Relevant and even prescient commentary on news, politics and the economy.

Confessions of a Late to the Party Russiagate Non-Skeptic

I don’t think this is worth your time, but when I post only at Dan pulls the post over here so I might as well.

Like many many (too many) people I am irritated by this article by Blake Hounshell.

Hounshell claims to have “doubts about whether Donald Trump colluded with Russia” I have a couple of questions. One is: Which word in “Trump colluded with Russia” didn’t you understand ? The answer is “colluded” which is repeatedly redefined in the article so it can dodge a hail of facts and remain unproven. The other (which I actually asked him) is:if in 2014, I had described the Trump tower meeting & preceeding e-mails (without inventing names) and asked “would you call this ‘collusion'” what would you say ?

There would be a follow up question if his answer were “no”: OK now describe an imaginary scenario which you would call collusion.

I think it is not possible for a mainstream journalist to note that we know that Trump jr colluded with Russians. The rules of journalism can’t all be followed in the present case in which the President regularly asserts things which are demonstrably false. There has to be a debate about whether collusion will be detected, even though it has been detected.

I think Hounshell’s interest is in whether there is a story out there. The fact that the story has already been told right in here where we all heard it is irrelevant. The question is whether there is a big scoop to be made. Another important question (which Hounshell barely addresses) is whether the collusion constitutes the crime of conspiracy and whether the colluders will be prosecuted.

I decided to write this post to point out an amusing contradiction in the article. Hounshell argues that, if there had been collusion, Trump would have openly blabbed about it since he has no self control & also that Donald Trump’s open collusion on TV isn’t collusion because it was open blabbing. I quote

Then there’s the Trump factor to consider. Here’s a man who seems to share every thought that enters his head, almost as soon as he enters it. He loves nothing more than to brag about himself, and he’s proven remarkably indiscreet in the phone calls he makes with “friends” during his Executive Time—friends who promptly share the contents of those conversations with D.C. reporters. If Trump had cooked up a scheme to provide some favor to Putin in exchange for his election, wouldn’t he be tempted to boast about it to someone?

[Biiig skip]

There is, of course, plenty of public evidence that Trump was all too happy to collude with Putin. “Russia, if you’re listening, I hope you’re able to find the 30,000 emails that are missing,” springs to mind …

The fact (noted six (6) paragraphs later) burns Hounshell’s argument to the ground, scatters its ashes to the nameless howling winds and sows salt where once it stood.

(I am quoting Noah Smith quoting someone else).

Hounshell’s argument rests on Catch 22 — Trump can’t collude without discussing the collusion in public and public requests can’t be collusion themselves or correspond to private collusion because they are public.

It may be true that Trump couldn’t conspire with Russia without talking about it in public, but it is certainly true that he talked about it in public. Again the question, can Hounshell imagine a scenario less devastating for his argument. Hmm well collusion typically involves a quid pro quo not just a request. Lets run the tape

“Russia, if you’re listening, I hope you’re able to find the 30,000 emails that are missing. I think you will probably be rewarded mightily by our press.” He said this after Papadopoulos was informed by the Russians that they possessed Clinton-related emails.

Hmm “rewarded mightily”. Note the “by our press” makes no sense. Respectable newspapers don’t pay for stories and disreputable ones don’t have access to sums which would seem mighty to the Russian government. I think the “by our press” locked the door after the horse had escaped the horses mouth.

So what’s going on ?
1) Maybe Hounshell is Fukayaming
2) Someone suspects he is stuck in the pose of a cynical sophisticate who has seen worse before and is amused by how excited the non-cognoscenti are.
3) He is thinking of whether a journalist should invest in an effort to find another smoking gun. It would be very hard to top the two which have been smoking away for months. So there is no professional gain there — only crimes against democracy.

I think point 3 is the important one. There will be no dramatic proof that Trump is a crook for the same reason that there will be no dramatic proof that water is wet.

Comments (37) | |

A Review of A Review of Enlightenment 2.0

I am very glad I read Alex Tabbarok’s review of “ENLIGHTENMENT 2.0: Restoring Sanity to Our Politics, Our Economy, and Our Lives,” by Joseph Heath. Tabbarok is a leading ultra pro market libertarian and “Joseph Heath is a Canadian philosopher who is unusually conversant with economics and also unusually capable of writing sparkling prose for a popular audience. ”

I have some thoughts. I advise the reader to click the link, and recklessly attempt a very brief summary of the book and the review.

1) Heath says our intuition and heuristics are unreliable, so we should try to use reason and logic.

Tabbarok says Heath focuses to much on logic and too little on empirical investigation.

2) Heath notes that advertizers gain by exploiting heuristics and biases, that we are constantly bombarded with alleged information and that this is making things worse.

Tabbarok notes that capitalism rewards rational producers. He hints at a market selection argument (I think he is careful not to spell it out *again* because he has done so too often for one lifetime).

3) Heath notes that the GOP has become extremely dumb and crazy.

Tabbarok discusses collective action problems a lot.

OK my comments

Heath also glosses over the fact that in the modern era measured IQ scores have risen, not fallen. IQ scores have risen especially in tests of abstract reasoning ability. Just the kind of ability that German social philosopher Georg Simmel argued was promoted by capitalism. Simmel argued that increasing familiarity with money as an abstract, representative unit of value improved abstract reasoning and rational calculation in other spheres of life. This kind of abstract thinking is also what makes people capable of understanding the idea of universal human rights.

First I note that another German social philosopher Karl Marx made the same point (no cite — its an aside on the origin of Protestantism in the Communist Manifesto). Here money abstracts value from the particular characteristics of goods and services — it encourages us to think of goods & services in isolation and treated equally. It encourages a view of people as individuals shorn of clan, guild, and (after centuries) sex and race. This has its costs but does help peopl understand the idea of universal human rights. I think history tends to support Simmel’s hypothesis.

Another possibility is that markets are unstable and disruptive — some rich people become poor and some poor people become rich. That churn disrupts traditional hierarchies. Hereditary rank ceases to be almost perfectly correlated with economic power. I’d tend to guess it was wealth => power => new political institutions => new habits of thought and new ideologies designed to praise and defend the new establishment.

On IQ tests I have another concern. The old Otis IQ tests includes simple logic puzzles. Simple algorithms for solving them can and have been taught. The aim of the test is to detect general cognitive ability. Our scores depend on what we have been taught (I would not score very well on IQ tests where the questions were written in Chinese). IQ like SATs affect lives. The problem, as always, is we can’t tell whether test scores correspond to that which the test was designed to measure. They can be increased by teaching to test. The new knowledge and skills may be narrow — useful for the test but not generally useful. In the extreme, the knowledge can be “the correct answer to question 3 is b to 4 is d …”, that is there is a grey area between teaching to test and teachers cheating. The test is valid so long as people don’t have an incentive to score high.

Tabbarok has a crazy idea of using betting markets to make public choice decisions. First he notes a fact

When pollsters ask Democrats and Republicans factual questions such as did inflation fall during Reagan’s presidency or were weapons of mass destruction found in Iraq, they answer in a highly partisan manner. But partisan bias greatly diminishes when voters are told that they will be paid if they answer correctly. Betting is a more reliable guarantor of objectivity than voting. Or, as I once wrote, “A bet is a tax on bullshit.”

Then he discusses a proposal to base collective decisions on betting markets.

But once we decide on values, let us then bet on beliefs. Do we agree that lower unemployment would be a good thing? Ok, let’s then bet on whether fiscal policy or monetary policy would best achieve that goal and let’s then go with the betting favorite. Economist Robin Hanson terms this “rule by prediction markets,” or futarchy. Is betting on policy a crazy idea? Perhaps so

I think the proposal is somewhere along the spectrum, serious proposal-provocation-joke, but I will take it seriously.

Here notice Tabbarok has no problem with rich people having more power over public choice (he likes markets and doesn’t mind inequality). The proposal is based on ignoring (not clearly rational) actions of rich people who use their money to get the political results they want. Now it’s campaign contributions and such, but with a betting market it would be direct. If there were a market on “the affordable care act will cause a dramatic decline in the number of uninsured” the odds paid for guessing yes would have been huge, because the Kochs would have bet however much they had to to make no the odds on favorite.

Here again, a measure is reliable exactly because it isn’t important, because it isn’t used to make a decision.

Now it may be that there is a collective action problem in betting in a way such that there are expected losses in order to affect the market odds. But in the real world, people do act collectively and serve a cause beyond their narrow personal interests. Also the actual market is not populated with huge numbers of people each of whom has wealth trivial compared to the total amount of money in a specific market. There are big players and they throw their weight around.

OK I feel a bit like a fool for taking a joke seriously. Why go half way ? click more for my crazy suggestion

Comments (1) | |

Mueller et al Declared that there Was Collusion

The post entitled “Trump’s Claim Mueller Found ‘NO COLLUSION’ Is Literal Nonsense” is not up to Jon Chait’s usual standard. Trump’s claim is, of course, nonsense. Chait accurately described Trump’s typical pathetic rhetorical trick “One of President Trump’s favorite methods to defend his innocence in the Russia investigation is to claim that any piece of evidence that does not explicitly assert his guilt is in fact evidence of his innocence.” and added some high quality snark “It is exactly like saying Trump was cleared by the Warren Commission because the Warren Commission report makes no conclusion about Trump and Russia.” However, he misread the indictment.

Chait (and many many others) concedes that the indictment didn’t declare that collusion has been detected “this particular indictment probably has nothing to do with collusion. ” In fact the indictment declared that collusion has been detected. it didn’t name all of the conspirators, but the grand jury did definitely claim to know of conspirators who were not named in the indictment.

I quote (with a pdf warning and my emphasis)

From in or around 2014 to the present, Defendants knowingly and intentionally conspired with each other (and with persons known and unknown to the Grand Jury) to defraud the United States by impairing, obstructing, and defeating the lawful functions of the government through fraud and deceit for the purpose of interfering with the U.S. political and electoral processes, including the presidential election of 2016.

The indictment explicitly states that only some of the “known” conspirators have been indicted. It doesn’t say whether any of the unindicted conspirators worked for the Trump campaign, but it definitely also doesn’t say that no crimes Trump campaign workers have been detected in the investigation into internet trolling (let alone the broader investigation).

The indictment explicitly states that there are known unindicted co-conspirators. It does not address the question of whether one is, say, named Donald Trump, even within the narrow limits of the investigation of “INTERNET RESEARCH AGENCY LLC” and its employees.

I mean which word in “conspired with … persons known … to the grand jury” didn’t he understand ?

update: This post by Marcy Wheeler is incomparably better than my little post. Read it.

One point from Wheeler– a US citizen conspirator who was not indicted has been named. Richard Pinedo was not indcted because he pleaeded guilty on February 7th. The indictment can’t be construed as stating that the Grand Jury does not have probable cause to believe Americans were wittingly involved in the conspiracy. At least one definitely is known. He wasn’t indicted because he confessed.

Wheeler wrote

In the wake of that indictment, the court unsealed a February 7 plea agreement with Californian Richard Pinedo, for identity theft (basically, selling bank account numbers; the information doesn’t identify the users who purchased the bank account numbers as IRA personnel who used them to set up “American” identities, but that is clearly what happened).

update 2: I have a thought. Wheeler wrote

Plus, Mueller likely obtained cooperation from one IRA employee, the unnamed person who traveled to Atlanta in November 2014 for reconnaissance. Had that person not cooperated, he or she would have been named in the indictment.

This is probably true, but I think that I have thought of another possibility. It is possible that Atlanta traveller wasn’t indicted because he or she didn’t commit a crime. Two women were indicted for lying on their visa applications saying they were in the USA for tourism not reconnaissance. But reconnaissance isn’t a crime. It sometimes called journalism and sometimes called market research. If Atlanta traveller claimed to be a journalist, he or she is in the clear. The US government can’t decide who is and who isn’t a real journalist (just as it can’t decide who is a real clergyperson) because everyone has a first amendment right to claim to be a journalist (the first amendment doesn’t say anything about citizenship — foreigners have free press rights too). IRA wasn’t seeking classified information — they wanted to know what Americans were saying about the election. The reconnaissance was journalism which is legal except for people who lied claiming to be tourists.

So maybe someone did something totally legal, can’t be indicted and wasn’t named because he or she has a right to some privacy while observing us.

Comments (22) | |

Behavioral Bitcoin

Bitcoin prices are an attractive topic for people who study behavioral finance. Behavioral means anything but rational expectations, Nash equilibrium and the Efficient Market Hypothesis. It is easy to argue that the fundamental value of Bitcoin is zero — it doesn’t yield income and there is no limit on the supply of cryptocurrency, because new cryptocurrencies can be introduced. I certainly consider the positive price of Bitcoin to be a failure of the efficient market hypothesis.

The case is also attractive, because the price certainly seems to roughly correspond to what one would expect based on one of the oldest beharioral models. For decades psychologists and experimental economists have noted that people tend to usually assume that time series are mean reverting — guessing that what recently went up will soon go down. People are then surprised if a variable goes up and up and interpret this as something other than random fluctuation, as a new trend. Basically, the story is that people treat a random walk as if it were a stationary process around a broken trend. This isn’t totally surprising as time series econometricians find it exceedingly difficult to distinguish random walks from stationary processes around broken trends.

There are many models in which agents switch back and forth from forecasts in which they predict mean reversion to chasing trends — forecasts in which they expect a variable to continue to change in the direction it has recently changed. In the models, agents switch based on the past performance of the two forecasting rules. Hands are waved so that all don’t switch at the same time and people never become firmly and permanently convinced that one rule is best.

I wrote a little MatLab program in which such agents trade an asset (with no particular fundamental value) & simulated the asset price a few times. Of a few dozen trials one looked like this ( remember you have to click more to see figures)

Hmm does that look familiar ? The price goes up to a peak around 100,000 times the initial price. It is stable for long periods and occasionally shoots up and crashes (you know bubbles). The periods of stability are periods in which most agents forecast mean reversion. The periods of extreme instability are periods in which 90% of agents extrapolate. The upper limit of 90% extrapolators is one of 5 free parameters I fiddle (the others are variance and persistence of unexplained shocks to price and sensitivity and persistence of beliefs about which rule works best).

As noted above (and as is obvious) the graph is cherry picked. It is the one of a few dozen which looks most like the actual time series of bitcoin prices (here fairused from

Comments (0) | |

Glenn Greenwald & the Nunes memo

I’m not sure this is of general interest, but I would like to argue (again) with Glenn Greenwald. In this tweet, he asks an interesting and important question

The FBI, and many Democrats, insisted vehemently that release of the Nunes Memo would endanger national security. Now that we’ve all read it, is there anyone who believes that this argument was even remotely true or honest?

Yes. This has been another episode of simple answers to simple questions. Now I will bore you by explaining at length.

1) One can’t conclude that something wasn’t endangered because, it the end, it wasn’t harmed. It is reckless to drive 150 mph drunk even if some people have done so and arrived alive. I don’t recall anyone saying that releasing the memo would certainly harm national security.

2) Importantly, the expressions of alarm (including the DOJ not FBI use of “extremely reckless”) came from people who had not seen the memo, who had requested a chance to examine it and whose requests had been denied by the committee. They didn’t know what was in the memo, but they knew it had been written by people who had access to classified information, that they didn’t know what was in the memo, and that it was proposed that it be released so everyone knew as soon as they did.

The argument that this violates normal procedures which are required to protect national security is clearly conventional — almost so conventional that it goes without saying. The procedures are followed not because every deviation has catastrophic consequences. That’t the way standard procedures are.

I don’t recall all the alarmed statements by Democrats, but many were made by Senators and such who had not been allowed to see the memo.

3) Finally releasing the memo clearly harmed US National Security. I get the impression that almost everyone but I has the impression that the memo didn’t contain information which was supposed to be kept secret (according to normal rules which are enforced on people who aren’t President or the majority of a House committee by the threat of prison).

There are two data in the memo which had been secret and which were kept secret for excellent reasons
a) October 21 2016 — the date of the application for a FISA warrant to surveil Carter Page. This was not a request for a renewal. Now, but not last week, I can infer that Page was not under FISA surveilance say in September 2016. I didn’t know that before. If I had conspired with Page on the phone during September 2016, was asked about it by the FBI and had to decide whether to lie to them, this information would be very useful to me. it is discussed only as proof that the fall 2016 FISA surveillance of Page was *not* surveillance of the Trump-Pence 2016 campaign, since he had severed all formal links with the campaign in September. But it would also be useful if there were someone who really wants to know that the FBI knows about his or her communications with Carter Page before October 21 2016.

b) The warrant was renewed at least three times. This is discussed becuase one of the requests for renewal was approved by Rod Rosenstein and because the fact that four requests were approved is strong evidence that the surveillance revealed Page’s participation in foreign intelligence efforts. But the information would be very useful to me if I had conspired on the phone with Page in December 2016. I would know that they know about it, so I would risk prison were I to lie and deny the activity.

In spy vs spy intelligence and counter-intelligence hiding all sorts of information from the other side is key. Who is being wiretapped is a closely held secret for obvious reasons. My point is that there are similar reasons to hide who has been wiretapped (including Page as was known before the memo was released) and when they were wiretapped, If one is under investigation, it very important to know what the investigators know. If one is not supposed to obstruct justice, one should not make that information public in the name of transparency.

OK so what is going on ? I think that Greenwald has become a knee jerk critique of Democrats. Also he has long had a very sincere extremely negative view of the FBI. I think his reflexive opposition to state surveillance has caused him to automatically reject arguments based on the idea that FBI investigations are sometimes in some ways socially useful.

He isn’t a consistent anarchist, but he seems to automatically oppose state power. Thus he seems to actually support complete trasnparency in investigations. This would make wire taps worthless. I think Greenwald automatically opposes them.

On the other hand, he is a brilliant lawyer. He should understand why the facts revealed in the memo had been kept secret. Almost everyone agrees with him that no secrets were revealed by the memo. I’d guess most people just don’t understand the issue. But I guess that he understands it and is so opposed to serveillance of any kind that he genuinely can’t see how anyone would see any disadvantage in hampering it.

Comments (25) | |

(88% of a Sample of) Republicans Helpfully Make it Clear That They Don’t Care About Accuracy

Jason Schwartz buries the lede in this genuinely alarming article in Politico “Study: Americans view media negatively, can’t agree on meaning of ‘fake news'”

I attempt to excavate it.

The study — the 2017 Gallup/Knight Foundation Survey on Trust, Media and Democracy, based on mailed-in responses from more than 19,000 Americans age 18 or older — asked people to rate whether four categories of information were “Always,” “Sometimes” or “Never” fake news.


Asked to rate “Accurate news stories casting a politician or political group in a negative light,” Democrats said 26 percent always, 50 percent sometimes, 22 percent never, while Republicans replied 42 percent always, 46 percent sometimes, 10 percent never.

10% of Republicans agree that news stories can’t be both “accurate” and “fake”. I admit the views of Democrats are also alarming, but, as a resident of the rest of the world outside of the USA I sure am glad those US Republicans don’t control nukes.


Also Orwell had a bit to say about the political goals achieved through the destruction of language. Unlike this post
Politics and the English Language
are worth reading.

Comments (0) | |


I am going to make a fool of myself by suggesting that a cryptocurrency might actually be useful. Bitcoin et al have negative social utility. They are pure speculative assets which enable people to gamble. Also bitcoin miners use as much electricity as Denmark. The problem is exactly the aspect which has made bitcoin famous and which bitcoin enthusiasts consider a strength — the enormous increase in the dollar price of bitcoin. This increase, and the recent sharp decline, make bitcoin useless as a means of exchange. Most firms don’t want to gamble.

So I (semi-seriously this time) propose botcoin which might have a more stable dollar exchange rate. The idea is to link the blockchain verification program to an official exchange.

Backing up, there are two very different sorts of web-servers related to bitcoin. One set, the bitcoin miners, implements the original idea using the Bitcoin shareware. They keep a copy of the ledger of all bitcoin transactions — the blockchain, race to create new blocks, and evaluate new blocks and add valid new blocks to the chain. The other servers are bitcoin exchanges in which bitcoin is traded for regular currency. They are not part of the original plan in which bitcoin would be traded for goods and services and function as a means of exchange. They have behaved badly with an unstable value of bitcoin (huge unpredictable Bitcoin deflation damages any use of bitcoin as a means of exchange as much as huge inflation would).

I propose linking the blockchain program to an exchange. So there would be an official botcoin exchange (this means it isn’t entirely free-entry shareware libertarian anarchism). If anyone were interested in a new cryptocurrency designed so that speculators can’t become rich (and pigs fly) there would be other unofficial exchanges.

The bitcoin program regulates the frequency of creation of new blocks to roughly one every six minutes. It does this by adjusting the difficulty of the pointless arithmetic problem which must be solved to make a new valid block. The idea was to limit the total amount of bitcoin which will every be created (to 21 million for some reason). This was supposed to make bitcoin valuable. So far it has succeded all too well (I am confident that in the end bitcoin will have price 0).

It is possible to make the supply of botcoin flexible so the dollar price doesn’t shoot up. I would aim at a price of, say, 1 botcoin = $1000. The idea is to make the pointless problem which must be solved to add a block easier if the dollar price of botcoin exceeds the target, and harder if it falls below the target. This should stabilize the price.

Now no one is really interested in cryptocurrency except as a way to gamble and take money from fools. But if anyone were, linking the blockchain program to prices on an exchange would make it more nearly possible to use the cryptocurrency as a means of exchange.

The system is vulnerable to a tacit agreeement to trade only on unofficial exchanges. It is necessary that the problem is also made easier if daily trading volume on the official exchange is zero. The problem is the price could shoot up on unofficial exchanges, but this would not affect the price on the official exchange if there were no transactions on the official exchange.

Comments (5) | |

Bitcoin Prime

Also know as collectors vintage bitcoins.

This is a crazy idea with no redeeming social value.

BitCoin is valuable only because people expect it to be valuable in the future. I am sure there is BitCoin bubble, but it wouldn’t be that strange for for something useless to be valuable. Cancelled stamps may be valuable. Old coins have numismatic value. Old coins hmmm … pity BitCoins are just numbers uh first recorded on a given date … hmmm

OK so I want to invent the Vintage BitCoin blockchain. The idea is that it is a subset of the BitCoin blockchain with only BitCoin awarded to miners before 2011. It would not be very hard to create this.

It is important that it remain a subset of the bitcoin blockchain so updated with edited down bitcoin blocks.

OK before going on here is the Bitcoin wiki (which is pretty much all I know).

To be even briefer, Bitcoin is just the bitcoin ledger, that is, a record of all bitcoin transactions. Each transaction has a payer, receivers and a number (an amount of bitcoin). Bitcoin is a quantity a measure — there aren’t identifiable bitcoins with serial numbers. In a valid bitcoin transaction the payer lists as “input” the transactionw by which he or she got the amount of bitcoin. These must not have been used as input before (you can’t spend the same bitcoint twice). A proposed transaction is sent out over the web and joins the pool of proposed (but not yet validated) transactions.

Bitcoint “miners” check transactions and form blocks of valid transactions. Then they do pointless math which is there just to slow down the creating of bitcoin blocks. The pointless math is that a block must be an approximate root of the bitcoin hash function which I will call F. I have to define a function G such that G(a,b) = F(a+b).

The block is a bunch of ones and zeros. the N+1st block starts with G(the Nth block) (this is important). then there are the valid transactions which were in the pool of proposed transactions. They are described with text and numbers, but it is all 0s and 1s on the web. The miner then claims a reward, so the block includes her identifier and the statement “receives 12.5 bitcoin”. So far the block is a huge number which I will call x. Then there is the “nonce” y which is any number such F(x+y) < epsilon. For very small epsilon it is very hard to find such numbers. It is very easy to check. Then the other miners all check the block. They check that it is made of valid transactions which have not yet been validated and that it is close to a root. The block is of the form (x,y) that is both the useful record of transactions and the pointless nonce are recorded separately. The amount of the award was originally 50 and is divided by 2 every 210,000 blocks. The difficulty epsilon is adjusted so that one block is added every roughly 10 minutes (after 2016 new blocks are made it is adjusted so that it would have taken an expected 2 weeks to make those blocks -- the hash function is well understood but not by me). The awards are the only way new bitcoin is created. To make a vintage bitcoin block, one has to check that transactions trace back to an award for a block made before January 1 2011. Sadly the vintage bitcoin inventor can't award bitcoin for making such a block. They can only be rewarded with a record of another quantity (I call it prime for the part of bitcoin prime which isn't bitcoin) which might or might not be worth anything. To make up for this, the difficulty of making vintage bitcoin blocks can be set to quite easy. Now I need to make a ledger of transactions of prime. I would put them in with the transactions of filtered bitcoin in the vintage bitcoin block (clearly marked). The modified program guarantees a fixed supply of prime which might be worth something. Also the vintage bitcoin might be worth more than plain ordinary bitcoin. This is totally pointless. Really a reducto ad absurdum of bitcoin. There is a rival blockchain to Bitcoin which forked off August 1 2017, so someone has been playing my sort of game (they have a justification, but I think they were mostly playing).

Comments (10) | |