Log in

No account? Create an account
Zvi Mowshowitz
[Most Recent Entries] [Calendar View] [Friends]

Below are the 20 most recent journal entries recorded in Zvi Mowshowitz's LiveJournal:

[ << Previous 20 ]
Sunday, March 19th, 2017
8:03 am
My current blog is Don't Worry About the Vase, at thezvi.wordpress.com.
I de facto abandoned this LiveJournal years ago, and wanted to make sure that anyone who has stuck with the format knows where to find me. My new blog is here.

If you would like to read my writings, would read them here but would not read them on WordPress, reply here to let me know that. If enough people do I will look into cross-posting.

Also, I noticed while looking back that a few old friends do still post here, and I will be trying to peruse those archives to see what I can find. If you are still posting here (or elsewhere) and would like me to be aware of that, let me know so I can add you to my RSS/Twitter.
Sunday, June 29th, 2014
5:09 pm
Friday, February 14th, 2014
7:19 am
SPARC for High-School Magic Players
I know the people who created and run SPARC. I've worked with them to help them develop the programs that have become CFAR and SPARC. Everyone I know who has gone has come back with rave reviews, including myself (I helped teach once), and I highly recommend getting involved if you qualify. In this case, they've decided to reach out to Magic players, which speaks to their understanding of how to find smart and talented people!

Winning in Magic takes a lot: creativity, problem solving, and psychology, all of which also help winning in life. A relatively new summer program called SPARC (http://sparc-camp.org/) targets quantitatively talented high schoolers to improving decision-making, both for humans and for computers (e.g. Bayesian statistics, machine learning, game-playing, and *practical* cognitive science). The program has been around 2 years, and has a highly competitive selection process that may be a good fit for Magic players with some quantitative background. SPARC seeks to take goal-oriented, mathematically talented problem solvers (possibly like you!) to approach life in general with a winning mindset -- it may even improve your Magic game.

The application is at http://sparc-camp.org/apply/ and the second-round interviews start April 15th (completing the first-round applications early gives more time and information for the part 2 of the process).
Friday, November 22nd, 2013
8:56 pm
Coalition: A Game for 9, 12 or 15 players (with rules for 3 or 6, but full design requires 9+)
The core concept of 3 players dividing a pool of 300 units of money is well known. http://www.overcomingbias.com/2013/11/the-coalition-politics-hypothesis.html is the direct source. The rest of the design is by Zvi Mowshowitz. Invention of core structure and first playtest session were 11/19 with the NYC rationalist group , and we had a lot of fun.

Alpha rules here, help test it out!Collapse )
Tuesday, June 18th, 2013
6:20 pm
...because then everyone will show up.
4000 players are coming to GP Vegas. My lord. What does this mean for you if you're going?

There will be two types of players at GP Vegas: In one corner, you have the great unqualified masses trying to break through. In the other, much smaller corner, you have the qualified players trying to score a top 8 birth. These two groups will be playing altogether different games.

For those who are unqualified, the majority of their tournament equity is the slot they can get by going 13-2. It doesn't matter what everyone else does: Show up, win 13 of the 15 rounds, and take home a blue envelope.

But what about making the top 8, or scoring some extra cash or PT points? It'll be tough, but what'll it take?

As a baseline, one must calculate how many extra players are effectively in the field due to byes. We don't really know the answer because we've never had a GP this big, but we can guess. Assuming, as Lucas did, that there are 25% additional players, is being highly optimistic. If we base our assumptions on GP Charlotte, we will have about 40% more 3-0 players than that calculation would suggest, so I'll be using that as my baseline from which to do projections, and cutting off those who miss the second day.

At the end of 15 rounds, with no draws or scoops or pair-downs (which is NOT what will happen), things would look like this:
15-0 0 or 1 players (21% chance for an undefeated)
14-1 3 or 4 players (3.2 on average)
13-2 22 or so players (22.5 on average)
12-3 72 or so players (71.9 on average)
So, the thresholds would be: Top 8 would have about 4 players out of 22 who are 13-2, top 16 will be 13-2 or better (with about 10 out of 22 of them missing), top 32 will have about the top 5 12-3 players, some of the 12-3s will make nothing, and if you have a fourth non-win you can drop. So far, so good. It'll be slightly harsher than that due to the "paired down" clause, where somehow the player with the better record who is playing for a much bigger reward tends to win the match most of the time. Funny how that happens.

Round 15 includes the usual end-of-tournament draws. If no one has drawn, the round will start with 6-7 players at 14-1 or better given this turnout and bye level, so if they can draw they will do so, leaving only 1-2 slots for the 13-2 players, so for most players 13-2 will mean no top 8 and they'll know it well in advance; this includes anyone making day 2 with two losses, who likely isn't even live for top 16.

If you're not qualified, there's nothing to think about until after Round 13: Play to win. In Round 14, if you have exactly one loss, and you can draw now but likely can't draw next round, you'll have to choose between 25% chance of top 8 and 50% chance of missing but qualifying, against 50% chance of top 8 but 0% chance of missing and qualifying. Is a GP top 8 twice as good as qualifying? I think it is, but I do not think it is obvious. In every other case, there's nothing to think about, and in Round 15 the only time you can draw is if you'll be 13-1-1, since you need to go 13-2 to qualify and 11-3-1 gets you nothing.

If you're qualified and don't care about going 13-2, however, things are more interesting; keep in mind that this group will be small. Your goal is 13-1-1, not 13-2, so the first draw is a win. Going 12-2-1 also isn't so bad: You get top 32. Yes, 11-3-1 is still terrible, but top 64 is a very small prize with only the top 5 GPs counting for PT Point status, so it matters very little (and 12-3 was often going home with nothing anyway). In addition, if there are still matches left in your pod, if you draw then there's a 50% chance each round you'll be paired down! This means playing against a worse deck than if you'd won, and on top of that you get scoop equity depending on the situation. You won't be able to get anyone with only 2 losses to fall on their swords due to the qualification slot, unless they are qualified, but those with 3 losses or that already have a slot (and know they can't make it to top 8 anyway) should be far more pliable if it comes to that.

This raises the question: If you can take a draw in Round 11 at 9-1, and you're qualified, do you take it? It allows you to 5-0 into a probable top 8, or 4-1 into top 32. If you play and lose, you're out, and if you win you still need to go 4-0-1... and depending on the pairings, there's a decent chance that you never get that draw. Maybe it would be better to take it now. To know, you'll need to watch the number of players very closely. This projection puts 384 players into day 2; if there are much more than that, 13-1-1 from behind is no longer safe (it's still not completely safe anyway, because of unintentional draws combined with pair-downs, plus variance, and other people pondering what you're pondering) whereas if there's less, it's a safe play but 13-2 starts to get more likely (although coming from behind there is still not going to happen). The closer you get to the end, the better your information will be, and the less chances you'll have to get an ID and the less chance you pick up a real draw later, which would be a disaster once you already have one.

And of course, if you can't benefit from going 13-2, and you can ID in the last round to 12-2-1, that will get you top 32 whereas 13-2 will be somewhere between 7th and 26th, so there's a good chance you can't get to 16th anyway and there's no reason not to draw, provided you can find a willing partner for it. If your breakers are bad enough that 12-3 is likely to miss entirely, it's an easy choice, and the moment you pick up a second loss, you should start looking for an ID since it's basically a free win for you.

Most important, of course, is to enjoy being at the biggest tournament of all time, with the best set ever and in Vegas, baby, Vegas. It pains me that I won't be there.
Monday, January 7th, 2013
12:58 pm
...because they lead to hearing loss
This weekend my friend Seth Burn and I went on a road trip to Baltimore to attend their wild card round playoff game hosting the Indianapolis Colts. It had been about 8 years since my only other time at an NFL game in Denver, at which we sat way up high, none of those in our group cared much about the Broncos even though we lived in the area, and I didn't even know what the point spread was. This time, we were down in the second row among the hardcore fans with season tickets and it was the playoffs. It was quite the experience.

There were three big takeaways from the game: The noise, the comradery and being able to actually see the field up close.

The noise was, I was told afterwards, about average for an NFL game. This is not a quiet level. It is, in fact, a level I am confident would cause season ticket holders hearing loss. Part of that is that it's fun for people to make a lot of noise; for the most part I stayed quiet except for joining some righteous chants of "BULLSHIT!" and by the end of the game had a headache of fully known origins.

The comradery was awesome. The entire stadium was a sea of purple, with tiny spots of blue. Everyone in the area knew each other and had the same seats as always; we were the interlopers who didn't belong, but once they knew we were on their side, it wasn't a problem. There was huge crowd pressure to get in the game as much as possible, and even new superstitions created. The woman on our left observed that us sitting down correlated with the Ravens offense working properly, so she called on all of us to sit down from there on in when the Ravens were on offense. I hope that one only lasted for the one game, but who knows? There were many celebrations, and late there were many of those previously mentioned righteous chants of "BULLSHIT!" against the referees, who the crowd turns on fast. To be fair, there was a taunting penalty, which is more or less the definition of a dumb call.

Actually seeing the field was a refreshing change, although the angle did mean that while some things were easy to see up close with high bandwidth, other things were at bad angles so when the teams went to the other side of the field I more or less reverted to watching the JumboTron. What I did notice, early in the first quarter, was that the Colts were not disguising their offensive plays. At all. Whenever you line up for a play, you're balancing having people in the right places for what you intend to do with the need to line up people in other places so the defense is kept guessing and has to defend against things you do not intend to do, and devotes less resources where it counts. The Colts were unwilling to do any of this. If you looked at the field up close, it was obvious in a way that it's not clear on TV that there was no real way for the Colts, on most plays, to do anything except the thing they ended up doing. Sure, they'd toss one guy off to the side - sometimes, but far from always! - but the formation was obviously throwing to a pair or trio of guys in one place, when it was throwing. When it was passing, they'd line up the receivers close enough to the rest of the line that they could function as linesman, but only in one direction, while on passing plays they got separation so they could get open.

I don't know if this played a large part in the Ravens victory, but it would not surprise me in the slightest if it did. There were some obvious (and a small number of creative) things yelled at Colts QB Andrew Luck, only some of which were obscene (hey it rhymed!) but no one seemed to notice him putting his cards face up on the table.

In terms of the experience, the biggest problem was definitely the noise level. Alas, there isn't any way to do anything about it. Defenses want lots of noise, so it's necessary for the fans to supply that noise, because fans care about winning and the whole experience is about going to the stadium and being part of the team largely by creating noise and getting mad at the refs whenever they don't cooperate. We must protect this house! The fans, if anything, have trouble obeying the signs that say "Quiet please, offense at work." There are multiple homes of "The twelfth man" because fans like to think they're louder than other fans. And no, ear plugs would not be something you'd want to be caught using. Thus, the experience is what it is, and I think that once every few years is about right for me.
Thursday, January 3rd, 2013
6:51 pm
...because they don't have good marketing
Recently I read the book Secrets of Question Based Selling: How the Most Powerful Tool in Business Can Double Your Sales Results. There's nothing like a little sales instruction to make one lose what little faith one has in humanity. Since I wouldn't wish such an act on more people than necessary, my summary of the book is before the fold.

Thank you, good to be here.Collapse )
Tuesday, January 1st, 2013
11:15 am
...because of a race for social points
Yesterday, I found a chart about US health care costs that claimed to show that our outrageously high costs don't kick in until age 50 or so, at which point we steadily go crazy at an increasing rate, with no jump at 65; I'm not resharing it here because doubts have been raised since then about the data involved. I showed it to Laura, as relevant to discussions we were having.

What she did then, of course, was to share it on Facebook.

I don't use Facebook except for viewing events and sharing contact information. It's toxic. It eats your life and gives little in return. However, it forces others to follow, which in turns forces others to follow. This is how everyday villains ruin lives. You see, because Facebook is what everyone looks at, because they're addicted and because it's what everyone else looks at, it means that whoever posts to Facebook is the person who found the new hotness. This results in the person scoring Social Points. This, in turn, means that if there is something that has yet to be shared to Facebook, if you tell anyone in any other form, all they will do is go ahead and post it to Facebook. Since that's what's going to happen anyway, if you want the credit, you have to post it there first, which reinforces the problem. The same thing happened with the issue of the letter grades on NYC restaurants; a day later I'm seeing people sharing it on Facebook. The evil genius is that if I share on Facebook first, then they share my link, which means I still get originator credit, whereas if I link here then I don't get the credit.

This, of course, means that all anyone ends up caring about is what the original person wrote because people reshare the link obliterating the discussion and what discussions do exist are in unthreaded unwieldy messes where everyone goes around worrying about how many people will "like" their comment and then going back to check on that and thus wasting their lives even more than usual.

Thus the norms are reinforced, and things get progressively worse and instead of creating nice things we're all posting links to Facebook to fight for social points and status, because if we don't do it than someone else will do it first and the whole point of life after all is to have us all go around liking things because no one should ever say anything negative about anything or anyone ever because that wouldn't score people any social points and even might make someone momentarily feel bad. And That's Terrible.
Thursday, December 20th, 2012
8:24 am
...because they're not "safe" and therefore illegal
That's right, a B-grade is a sign of a quality restaurant. I knew it. I freaking knew it.

Score one for my naive empiricism. Anecdotal evidence wins!

I'd noticed some curious B-Grade ratings. It started with Billy's Bakery and RUB (Righteous Urban Barbecue) which are two of my favorite places to eat. It didn't make sense that two of the places I loved most were being given B grades, especially when I couldn't find any places I didn't like that had that rating. I kept a close eye on that, and kept noticing that the places with such ratings seemed to be above average on every other metric, and never once did they appear to have any actual health issues.

It's old news to most of you by now, I'm sure, but now we know why. The next time someone uses this as a good example of government intervention, they need to look elsewhere.

The flip side to that is that I've seen a few C or worse grades, and universally I understand where those are coming from without even having to go inside; I did once out of curiosity, and am curious no longer. This is likely due to a signaling equilibrium. If you get a B rating, it could be a slip on something trivial, or it could be you actually fighting to offer quality food instead of freezing and zapping everything into oblivion. If you get a C rating, it often enough because you're not running a clean kitchen that nice places will do whatever it takes to avoid getting a C, which makes a C a very bad sign, thus forcing a lot of freezing and zapping into oblivion, and/or additional costs that are passed along.

And of course, the one time I saw a notice that the place had failed and was being closed, it was a Domino's Pizza, so that's an example of good government intervention. Alas, like all good things, it came to an end, and they were allowed to reopen. Bastards.
Friday, December 14th, 2012
1:10 pm
...because we don't plan for them in advance
I spend the bulk of my time in my apartment. I love this place. When I last set out to move, I knew what I wanted:

I wanted a desk outfitted with my giant monitor setup in the bedroom, room for a carefully selected queen size bed, and a television that could be seen from both locations, so I could watch it at my desk or lying in bed. Only later did I realize how important it was that I be able to see the monitors from the bed as well; the current setup is good at this, but not great.

I wanted the living room to serve a few different purposes. The television had to be centrally accessible, across from the best couch I could find. The area had to be able to function as a hosting location for Magic drafts. Finally and perhaps most importantly, I wanted the place to work well for dates, and potentially work for two people at some point in the future.

And of course this was New York City, so the place had to work with a minimum of space.

Overall the place did an excellent job, but times have changed and in a few months I'll be moving again, although staying in NYC. The new design will have many of the same requirements, but also new ones as well as a lot of lessons learned. It'll need to be a place for two people from the start, the other person in question bringing the desire for a lot of stuff, and room for additional new people. That's by far the biggest change. But some other requirements/lessons:

1. It's very important that all desks and other places people sit face towards the center of the room. It is horrible to spend time facing a wall, and it cuts you off from anyone else. The extra space is worth it. Speaking of desks, the new place will need two.
2. Keep the things you want on your mind visible, other things invisible. The games closet shouldn't be hidden.
3. The main table needs to be big enough for three on each side; even though in theory you can play the third match somewhere else, this is kind of terrible. This time there should be room for that.
4. Not having bookshelves was a noble goal but it didn't play well with others, and won't play well with the need to surround children with the right books. Planning for these will be necessary.
5. Wires are far more unsightly and annoying than you'd think. One must plan ahead for these from the beginning.
6. Take care of the blinds right away, and make sure you can actually make the room outright dark.
7. One of the main goals of a main room is to fit as many people into a conversation or two as possible, and give them all places they can reasonably sit down.
8. Have recharging stations... everywhere.
9. Grand plans early, or they won't happen, such as the idea to have duplicate monitors in other rooms. If it's to be done right, it needs to be there at the start.
10. Things to check: Water pressure and available temperatures, available power (I've blown through some circuits), power outlet locations (see wire issue), A/C and heat, water filter (on refrigerator if possible), bath size, time to outside (plus general area of course), maid and laundry services, package handling.
Let's see, what else...
Monday, December 10th, 2012
10:49 am
...because scope insensitivity doesn't cancel out
There is the problem of people pushing downside risks, especially unlikely but large ones, onto taxpayers, shareholders or others. If you win, great, you've won, and if you lose big, whoops, oh well, sorry guys.

There is the problem of people not capturing their own upside, since they don't own the business they are working for, and since people can't really properly exploit true big wins, but facing a big loss when being blamed for things.

Most people in large groups, in fact, face both at the same time, and they very much don't cancel out. This is scope insensitivity at its most pernicious, and prospect theory having its revenge on us all. Gains are capped in perception of others, and capped for real somewhat higher than that, thus they are capped in reality. Small losses can lead to the largest possible loss, so any loss isn't that different from the largest one.

These are people's actual incentives. It's not you having a "bias" when you're properly predicting your payoffs! The system is biased, and the people are adjusting to that, which is exactly what they should be doing.

We can't fix the downside risk problem because our society has decided it's unethical to inflict large downside risks on people. We don't boil the cheaters in oil these days, let alone boil their children along with them, because that would be wrong. We can't grant the upside risks either, because that would be increasingly inefficient, and even when we try to do it anyway, all it usually means is that a bunch of money is being spent to accomplish very little other than victories in status competitions.

I keep updating towards this being a bigger and bigger problem and have very little idea what to do about it.
Sunday, December 9th, 2012
11:15 am
...because we need to show we can wait for them
Children who can postpone gratification do better in life. There are many possible explanations for this. It's possible that having discipline leads to success, it's possible that other factors that cause success also cause or correlate with discipline. Some amount of both is almost certainly true.

Alas, the consequences go well beyond losing half your marshmallows.

Earlier this week, I bought blackberries, which were requested so that they could be put into yogurt in the morning. Alas, this never happened, because the person in question was sufficiently short on time and impatient that the berries were consumed on their own before the morning came.

The quote came unprompted: "This is why I can't have nice things."

As a result, there will be no more berries, because if that's what's going to happen to them, they're not worth buying.

What happens when the child goes to interview to get into the best possible primary school, and they put two marshmallows in front of her? The payoff isn't just one versus two. It's one marshmallow versus two... and a leg up on her entire future. This will happen again when she interviews for a job, tries to make a friend, goes on a date, you name.

Once discipline becomes a signal, discipline becomes far more important than it ever was before, thus forcing us to engage in bigger and bigger feats of discipline, so we can send a costly signal that we have discipline, and suddenly that's eating all of our consumer surplus... from life, or at least childhood.
Saturday, December 8th, 2012
9:35 am
Two days ago, on twitter, I asked the following question:

Zvi Mowshowitz ‏@TheZvi
Survey question: Randomly selected 12yo American decides to devote next 5 years to getting into Harvard. Tries for real. Chance of success?

This was after asking Michael Vassar as a similar question, although there I worded it “What do you think is the chance that, given a kid at age 12 (let's say) who decides to dedicate their life to getting into Harvard, that they get in?”

[You may want to think about your own answer before looking at his.]His response was: VERY high, assuming any creativity, decent information, and an IQ of 120+, but them getting in would both demonstrate and create lots of human capital.

That got me curious enough to ask on Twitter.

Here were their answers:
It's long, and the conversation is in reverse orderCollapse )

[Summary]There were answers of 0.1% that thought sure, maybe it’s 1%, then there were a bunch of people at 75%+, even one at 90%, plus a few who think that the meaning of try is too ambiguous in context to answer the question.
It’s great fun when the range of answers is so wide, and both sides stick to their guns with high confidence.
It might also be telling me to stick to Twitter. Hard to say.
Friday, December 7th, 2012
10:58 am
My friend Oscar always says that I'm a big turkey
One of Taleb's best metaphors is that of the Thanksgiving turkey, and how important it is not to be one. I love the metaphor, and want to take it farther.

A turkey is born. Every day, it and its fellow turkeys are fed, housed and otherwise treated well. As the days go by, the turkeys become more and more convinced of the good nature of their situation. The turkey population rises and grows fat and happy.

Then one day, as Thanksgiving approaches, the Turkeys are killed and become dinner.

Absence of evidence is not evidence of absence, except when it is evidence, and in the case of very rare but large events it is poor evidence indeed. How, then, Taleb asks, can we be the antifragile people who when the special day comes dine on delicious meat, and avoid being the turkey, fragile to the unpredicted rare event of Thanksgiving, for which it most certainly will not be giving thanks? How do we prepare for unknown unknowns?

I can't argue with that...Collapse )
Thursday, December 6th, 2012
9:12 am
...because people don't understand statistics.
Prompted by one of my commentators, let's cut to the chase of The Black Swan, since that's necessary before one can tackle Antifragile. In fact, one of the great reliefs of reading Antifragile is that it assumes you've read or understood the previous two books; I am so sick of not finding properly advanced reading material that builds upon rather than repeats!

Arthur summarizes the point of The Black Swan this way: The technical reason is that you can't tell - even given a somewhat large sample - if there isn't an unseen mode many standard deviations away. Assigning a probability of o(1/N) for X=x is foolish for sample sizes O(N).

One can never rule out, from a finite amount of data, anything that was (or should have been) in the prior. The less likely something is, the less useful data is at reducing that probability. It is a mistake to fit the model to the data without retaining the prior and giving probability mass to rare events, even if most of them will never happen. One should assume that over a similar amount of time, it is likely that something will happen that is more extreme than any so-far observed event.*

People make this mistake all the time, and it causes a lot of damage when the thing that Can't Happen, happens, with this problem becoming bigger over time. Some of them have incentive to make this mistake because they don't have to pay for the damage, and some of them do have to pay but make the mistake anyway.

So far, so good, but there's a huge difference between people's naive prediction algorithms being wrong, and prediction being impossible or useless. Predictions are hard, especially about the future. Prediction is also both vital and inevitable. One must have some probability distribution for events whether or not one wants to admit that or say it out loud.

It is very important how probable a "Black Swan" is, if you care about whether or not there is a Black Swan. Taleb's answer is more or less that the answer is "more." It happens more than anyone is willing to admit. Fine, perhaps that is mostly true. It doesn't tell you the right course of action. How much more likely? Of what form? Taleb's actions guard against or profit wonderfully from certain types of possible rare events, but break when faced with other events that seem even rarer, leaving him even worse off than everyone else; he admits this, in saying that things that are Antifragile are only Antifragile up to a point.

Betting that the stock market will collapse is a strange bet, because the events that collapse it most effectively also mean you won't get paid your winnings, or won't have any way to spend them. Taleb, as it turns out, won such a bet and made a lot of money, but he doesn't have any way to usefully spend it. He gets more utility out of having been right, than in having the money, and in fact in Antifragile he illustrates this with stories of other people in the same position. At that point, how good were the odds you really got, and does it matter whether you had a "good bet" in terms of dollar Alpha/EV? It seems very easy for Taleb to gain utility by making bad bets on volatility, and it seems very easy for him to lose utility by making good bets.

This is, of course, like everyone else he's criticizing for three books. They make bad bets in dollar terms that, given the person's situation in life, are actually good bets. Sometimes they do it on purpose, sometimes not. Those who he praises often are doing the same thing.

* Note that power law distributions are common, and often very good at probabilistic prediction of events that Taleb calls Black Swans. Explicit examples from his books include 9/11 and the recent large earthquake in Japan; both were, as Nate Silver shows in The Signal and the Noise: Why Most Predictions Fail but Some Don't, right on the otherwise expected distribution lines. Taleb calls the war that destroyed his home in Lebanon a Black Swan, but it was the 8th time the place had been turned to rubble. This was the Middle East, after all.
Wednesday, December 5th, 2012
12:57 pm
Why We Can't Have Nice Things: Because They're Fragile?
Nassim Nicholas Taleb’s Antifragile: Things That Gain from Disorder, the conclusion to the Black Swan trilogy that started with Fooled by Randomness: The Hidden Role of Chance in Life and in the Markets and The Black Swan: Second Edition: The Impact of the Highly Improbable: With a new section: "On Robustness and Fragility" is a deeply, horribly flawed but fascinating and valuable book.

The progression is striking.

In Fooled by Randomness, Taleb has a well-defined, true and important thesis to prove, and proves it: That most of the time, those who claim to have beaten the market, or the odds in general, only succeeded due to blind luck. Past performance isn’t only no guarantee of future success, it’s not even correlated. Back testing a model doesn’t mean much for the future, and everyone is cooking the books to look good and attract investment by exposing themselves to large downsides. The book is concise, tight and on point. If you’ve studied a bunch of rationality then You Should Know This Already, but even then it’s fun and somewhat useful to watch him go through the details.

In The Black Swan, Taleb gave himself free rein to ramble, tell personal stories and espouse points of view without statistical evidence for them, while claiming to be smarter than everyone around him. Instead of a making a narrow claim backed up with very strong evidence, he instead made a much broader claim, that the important things in life are rare, unpredictable events that he calls Black Swans; the origin story of the term is that for a long time everyone thought all swans were white, until someone saw a black one. Rare and unexpected events are both far more likely than people think, and far more important than people admit. He starts labeling everything a Black Swan. The book is a mess, aggravating at times, and clearly overreaches, but provides a rare perspective that’s worth taking into account. Whether or not it’s a good idea to read the book is less clear.

The core concept of Antifragile is that since random events and disorder are inevitable (Hail Eris!) the most important thing for anyone and anything is to be antifragile, benefiting rather than being harmed by large random events, or at least be robust and indifferent to them. Anything that is instead fragile, that is harmed by random events, will eventually hit the wrong one and die no matter how strong it appears, whereas those things that are antifragile use adversity and random events to grow stronger.

Antifragile is to The Black Swan as The Black Swan is to Fooled By Randomness. Antifragile is maddening. It is arrogant. It is completely unfair. It is hysterical and repetitive. The editor has clearly been fired, and there’s even a story explicitly about that. It has huge gaps of reasoning, and calls people losers and suckers in situations where on reflection they are clearly winning. It lays bare the author’s true motivations and utility function for all to see, and they aren’t pretty. It reveals a severe and seemingly willful misunderstanding of rationality and probability; its core epistemology doesn’t even make any sense.
Despite that, I am very happy I am reading it and have the urge to spend time writing and talking about it, because the good stuff here is worth it and the bad stuff is there undisguised for all to see, analyze and learn from. It feels necessary to write down and talk about the ideas here in order to properly process, understand and benefit from them.
Tuesday, December 4th, 2012
6:22 am
Why We Can't Have Nice Things: Educating Those At Risk
"F*** you, paper author, you do not share my utility function."

Bingo. First comment wins.

I don't share the comment author's utility function either. That's the point.

Sexual activity is a nice thing. People are telling other people not to have this nice thing, by convincing them it would be bad, because it involves becoming At Risk. And That's Terrible. We need to teach people exactly how At Risk they are, by giving them sexual education that explains the risks, and then they'll only do it when it's appropriate!

Drinking is (according to many people's reports, anyway) a nice thing. People are telling other people not to have this nice thing, by convincing them it would be bad, because it involves becoming At Risk. And That's Terrible. We need to teach people exactly how At Risk they are, by giving them education that explains the risks, and then they'll only do it when it's appropriate!

Eating calorie-rich food is a nice thing. People are telling other people not to have this nice thing, by convincing them it would be bad, because it involves becoming At Risk. And That's Terrible. We need to teach people exactly how At Risk they are, by requiring calorie labeling, and then they'll only do it when it's appropriate!

The person you are trying to reach doesn't share your utility function. Given their utility function, better information will usually cause them to make better choices. If giving them better information doesn't cause them to make the choices you want them to make, perhaps they're not maximizing what you're maximizing.
Monday, December 3rd, 2012
9:05 am
Nice Things: Why Can't We Have Them?
Nice things, whether or not they are physical objects, are awesome. They are much better than things that are not nice, or not having things at all.

There are some people who dispute this. They tell us we cannot have nice things.

There are religious people, many of whom claim that God doesn't want you to have particular nice things, or even any nice things. Sometimes, they promise you nice things later on in exchange, but tell us to take their word on that.

There are the ascetics and the moralists, who think that nice things distract one from the true path.

There are the socialists, who think nice things are created regardless of incentives. There are the victims of akrasia, the lazy, and others who think that nice things should happen and be shared with them, without their help.

There are the egalitarians, who think that if you didn't bring enough nice things for the whole class, then you can't have them either.

There are the advocates of social justice, who won't let you have nice things unless everyone is happy with every element of them, and trade would be cheating.

There are the regulators, who think nice things are against the rules.

There are the central planners, who direct all the things however they think is best, usually destroying most of them in the process.

There are the thieves and murderers, who think that they want your nice things, so they're going to take them.

Then of course there are the oligarchs, whose most important goal in life to make sure no one has nice things but them and their children. There are the status seekers, who want more nice things than you and don't much about the absolute numbers. There are those who think that if you have nice things, it means that they don't and are losing, and that can be fixed by destroying the nice things, after which they win. Mostly, of course, there are those who simply don't care at all about nice things that aren't theirs, or plan for the future, and ruthlessly turn to extracting what nice things they can, for themselves, no matter the consequences.

Know thine enemy. Know thyself.
Friday, December 10th, 2010
7:40 pm
Main Game Fall 1904: War Continues
All west as expected, and now it's West vs. East as the race to Italy begins. Chances are that one of the two sides breaks but it could be a long time. I won't comment on this game until something interesting happens.
Wednesday, December 8th, 2010
5:59 pm
LW Main Game: Spring 1904: A Lot More Straightforward Than It Looks
We now have two clear sides: G/E/I/R against A/T/F... for now. The game master is being coy about the situation but it's hard to be clearer than that. Turkey can take the Ionian Sea and Austria can knock out Venice with the risk being Germany sneaking Trieste. I don't expect either side to try anything on the eastern front for a while. The question is, will Italy pick up enough from his new friends to survive? France is toast and will be down to only Liverpool after this year, at which point he's only a nuisance. England and Germany's task now is to get as far into the Med as possible and meet the Turks as far east as they can. Granting Italy Spain and Mar for this turn is potentially a good idea, but trust is a huge factor there, but sending the extra unit west was costly; as E/G I'd much rather he stayed behind and held the Ionian. A/T has grown large and it's not going to be easy to take them down. This game is probably going to be a long, hard slog sooner rather than later.
[ << Previous 20 ]
Star City Games   About LiveJournal.com