, it's easier to see it in action in larger groups, but small groups don't seem to dodge the bullet easier.  Smart individuals can do so if they're willing to do so and the culture isn't punitive to admitting fault.  As +Daniel Cook says, the first step is to empower people rather than having people follow orders and smile and nod when things are going to hell.
 
On a smaller scale, I am currently seeing this happen with the team I used to work with. 
 
Except, of course, that companies can learn if the people at the company adopt a mature process that is both repeatable (and flexible) and ensures a rigorous approach to ensuring the quality of their product. This takes time, and so is antithetical to the current model of "squeeze for all they are worth to hit an arbitrary deadline and fire them when they can't give any more." As a result, and this is true at EA as well as much, much smaller shops (I've seen both from both ends) process is resisted at high levels for taking too long and at low levels because the concept of "everything in its own time" is derided as anti-creative, when the reality is that it fosters more creativity, it just deters changes from happening on a whim (a designer's, a VP-EP, whatever).

Middle folks, like EPs, don't want a rigorous process because they believe it constrains them from making last-minute changes; it doesn't, but again, they don't want to be constrained to only making the changes a review board considers good. So they continue to make changes at the expense of their teammates' lives (sometimes that means a real sacrifice of health) and to their benefit, since EP bonuses are_real_ money in most cases... Especially at EA.

A little discipline would go a long way toward making mistakes like those plaguing SimCity avoidable.
 
My instinctive response is an extreme one: we should not allow such large organizations in our society. The economic advantages that scale (and the attendant power) brings do not seem worth the societal costs that are involved.

However, I am also aware that scale is not everything: small organizations can be dysfunctional too. I am also aware of the sheer impracticality of this proposal.

So, has anybody got any suggestions for systematic mechanisms that we can put in place to reduce the aggregate dysfunctionality of our society?

Perhaps we could lobby for legislation that exerts a gentle downwards pressure on the size of organizations?

Maybe we could implement tools to help large organizations build effective institutional memories?

Maybe we could build tools and services to help small organizations work together to exert more power & influence in the marketplace? (A modern, highly automated version of industry groups)?

Any suggestions?
 
"- Avoid screaming.  If you wouldn't say it to your trusted teammate, don't say it to a developer.  They are human too.  As humans, we tend to tune out weirdly angry people when engaged in polite conversation. "

NOTE: There is a lot of scientific support that states anger causes better decision making. As a developer, considering 'angry' arguments is an important skill. Anger is only bad when unproductive. You can turn anger into positive feedback.

Not a good link but all i could find at hand: http://www.livescience.com/1583-anger-fuels-decisions.html
Add a comment...