Ask HN - Review My Idea: Make Work More Like Games | Hacker News Reader

Ask HN - Review My Idea: Make Work More Like Games

thezukunft.com

45 pointsjtemplestein53 comments

If only there were some sort of tokens employers could give employees as an incentive. The tokens could then be exchanged for goods and services. Perhaps the production and circulation of tokens could be managed by the government to prevent abuse..

There are few times I care about voting, but for this I'd vote 5 times ;)

Except those tokens are currently handed out by a person instead of a unbiased machine who could dispense them based on measured progress for some task. Still made me laugh, you get my vote :)

The problem with that idea is that eventually people feel entitled to win tokens whether or not they made any achievements. They get so angry at not automatically being given tokens that the government mandates a minimum amount of tokens every employee must win per hour.

However, empirically we've seen that touching token allocations causes huge amount of resentment and is very, very difficult to do in an elastic manner. Demands of the business can arise in hours or days, token allocation tweaks take years to take effect. We've seen that people will react -- within seconds -- to non-token incentives to do things that certainly resemble work. Blizzard can make a database entry with a few purple pixels attached and wham several million people start doing very repetitive tasks to achieve it. Think of the business value that could be harnessed if a little alert on my dashboard said "Hey Patrick, that funky Mersenne Twister re-implementation that has been in purgatory for the last 12 weeks and is now blocking delivery now is offering purple pixels for unit testing!"

I know you can game any system for rewards -- it has been a near obsession of mine since a kid, when I found that 89.5% is still an A- and that meant I could still miss 3 homework assignments and not suffer any consequences as long as I did really well on the tests. Like was mentioned in the video posted yesterday, everyone knows that grades are a game... its just an unsophisticated game that sucks, but nonetheless seems to converge on education for many people. Grades as a game have always been Pong in a world where Pacman exists... now they're Pong in a world where WoW exists. (Or even where Farmville exists.) Holy cow, we can abuse that better. ("Little Timmy learned 7 * 8 = 56. Ding, Mathematician Level 14! Gratz Timmy!")

token allocation tweaks take years to take effect

This is mostly a function of a particular method of token allocation, rather than a property of the tokens. There are entire lines of work where the allocation is quite elastic - people who work on commission, people who make a significant part of their income based on performance bonuses, etc. Even at start-ups, in my experience, bonuses for important milestones are not all that uncommon, even if they often are, well, token.

There is a distinction between monetary rewards and the kind of peer-recognition the author is discussing here.

"A raise is only a raise for three months, then it's just your salary."

The argument here is that there are many, many other game mechanics besides simply giving out "tokens" that have been demonstrated to be compelling motivators.

Also, obligatory-TED-video-exposing-counterintuitive-point: http://www.ted.com/talks/dan_pink_on_motivation.html

Providing higher rewards decreases performance.

Actually providing rewards decreases performance in creative problem solving. He actually said that it worked really well ("kicked butt") in getting people to perform simple mechanical skills based tasks.

Besides, the point isn't about increasing performance in situations where people are highly engaged regardless, it's about increasing engagement where motivation can wane, especially in more menial type of tasks/jobs.

There is probably a pretty large body of work on employee motivation and reward systems in the academic literature. I'd suggest familiarizing yourself pretty closely with all of the research before going anywhere with this idea.

There is a large body of work on this. The short, short version: this is a Very Bad Idea.

For an intro into the subject, check out "Punished By Rewards" by Alfie Kohn.

There was a time when I played WoW religiously, and actually had my own guild. This was back in the day, when guilds spent 30+ hours a week clearing dungeons that required 40 people to memorize and execute intricate strategies. Even though my guild was relatively successful, motivating people to show up to these raids was always a chore. Many players consistently left dungeons early, or avoided them altogether. Interestingly enough, these same players didn't mind spending hours grinding or fishing or hanging out in the city, doing nothing in particular. They did this despite the fact that raiding dungeons was more productive, stimulating, exciting, challenging, and potentially rewarding.

I suspect one of the appeals of doing repetitive, mindless things like fishing or grinding in games is because... they're repetitive and mindless. There are real-life equivalents: clicking around on Hacker News, slumping on a couch for hours watching television, etc. These activities are much less mentally exhausting, much less stressful than doing actual work.

Anyway, I think your idea is a great one, but there are some hurdles. If you reward people for menial activities, then they'll waste tons of time on it because it's mindless and easy. Chances are this type of work is not what any company would want its employees doing. If you only reward people for actual demanding work, then chances are you won't really be adding anything.

Unfortunately, there are a lot of jobs that are pretty mindless. Let's help these people out :)

You have to be very careful that people don't 'game' the system. Metrics like these are rife for exploitation - especially likely if the metrics and rewards are designed by people who don't have a deep understanding of rule systems.

"Ding! 10,000 lines of code reached!" -- A recipe for a very bloated product

Yes, that's one thing that has to be carefully watched. When you incent someone to do something, they tend to do it. It's a bit of the "be careful what you wish for, you just might get it". Along with that is the disincentive which he did mention. He mentioned it more from a "don't embarrass someone" point of view but again - some people see it as an incentive. Some people are completionists and feel compelled to "get them all". You have to be careful not to encourage bad behavior.

Well, that's not that difficult. Have negative tokens for bugs filed, similar to loss of experience for dying in an RPG.

It seems someone needs to read Joel's article on metrics: http://www.joelonsoftware.com/news/20020715.html . Relevant quote:

Software organizations tend to reward programmers who (a) write lots of code and (b) fix lots of bugs. The best way to get ahead in an organization like this is to check in lots of buggy code and fix it all, rather than taking the extra time to get it right in the first place. When you try to fix this problem by penalizing programmers for creating bugs, you create a perverse incentive for them to hide their bugs or not tell the testers about new code they wrote in hopes that fewer bugs will be found. You can't win.

While I find Joel's articles interesting and often enlightening, I don't think "someone" needs to treat them as the Word From On High of software development or program management.

Anyway, that's still not impossible to solve. After the design phase, reward for every module shipped, then deduct points for every bug reported by users. If the programmer is able to hide bugs such that they never manifest in production, I say all power to him.

Now the programmer has an incentive to make their subsystem hard to use, so users don't go there and report bugs. C'mon, add some more epicycles.

Why would the bug reporting mechanism be in a subsystem? Unless every subsystem had its own bug reporting mechanism, which makes zero sense.

[Edit]

Oh ok, I think I see what you mean. Well, firstly, if a subsystem weren't integral to the app, why develop it in the first place? Secondly, you could always track the usage of the subsystem and provide additional incentives for its success. Lastly, who says subsystems have to be directly accessible to the user? If tasked to write a module that runs in the background, one can hardly make it deliberately obtuse.

That would lead to a black market for defects. I've heard the stories.

That's why you need the game master :)

it's too difficult to build a "one size fits all" application. as an example, target has one goal with their cashiers: a fast checkout process. so they start a timer when the first item is scanned, and that timer only stops once the receipt has been printed. i was just at a target last week and the cashier told me she had to pause the transaction while someone found something i had asked to be put on hold; she was actually worried about receiving a low score. the score looks like this: 92% A, or 85% B, etc.

so target's "reward" mechanism is partially built into their POS machines, while the other part is driven by their management. if you sell more than 3 target cards in a day, you can pick out anything in the store <$20 for free. that could be built, but it'd have to be tied into target's hardware to work. and how are you going to give those badges away in that case?

i guess i'm not sold on the idea because i don't see any specific examples of where it would be compelling. maybe web-based companies, at that point your main goals are things like code written or commits. so you could monitor all employee's github accounts (as an example) and send out a company e-mail with badges based on that. just an example.

Presumably Target would know to throw out the outlying data points, but that doesn't mean the cashiers know they're allowed to take a few minutes to help the customers.

I think the most appealing market are government agencies and large corporations with lots of paperwork, such as insurances.

The process of managing a customer request in an insurance company goes like this: file the request, create a template mail, press "send", have it printed. Why not stir that up a little?

This is a great idea and has crossed my mind many times. In fact, it was one of the motivations behind our current startup. (http://blog.iactionable.com/ & http://news.ycombinator.com/item?id=1126780)

I think (and hope) the achievement/badge trend will continue to grow and show up in all sorts of new places. The biggest hurdle we are facing is finding ways to 'hook' into the events which get tracked and converted into badges. Its pretty simple for websites, but for a company to setup or modify existing systems to automatically ping events is no small undertaking for that company. Maybe there's an opportunity for consulting here too?

Then there's trying to decide what things are valuable for company X to track and award badges for. All this takes time and money. With a little backing info from research studies I could see companies giving this a shot. There are existing services which allow employees to give 'kudos' points to each other for things they do. This would be taking that to the next level and fully automating it.

Have you looked into workflow engines at all? If a company had any of those deployed, tracking the metrics would be easy.

I don't think the problem lies in tracking the metrics. It lies in managing and organizing the "game".

I have tried doing this repeatedly for me individually and every time it has failed. The reasons:

1. Hard to define measurable goals: How many points for a bug fix. What if a bug fix requires 5 mins other requires 4hrs. Its hard to assign points to each individual work item and track 'em.

2. How to set good milestones.

3. Avoiding Karma whoring: How to stop from racking companionship points by having extended lunches

In a company scenario it would be hard to sell this idea to all employees and might deter some prospective employee to join as it might set the corporation's image as childish in the outside world.

If your job is like fishing in WoW, it should be automated so that you can actually do something of value with your time.

A better approach: http://www.paulgraham.com/hiring.html

My favorite talk from TED2010 was by Jane McGonical and her theory was that kids around the world spend enormous amounts of time and effort playing games. They should not be considered losers for gaming so much, everyone else is the loser for not knowing how to tap their energy.

Her idea is not so much applying game dynamics to real life, but changing existing games slightly so that people solve more important problems (think ending world hunger by leveling up).

You can find several games she created on http://www.avantgame.com/

The talk is not yet online, but I"ll post it on HN when it comes out.

I used to work in a tech support call center. They tracked every metric you can imagine. A large percentage of the tech support staff were WoW addicts, so I could definitely see this working.

Ha! I seriously laughed when I read your comment. Only because I actually work for a company that sells SaaS Call Center software and this has DEFINITELY been on my mind. ;)

A company spanning system is a pretty lofty goal. Perhaps you should look into something smaller, say something at the application level. A good example would be Ribbon Hero for MS Office.

I like it. I've worked at a few big corporations where you'd get special plaques or pins for years of service, good ideas, attending certain training courses, and so on, and although I personally didn't understand the appeal -- I'd rather have challenging and meaningful work -- the people who earned those awards seemed to display them proudly. I always wondered why that kind of "rep grinding" wasn't more prevalent than the usual winner-take-all "employee of the month" stuff.

Can't believe this wasn't posted, I believe the following video was a heavy source of inspiration for that blog post... http://g4tv.com/videos/44277/DICE-2010-Design-Outside-the-Bo... Brilliant 30 minute talk by Jesse Schell (funny too), I was looking for it before everyone started posting billions of comments but Chrome's history acted up in a funky way. Oops :)

If you want to find out more about this guy, I'd recommend checking out his website: http://artofgamedesign.com/

His book is also excellent: http://www.amazon.co.uk/gp/product/0123694965?ie=UTF8&ta... - I checked it out of the library recently and it really is a bible for game design. It literally covers everything you'd want to know about making high-quality games.

Your mind always knows if it's working or having fun. Every way to trick the mind in confusing the two fools it at most for a short time, then it catches up. I guess that's because your mind is about as clever as you are yourself (and it knows what you're trying to do!).

Also there are already reward systems which are real. I like to be proud of what I do. I like to be rewarded with fair payment. I think stuff like that works better for me than an artificially added point-system.

There was a recent VentureBeat article by the authors of Total Engagement: http://games.venturebeat.com/2010/01/15/the-lesson-of-gaming...

We're working with several companies to implement game mechanics for both training and job performance: http://www.bunchball.com

I don't think this will work in the tech world. Don't you think developers will try to game the system just to make more points?

Joel already discussed this kind of renumeration, he gave an example of bugs. If you pay your developers more money if they produce bug-free code, they won't report bugs and it'll result in a worse result.

Point taken. But the market of lowly routine workers is much bigger anyway, right? :)

There is a HBS Press book that goes into this in detail. http://www.amazon.com/gp/product/142214657X?ie=UTF8&tag=...

I think a better solution would be to find an innovative way to add interaction like "The Fun Theory" people:

http://experimentgarden.blogspot.com/2009/11/fun-theory-how-...

This idea is similar to:

http://chorewars.com

Tangent: This reminds me of a free book I downloaded on the kindle: Metagame http://www.amazon.com/MetaGame-ebook/dp/B002AJ88LC

I think a far better idea for the hypothetical Bor Ingsurance is to hire only WoW addicts, skip the ball-point-pens\star-stickers, and reward them with in-game items and gold.

Too Dilbertesque for my tastes, sorry.

Just what I was thinking: http://www.youtube.com/watch?v=eD4-EON1RhI

you forgot one thing: not everyone likes to play games.

I believe that this would only be really beneficial for a McJob-like work environment where the employees believe their work has very little importance or impact.

I would imagine that working at a start-up or at Google or at some other fast-paced innovative company would negate the need such a system and would possibly devalue (within the employee's mind) the importance of their work.

Probably true, but that doesn't mean there isn't a market for such a idea. What do you think the ratio of McJobs to Startups/Google jobs is?

Exactly. The overwhelming majority of all jobs in any sector probably suck.

I thought that Google had a system spiritually very close to this where incentives align in a semi-automated manner to encourage people to do what the business needs doing? (For example, they're apparently able to devote huge portions of their development resources to AdWords -- Google's only product that makes an appreciable amount of money -- despite e.g. report generation for AdWords being much less intrinsically sexy than working on their latest, greatest delivery vehicle for the advertising warhead.)

Edit: Yep, found the article I was thinking of: http://steve-yegge.blogspot.com/2006/09/good-agile-bad-agile...

Search for the word "incentives".

Reply on news.ycombinator.com