The Art of the Overbet

Pre­vi­ously: The Kelly Criterion

Last time I said, never go full Kelly.

In prac­tice, I strongly agree with this. Either one should go far over full Kelly be­cause the core Kelly as­sump­tions have bro­ken down and you want to throw the rules out the win­dow, or you are try­ing to re­spon­si­bly grow a bankroll and full Kelly bet­ting on what you be­lieve your edge to be would be mas­sively overly ag­gres­sive.

This time we go over the sec­ond cat­e­gory with prac­ti­cal ex­am­ples. What are situ­a­tions in which one should en­gage in what ap­pears to be mas­sive over bet­ting?

Four main sce­nar­ios: Not win­ning is catas­trophic, los­ing is ac­cept­able, and los­ing is im­pos­si­ble, los­ing is in­evitable.

You Win or You Die

When you play the game of thrones, you win or you die. If play­ing it ‘safe’ with your re­sources means you don’t win, then it means that you die. Either don’t play, or don’t hold back.

In Rounders (spoiler alert, movie is recom­mended), the star finds him­self ow­ing $15,000 to the Rus­sian mob, and his best efforts could only get him $10,000. Without bet­ter op­tions and un­will­ing to run for his life, he walks into the poker club of the man he owes the money to, and says “I owe you that money to­mor­row, right? I’ve got $10,000. I’m look­ing for a game.”

Fa­mously, early in the com­pany’s his­tory, the founder of UPS once found him­self with­out the funds to make pay­roll. He knew that if he missed pay­roll, that would be the end of UPS. So he flew to Ve­gas with what funds he had, bet them all on black, won, made pay­roll, and now we have a UPS.

Magic play­ers of­ten se­lect ‘safe’ decks in­stead of ‘risky’ decks. This is ex­actly back­wards. If you flame out of a tour­na­ment and lose all rounds, you get noth­ing. If you win half of them, you still get noth­ing. If you win three quar­ters, you usu­ally still get al­most noth­ing rel­a­tive to win­ning. Ex­tra var­i­ance is great.

If you be­lieve that the fate of ev­ery­one de­pends on cross­ing a thresh­old or hit­ting an im­pos­si­bly pre­cise tar­get, no mat­ter how bad the odds, you de­ploy ev­ery­thing you have and take your best shot.

There’s a dead­line. We don’t have time for Kelly.

Win or Go Home

Los­ing not be­ing so bad, or con­tin­u­ing to play not be­ing so good, is effec­tively very similar to los­ing be­ing catas­trophic but not safely avoid­able. In both cases, sur­viv­ing is not a worth­while goal.

A clas­sic mis­take is the gam­bler who for­gets that they are us­ing their bankroll to pay the rent. On fo­rums I would of­ten read sto­ries of hard work­ing sports gam­blers with mid-five figure bankrolls. They make sure to make only ‘re­spon­si­ble’ wa­gers on sport­ing events, risk­ing only 1-3% of their funds each time. When they have a good month, they could eat and make rent, but that ate up most of the prof­its.

What they con­tin­u­ously failed to re­al­ize was that this was not a sus­tain­able situ­a­tion. Be­ing ‘re­spon­si­ble’ only en­sured that, even if they were good enough at pick­ing win­ners, they would never suc­ceed due to their fixed costs. They were be­ing ‘re­spon­si­ble’ with their siz­ing rel­a­tive to their bankroll, but com­pletely ir­re­spon­si­ble when siz­ing rel­a­tive to fixed costs of their time.

When I first started gam­bling on sports, I put aside a fixed bankroll I com­mit­ted not to re­plen­ish­ing, but then acted what any Kelly-style for­mula would call com­pletely ir­re­spon­si­ble with that bankroll un­til af­ter my first few dou­ble ups. As the op­er­a­tion be­came clearly worth my time, I scaled back and did more ‘re­spon­si­ble’ siz­ing, know­ing that I wouldn’t be eaten al­ive by fixed costs.

Later, when ex­plor­ing if it was worth­while to re­turn to wa­ger­ing, I did a similar thing, strug­gled, and even­tu­ally lost ev­ery­thing I’d set aside. This was very good. It let me move on. To this day, we can never know for sure whether I had an edge dur­ing that sec­ond at­tempt – I sus­pect my luck was quite per­verse – but I do know it wasn’t the best use of my time to find out.

Fail fast is a well-known prin­ci­ple for start-ups. I broke this rule once, with Me­taMed, and it was an ex­pen­sive mis­take. Even if you have seg­re­gated your po­ten­tial losses in dol­lar space, you need to con­tain them in time space, and emo­tional space, and so­cial space.

There’s no dead­line. But we don’t have time for Kelly.

Bet You Can’t Lose

You can’t lose ev­ery­thing if you can’t risk ev­ery­thing.

Most peo­ple’s re­sources are mostly not money, or even things at all, most of the time. When some­one ‘loses ev­ery­thing’ we talk of them los­ing their friends, their fam­ily, their rep­u­ta­tion, their health, their abil­ity to work. And for good rea­son.

There is a scene in Defend­ing Your Life where we flash back to the main char­ac­ter pay­ing over $2,000, a third of all the money he has, to avoid a mid­dle seat on an in­ter­na­tional flight. In con­text, this is to his credit, be­cause it was ‘brave’ to risk spend­ing so much. In a sense it was brave, in a more im­por­tant gen­eral sense it was stupid, but in the most im­por­tant sense he was spend­ing a very small frac­tion of his re­sources, so the rele­vant ques­tions were: Was this trans­ac­tion worth it? No. Did this put the char­ac­ter at risk of hav­ing a liquidity crisis where not hav­ing any cash could be ex­pen­sive? A lit­tle.

The best rea­son to do ‘safe’ things with money, and to main­tain liquid sav­ings, is to avoid pay­ing the cost of need­ing liquidity and not hav­ing it, or pay­ing the costs of avoid­ing sce­nar­ios where that liquidity might be­come nec­es­sary. This in­cludes the abil­ity to take ad­van­tage of op­por­tu­nity. Avoid­ing ex­tra costs of hav­ing to bor­row money, which is ex­pen­sive in time and in emo­tional well-be­ing and so­cial cap­i­tal, not only in money, is by most peo­ple largely un­der­es­ti­mated. Slack is vi­tal.

There is also a benefit to hav­ing no cash. For some peo­ple, and in some fam­i­lies and cul­tures, one who has cash is ex­pected to spend it, or in­evitably will quickly spend it. Some­times this is on one’s self, some­times on oth­ers, but the idea that one can have money and con­serve it by spend­ing only re­spon­si­bly isn’t a thing. The only so­cially ac­cept­able way to not spend money is to not have money. Thus, there is a high effec­tive ‘tax’ on sav­ings.

In such situ­a­tions, not only is risk not so bad, it can be ac­tively great. If your risky bet pays off, you can have enough money for big pur­chases or even to es­cape your poverty trap. If it fails, you would have wasted the money any­way, and now you can con­serve slash mooch off oth­ers. Thus, your goal is to find a way to trans­late cash, which one will in­evitably lose, into in­alien­able prop­erty that can be pro­tected, or skills and con­nec­tions and re­la­tion­ships, or at least ex­pe­riences one can re­mem­ber.

This isn’t just for poor peo­ple. Start-up cul­ture works this way, too, and the only way to get things at rea­son­able prices, in­clud­ing la­bor, or to be able to raise money, is to clearly have spent ev­ery­thing you have and have no spare re­sources. This is a large por­tion of why start-ups are so stress­ful – you are not al­lowed to have any slack of any kind. I’m likely about to start a new com­pany, and this is the thing that I dread most about the idea.

You Bet Your Life

This is all in con­trast to the grand pro­ject we have been as­signed, of ‘sav­ing for re­tire­ment.’

Older peo­ple sav­ing for re­tire­ment in mod­ern de­vel­oped coun­tries, or at least some of those sav­ing for re­tire­ment, face a spe­cial situ­a­tion. Un­able to earn ad­di­tional funds, and with­out fam­i­lies or friends they can count on, their sur­vival or at least qual­ity of life de­pends en­tirely upon their abil­ity to save up money ear­lier in life to spend later.

If they run out of money, they’ll still get some amount of gov­ern­ment sup­port, and per­haps some amount of fa­mil­ial sup­port, but any re­main­ing years are go­ing to suck. If they die with money in the bank, that money can be passed on to oth­ers, but this is mostly a small benefit rel­a­tive to not run­ning out of cash to spend.

Com­pound this with large var­i­ance in re­main­ing life span, and highly var­i­ant needs for health care and as­sis­tance dur­ing those re­main­ing years, and you have an en­tire pop­u­la­tion pres­sured to fu­ri­ously save ev­ery dol­lar they can. Any other use of one’s re­sources is looked at as ir­re­spon­si­ble.

This is profoundly weird and profoundly messed up.

It is han­dled, by most peo­ple and by those we trust to ad­vise us, mind­bog­glingly badly.

That’s true even in the cases where the prob­lem defi­ni­tion mostly ap­plies to one’s situ­a­tion, if one is will­ing to as­sume the world will con­tinue mostly as it is, and one is un­able to in­vest in other types of re­sources and ex­pects to have lit­tle as­sis­tance from them.

It also leads us, as a so­ciety, to treat our sav­ings as the sav­ings, the re­tire­ment sav­ings, and to ap­ply the prin­ci­ples of that prob­lem to all sav­ing prob­lems, and all risk man­age­ment prob­lems. Young peo­ple take ‘risks’ and buy stocks, older peo­ple play it ‘safe’ and buy bonds.

If the world were a much more cer­tain place, where we knew how long we would live, in what state of health, and how the econ­omy and world would do, and ev­ery­thing else we might need money for, and how much money we needed to en­gage in var­i­ous ac­tivi­ties and con­sume var­i­ous goods, we could kind of have a tar­get num­ber of dol­lars. There was an ad a few years back that was liter­ally this. Each per­son had a gi­ant seven-figure red num­ber they would carry around from place to place, with an oddly ex­act num­ber of dol­lars they ‘needed’ in or­der to be ‘able to’ re­tire. Im­plied was that less than this was failure and would be ter­rible, more than that would be suc­cess and no fur­ther funds re­quired. Now you can rest.

In­stead, at best we have a prob­a­bil­is­tic dis­tri­bu­tion of how much util­ity one would be able to get from var­i­ous amounts of cap­i­tal, in var­i­ous sce­nar­ios in­volv­ing one’s life and how the out­side world is work­ing. Even in this sim­plified prob­lem, how does one then ‘play it safe’? At best one can re­duce var­i­ance from some ‘nor­mal’ shocks like a de­cline in the stock mar­ket, while still be­ing ex­posed to oth­ers, and likely not pro­tect much at all from big­ger risks. No mat­ter what, the bulk of what you have will prob­a­bly go to waste or be passed on to oth­ers, or else you are risk­ing dis­aster. At a min­i­mum, you’re ‘bet­ting’ on what to do with and where to spend the rest of your life, and there you are very much all-in. Odd for some­one who is about to die to even try to ‘play it safe.’