Failure Modes sometimes correspond to Game Mechanics

If you want to carry a brim­ming cup of coffee with­out spilling it, you may want to “change” your goal to in­stead pri­mar­ily con­cen­trate on hum­ming. This is an ex­am­ple of a gen­eral pat­tern. It some­times helps to fo­cus on a nearby ar­tifi­cial goal rather than your ac­tual goal. Let me call that strat­egy “gam­ifi­ca­tion”. There is a busi­ness strat­egy, also named “gam­ifi­ca­tion”, of adding game me­chan­ics to a web­site in or­der to achieve var­i­ous busi­ness goals. This is re­lated but differ­ent. Here I’m refer­ring to a strat­egy for prob­lem solvers.

We some­times fail, and some­times one failure is very similar to an­other failure. That is, there are char­ac­ter­is­tic ways that we fail. One of the pri­mary ways that we can im­prove is to learn our failure modes and cre­ate ex­ter­nal struc­tures (pieces of pa­per, soft­ware tools) that check, pro­tect against, or head off those forms of failure.

For ex­am­ple, imag­ine this plan of check­list im­prove­ment:

  1. Change your nor­mal way of work­ing to in­clude an ex­plicit check­list (that starts empty).

  2. When you make a mis­take:

    1. An­a­lyze what went wrong

    2. Try to gen­er­al­ize the par­tic­u­lar in­ci­dent to a category

    3. Add an item to your check­list.

This is very sim­ple and generic, but it is rea­son­able to be­lieve that if you care­fully and dili­gently fol­lowed this plan, your re­li­a­bil­ity would go up (with diminish­ing re­turns be­cause your er­rors are also your op­por­tu­ni­ties for im­prove­ment). I have not read Mayo, but her “er­ror-the­o­retic” philos­o­phy of sci­ence might be ap­pli­ca­ble here.

We can try to build a cor­re­spon­dence be­tween failure modes, and game me­chan­ics that at­tempt to cope for that failure mode.

Failure modes and their cor­re­spond­ing game mechanics

  • Over­whelmed­ness and Bore­dom/​Apathy

In Mihaly Csik­szent­miha­lyi’s flow di­a­gram with axes of challenge ver­sus mas­tery, flow ap­pears when both the task challenge (or difficulty) and the in­di­vi­d­ual’s mas­tery of (or skill in) the task is high. If challenge is too high, you may feel over­whelmed and fail to make progress. An ar­tifi­cial, very easy goal may help. If the challenge is too low, you may feel bored or ap­a­thetic, and fail to make progress. An ar­tifi­cial, more difficult goal may help. Quests (see Wor­ldOfWar­craft or al­most any fan­tasy RPG) and Achieve­ments are game me­chan­ics that fit this gen­eral pat­tern of GoalSet­ting; easy achieve­ments (see Achieve­men­tUn­locked) cor­re­spond to the over­whelmed­ness failure mode, and difficult achieve­ments (see Nethack con­ducts) cor­re­spond to the bore­dom/​ap­a­thy failure mode.

  • Failures of bal­ance.

Some­times it can be hard to main­tain a good bal­ance among mul­ti­ple ac­tivi­ties. For ex­am­ple, it is im­por­tant to no­tice new good ideas. How­ever, I tend to spend too much time pur­su­ing nov­elty, and not enough time work­ing on the best idea that I’ve found so far. There is a tra­di­tion of browser games (see King­domOfLoathing) that en­force a kind of bal­ance us­ing a vir­tual cur­rency of ‘turns’. You ac­cu­mu­late turns slowly in real time, and es­sen­tially ev­ery ac­tion within the game uses up turns. This en­forces not spend­ing too much time play­ing the game (and in­creases the per­cieved value of the game via forced ar­tifi­cial scarcity, of course). If I gave my­self ‘ex­plore dol­lars’ for do­ing non-ex­plo­ra­tion (so-called ex­ploit) tasks, and charged my­self for do­ing ex­plo­ra­tion tasks (like read­ing arXiv or wikipe­dia), I could en­force a bal­ance. If I were also prone to the op­po­site prob­lem (“A few months in the lab can of­ten save whole hours in the library.”), then I might use two cur­ren­cies; ex­plor­ing costs ex­plore points but re­wards with ex­ploit points, and ex­ploit­ing costs ex­ploit points but re­wards with ex­plore ponts. (Vir­tual cur­ren­cies are ubiquitous in games, and they can be used for many pur­poses; I ex­pect to find them able to be placed ac­cross from many differ­ent failure modes.)

  • Laps­ing from a habit

I might fail by laps­ing from a good habit, of ex­er­cise, prac­tice or mea­sure­ment. A chain­ing ap­point­ment me­chanic might help pre­vent that. For ex­am­ple, Far­mville has a me­chanic where play­ers plant crops, but can­not har­vest them un­til a longish real-time in­ter­val later (e.g. 24 hours), and ripe crops rot if they are not har­vested. Don­tBreakTheChain is (see http://​​dont­break­​​) might be an even bet­ter cor­re­spon­dence. Some­times peo­ple re­fine the chain me­chanic with a Mario-style “two hit point” me­chanic. (Mario is ei­ther big or small. If he is dam­aged while big, he be­comes small. If he is dam­aged while small, he loses.) That is, a one-day break in the chain is re­cov­er­able, so long as it is only one day, and fol­lowed by “suffi­ciently many” con­sec­u­tive suc­cess­ful days.

  • Dis­tract­edByCon­tex­tSwitch and YakShaving

There are var­i­ous failure modes re­gard­ing the sub­goal stack. On the one hand, there is a dis­trac­tion failure mode when I switch con­texts to achieve some sub­goal, but be­come dis­tracted, and for­get, at least for a while, to pop the stack and re­turn to my origi­nal goal. On the other hand, YakShav­ing is fol­low­ing your goal/​sub­goal stack too rigidly and too deeply. In ei­ther case, it may be valuable to keep your en­tire cur­rent goal/​sub­goal stack in a nearby ex­ter­nal mem­ory de­vice (for ex­am­ple in­dex cards or a todolist tool). Many games have a UI com­po­nent that dis­plays the next step of the the cur­rent quest. I don’t know of a game with a UI com­po­nent in­tended to pre­vent YakShav­ing, but it is com­pre­hen­si­ble—sim­ply dis­play the top of the stack as well as the bot­tom.

  • Dead­line-sen­si­tive work style

There are failure modes as­so­ci­ated with dead­line-sen­si­tive rates of work­ing. Peo­ple tend to work differ­ently when they per­cieve time pres­sure than when they do not per­cieve time pres­sure. You might call this failure mode Park­in­son’s Law when you’re think­ing of the pres­sured work as more effi­cient, HalfAssedCom­ple­tion when you’re think­ing of the pres­sured work as low­ered qual­ity, or Dead­lineBrinks­man­ship when you’re think­ing of the ten­dency for all tasks, no mat­ter how much slack they have, to have the same risk of cross­ing the dead­line.

This can make schedul­ing and plan­ning (which are already difficult) even more difficult. For ex­am­ple, a plan­ner may pre­fer an­nounc­ing an op­ti­mistic sched­ule, be­cause they think that even with in­evitable slip­page, the op­ti­mistic sched­ule will be com­pleted faster than a sched­ule con­ser­va­tive enough to avoid slip­page.

Of course many games have timed goals, and the po­modoro tech­nique and other forms of time­box­ing are stan­dard pro­duc­tivity con­cepts. The only re­fine­ment that this “game me­chanic” sug­gests is mak­ing the count­down more visi­ble and salient.


  • ac­cu­mu­lat­ing an ex­plicit repos­i­tory of failure modes can be valuable

  • game me­chan­ics of­ten cor­re­spond to failure modes

  • ex­plic­itly struc­tur­ing your work habits with game me­chan­ics can be fun, and per­haps pro­duc­tive.

Fu­ture work

There is an in­de­ci­sive­ness failure mode that I think of as “Buri­dan­sAss”. The parable is that the ass is as hun­gry is it is thirsty and is ex­actly po­si­tioned be­tween wa­ter and food, and so dies with­out choos­ing. In my case, it is more cycli­cal—when I am try­ing to work on one thing, work­ing on some other thing seems more valuable or more en­joy­able. I don’t know of a game me­chanic cor­re­spond­ing to this failure mode.

Another failure mode I’ve no­ticed is “Blub” (see Paul Gra­ham’s es­say). This is when I see some­thing I don’t un­der­stand at all, and in­stead of pay­ing it closer at­ten­tion and/​or stor­ing it away care­fully, I ig­nore it.