Don't get Zhou Tonged or How to Run a Margin Shop

It's pretty simple really.  People who are trading on margin are gambling.  If you run a casino, you are trying to maximize  your profit.  If you run a margin shop, you have a list of all the current folks and their bets, including knowing various points at which they will have margin calls and liquidations.  It's not too hard to run a script that determines where you need the price to go to maximize those liquidations.

If a lot of people have recently shorted the LTC/BTC trade pair, you want LTC to go up so they will lose their bets.  The only question now is how to best move the market.  I say "the market", but really I mean your market.  It's your shop after all.

How you move the price on your market depends on how much of a whale you are.  Chances are you didn't get into this game solely with a handful of pocket change.  In fact it's much more likely that you approached wealthy backers to put some capital down.  Why not?  The returns here are sensational.

So your script is telling you that if the LTC price goes up by 20% you can give a fair number of your gamblers a haircut.  In comes the stash of BTC and orders are placed, buying up the LTC.  Momentarily there is a spike on your exchange, and not on others.  Ever seen this behavior in markets before?  This thing is simple enough to fully automate.  Automatic market making.  Easy money.

Hypothesis: This behavior of margin shop operators is what drives short term prices in ALL markets from commodities to stocks to currencies.

So far we have a nice straightforward operation, being run with real capital and real orders.  However I've barely scratched the surface here.  Remember that a lot of traders are leaving funds with you.  This gives you additional leverage.  Accounts can be created on your exchange which hold funds that aren't real, i.e. can't be withdrawn.  Your enemies here are the flying fish who can arbitrage your market moves on other exchanges.  Determining exactly how far you want to push this seems like a dark art.  It's my guess that running this thing with full reserve is already profitable enough and not many risk going further.  However, if you see one margin shop's price get substantially away from other exchanges without immediately bouncing back, you know the shop has been caught rigging too far.  Chances are they are suspending withdraws and pulling out the cold storage / rich friend phone calls in desperation.

Don't get zhou tonged!



Top 5 errors in bitcoin reportage - volume 2

A little over a year ago we looked at the top 5 things reporters get wrong about bitcoin.  Things have improved somewhat, but not a lot.  You can still find a few high profile pieces a week written about bitcoin by people who clearly never used the currency regularly or gave much thought about it other than imagining that it could be a topic for their literary portfolio based on their own virtues unrelated to knowledge of the topic itself.

So without further ado let's take a look at the common errors in 2016.

5)  Bitcoin is Energy-Inefficient

This is total nonsense.  The network has no minimum requirement energy.  It could run perfectly well if the whole world only solved a single SHA256 hash every minute (about 10^20 times slower than the current hashrate).  Claiming bitcoin is energy inefficient is like saying that using fuel is wasteful.  Well sure, if you are doing it wastefully!  Bitcoin is designed to run on whatever excess power is available, and what people choose to use to mine bitcoin is up to them.  If they do it inefficiently they lose money and will have to stop, unless of course they are issuing themselves fiat currency and using that to purchase electric power from morons.  On this topic the statement is extra-misleading because it implies that the fiat system and it's gross inefficiencies might be somehow more efficient.  It absolutely isn't!

4)  Bitcoin is dead

This one still making the rounds!  Yup, "interest has died", people have moved on, bitcoin is dead.  Sounds good?  Well the trouble is that anybody can easily check if it's dead because it has a heartbeat.  The heart beats once every ten minutes or so.  Being a public coin, verifiable transactions are public and so we can see the interest and the vital statistics.  They have only showed remarkable continued vitality and growth.  There has been growth in the amount of money transferred per minute, the network fees, the hash rate, the value of a coin, the number of exchanges and businesses using it, in fact any metric I know of.  This is perhaps the highest growth phase of the thing and you want to claim it's dead?

3)  Satoshi was found!

Really, this one doesn't matter in any way, apart from the financial implications if Satoshi does indeed have a stash approaching 1M btc (unclear), but because it has personal interest reporters love it.  There are some who have suggested a scheme in which a fake Satoshi could be trotted out and then convince everybody to convert their bitcoins to Cosbycoins.  This doesn't seem like a worthy scheme to me, but hey people have worked on stranger misdirection ops.  In any case, no Dorian is not Satoshi, Craig is not Satoshi, etc.  Expect more of this nonsense from sensationalist press, including more elaborate hoaxes.

2)  Bitcoin Developers

Many people talk about "the bitcoin developers" as if this specified something or other.  In fact bitcoin already exists and is immutable.  People can develop full node clients, wallets, 3rd party payment channels, and a variety of software and hardware.  However none of these people are developing bitcoin.  The 2008 paper still describes what bitcoin is.  Speaking of "bitcoin developers" therefore makes you look like you don't really know what you're talking about.  It's really even worse than talking about bitcoin accounts or the bitcoin CEO.

And the number one error in bitcoin reportage:

1)  Bitcoin doesn't scale and can't compete with VISA

The basic idea is that the bitcoin network only supports three TX per second while VISA lets you do thousands.  So, what's wrong with that, it's true isn't it?  Well for one, VISA is not a currency.  It's a third party payment service and as such can work with bitcoin just as easily as with any other currency, so the comparison is already broken.  For two, bitcoin supports three publicly verifiable TX per second.  The US dollar supports 0 per second, and VISA happens to also support 0 per second.  Off chain of course you can do as many transactions in bitcoin per second as you like.  Exchanges report hundreds to thousands of transactions per second.  Yes, valued in bitcoin.  This kind of claim is just completely broken any way you look at it.  It's like saying Horses have 4 legs, but Ford Motor Company has no legs.  Hardly a comment indicative of study of modes of transportation is it?

Be sure to tune in next year to see how reporters continue to expose their ignorance.



C sharp some more

We looked at some C# code about 6 months ago.  I was very happy with the way it turned out. I learned the program and I run it regularly, the thing sits nicely in permanent ROM.  Turns out it isn't very difficult after all, even for an amateur noob like me.  So, why not try another?  How about this little C# number?  It has fewer sharps in the key signature (because it's C# minor), and it looks less dense.  Probably easy right?



This thing is nasty.

A little background.

JSB didn't so much compose music as he did discover it.  The pieces go that way not because of a desired effect, tradition, or the will of the composer, but because that's the way the pieces go.  Bach would pry open a door to the other world, step through, grab something, and come back with it.  In the first C# major piece we discussed, he does this in a playful rhythmic way - bouncing back and forth between worlds, skimming the surface of the unknown, contextualizing it to our world, and repeating.  It's not that hard to catch the gist of it.

In the Andante Con Moto in C# minor however, shit gets real deep real fast.  He finds a way to slide into the crack and follows it down down down.  Where the fuck are we?  Everywhere at once, and always moving.  There's no explaining this shit away with "oh it's just an alternating series of exploring the 10th interval in descending 5ths alternating between the left and right hands".  I mean sure, you can try - and there is a good deal of alternation between the hands, as well as some obeissance to Canon and Fugal forms, but you barely notice that.  And you can't deny the raw gnarly nature of this thing - this thing is still "correct" in it's contextual representation of what it is, and in affecting the timed arrivals necessary to stay together and alive, but appears even more outside the borders of the describable.

Or maybe in another 6 months it will seem like old hat after I get it down.  Either way I'm happy.  One interesting thing is that merely listening to this piece doesn't seem to produce the same effect as playing it.  To some extent that is always true but with this piece especially.




Smart Contracts could save the Legal System - but not in the way you think

Smart contracts have gotten a lot of airtime lately.  Ethereum-coin's ability to do interesting things with smart contracts, perhaps (but not clearly) easier to do so than with bitcoin,  has been coincident with a large rise in the market cap of ethereum even though there are substantial issues with the distribution economics and a less tested and more complex codebase as well as centralization due to a supposed corporation being to some degree in charge of the thing.

The basic idea of a smart contract is that the terms of a contract can be executed automatically.  A well designed contract can foresee all possibilities and describe what should be done for each of them.  Then, if it is a smart contract, it can sense which possibility ensued and execute the relevant contingency plan.

The first problems associated with this idea as it came towards fruition appeared in the concept of an "oracle".  The oracle is that entity which can enable events to be accessible to the smart contract.  For example if the contract stipulates that if it rains, you get paid - who decides if it rained?  Somebody has to make the call as to what happened in order for the automatic contract to proceed.  Having a someone that decides takes a bit of the gleam off the idea that the contract is fully automated.

But no matter!  The narrative goes that smart contracts and code are so powerful that they can become law itself - "code as law".  Nice idea eh?  It follows in the footsteps of Hammurabi, of judges following "the letter of the law", which has been so effective at promoting injustice and inefficiency.  But I am getting ahead of myself here.

Can smart contracts really help us with our legal system, which everybody agrees has serious problems (in every place on Earth)?  Yes they can, but not in the way you think.  The way they can help is by failing completely, and forcing us to rethink the whole concept that decisions should be fore-written, cases fit into nicely labeled cubby-holes, and dealt with without examination of extenuating circumstances.

Smart contracts and their failures will teach us that the map is not the territory.  They will teach us that a contract or a legal structure is a guide - a map if you will - and not something to be blindly followed without opening the eyes and seeing the world, any more than you would take a left turn into a brick wall just because your GPS was telling you to.

There is no way to foresee all possibilities, and no way to write down a course of action which future people should follow without exception.  No written instructions should be sacrosanct over the behavior of the living.

HOWEVER, there are exceptions.  Shall we let some teenagers move the control rods in and out of nuclear reactors by hand?  Should we let bankers and politicians issue currency that we use on day to day basis?  Lol right?  What kind of idiots would do these things?  Automation has its place, but this place is decidedly not in the realm of conflict resolution.

And what is a contract or a law other than a method of conflict resolution? Ideally, if all parties agree to the current actions, there is no need to look at the contract or the law.  I mean sure, the million or two political prisoners in occupied North America might not have had any conflict to resolve.  But the original stated purpose of the laws these prisoners have allegedly broken, even under its current auspices, was to be a conflict resolver.  Conflict resolution abilities come through experience, and those who master it need to earn respect of their peers.  Sure, studying previous cases can help.  We need good, far-seeing, adult merciful and empathetic judges.  Smart judges, not smart contracts please!

I digress.  The take-home point here is that like public coin, smart-contracts can give us something to think about and push us to re-evaluate certain traditions.  This is sorely needed.



The rise of the bitcoin hash rate - an analogy for hash smokers

How can we visualize the rise of computing power on the bitcoin network?  We need something more common in people's lives than hashing power.  So, how about: hash?  That's right, hashish.  Lets go with it and see if the analogy helps. We'll look at the hash rate of the network and compare it to a smoking hash rate.


In 2009, bitcoin started smoking hash.  It had a modest habit, 1 gram per week.  Nobody would consider this to be a lot, maybe a joint or two one weekend night and a little touch here or there during the week before work.

Jan 2009 - 1 gram per week -  5 MH/sec

This went on for about a year, from Jan. 2009 to Jan. 2010, at which point bitcoin started slowly smoking a bit more.  By spring of 2010, bitcoin was still smoking modestly, but had picked up pace quite a bit puffing almost every day before work and also more frequently on weekends.  By early summer bitcoin was smoking 4 or 5 grams a day.

June 2010 - 30 grams per week - 150 MH/sec

In July bitcoin met a friend named Artforz that taught him how to use a bong.  This allowed bitcoin to seriously pick up the pace and start consuming a more generous portion of hashish.  Bitcoin always shared with his friends who were more numerous now as well.

Sept 2010 - 600 grams per week -  3 GH/sec

At this point bitcoin was basically chain smoking hash all day, in joints and bongs.  It might sound like a lot but this is really only about an eighth of an ounce per hour.  Bitcoin was definitely a heavy smoker, but still didn't stand out too much from the crowd.  It doesn't take that much talent to smoke an eighth in in a hour.  However, this was only the beginning.  Bitcoin continued to collect more bongs, and more smoking friends.

Jan. 2011 - 18kg per week - 100 GH/sec

At this point bitcoin was up to roughly a hundred grams per hour.  Bitcoin was definitely high all the time, but still not up to the standards of some celebrities like Lil Wayne.  Bitcoin was rolling up massive dubies and filling bongs nearly simultaneously, as well as using vacuums to "hot box" large rooms.  However, this was still not enough hash for bitcoin's dome.

July 2011 - 1800kg per week - 10,000 GH/sec

At this point bitcoin reached a saturation.  Bitcoin was now smoking 3 grams a second.  Imagine this video every second or so.  Bitcoin was the biggest supersmokeuter in the world by a very large margin at this point.  For the moment, the dome was full.  Bitcoin kept this up for a year or so, at which point a friend came over with a new toy: A sick oil rig.  With this "a sick" rig, new amounts of hash could be consumed.

July 2013 - 18,000kg per week - 100 TH/sec

At this point bitcoin was smoking 30 grams a second.  That sounds like a lot of hash doesn't it?  Even Cheech and Chong would be impressed.  But bitcoin found new ways to increase it's dome further, with better and better "a sick" oil rigs coming out, bitcoin always got the very best vaporizers.  At this point bitcoin moved into the superdome and kept it hotboxed 24/7.

July 2014 - 100,000 kg per hour - 100 PH/sec

A hundred thousand kilos per hour is a lot of hash by anyone's standard.  And it all started with a modest 1 gram per week habit.  At this point bitcoin stood up and stretched and said "I'm not high at all".  More oil rigs were ordered and new grow operations initiated.

July 2015 - 300,000 kg per hour - 300 PH/sec

That's about 85 kilos per second.  Seriously puffing tough here, bitcoin was going through the total annual output of Afghanistan's famous black hashish in just a couple days or so.

June 2016 - 1.5 million kg per hour - 1.5 EH/sec

Bitcoin at this point is puffing a few grams every microsecond.  Many planets of course are needed to supply this kind of cannabis habit.  At this point every person on the planet is puffing with bitcoin.  With 6 billion people, that's about a quarter gram per person per hour.  No sleeping!

Lets recap: bitcoin went from rolling itself a few joints a week, to rolling itself and 6 billion people a few joints an hour.

Or as Bob Marley didn't say:  "I smoke two joints, then I smoke two joints, then I smoke 2 billion more in a few seconds".

We could have just said bitcoin went from 1 MH/s to 1 EH/s.  It's only 12 orders of magnitude.  From one gram a week to one trillion grams a week.  What's 12 orders of magnitude amongst friends?


It's all genetic no need to worry

A while ago we showed that intelligence is caused by big feet and obesity by genetics.  In today's lulz we have more "correlation is causation!", in splendid comic form.

Take a look at this post from a "genetic counselor".  This thing struck me as noteworthy for a couple reasons.  The first thing you are likely to note is the avoidance of the subject at hand.  This is de rigeur in the modern medical establishment.  The author has been guided into this world of "the reader, the patient, the customer, are all illiterate scum who could never and should never try to understand anything".  This is why there's no discussion of what DNA is, what the procedures of sequencing are, how they can be interpreted, etc.  Because you're a fucking moron, get it?  Go ask your doctor, idiot.  The fact that you tried to even read this thing is already disturbing.  Isn't there a Sports Illustrated lying around or perhaps a movie or color matching game on your phone?

So instead of talking about the subject at hand, we have some nice people-friendly anecdotes:

We studied their DNA.  Nothing.

Gotta love that suspense and educational material there!  Haha.  Anyway if you're interested you can read this paper, the publishing of which appears to be the "success" referred to in the blog post.  Here's a highlight:

In summary, our findings further document a relationship between the 40 kb duplication upstream of GREM1 and HMPS in Ashkenazi Jews. Genetic counselling and appropriate molecular testing can lead to successful management of HMPS patients.
Declaration of interest: None
Note the goal here:  management of patients.  Note a tentative loose assignment of chumps for future milking.  Note the finding: "a relationship".  Gee, you think big feet are also related to intelligence amongst Pennsylvania-Dutch?  Obviously I'm going to have to heavily obfuscate that study to get it considered for publication here.
Anyhew, lets cut to the chase.  What else is correlated with colon cancer?  What is indeed the most strongly correlated factor with all cancers?  What factor was forcibly removed from the Danish population during WWI which led to a substantial drop of all cancers?  What dietary factor is found to be a requirement for laboratory animals to develop tumors even after being given known-carcinogenic aflotoxins?  What can you tell about this kid's diet based on the photo (male pattern baldness)?  If you don't get it, there's another photo at the bottom of the blog post to give you a hint.
Yup! It's GREM1 genes my friend, go have another FDA approved roast beef sandwich.  Patient successfully managed.  What epigenetics?

The Professor

"The Professor" (1986) (Giuseppe Tornatore) is a film that..  just kidding!  What you think I'm going full orc and doing movie reviews here?

No no no.

The professor that I am reviewing is a 5x5x5 cube made by ShengShou called the LingLong.  It's a white stickered cube and I picked it up because it seemed like a good cube available dirt cheap (less than 2 bitcents) and as such it was my first 5x5x5.

First off, the thing is smooth as fuck.  While I imagine there are better speed cubes out there, and of course a stickerless design is always more durable, this thing has greatly impressed me.  No pops, no sticks.  If you don't have one, why not?

The 5x5x5 being an odd cube has fixed centers, and avoids some parity solving issues which can be annoying on the 4x4x4.  It is also not a supercube, which means that chances are it will never return to the same exact position it started.  This is because some pieces in a face can swap each other out, so that the cube will be solved but the actual position of the original pieces is now different.  Make some markings on a cube to verify this yourself.  Neato!

Solving larger puzzles provides new interesting content in a couple of ways, but doesn't get exponentially harder with puzzle size like e.g. juggling does.  First of all there is the "hunting" aspect, which is good for getting you to scan areas looking for that one piece you are looking for, and which cause bigger puzzles to take so much more time.  Time likely is exponential with puzzle size, but not difficulty.  For me, the 5x5x5 seems a great medium.

Then there is the fun of getting the last faces done.  Lets back up a bit: the standard route for these things is to get the faces first, then getting the edge pieces matched (not in the right place, but matched together forming complete edges), after which you have something equivalent to an unsolved 3x3x3 and you can proceed to solve it.  The first face is easy because you have lots of freedom to move pieces about.  Just like with the 3x3x3 however, once you have gotten some pieces where you want them, this takes some freedom away to move other pieces about.  The last two faces must be solved simultaneously, every move balanced by a countermove to return the solved faces to where you want them, and it's always a pleasure to see how they wind up falling into place.

The crux and perhaps the most fun bit is getting the last few edges matched.  It turns out you will be getting the last three edges matched simultaneously.  This is easy to see on the cube but hard to explain.  Basically you need to put a piece of one not-yet-solved edge (1) into another not-yet-solved edge (2) to create a solved edge.  Then you need to get that solved edge out of the way, replacing it with a not-yet-solved edge (3) before undoing the move which you used to solve the edge - said undoing now not only fixes the faces you have messed up in the procedure but also all edges.  The geometry is beautiful here, very often you will see things just fall into place magically as there's nowhere else for the pieces to go after all.  In fact this part is nice enough that you might be tempted to not finish the cube after this step, instead messing it up again 🙂

Anyway if you can solve the 3x3x3 you shouldn't need any new algorithms for this cube, until it's time to start shaving tens of seconds off your time that is.  It's pure group-theoretical geometrical fun with over 10^74 permutations. Happy cubing!