Rationalistic Losing

Play­ing to learn

I like los­ing. I don’t even think that los­ing is nec­es­sar­ily evil. Per­son­ally, I be­lieve this has less to do with a de­sire to lose and more to do with cu­ri­os­ity about the game-space.

Tech­ni­cally, my goals are prob­a­bly shifted into some form of meta-win­ning — I like to un­der­stand win­ning or non-win­ning moves, strate­gies, and tac­tics. Ac­tu­ally win­ning is ic­ing on the cake. The cake is learn­ing as much as I can about what­ever sub­ject in which I am com­pet­ing. I can do that if I win; I can do that if I lose.

I still pre­fer win­ning and I want to win and I play to win, but I also like los­ing. When I dive into a com­pe­ti­tion I will like the out­come. No mat­ter what hap­pens I will be happy be­cause I will ei­ther (a) win or (b) lose and sa­ti­ate my cu­ri­os­ity. Of course, learn­ing is also pos­si­ble while watch­ing some­one else lose and this gen­er­ally makes win­ning more valuable than los­ing (I can watch them lose). It also pro­vides a solid rea­son to watch and study other peo­ple play (or play my­self and watch me “lose”).

The catch is that the valuable knowl­edge con­tained within win­ning has diminish­ing re­turns. When I fight I ei­ther (a) win or (b) lose and, as a com­pletely sep­a­rate event, (c) may have an in­ter­est­ing match to study. Ideally I get (a) and (c) but the odds of (c) get lower the more I dom­i­nate be­cause my op­po­nents could lose in a known fash­ion (by me win­ning in an “old” method). (c) should always be found next to (b). If there is a rea­son I lost I should learn the rea­son. If I knew the rea­son I should not have lost. Be­cause of this, (c) offsets the nega­tive of (b) and los­ing is valuable. This makes win­ning and los­ing worth the effort. When I lose, I win.

Per­son­ally, I find (c) so valuable that I start get­ting bored when I no longer see any­thing to learn. If I keep win­ning over and over and never learn any­thing from the con­test I have to find some­one stronger to play or start los­ing cre­atively so that I can start learn­ing again. Both of these solu­tions set up sce­nar­ios where I am in­creas­ing my chances to lose. Math­e­mat­i­cally, this starts to make sense if the value of knowl­edge gained and the penalty of los­ing com­bine into some­thing greater than win­ning with­out learn­ing any­thing. (c—b > a) My hunches tell me that I value win­ning too lit­tle and cu­ri­os­ity is start­ing to curb my de­sire to win. I am not play­ing to win; I am play­ing to learn.

skip to sum­mary

Los­ing is good

To be fair, I am speci­fi­cally talk­ing about win­ning within or­ga­nized sys­tems of com­pe­ti­tion. This gen­er­ally means some­thing like Magic: The Gather­ing, Go, or Mafia. Trans­lat­ing this onto Life is harder be­cause I have more emo­tional in­vest­ment in the out­come. The penalty of los­ing is stronger. If I lose a Game I know that the next round is en­tirely in­de­pen­dent of this loss and there are min­i­mal long-term effects to worry about. The con­se­quences of los­ing will be much more se­vere if I screw up an in­vest­ment port­fo­lio or fail at at­tempt­ing the perfect mur­der. To draw a finer line be­tween life and gam­ing: If the win or loss means plac­ing in a tour­na­ment with cash prizes the in­cen­tive for win­ning jumps well be­yond the in­cen­tive to learn some­thing new and I start play­ing to win. But is the pain of a loss in­versely pro­por­tional to the value of a win? No, not nec­es­sar­ily.

To map a tour­na­ment into pay­outs, say it costs $10 for en­ter­ing and the prize for win­ning is $20. Los­ing at any point in the tour­na­ment has no mon­e­tary costs as­signed to it. The $10 is a sunk cost and the $20 is only el­i­gible to win­ners. Los­ing has some in­trin­sic emo­tional penalty but, other than feel­ing icky, the loss sim­ply gives an­other op­por­tu­nity to learn. The value of win­ning is greater than that of los­ing, but los­ing still has value. Be­cause it has value, los­ing is good.

Life works the same way. If I am ap­ply­ing for a com­pet­i­tive job I should be play­ing to win be­cause the pay­outs are enor­mous. Los­ing is a bum­mer, but the value in the loss is learn­ing new in­for­ma­tion within the game-space of ap­ply­ing for jobs. Namely, this could mean prop­erly build­ing past ex­pe­rience or learn­ing to sell your­self well. Be­cause you learned some­thing, you are bet­ter off than when you started even though you lost. There­fore, los­ing is good. Win­ning is bet­ter, but los­ing is still valuable. The cost of los­ing is not the same as flip­ping the value of win­ning nega­tive. You did not “lose” the job be­cause you never had it. You failed sim­ply failed to win.

Ir­ra­tional winning

Win­ning is great, but if there is no value in win­ning other than sim­ply be­ing the win­ner, los­ing may be worth more. Win­ning for the sake of win­ning is no­ble but use­less. In such a sce­nario, play­ing to win may not be the most benefi­cial course. Play­ing to learn can re­sult in a gain even if it means you “lose” the game. Look­ing at it ra­tio­nal­is­ti­cally, “los­ing” is win­ning.

If I play Car­cas­sonne against my op­po­nents and whomp them thirty times in a row by re­peat­ing my best known strate­gies I will have gained noth­ing. If I de­cide to use the game as a learn­ing ex­pe­rience to test new strate­gies I can cre­ate an op­por­tu­nity to learn, but am no longer play­ing to win. If I play just strong enough to win I can learn and win but this is less valuable than sim­ply learn­ing as much as you can be­cause the win still means noth­ing. And if it did than you should have played to win.

A bet­ter ex­am­ple would be a movie ticket that you pur­chased for $10. The game-space re­volves around whether or not you get more value out of watch­ing a movie than what you spent on the ticket. If you pur­chase a non-re­fund­able ticket in ad­vance but, on the day of the movie, you do not feel like go­ing to the movie the “win” would be stay­ing home which is ac­tu­ally a “loss” in the origi­nal game. The los­ing sce­nario has changed be­cause “los­ing” now has more value.

Note: This ex­am­ple is di­rectly bor­rowed from Z_M_Davis’s Sunk Cost Fal­lacy ar­ti­cle.

Minor point: In this ex­am­ple, the value of “los­ing” could be learn­ing to not pre­pay for tick­ets or check­ing the weather first or learn­ing from what­ever caused you to mis­pre­dict your mood.

Ra­tion­al­is­tic losing

Ra­tion­al­is­tic los­ing is es­sen­tially ac­knowl­edg­ing and play­ing a su­per-game so that no mat­ter what hap­pens, you win. In this su­per-game, play­ing to win does not mean win­ning the con­test. It means get­ting some­thing valuable from the con­test. In the above ex­am­ples, even though the con­tests were lost, the ra­tio­nal­ist should still win by learn­ing the in­for­ma­tion available. Los­ing should be good. If it wasn’t, some­thing went wrong be­fore you got to this point. Do not rob your­self of the value of los­ing by fo­cus­ing on the lost win.

This prin­ci­ple is hard to see when it ap­plies to some­thing you re­ally, re­ally wanted to win. If you re­ally, re­ally wanted that job than los­ing feels like los­ing the job. If you skip the movie it feels like los­ing $10. But you never had the job and you already spent the $10. The only thing left to lose is in­for­ma­tion. You can still win the su­per-game if you are able to gather this in­for­ma­tion. This is ra­tio­nal­is­tic los­ing.

There are many ex­am­ples of, and ex­cep­tions to, this prin­ci­ple but the whole point can be boiled into this:

  1. As­sume a con­test where you can win or lose.

  2. If there is value in win­ning the con­test, play to win, oth­er­wise play to learn.

  3. If you play to win and lose, learn why you lost.

  4. If you already knew why you lost, you were not play­ing to win.

  5. Learn­ing why you lost is valuable.

  6. Learn­ing af­ter a loss means the loss was valuable.

  7. If the loss was valuable, the loss was “good”.