Formatting bug: something that should be a link around the text "Moore's Law" is instead the text OF the html for a link.

Expand full comment

I've had video game addiction to League of Legends for the past few years, and I relate very much to what was described. The key to the whole addiction really is 'dark flow' as the author described it. It was very easy to start a game, then go semi-brain dead for 30 minutes until the game ended which let me ignore reality like bad marks or a lack of social life, then after the game ended I just hop back in queue for another round with a couple minute break where I could browse reddit as match making occurs. For me, the only thing I could do to break addiction was to have the game entirely uninstalled. If it was on my laptop, I was basically guaranteed to play many, many hours a week.

As for why Duolingo, Khan Academy, fitbit don't inspire addiction, it's because they don't allow for 'dark flow'. League of Legends is a skill based game and to do well you need to think hard, but it's very easy to do mediocre with your brain half off. But it's not possible to learn a new language with your brain half off. On Duolingo, if you're actually learning, then your brain is too active to be in a dark flow state.

Expand full comment

this is my favorite review yet, though I haven't read all of the others.

My one issue is that, in my undergrad, we learned about early Skinner boxes, and mice did not work harder for random rewards than constant rewards. It's a small nitpick, but it's there.


Expand full comment

Reuven Dar has some surprisingly compelling evidence that cigarettes and tobacco are addictive for social and ritual reasons rather than because of the biochemistry of nicotine. I'm not fully convinced, but if true it would clearly have implications for addiction more broadly. I've always sort of wanted to see Scott review this literature.

Expand full comment

Reading ACX is part of my dark flow. I'm reading it right now to procrastinate getting out of bed.

Expand full comment

The question is, what is a Game? The answer is in this Google Doc: https://docs.google.com/document/d/17Qu2pki_gphSzZCFBBOGOWKKN5-XrVRMVlgGCNhU8fE/edit?usp=drivesdk

Expand full comment

A well written review, which I think found the right balance of summary and personal feedback.

I did this stuff for a while, so some thoughts.

1. All the tricks described are real.

2. I don't know of any genius psychologists working on these things. Features are implemented mostly on a "copy this successful thing, but a bit different" approach. Not necessarily less unethical, but certainly less sinister.

3. The book makes these games seem like alien technology directly hacking all our brains pleasure sensors. Most people will find these games boring and stupid. It's still tragic that people get addicted despite this. Perhaps more tragic.

4. The book attributes the success of the machines to their sheer addictiveness. In reality factors like density, efficiency and automation are probably more important.

5. Honestly, I'd say people find games way more addictive. The things I hear from people working on micro-transactions are, frankly, scary. Khan Academy and Duolingo are very successful, and they have to optimize for two things rather than one.

Expand full comment

> If designers optimized gambling machines for addictiveness, why can’t they do the same for these apps?

There has been quite a lot of research into the neural correlates of addiction, and while I'm certainly not an expert here, I don't think any discussion of addiction is complete with considering the role of dopamine.

Addictions are built on top of activities that activate the pathways of reward, craving, and seeking in the brain. No matter how much gamification is thrown into the mix, you (or most people) aren't going to get addicted to being punched in the face, because this isn't the sort of thing that primes the pump of reward, habituation, and craving.

Yes, there's a reason that only vices are addictive, but it might be more accurate to say that there's a reason we call them vices: they are fundamentally attractive in ways that make people feel they need to resist or control them via moral codes or other mechanisms. If we weren't drawn to them naturally, we wouldn't be concerned with them. Activities (or chemicals) that take advantage of that natural attraction are the foundation that game designers are building on top of.

(Notably, a lot of really enjoyable activities don't involve craving. I love to ski, but I don't crave it in the way I crave caffeine, and while skiing can get me into a flow state in a way that a cup of coffee can't, coffee is addictive in a way that skiing just isn't.)

Expand full comment

Nice review. My two cents on "virtuous addiction": I think this actually happened to me for a few months when Pokemon Go came out. The core gameplay was mostly "find the part of town with the most Pokestops, then walk in circles there until your legs give out." It had all the hallmarks of the flow state, and perhaps most importantly, you could do it with your brain half off.

Expand full comment

In the context of social media, thinking about what exactly users are "addicted" to is also important. Great social scientist Gary Becker believed that (one of) the most addictive factor to people is...people. https://freakonomics.com/2008/11/11/gary-becker-thinks-the-most-addictive-thing-is/

Expand full comment

I design casino games including slot machines, and in the USA there are regulations against faking the reels to give near misses. The frequency with which symbols appear has to be consistent whether they are part of a winning configuration or not, so that in theory you can reverse-engineer the math and figure out how often you should win. (Recently certain states have begun to relax this regulation.)

The thing to remember is that all the slot machines will give similar “RTP” (return to player, on average 88%-95% of the player’s bet comes back as a prize and 5%-12% is kept by the casino, the higher RTPs usually go with the higher-bet-amount machines of course). The designer is trying to keep the player playing HIS game rather than a different game, and players generally come in with a certain amount of money which they will usually lose but which they will get several hours of enjoyable play out of if the game is well-designed. The money lost is considered part of the “entertainment budget”.

A key metric is the doubling percentage—the game is volatile enough that a player who quits when he either doubles his money or loses or it will walk away a winner 20-30% of the time, enough that they sometimes get the satisfaction of beating the casino THIS TIME. If they always ended up broke they wouldn’t enjoy playing.

There is also a distinction between “destination” casinos and “locals” casinos. Players on the fancy Las Vegas Strip casinos are there to splurge and the games are therefore greedier. But no st casinos make their money from regular players whom they need to keep coming back, so they have to offer better odds.

My recommendation is, if you must play a machine, the Blackjack and Video Poker machines give way better odds, if you learn some strategy it is easy to keep the House Edge under 2%. On the casino tables, Baccarat and Craps also have good odds and don’t require any strategic skill (except to avoid the sucker bets and just make the main bets).

Expand full comment

I have a few more hypotheses re: the gap between the addictiveness of computer gambling and "virtuous addictions":

1. Orthogonal constraints. A "virtuous addiction" app designer needs to make the app both addictive AND good. This constrains their options. On the other hand, a purely-addictive app designer simply needs to make the app addictive. If something's non-addictive, it goes. It's not surprising that a group optimizing for a single objective can hit it harder than a competing group operating under a major constraint.

2. Not targeting addiction. "Gamification" apps may, for institutional reasons, shy away from the factors that could make them maximally addictive. In your essay you describe the addiction-promoting tactics as increasingly ethically dubious. Apps designed to promote some good behavior may be more likely to avoid ethically dubious practices. (Not super high confidence on this but it's possible.)

3. Reality. When you gamble money you are actually losing money. This actual outcome may tie in to risk-reward systems much more powerfully. Gamification apps have shied away from this. (I can't bet on the Duolingo Owl.)

4. Addiction rarity. We all live in a world with gambling machines; while they are powerfully addictive they only seem to affect a relatively limited total slice of humanity. (You can put forward a Rat Park or a genetics or whatever hypothesis you'd like for this.) That means that the total baseline propensity for addiction (or for any particular addiction) may not be that high; and so the member of the general public who picks up gameified tech—thinking NOT, "let's look for an addiction!", but instead" let's learn French!" or whatever—may just have a relatively low likelihood of getting addicted. (This pairs well with other hypotheses; if a small fraction of humans are addiction-prone, and slot machines are more powerful than the owl, then it may not be surprising that everyone is with slot machines and nobody with the owl.)

Expand full comment

🍎🍒🍓 Winner, winner, chicken dinner. This has been my favorite review.

Expand full comment

> If designers optimized gambling machines for addictiveness, why can’t they do the same for these apps? If bad machines can be made addictive, then why can’t good machines?

The problem is that (for lack of a better term) simple addictions are easier than complex addiction. In the same way that slots are more addicting than poker (faster reward loop, more stimulating), it's also more stimulating than "good" addictions. A good addiction is something like an app that helps you keep a habit running every day. even as it is gameified, only a tiny portion of the habit is gamified. Running every day still is the same experience it always is, and it will be the majority of the "use gamified app, run" loop. maybe you spend 5min doing addicting app stuff, and 20 minutes doing running stuff (which is exhausting and sucks).

However, there is a parallel experience - addiction to self help *content*. Stuff that tells you to get up, go, solve your problems through the power of will. You read it, you're stimulated, and instead of getting up and solving your problems (hard, long reward loop), you seek out more of the stimulant - more self help. You can find tons of youtube vlogs and articles and blogs of people talking about their addiction to self help, and how it didn't actually help them.


Expand full comment

I wonder what observational evidence Schull (or the author of the review, or society at large) would or should view as supporting the opposite idea, namely that the participant in the activity is at fault for engaging in it self-destructively. I’m not saying that this can’t be answered, only that it strikes me as a difficult question, and perhaps an important one before we dismiss the “fault” hypothesis too quickly and categorically.

Expand full comment

>>Third, maybe all designers worth their salt work for casinos, and virtuous apps are scraping the bottom of the barrel when they hire designers. The incompetent designers who work for virtuous apps just aren’t smart enough to create addictive products.<<

I suspect that this hypothesis comes closest to the truth. There isn't as much incentive or profit, on the whole, in the "virtuous" ones, and the lack of incentive causes them to be, quite simply, lower quality.

I have ADHD, and like many people with ADHD, I am reasonably prone to behavioral addictions. It requires constant vigilance to prevent them, and I still don't always succeed. I've been addicted to certain "free-to-play" games in the past, for example. I also use DuoLingo regularly, and I can say there's frankly no comparison between the two, in terms of addictiveness.

There are a few elements in DuoLingo that show "promise" in that regard (such as achievements or advancing your rank and league) but not many. Take in-game currency, as an example. Free-to-play games typically have at least one (if not more than one) in-game currency and a dizzying array of things to buy with it. Typically, many items or actions flatly require it. DuoLingo does have in-game currency, but the problem is, there's almost nothing you can buy with it. So, once you've bought the few available perks (as I have), there's very little incentive to earn more currency or keep going endlessly. Similarly, FitBit was uncomfortable to wear, and I found the app rather clunky and unintuitive to use. It wasn't really clear to me what was even *supposed* to be addictive about it.

So, those are just one or two examples, but it's illustrative of my point that the quality isn't generally there with the virtuous stuff. I truly wish it were, so that when I indulge bad habits, as inevitably happens from time to time, I could at least get something worthwhile out of it, but it's just not.

For what it's worth, Habitica (a to-do list gamification), came the closest for me, but it wasn't effortlessly social enough to hold my interest. I didn't see much point in fun new digital outfits or backgrounds, when no one was seeing them but me, and questing alone wasn't as motivating. But I probably could have tried harder to find a guild too. *shrug*

Expand full comment

The part about gamblers and flow reminded me of Edward Ugel's "Money for Nothing" (a memoir about gambling, high-pressure commission sales, and the business of buying out lottery annuities). Ugel suggests that one thing setting apart problem gamblers is that the primary appeal of gambling becomes the thrill of risking loss, not the possibility of winning, so no amount of actually losing detracts from that primary appeal.

Expand full comment

An overall great review but the push back against capitalism as the root of evil is perfunctory and poorly argued.

Modern China is very capitalist. It's no more a communist country right now than north Korea is democratic. If it walks like a duck, quacks like a duck, why are you paying attention to the faded lettering saying, "cat?"

There have been capitalist states with one party systems. To go for a very applicable example, Taiwan before 1987 was a one party dictatorship which defined itself then and now in capitalist terms.

Communist states tend to be one party authoritarian states, but that doesn't mean that one party authoritarianism is equal to communism.

Also, even if we extend charity and accept the PRC as still being communist, Macau is not a good case. Gambling is allowed in Macau because it is explicitly and legally separate from the rest of China as special administrative zone.

Again, separating gambling from society is a long and storied tradition. Arch capitalist Singapore has a very successful casino, but Singaporeans and permanent residents are required to pay a 100$ cover charge to enter the door. It is meant to fleece tourists not residents.

I think the idea that "capitalist society in particular creates people who are susceptible to gambling addiction" is still very much plausible.

Expand full comment

Formatting bug: 10,000 (222222) should have some multiplication symbols in between the three instances of the number "22".

Expand full comment

A really outstanding review. Described the book in enough detail so that you knew what it said, and could decide whether it was worth actually reading, while also presenting a nice separate critique, so that if you didn't want to do a deep dive (by reading the book) you were at least informed of the terms of the debate and the ideas present, and could riff on those.

Also, just plain good English writing. Very well done indeed.

Expand full comment

Some people previously complained that they couldn't tell which book reviews were by me vs. contest entries.

You can tell this one isn't by me, because I would have made a bigger deal about the kabbalistic implications of slot machine reels having 22 symbols.

Expand full comment

185% dead link

Expand full comment

re: virtuous apps

one difference is the lack of randomness and losing.

Like, if you get an answer right on Duolingo, you get a happy bell and animation - you don't lose points the majority of the time and then gain big one of the times

fitbit similarly has all the rewards, but none of the randomness - you set yourself a step goal, you walk that number of steps, you achieve it. Maybe if every 500 steps you either won or lost points, you'd feel compelled to walk more

Expand full comment

Kind of tying back into the Rat Park concept, I think the life-ruining aspect of a lot of the most addictive activities and substances is probably important.

For myself, procrastinating is most tempting when I already feel hopelessly behind. And of course the marginal minute of procrastinating makes me even farther behind.

Likewise being broke probably increases the appeal of the sort of “dark flow” that slot machines provide and the fact that they makes you more broke seems like it would be reinforcing.

That would predict that virtuous activities are only addictive once they start ruining your life, but that definition has an unsatisfying circularity to it.

Expand full comment

Maybe there is some insight to be gained on virtuous addiction from "big boy" gambling- the stock market. The common wisdom and evidence say that you can't beat the market- but the market is still going to increase ~15% per year, iirc. I know I've fallen down rabbit holes of various types of analysis/straight up dumb guessing on the market, similar to how I treat my more traditional addictive vices (video games), and I'll gain and lose money locally, but I end up ahead by the years end, at least so far; the "addiction" keeps me funneling a large part of my excess income into my "casino" account instead of my safer account that is all ETFs that I let money sit in.

My gut instinct is to say this works because the background steady state is positive (market generally goes up), and the wheels of the activity are only marginally correlated; another poster suggested Pokemon go works here because you are walking in pursuit of the game goal. I don't think you can effectively background e.g. language acquisition, that or it takes too long and the "fun" parts aren't that fun. Although as I write this, it occurs to me that on duolingo I've been in a holding pattern of mostly just doing the review exercises to keep my streak up for about a month, and I *think* I'm trending upwards in my language skills (I solve the problems faster/understand the sentences better).

Expand full comment

About DuoLingo: it certainly can be addictive. Last summer I found myself spending a 1–2 hours on it each day. I realized it was becoming a problem when I was reluctant to join a real-life conversation in the language I was learning, because I was busy doing DuoLingo. After that, I've cut way back on it.

Expand full comment

I thought this was pretty good and put me in mind of the one time I ever visited Vegas. I liked it there a lot more than I expected to because it was so forthrightly superficial, I couldn't help but respect it.

As for the why of gambling apps and not duolingo, I think it's related to the second point. But if vice is intrinsically alluring, that just pushes the question back another step, why? The pain of the loss is key I think, without that, there's no joy in gambling and not much in any kind of game. See for example: https://www.youtube.com/watch?v=Su1Rdw3SRHY

1st Ol Trbetr

2nd Qba'g pbagenqvpg Tnyra

3rd Ur pnyyrq uvf Whzob

4th Tnaqnys naq Rmrxvry

5th Obbx fgnoovat gvzr

6th Ab, ohg jr pna yrnea n ybg fgvyy

7th Orrc obbc V nz na nqqvpg

8th Ebzr jnfa'g qrfgeblrq va n qnl

9th *Qha Qha*

10th Jung qvq gur theh fnl gb gur ubgqbt iraqbe?

Expand full comment

Great review. One of my favorites.

Expand full comment

Re. the NRA: So what you're saying is, if most people hear a story about a kindly old grandmother who picked up a shotgun and blew someone's head off, and became a murderer, they blame the gun for causing her to murder. Without the gun, she would still be that kindly old grandmother.

Expand full comment

"maybe vices are just inherently more addictive than virtues". what is addictive are freebies. easy wins with no effort. case in point my roommate uses a fitness app where you can enter a draw for every 10 push-ups you do. he freely admits he cheats on the app daily.

Expand full comment

It seems to me that, no matter the _quality_ of addictive gambling/apps/whatever, there is always the choice of whether they shall be present in one's life in any nonzero _quantity_.

Framed this way, it is clear that highly-addictive outlets of all kinds, while indeed "vices", are nonetheless in a position to provide a unique service to the rest of the world.

The uniqueness is an unfortunate byproduct of modern society's ever-increasing desire to stamp out individual adversity even at the cost of great amounts of total adversity.

In short, this service is colloquially known as an "idiot tax".

Other such providers of this service have included charlatans, quacks, snake-oil salespeople and con-people. Their dwindling is a great loss to the rest of us who are capable of peacefully coexisting with them without being mysteriously moved to donate them anything.

Expand full comment

Fitness apps, by and large, don't have randomized rewards.

Expand full comment

Good review, but it’s pretty hard to argue China isnt a capitalist society.

Expand full comment

The problem with identifying behavioral issues like gambling, sex addiction etc.. etc. as addictions is less the question of whether or not they qualify as 'addictions' in some definitional sense (after all they are all dopamine releasing activities and have similar behavioral responses) but the implicit moral categorization. Yes, all these things share this dopamine reward pathway but plenty of other things we see as good and desirable do as well. Many of us feel that way about successes at work, others get that thrill from meeting new people and maybe impressing them or making a good point in an online community.

The problem is that the real distinction we are making when we call something an addiction is really between problematic/harmful behavior and beneficial/acceptable behavior. The guy getting his dopamine hits starting companies, proving theorems or making friends isn't harming themselves in the same way the problem gambler might be so we don't call it an addiction.

That's fine but the word is deployed to create the impression that something more medical and technical is going on than a judgement about the desirability of some behavioral pattern and that is misleading to the general public.

Expand full comment

I read Addiction By Design when I was writing my book on luck. I thought it was harrowing and eye-opening. For me the most fascinating parts of the books were less about the casinos’ strategies to achieve addiction and player extinction, but about the experiences of the players themselves—particularly the ideas of flow, assimilation into the game, and indifference to winning. The review of the book is good and fair.

Expand full comment

Aren't we down on rat park? It's one of those typical too-good-to-be-true sociology experiments of the postwar era IMO. Also Scott: https://slatestarcodex.com/2017/10/25/against-rat-park/

Expand full comment

having worked in IT for casino: the casino is not allowed to adjust the odds of a particular slot machine, but it is allowed to adjust the odds of a bank. if someone relatively new to the casino begins playing you want to sit in that same bank of slots.

Expand full comment

Virtuous apps aren't addictive / flow-inducing because there are no immediate status gains to be anticipated. We don't care about levelups and badges, we care about levelups and badges other people care about.

Expand full comment

Pokemon Go is absolutely addictive in a way that Fitbit isn't. I'm not sure if Niantic is just better at this, or just more committed. I've planned activities throughout an entire week to bump up my walking to the magic 50 km. It's somewhat self limiting in that there's only so much walking you can do in a day, and there's really not much to do that doesn't involve walking. But yeah, humanity can make addictive virtuous experiences, we just don't.

Expand full comment

Great review. Thanks for writing it!

One thought I had is that both you and the author are drawing to strong a division between “unhealthy, addictive behavior” that incorporates activities that are basically empty calories (gambling, social media addiction, etc.) and healthy, behavior where you do lots of good things (???).

I’m not on social media. I’ve watched you all from the sidelines. You’re all insanely addicted. I’m 47, so I clearly remember when all of you USED to do other things, but now you stare at your phones and little else. You even talk about Facebook posts while we’re out to dinner even though you know I’ve no fucking idea what you’re talking about. It’s real, and it’s a serious bummer.

And I don’t participate. Unfortunately, what I do instead is stare at my phone reading Substack articles, all day long (and before Substack, blogs). I’ve subscribed to a ridiculous number of Substacks at a cost that I’ve hidden from my wife, and there are probably eight longform articles for me to read a day. God forbid I miss the third 6,000 word article of the day written about cancel culture.

Maybe the problem is the PHONES. Maybe the problem IS the pills, regardless of what the rats tell us. Maybe there are objects that cause addictions. Facebook existed before smartphones, and you all incorporated into your lives just fine. But when it was put on your phones, wow.

I’m not convinced that me reading eight Substack articles a day is materially more beneficial to me, my family, or society, than you staring at your phone all day owning the Xs on Twitter. It’s all just too much time, too much busyness. Anything that distracts us from our duties (to our families, our community, our jobs) to this extent is bad. So put me down for blaming the phones and the pills.

I wrote this way too quickly - I've got run. When you read it, please pretend it's more polished.

Expand full comment

I'm wondering whether having a taste for dark flow is connected to something. Maybe depression?

Expand full comment

This was a great piece to read!

Perfect for temporarily escaping the stress of reality.

Thanks! ;)

Expand full comment

There have happened a few things since 2012, perhaps especially in Europe. The state possessing the regulatory power creates an interesting back-and-forth. The state wants gambling (this provides income, either directly through having a state lottery, or indirectly through concessions and taxes), but it doesn't want addiction - addiction has significant social and PR costs. The lotteries would kinda want properly addicted players, but realize that because the state has the regulatory power, and can be influenced by the public's perception, there are limits to how far they can push things, and ideally they would like to pre-empt actual regulations by their own systems. Therefore, both sides like a pretty high amount of spending from each person, while trying to reduce social costs (an interest of the state) or the risk of harsher regulation due to social costs and public relations (the game providers).

For instance, the Norwegian state lottery runs gambling machines with a number of restrictions on them (some of these come from above, from the political parts of the system, and others are created by the state lottery itself in order to seem responsible to its superiors). These include a player card linked to your identity (this being Scandinavia, it's easily achieved and hard to bypass), enforced loss limits on a daily/weekly/monthly level (well under $1000 monthly), making it illegal to put the gaming machines in establishments that serve alcohol, capping individual wins to just a few hundred dollars, and forbidding some of the methods discussed in the review, like near-misses or features that encourage a "one more round" mindset. The whole system also runs on a centralized network, making it even easier to control and analyze play. These kinds of "responsible gaming" systems are increasingly common throughout Europe.

Another thing that has happened since 2012 is the increase of web/app-based gambling (obviously accelerated by the Corona pandemic). But what's strange here isn't that it has happened, but rather that it hasn't happened _more_. For what else that you could do online would you even consider going to a specialized machine outside your home to do the same thing on _its_ screen? With a casino, there's an obvious explanation - the surroundings make for a different experience, and it's easy to see how it would encourage flow. But a lot of gaming machine play happens on fairly simple machines placed in numbers of one to three in coffee shops and 7-Eleven type stores that certainly don't bring that kind of ambience.

Expand full comment

I have a pretty hardcore internet pornography addiction I've been working to get over, it was created over the decades and I don't know it was a problem until I tried to quit. But I've also been working on getting rid of it and studying math and computer science to hopefully switch careers.

So here's my take on the different flows. I like math and computer science, out of all things in my life its what I enjoy the most. However I usually drag my feet to start studying for the day and it usually feels like I am fighting myself to do so even while I know I will enjoy it and feel good when I am done. When I am working on a programming problem getting into a state of flow is easy. There is a big difference between doing and studying, studying requires 100% conscious effort to get things right and correct my errors, programming takes half my mental effort as I have some best practices ingrained so I can usually iteratively try things to see what works. I enjoy both those activities, but they are draining for me mentally.

Porn is a hit of a wonderdrug, whatever bad feelings are usually wiped away and I am wired in a state of dark flow for hours. It's like a lightswitch in my head with an on/off, if I screw around on the web or go anywhere I am not supposed to and whatever neurons get activated I know my day is shot. I can feel the change instantly, I feel my head constricting and I feel more locked in to my mind rather than controlling it. impulses get amplified 100x. A true rogue agent seems to take over.

It would be nice if I got to have the controls and just avoid content I don't want to see usually. But its not that simple, Something I call adjacency cascade usually always leads me back to porn. As an example lets say I watch YouTube, I look up meme videos, I scroll through those until I see I see one that's slightly sexualized, I tell myself I shouldn't but If I am already feeling bad my ego is already depleted. It keeps hopping down those lines until I find a rare uncensored porn video on youtube.

My rouge addiction agent uses all latent intelligence I have to optimize it's one goal, and Its really really good at breaking down all barriers I create if it gets power. I am smarter than myself. (as a fun corollary my addiction helped me understand the orthogonality thesis and deceptive agent AI research easily and intuitively.) My solution was the same but more hardcore that the OP. I blocked all the internet except for sites I use to study. No searches on google, if I don't know something I have to find it another way. and congrats you made the cut.

Addiction doesn't feel like 'you' from the inside, maybe that's what's so enticing, but once you get back in the drivers seat and see all the time the rouge agent has wasted just to hurt you, you wish you never gave up the seat in the first place... Until it whispers those sweet nothings of escape in your ear again.

Expand full comment

Question: how much of casino's profits comes from people with gambling addictions? I'd have thought lower in Vegas than for local casinos, but are there stats on this?

Expand full comment

I just wrote a post on LessWrong referencing Addiction by Design… if only I had waited, I could have linked to this excellent review.


Expand full comment

Are gym rats and workaholics addicts? There is certainly a flow aspect to their behavior.

In terms of the rat park experiment, I recently went to a fancy spa resort that had a strict no cell phone policy. I was worried I’d have withdrawal. But with so many activities it was an absolute delight.

Expand full comment

"Good flow" = when studying or solving a problem or working on a project or exercising; you are in the zone and you do something useful.

"Neutral flow" = when playing video games; you are exercising some skills, but ultimately nothing useful, and probably procrastinate from something useful.

"Evil flow" = when gambling; the zone is about destroying your wealth and possibly your life.

The question how much are addicts responsible for their addiction seems obvious from the perspective of multiple agents in the mind. There is a part of the mind that wants playing on the machine and losing the money. There are other parts of the mind that would prefer to keep the money for something more useful. The machine is designed to provide advantage to the gambling part of the mind in the internal fight. Yes, the addict had a dark side, but it was not obvious that the dark side was going to win against the rest of the brain, perhaps the rest of the brain would win. The machine did whatever it could to support the dark side, and then the owners said "well, it was a dark side of *your* mind, why do you blame us?"

As an analogy, imagine a country, where a certain small fraction of population (let's say 1%) are some kind of Nazis. A foreign government contacts them, sends them tons of money, weapons, provides training, helps them hack government computers, and so on (the only thing the foreign government does *not* do is send their own people). There is a revolution, and the country becomes a Nazi dictatorship. And the foreign government says: "hey, why does anyone accuse us of interfering? it was *their* citizens that did the revolution, how could that be our fault?"

Expand full comment

On Khan Academy games and Duolingo: they are absolutely addictive. For around a year I logged on to Duolingo every day and dutifully logged my 30 points so that I could keep my streak. I don't know whether that's because I've grown up online so I'm just predisposed to any form of internet addiction but in any case, yes. Obviously those aren't the same as other types of addictions (it's five minutes of a language exercise, in my opinion it doesn't have the same capacity) but they do create a compulsion to go through with that language exercise to the point where your autonomy can disappear.

The reason why I'd also say that Duolingo and Khan Academy-esque academic games aren't as addictive as Instagram or TikTok is that they don't have any element of humanity to them. After a certain point the experience of logging your points just becomes stale and unnecessary--at least in the face of the "one more video" or "ten more minutes" impulse that you'd otherwise find in social media.

Expand full comment

In terms of what a "virtuous addiction" would look like, I submit that I am addicted to air, water, and food. I consume these compulsively and suffer from serious withdrawal symptoms if I attempt to give them up. Luckily, they're good for me (with the exception of some of the food) so it's just a virtuous addiction.

Or if that's a bit too biologically primal to count, how about this? I am addicted to showering every morning, brushing my teeth once or twice a day, and going to work five days a week. Much like any other addiction I know I physically _could_ stop if I wanted to but it would make me feel bad; nonetheless I'm pretty sure that if I could get out of these habits then I'd eventually stop feeling the urge. Anyway, these are all virtuous (or at least beneficial) things that I feel strongly compelled to do, they just don't seem like "addictions" because they're so commonplace and beneficial that nobody uses that particular frame to analyse them.

Expand full comment

This is my favourite piece so far by far. Love the author's writing style.

Expand full comment

Very nice review, winner for me so far. I noticed myself circling down the TikTok drain the last 2 weeks and convinced myself I was getting something useful out of it. Just deleted it, thank you.

Expand full comment

My virtuous app is my radiology reading worklist. I switched jobs from an academic to more private practice environment and have been able to get into a flow much more easily due to the larger volume but lower complexity of cases. I think many of my peers would call the academic to private move as a downgrade, but it has definitely improved my work enjoyment.

Expand full comment

This is my favourite review so far - by a long shot. In fact I'd wager that it is the review Scott predicted with 60% certainty would win the book review contest. [And I also notice Scott joining in the comments with something positive vibey]

Having said that, the first things that come to mind as I comment, are criticisms. That fact was surprising enough to me that I stopped to wonder why it might be. A couple of answers presented themselves. Firstly I recently re-read Scott's "I can tolerate everything except the outgroup" and there was enough overlap between this reviewers worldview and my own for the differences to cause a certain amount of grievance in me.

Secondly, I'm a recovering compulsive gambler and therefore take it as read that non-gamblers can't possibly have an understanding of the lives, motivations and experiences of people that do (or did) gamble compulsively.

However, having noticed the arousing of my completely gratuitous prejudices, I'm left with very few criticisms at all. I was annoyed at the ending, but of course the problem was entirely mine. I was enjoying the flow of the review in the way I would a good Tom Hanks movie - I was in good hands, utterly trusting, and then suddenly I was thrown back into my everyday life with its uncomfortable unpredictability.

And the teaser at the end - that the book had numerous personal stories from gambling addicts, but the reviewer didn't have time or space for them, annoyed me for a few seconds. Then I realised it was an extremely good decision to omit them - but to let the reader know that they were in the book should the reader be interested. They would both take up a huge amount of space for very little addition to the review, and also wouldn't be a particularly good fit for typical SSC/ACX readers. Too much anecdote, too little persuasive argument.

The writing was deceptively good - another Tom Hanks anology would probably be appropriate. There weren't any literary pyrotecnics, just polished, unfussy clarity. What I liked most was the intellectual and general humility of the review, especially because it is something I lack.

A good example of some of the qualities I admired is how the reviewer dealt with the authors contention that capitalism is somehow responsible for the prevalence of gambling addictions. I don't know if that seems more ridiculous to me than it does to the reviewer, but I would have been launching tirades of sarcasm and ridicule. What I read was a calm but convincing rebutal. Most impressive.

I especially liked that the reviewer managed to convey something central about gamblers - or at least the vast majority of those I met in rehabs and in the meetings of gamblers anonymous. Which is (contrary to the reviewers and most peoples initial understanding) that gamblers are not primarily motivated by trying to win more money. Yes, superficially it looks (and can feel) like that. But it is, as was said, much more frequently, and more deeply, about escaping to a safer, more comforting place. Which is indeed why winning can be such a depressing experience.

The book reviewed was specifically about slot machines and it is quoted as saying that "Machine gambling is not like other kinds of gambling" Of course, a book devoted to one kind of gambling is prone to making such distinctions, but although Iim maybe out of touch with the addictive power of modern gambling machines, I find that hard to believe. I'd bet that the majority of money gambled on the high streets of Britain is still spent in betting shops and not on gambling machines of any kind. Vegas, of course, is a different world entirely..

Expand full comment

probably was discussed before in the comments, but still:

I think the reason apps like duolingo aren't as addictive is because duolingo has to teach you a language, at least with mild efficacy (and a fitbit has to get you fit and so on).

Under these restriction, it isn't necessarily the case that gambling machine engineers could have made a machine addictive. Learning a language is by default hard, or should I say broccoli, otherwise everybody would do it all the time just for fun and there wouldn't be a need for apps to gamify learning. Gambling machines have no such limitations and can focus on purely chocolate aspects of gaming. This doesn't map perfectly with vice and virtue but the correlation is pretty obvious.

So some activities have inherent broccolian qualities to begin with.

Expand full comment

>Designers replaced mechanical levers with buttons and physical reels with video screens. This made the games three to four times faster. The quicker each game, the more money gamblers can spend during their gambling session. Quicker games are also more addictive. Bally, a casino company, targets 3.5 seconds per game

This is an important thing to understand. Casinos don't particularly optimize towards taking all your money. They optimize towards making you play as many games as possible.

You know card-counting? People often wonder "why don't casinos take easy steps to stop you doing that? Such reshuffling the deck after each hand, or dealing from four mixed decks, etc?"

They don't because it would waste time and slow the pace of games. Fewer games means they make less money. It's not worth it just to catch the occasional decent card counter (as opposed to random idiots who rush to casinos after seeing it in a movie). They more than cover their losses by speeding up game-time for the remaining 99%.

Expand full comment

There's one thing that's common in addictive activities that I think can't be easily replicated by self-improvement ones: the conjunction of feelings of "just one more time" (which hacks or hedonic treadmill) and ex post regret. When it comes to self-improvement, you'll either give up or achieve a milestone and stop to give yourself a rest.

But then... NUCA ZARIA: Effective Addiction

We build an addictive industry and transfer all earnings to effective charities. Then we send a message to our victims, I mean, customers, saying "Thanks for saving 20 kids from malaria. Please keep going," so they'll feel proud and be cured.

Expand full comment

A question to pose to casinos: If addiction is purely a moral failing, why did people get less addicted to older slot machines?

That doesn't prove that it's 0% a moral failing, but I think it disproves the idea that it's 100% a moral failing.

Expand full comment

About the part where virtuous apps don't addict as easily as vices, couldn't this just be because vices like slot machines don't take much effort while the Duolingo's and the Fitbit's do take effort? Like, pushing a button for the dopamine hit is going to be more addictive then spending mental effort to learn more of a language or spending physical effort to workout.

Expand full comment

Hmmmm, we know there are certain factors that make substances more/less prone to trigger addiction/compulsive use.

The ones I remember are;

- must have a pleasurable effect (this is why we don't get addicted to Tylenol)

- the faster the pleasurable effect hits, the more addictive (this is why crack is more addictive than snorting cocaine which is more addictive than taking it in tablets)

- the shorter the pleasurable effect, the more addictive

- if there is a down after the pleasurable effect wears off (ie, you actually feel worse than you did before taking the drug), more addictive (all stimulants do this, some more than others)

Then we add in factors like how often the person is using the drug (cigarettes; hundreds of tiny hits a day! Each drag counts!) etc

It seems to me that similar factors might explain why game-ifying around good/desired habits hasn't worked that great. Slot machines are pleasurable, fast, wear off fast, and we can take hundreds of 'hits' in a short time. FitBits can't 'reward' us for every. damned. step. Although even without that, I've known of a couple of people who get VERY compulsive about their step total.

Expand full comment

The 'dark zone' certainly fits with the stress factor; people who are addicted to substances or who have compulsive behaviours such as gambling, video-gaming, shopping etc are much more likely to indulge or over-indulge or relapse when their stress is up or they are unhappy but avoiding.

Then of course the consequences of the addiction/compulsive beh create more stress and unhappiness .....

Expand full comment

We know that people are much more likely to get addicted to something that goes well the first couple of times they try it, whether it's a substance (felt great and laughed a lot using cannabis the first time, or had a panic attack) or activity (won the first few times you gambled, or lost?).

So now I'm wondering whether gaming machines/online gambling takes advantage of that. When a gambler can be identified (by the card they use to play?) as trying a specific game or site for a first time, make them win. Or a few smallish wins within the first few times they play. Then set the winning ratio back to normal .....

This would work particularly well on those sites where you can play poker for free. Free games? Win ratios much higher than usual. Once they're playing for $, back down to normal.

Expand full comment

I used to manage an arcade, and there was a game called "Stacker" that used some of these tricks. https://en.wikipedia.org/wiki/Stacker_(arcade_game)#/media/File:StackerMachine.jpg

In this game, players had to time a button press to stack 1-3 blocks on top of previous rows. If they got to the penultimate row, they could cash out for a small prize. If they kept going, the top row had the potential for expensive prizes (e.g., a 3DS).

The top row was a lot faster, and the player could absolutely mess up the timing. However, the operator menu in the game also let the arcade owner set the rate that big prizes were allowed to be won. If it wasn't time for a win, even a perfectly timed player would get a "near-miss" when the game cheated and moved the block one space over. If you watched carefully, you could see this happening.

Conversely, if it WAS time for a win, the top row would move a bit slower. If you payed attention and noticed this happening, you could watch someone lose because of bad timing, walk over there, and win the big prize yourself. A few of the regulars at the arcade used to do this and then go sell the 3DS's to Gamestop, which I had no problem with. They'd spend that money on the DDR machine.

I'm pretty sure newer versions of the game have removed this exploit, but not the cheating "near misses". There's plenty of other redemption games with similar behavior.

Expand full comment

Speaking of dopamine drips, interesting that this particular substack doesn't have a "like" function.

Expand full comment

> I think this is the biggest gap in Schüll’s theory. If designers optimized gambling machines for addictiveness, why can’t they do the same for these apps? If bad machines can be made addictive, then why can’t good machines?

It seems to me that you answer this question in your review:

> When it comes to machine gamblers, my theory is completely incorrect. People who spend hours and hundreds on machine games are not after big wins, but escape. They go to machines to escape from unpredictable life into the “zone.”

I suspect that given the nature of virtuous tasks, you can't design a system that lets someone escape into "the zone" the pursue them. If you're trying to get better at picking up the dishes, or exercising, eventually you have to stand up and actually do the work, which will shove you right back into "unpredictable life".

Expand full comment

I think this was one of the best reviews so far, reading it was rather... addictive. I think the mention of 'dark flow' was one of the most interesting parts - the idea of creative/productive flow and the brain-deadness of procrastination having something in common has been in my head for a while now, but this helped my clarify what the difference could be.

Someone in the comments mentioned randomness, which I agree would make a big difference (e.g. Duolongo points are hardly random, you only get them when you remember vocabulary or something). But TikTok and other social media aren't that random either - perhaps the key is the lack of 'challenge-skill balance'? If there is very little challenge but otherwise still a lot of engagement, maybe that leads to a dark flow.

Expand full comment

Good review! The gacha mechanism which was developed in Japan has found its place in tons of mobile games during the last 5 years especially, and kids are used to the idea of gambling as a normal risk to take, from a very young age. If you walk past a pachinko parlor in Japan in the evenings, you can observe thousands of people who spend 2-3 hours sitting there immediately after work, completely immersed in the "dark flow" outlined in the review. The appeal of games like Candy Crush and Clash of Clans among older demographics is a well-known phenomenon as well. Games have addiction circuitry to begin with, and if you add a gambling component, I don't really see how it could possibly be overcome by most individuals. After dumping about $100 into a gacha game a few years ago and then quitting it a few months later due to feeling like it was a chore, I play multiple ones with a no-spending rule purely for art/game mechanics. My outlook is that this issue is going to become a full-blown catastrophe among swaths of young adults in the next decade and it will become a societal concern just like cigarettes and hopefully, meet its end in a similar way too.

Expand full comment

Definitely the best review so far. I simply deny (refute and repudiate) the idea that addiction isn't 90% or more circumstances over genetics. (I'm not quite willing to say it's 100%, but I'm close.) That doesn't mean it's just "willpower" or "just" a personal failing. Let me put it this way: people who feel a calling and whose lives are in order simply don't waste their lives gambling or getting high or drunk enough they destroy their lives. Said another way, the addiction doesn't destroy men's lives; men's lives are destroyed and so they become addicted.

I admit this would be a hard thing to argue with me about, for if you pointed to a guy and said 'his life was good until he got addicted', my response would be to deny the premise: 'it obviously wasn't.'

Expand full comment

Great review. Small nit for the author: While I agree that the connection between late stage capitalism and gambling isn't strong, Macau *is* actually a hotbed of capitalism, one of the "one china, two systems" areas that stayed westernized and capitalist during the latter 20th century, just like Hong Kong.

Expand full comment

I've derived the my sole income from professional gambling for the last five years (things kinda got awkward at my old programming job at a startup after I came out in support of brexit and trump, which was a factor in my deciding to quit and move to vegas)

Professional APs tend to hate the casinos because they knowingly exploit addicts and delusional system-players while barring anyone who has a real edge as soon as they're detected. I'd guesstimate that addicts are only 25% of casino patrons in Vegas but the majority of casino revenue (they're outnumbered by casual tourists who only lose a little). Genuine professional APs are fewer than 1 in 1000 casino patrons (not counting poker players who sit around waiting to pounce on drunk tourists, or small time slot hustlers scavenging the obviously scavengable slots like Ocean Magic. I mean people who beat the house for a living).

It's interesting to see doctors and lawyers punt money away on gaming decisions that are mathematically silly. They may be rational in their domains of expertise, but it doesn't transfer. They throw money away on the basis of superstitions.

To the ideal advantage player, it feels good to be playing with a positive expectation, and feels bad to be playing with a negative expectation, regardless of the actual outcome. Kind of like how the ideal utilitarian might be able to suppress his feelings of ickiness or warm fuzzies, to just shut up and calculate his expected utility. At this point I'm so jaded that I can win or lose $50k in a few minutes and feel nothing, so that helps.

A minor correction to the OP: money on the card did not directly replace coins. TITO (ticket-in-ticket-out) replaced coins, and it is still the dominant system.

I was already addicted to video games like world of warcraft long before I tried gambling, so I never found it particularly compelling as entertainment. All the gambling games are really lame compared to the latest video games. For me it was just all about the money. I calculated that I could make more per hour from AP than I could make from other things, so I did that.

I theorize that a good treatment for gambling addiction might be to get someone addicted to the most exciting video games instead, so that they start to find all the slot machines to be super lame by comparison. And then you can gradually taper down to less dopaminergic video games such as turn-based RPGs from the mid-90s, and then quit entirely. Perhaps the sequence could go TF2 deathmatch, TF2, Portal, Escape Velocity, Civ4, Realmz, nada.

Expand full comment

Brief review-of-the-review:

I liked this one a lot. Clear, cogent, readable, and interesting. It presents the book's core argument clearly and charitably but not uncritically. The questions the reviewer raises don't seem all that mysterious to me. Most patterns of behavior, especially simple ones, are unhealthy when pursued compulsively. And I sometimes gets compulsive over the video games or TV shows I use to unwind, so I can easily understand that there's a spectrum from "I'm doing this because I enjoy it" to "I'm doing this because it's designed to hyperstimulate certain processes in my brain". Besides wishing it went into a little more depth, I can't find anything to fault here.

Expand full comment

This probably sounds silly, but... if gambling addicts plan and budget around their need to gamble, and they deliberately seek to gamble, and gambling away money is the thing they prefer to do with their time... then the "seatbelts you can refuse to wear" won't work, but then, why do we have the obligation to stop them from doing the thing they want to do with their resources? For all the description of how insidious the machines are to trick people into thinking they are close to winning, it goes on to say that the gamblers don't really care that much about winning and do not behave like people who have been tricked. They have a weird-ass hobby, one that they can take to a self-destructive level, but there's a lot of hobbies like that. Why do we have the obligation or the right to stop them?

How many people do you know who are by the "behavioral addiction" standard "addicted" to football or basketball or hockey? Okay, less than the national average given the audience here, but it's still a thing you know about. When do we have the obligation to step in and make their hobby stop?

Expand full comment

It's worth noting that physical machines are vulnerable to manipulation as well. For example, modern claw machines can be programmed to raise or lower the claw's gripping strength in order to manipulate the win probability, limit payouts, produce near misses, etc.

Expand full comment