I watched episodes 1 and 2 of the Secrets and Lies tv show.

so, this guy finds a (very recently) dead body on his morning run in the woods and then starts answering questions from a detective – where was he last night, did he move the body, why did he go for a run that morning, etc

the detective even starts tripping him up on the details. did he get home at 2am or 3am – which was it? did he have a couple drinks like he said he wanted to run off, or more?

he goes along with all this, is helpful and tells her stuff, and then is taken off guard that he’s a suspect. he needs a lawyer to advise him to stop volunteering help and instead tread carefully.

people are very naive about this stuff. DON’T TALK TO THE COPS, YOU WILL NOT BENEFIT. and if you find a body or were near a crime, YOU ARE A SUSPECT.

and lots of people are bad at their jobs, incompetent, stupid, etc. which includes cops. i am not saying cops are especially bad. but if they are just the normal amount of bad you find everywhere, TREAD CAREFULLY. they might think you did it for no reason or a stupid reason, or just because of your body language or tone of voice or they don't like your subculture's linguistic style.

and it’s so easy to accidentally contradict yourself in minor ways when you answer questions about the same thing multiple times. especially if you say anything before going over it carefully in your head for hopefully a few days and talking about it with a lawyer. people are not in the habit of being 100% precise and never contradicting themselves, it isn’t required in most social situations.

hollywood, by presenting talking to the cops as just what normal innocent non-weird people do is sending a really bad and dangerous message. they don’t preach like “you should talk to the cops, do your civic duty”. instead they just frame it as completely normal and something to take for granted. instead of trying to debate the issue, they send a message without raising it as an issue to debate. be wary!!

don't try to be polite. if a crime happened, stay out of it. don't talk to cops without a lawyer. don't try to be helpful. you're putting yourself at risk. if you have important info, call in an anonymous tip.

even if you think you're completely safe, e.g. you were home with your family and saw something out the window or heard something, and you successfully keep your story 100% straight with no contradictions, you are not safe talking to the cops. if anyone else says something contradictory (by accident, or because they are guilty and lying, or because they suspect their friend might be guilty and want to cover for him, or whatever else), you become a suspect. anything you say can be used against you. the only way to avoid someone else contradicting you, and raising doubts about you, is to say nothing.

oh and, of course, don't let the cops into your home if they don't have a warrant. seriously. your life is at stake. yes the risk of getting randomly involved in a crime you didn't commit is low in general. it doesn't happen every day. but by the time the cops are trying to talk to you, the odds aren't so low anymore, so take it seriously.

later in the episode, the detective says she wants to ask him some more questions but is actually just trying to get him out of the house. then while he's gone, his wife gives the cops permission to come in and search the house, without a warrant. sigh :(

then because he's being harassed by aggressive reporters who make his child cry, instead of calling the cops of them (which is what I'd suggest, especially considering they went on his property, but even if they hadn't), he gives the detective a DNA sample to try to prove he didn't do it in order to be left alone. very bad strategy.

then he asks his wife about letting the cops in. she says 1 police officer came by and next thing she knows there are 10 of them. he asks how come she didn't say no. she says, "how could i do that? they were investigating the murder of a child". BAD REASONING. DO NOT LET COPS SEARCH YOUR HOUSE WITHOUT A WARRANT. EVER. PERIOD.

(and keep in mind there are so many things that could go wrong with cops doing searches that aren't even related to the case. like maybe your kid has some drugs hidden in his room that the cops find. in the show, letting the cops in to search visibly upset the family's children – the mother failed to protect her kids and let harm come to them.)

as the show continues, the guy's life is getting screwed overly merely for being investigated from the crime (not charged with any crime, certainly not found guilty). the community starts shunning him. he does work like painting houses but no one wants to hire him anymore. being innocent does not protect him from this. and if the detective was being biased and unfair, or incompetent, he'll never be paid back for the harm done.

and even at this point, he agrees to answer questions from the detective without having a lawyer present.

near the beginning of episode 2, he's asked to go answer police questions at the station again. he asks don't they need to go through his lawyer? they cop says they don't because he isn't under arrest.

don't be fooled by crap like this. you don't have to answer police questions without a lawyer.

then the main character offers to take a polygraph to try to prove he's innocent. this is the worst idea yet. NEVER TAKE A POLYGRAPH. they work badly and are unreliable. (you can google info on this, and on talking to the cops. if it ever comes up, at least don't talk to them until you have time to google more info on these topics for a few hours. i don't expect to 100% persuade you, but i hope i get you to think twice enough to not answer initial questions and then look for more info to decide how to handle it.)

people think by being polite and obeying social norms, it will ensure the cops treat them decent in return. it won't. issues like crimes should be handled objectively, with standard procedures, not by social convention, and to a reasonable extent, they are. stick to standard procedure yourself – cooperate in ways you are legally obligated to and that's it. you get no official bonus points for being extra helpful, and you won't get you out of any legal obligation.

cops are not on your side. they are not there to help you. they do not work for you. you aren't their boss. you don't pay their salary. don't be naive.

There are rare exceptions if you know what you're doing, for example if you see someone discard a weapon that was used against you into a trash can, you might want to point the cops to that trash can so they can find it before the trash gets taken out. see http://blog.suarezinternational.com/2014/08/afteraction-discourse-what-to-say-to-the-cops.html even this kind of thing can be dangerous, e.g. if you say "i thought i saw him throw a knife on that roof, it was dark though" and the cops check and don't find it, now you look like a liar (even though you hedged). and if you say anything you can easily make mistakes. life gets dangerous and scary and risky once cops are involved like this no matter what you do. be careful, be slow and thoughtful. maybe write stuff down and read from that if you want to say anything at all, so you have an exact record, or print it out and give it to them. but, really, read about these topics, and "don't talk to the cops" is basically the main thing to know. i kind of don't want to mention any exception whatsoever because 99% of mistakes people make are in the "talk too much" direction.

Marth's reverse Dolphin Slash (up-B) is an important technique which people tell you to learn how to do. They're right. But I tried to do it, and I couldn't. There are a couple key things I figured out that really helped. I want to share them.

The inputs are simple. You do up-B, and then during the startup frames (a very small time window), you press left (if you were facing to the right). This press to the left has to be done very fast. I won't discuss why this technique is useful, other people have done that. I just want to talk about how to do it.

Also, just to be clear, you can face right and hold up-left, and then hit B, and you will do the Dolphin Slash behind you and turn around. None of the information I've read is really clear about this, but I'm pretty sure reverse Dolphin Slash is different and requires doing it the hard way of up-B first and then press behind you second, separately.

At first I thought the problem is that my hands are slow. I'll just try it more and try to do it super fast, and then hopefully I'll get it. Well, I didn't get it. I went in Training Mode and tried in slow motion to make sure I was doing the inputs right. It worked. But at regular speed I was hopeless.

Then one day, I had a thought. You know what would save time? Don't push the dstick all the way up.

So I tried doing up-B, all by itself, without pressing the dstick all the way. And I found you only have to press it a tiny bit further than for up-tilt, but really not very far. Only a fraction of the way up is far enough.

The main reason I couldn't do it is because I was pressing the dstick all the way up, then pressing it to the side. And that takes too long. Maybe if you play in tournaments and you're really good, you could press it all the way up and still be fast enough. But I sure can't.

Well, once I had this insight, I was able to do reverse Dolphin Slash successfully about half the time in only 10 minutes of practice.

But I didn't just start doing it. I practiced an intermediate step that I think was a really good idea. If any guide had told me to practice it this way, it would have really helped me.

Press the dstick up half way. Hold it there. Now if you hit B, you will Dolphin Slash. Try it. So now instead of pressing up-B for dolphin slash, you start with half the work done, you just have to press B. Now do this: press B then, almost at the same time, press left (if facing right. press behind you).

When I just tried to hit up-B then left, it was so hard, I couldn't do it. But when I held up and then tried to hit B and left, it was so much easier, I could do it pretty much right away. It's not that hard to do one thing with your right thumb and one with your jump thumb, and do them very close together. Doing two things with your left thumb and something with your right, and coordinating the timing, that's hard. But only one thing with each thumb isn't too hard.

So practice that a bunch and you can learn the timing of when to hit left relative to when to hit B. Without a bunch of stress and failure. You can learn part of the technique by itself without having to be able to do the whole thing.

Once you're good at that, then practice the dstick motion without B. Press it up only a little of the way, definitely not all the way up, and then jam it left hard and quick. And practice it to the right also.

When that feels OK, then try another small step. Press up a little ways, pause for a split second, then press B and left. So it's like doing it with up already pressed, but instead of just holding up and not thinking about it, you do the up press only a moment early, so it isn't totally separate.

Once you can do that, then try to do the whole thing. And because of all the little steps you did, I bet you'll be able to do it sometimes. Not all the time, but sometimes. And once you can do something 5% of the time, then you have a good start and you just practice more and increase that percent. Whereas if you can't do it at all, it's hard to get started and you'll need some easier steps.

So you press up a little ways and B, and then hard left. It won't work every time. You'll get some neutral B (Shield Breaker) and some side B (Sword Dance) at first. But now you should have a good enough idea of how to do it that you can practice until you get it consistent. These little steps to work up to it will get your foot in the door and make the technique approachable.

Again, I'd like you to learn not just how to reverse Dolphin Slash, but also how to approach learning anything that's hard to get started with. This is both a specific example that will help Marth players, but also it's about the method of how to learn.

---

For part 1 at my blog, click here.

For all parts, and some people's helpful replies, see my thread at Smashboards.

Super Smash Brothers Melee (SSBM) is hard. And it's hard to get started. I've read a lot of guides and tips. A lot of the info is very helpful. But I think most of it is way too advanced for most players.

I'm not very good at SSBM, but I think most people are probably a lot worse. No offense. I've played games from a young age, I've played a lot of games, I've played a large amount, and I've been very very good at some games. And I started playing Smash before SSBM came out. Not very well, but I've been familiar with Smash for a long time, and followed it much more closely than most fans.

I've been practicing SSBM. Mostly tech skill, alone. I like the game, I like understanding how it works, I like seeing how hard it is and facing a challenge, and I like having a better understanding of what the pros I watch in tournaments are doing, what it's like for them.

I have figured out some ways to practice that are more basic than are usually taught, and I think they could really help people. For example, people say to practice Marth's SH (short hop) double fair (forward air attack). But I can't do that. It's really hard. To some people, it's just the basics. But to me, it's an advanced skill that's going to take a lot of work. My hands have sped up a lot from practice, but I still have a long way to go to SH double fair.

So how do you work your way up? What's in between nothing and SH double fair? My main point in this post is to show you how to break down a technique, like SH double fair, into a bunch of intermediate steps you can practice one by one. Even something pretty simple can be divided into a lot of different things to practice, instead of just being all-or-nothing.

(And for my regular philosophy audience, take note: you can apply similar methods to many other topics outside of gaming. Treat this as a detailed concrete example which illustrates an important philosophical method, and see what you can learn about philosophy.)

- SH

Start with SH alone. To SH, just hit jump and let go fast (before you're in the air). Don't feel bad if you suck at it. I would stand there and hit jump and do nothing else, and Marth would full hop. It took me a ton of practice just to SH. Actually, first I learned to SH Peach, who has an easier one than Marth. Marth is 3 frames, Peach is 4, Fox is 2. Almost all the characters are in one of those three categories. If you have trouble, practice with a 4 frame SH character first. Here's the list of how many frames each character has for short hopping (smaller numbers are harder, meaning you have to let go of jump faster).

One of the cool things I found is, after I practiced Marth's SH a lot, even when I still wasn't very good at it, then when I went back to Peach she became easy. And then once I practiced Sheik's 2 frame SH, and went back to Marth, then Marth felt easier. But you can't move up too early, just starting with Sheik wouldn't have done me any good if I can never get it at all.

- SH While Distracted

As an aside, let me say that being able to stand still and do a SH, and being able to do it while playing the game against an opponent, are different things. As one example, once you can SH ok, try to run forward and SH. You'll miss some because of the distraction. Once you get better at that, try shield stop SHs. That means you dash forward, then very quickly hit shield, then very quickly after that, short hop. Even once I was good at SHing in place, I couldn't do shield stop SHs without some practice. Learning to link together the things you practice makes them harder.

The point is, don't get frustrated if you thought you could SH, but then you try to do SH and something else, and suddenly you can't SH. It's going to happen. It's no big deal, you just need more practice until your ability to SH is less barely and more solid.

- SH Nair

Once you can SH, try to SH Nair (neutral air attack, meaning A with no direction). Hit jump then A. You'll probably miss some SHs from trying to hit A also. Don't worry, practice, you can learn this.

Now to the main point: if you jump and then hit A fast enough, you will land without going into a recovery animation from the nair. The best way to see this is get the 20xx Hack Pack and turn on the flashing red and white for failed and successful L cancels. If you SH nair and you hit A slowly, you will see Marth flash red. If you do it fast enough, Marth will not flash any color.

When I started, I couldn't do this. Marth would flash red. Maybe I could get it 10% of the time. But, again, you practice and you get better. This is a hell of a lot easier than SH double fair. It's a smaller step forward. This will get your hands faster while being a smaller and more achievable goal.

- SH Fair

Next, try to SH fair. If you do this quickly, Marth won't flash red. You have to be a little faster than with SH nair. (If you don't have 20xx hack pack, you'll have to try to watch Marth and visibly see the difference between whether he does his recovery animation from landing during fair, or not. Which is a skill that takes practice. You can learn it early if you have to, but I'd really recommend getting the 20xx pack.)

- SH Uair

Next, try to SH uair (up air attack). Again, you'll have to be a little faster. You'll also have to learn to press the dstick (directional stick, the joystick used for moving) lightly so you don't double jump.

- SH Bair

If you can go even faster, you can do a SH bair (back air) with Marth and land without flashing red. If you do it successfully, Marth will turn around (so this one is easy to tell if you succeeded even without the 20xx Hack Pack).

- C Stick

Then go back through and practice all of these using the cstick (the little yellow joystick) instead of A. (Except not nair, you can't nair with cstick). Again this makes it harder. But it's possible, and with practice your hands will get faster. (As I write this, I can just barely bair with c-stick on a small percentage of attempts. And one really interesting thing I noticed is I can do it a lot easier to the left than the right. After hitting jump, I can press cstick left faster than right. The only reason I can tell the difference is because when doing the SH bairs, that tiny difference actually affected my results because I was so borderline on being able to do it at all. I think that's pretty cool to find that out, and gives me useful information, and potentially something to practice. For example, once I can start to do some SH double fairs with cstick, I'll have to practice to the left first which will be easier so I can have success sooner. And once I can do that a little, I'll have to practice to the right also. Doing it to the left first will be a little easier, another step I can practice before doing it to the right.)

- SH, Fair, Double Jump

Next, try to short hop fair, then as soon as you start the fair, start mashing jump. If you're fast enough, you'll double jump instead of landing. You can also try to learn to press jump at the right timing instead of mashing.

Once you can do that (I can only do it 10% of the time as I write this), try to SH fair with cstick and then get the double jump (I can't do that yet).

- FH Double Fair

Practice doing full hops and then doing fair twice. The point here is to learn the timing for how soon you can do the second fair after the first one. It's not something that's hard, but you do need to practice and learn that timing. Practicing it separately will be helpful. You should also practice other aerials this way just to learn really accurately when you can do a second one. Learning how long your moves last is important and worth practicing for each move individually.

- SH Double Fair

Then, finally, after you progress through all those steps, you can work on SH double fair. That means you do a SH, then you do fair twice before you land. To succeed at this, you need to do the first fair extremely fast after jumping, even faster than any of the things you practiced above. Then you have to do the second fair with good timing as soon as it's possible.

To do a SH double fair correctly, you need to be 6 frames faster than SH, fair, double jump. Fair can hit the opponent on the 4th frame through the 7th frame. Double jump comes out in 1 frame (I think). So suppose you SH, fair, and then you double jump on your last frame in the air. To do a second fair instead, you'd need to be 6 frames faster so you'd have 7 frames of airtime left instead of 1. Then you'd be able to replace the double jump with the second fair and have enough time for it to fully complete the part of the move that can hit the opponent.

---

The point here isn't just to teach you to SH double fair with Marth. The bigger point is to show you how to practice things step by step and work your way up, a little at a time. Instead of failing to SH double fair over and over, it's better to gradually start with something a lot easier and then keep progressing to slightly harder things. It's a lot more fun to practice when you're learning new things, successfully, as you go along.

Whatever you want to learn, for whatever character, try to figure out a series of small steps that can help you build up to it. Commonly people recommend pressing the buttons slowly at first and then speeding up. That is great advice but there's other ways to practice too.

All the information in this post, I basically had to figure out myself (except the frame data). No one told me to try practicing bairs fast enough I would turn around. But I find it really helpful as an intermediate step. I hope some Marths find this helpful, and also everyone understands the method of creating a gradual progression of small steps to practice. Most melee training information doesn't cover little things this basic, like I never ever heard anyone say "practice doing SH fair fast enough you land without going into recovery from attacking", but I think it's a really useful idea. So hopefully this will encourage a lot of really new players who are struggling. By breaking things down into smaller steps like this, you'll be able to see your progress and succeed one step at a time.

---

For part 2 at my blog, click here. It provides another example with the same philosophical point.

For all parts beyond 2, and some people's helpful replies, see my thread at Smashboards.

Israeli Prime Minister Benjamin Netanyahu said:

Today the Cabinet will be briefed on the security challenges developing around us, first and foremost Iran's attempt to increase its foothold on Israel's borders even as it works to arm itself with nuclear weapons. Alongside Iran's direct guidance of Hezbollah's actions in the north and Hamas' in the south, Iran is trying to also to develop a third front on the Golan Heights via the thousands of Hezbollah fighters who are in southern Syria and over which Iran holds direct command. The fact that Iran is continuing its murderous terrorism that knows no borders and which embraces the region and the world has, to our regret, not prevented the international community from continuing to talk with Iran about a nuclear agreement that will allow it to build the industrial capacity to develop nuclear weapons.

... The agreement that is being formulated between Iran and the major powers is dangerous for Israel and therefore I will go to the US next week in order to explain to the American Congress, which could influence the fate of the agreement, why this agreement is dangerous for Israel, the region and the entire world.

This is very important. Obama wants Israel to be destroyed, and is actively pursuing that agenda, and most Americans don't recognize it. And Obama is far from alone in this matter.

I look forward to Netanyahu's speech and seeing the reactions. Maybe he can talk some sense into America. I really hope so.

Most statements which are sufficiently deviant (from cultural norms) are assumed to be jokes by default. This is a way of protecting everyone from admitting that serious disagreements exist.

For example, if you say, "Thank you so much, you've persuaded me and I've learned a lot. I will completely rethink all my values and take on board the moral values you've shared with me." that reads as likely sarcasm because it would be much more rational than typical people in our culture.

And if you sound significantly less rational than the typical person, it again doesn't read as serious. For example, "I hate you for trying to use logic to share ideas with me that you think would help me. I'm very mad that you could be so arrogant as to think you could know anything useful to me. Did it ever occur to you that I don't want to think?" People will assume someone doesn't really mean that and is just making a joke, perhaps an exaggerated parody to imply the other guy is wrongly treating him like the person in the parody.

Statements which are reasonably normal are taken at face value, but statements which are deviant are frequently not treated as real statements in the usual way. This is a way of denying the existence of deviance and generally suppressing disagreement and pretending it doesn't exist. It's a strategy which helps people irrationally refuse to consider many disagreements and criticism.

Capitalism: A Treatise on Economics is free at this link.

The theory of marginal utility resolved the paradox of value which had been propounded by Adam Smith and which had prevented the classical economists from grounding exchange value in utility. “The things which have the greatest value in use,” Smith observed, “have frequently little or no value in exchange; and on the contrary, those which have the greatest value in exchange have frequently little or no value in use. Nothing is more useful than water: but it will purchase scarce any thing; scarce any thing can be had in exchange for it. A diamond, on the contrary, has scarce any value in use; but a very great quantity of other goods may frequently be had in exchange for it.”

The only explanation, the classical economists concluded, is that while things must have utility in order to possess exchange value, the actual determinant of exchange value is cost of production. In contrast, the theory of marginal utility made it possible to ground exchange value in utility after all—by showing that the exchange value of goods such as water and diamonds is determined by their respective marginal utilities. The marginal utility of a good is the utility of the particular quantity of it under consideration, taking into account the quantity of the good one already possesses or has access to. Thus, if all the water one has available in a day is a single quart, so that one’s very life depends on that water, the value of water will be greater than that of diamonds. A traveler carrying a bag of diamonds, who is lost in the middle of the desert, will be willing to exchange his diamonds for a quart of water to save his life. But if, as is usually the case, a person already has access to a thousand or ten thousand gallons of water a day, and it is a question of an additional quart more or less—that is, of a marginal quart—then both the utility and the exchange value of a quart of water will be virtually nothing. Diamonds can be more valuable than water, consistent with utility, whenever, in effect, it is a question of the utility of the first diamond versus that of the ten-thousandth quart of water.

I'm not satisfied with the way this passage explains the issue (I don't think I have a disagreement with Reisman about economics here, though). It doesn't mention supply, demand and competition, which I think are crucial. It leaves a reader to wonder: why not charge the value of people's first quart of water? Sure you wouldn't sell the marginal 10,000th quart at that price (they'd find the price higher than the utility), but you could potentially make more profit, with less water inventory, at a higher price. The reason is because other water sellers would compete with you and that keeps prices down, not because of marginal utility. This answer's Smith's question about how something so valuable can be cheap.

Water isn't just cheap because of mariginal utility (the utility of the 10,000th quart is much lower than of the first quart, lowering demand after prior quarts of water are supplied), it's also because a large supply is cheaply available from nature. If the question is about starting a new business to acquire and sell additional water, than the marginal utility of water is the relevant issue. But if the question is about the cheap pricing of current water, I think prices would be raised (despite that meaning fewer quarts sell) if not for competition. (BTW I'm ignoring issues like government regulations and the real-life water situation, and just treating it as a free market commodity.)

If I'm mistaken about any of this, then I'd still say the passage's explanation is unsatisfactory, because in that case it apparently didn't adequately clarify matters for me.

Very soon thereafter, the whole Circle Bastiat, myself included, met again with Ayn Rand. We were all tremendously enthusiastic over Atlas. Rothbard wrote Ayn Rand a letter, in which, I believe, he compared her to the sun, which one cannot approach too closely. I truly thought that Atlas Shrugged would convert the country—in about six weeks; I could not understand how anyone could read it without being either convinced by what it had to say or else hospitalized by a mental breakdown.

The following winter, Rothbard, Raico, and I, and, I think, Bob Hessen, all enrolled in the very first lecture course ever delivered on Objectivism. This was before Objectivism even had the name “Objectivism” and was still described simply as “the philosophy of Ayn Rand.” Nevertheless, by the summer of that same year, 1958, tensions had begun to develop between Rothbard and Ayn Rand, which led to a shattering of relationships, including my friendship with him.

That Rothbard letter can be found here. I think it may be Rothbard's most interesting writing.

I agree with Reisman that Atlas Shrugged should have persuaded the whole country in about six weeks. That it didn't is one of the largest and most important unsolved philosophical problems. (Note I'm thinking of this problem broadly. Why didn't Popper's work persuade more people? Mises? Szasz? Deutsch? I consider those the same issue. Atlas Shrugged is the best, but there's a lot of good work which should have persuaded a lot of people but has only had limited success.)

13. Cf. Murray N. Rothbard, For a New Liberty (New York: Macmillan, 1973). In that book, Rothbard wrote: “Empirically, the most warlike, most interventionist, most imperial government throughout the twentieth century has been the United States” (p. 287; italics in original). In sharpest contrast to the United States, which has supposedly been more warlike even than Nazi Germany, Rothbard described the Soviets in the following terms: “Before World War II, so devoted was Stalin to peace that he failed to make adequate provision against the Nazi attack. . . . Not only was there no Russian expansion whatever apart from the exigencies of defeating Germany, but the Soviet Union time and again leaned over backward to avoid any cold or hot war with the West” (p. 294).

I already had a very low opinion of Rothbard. He has a lot of really awful views, such as anti-semitism and children-as-property. I didn't know this specific thing. (Some of his writing about economics is actually pretty decent.)

It is the division of labor which introduces a degree of complexity into economic life that makes necessary the existence of a special science of economics. For the division of labor entails economic phenomena existing on a scale in space and time that makes it impossible to comprehend them by means of personal observation and experience alone. Economic life under a system of division of labor can be comprehended only by means of an organized body of knowledge that proceeds by deductive reasoning from elementary principles. This, of course, is the work of the science of economics. [emphasis mine]

I disagree with this epistemology, which thinks you have some foundations and deduce the rest. For info on my epistemology, see my David Deutsch and Karl Popper reading recommendations, and my own writing on my websites.

Here is a conversation I had on Twitter, in order. This is fine:



And here's my main Twitter feed. Notice how Twitter has taken the same conversation, shown it as four separate parts, which are out of order and incomplete. And Twitter put things together which don't go together (my reply to @Spiff is linked to @asymco's tweet). It's extremely confusing and terrible.



I've noticed other major problems with Twitter threading too. (Threading is how it puts together multiple tweets into one conversation. An example of an OK method is chronological order, like my first screenshot. Split into parts, out of order, with some messages omitted, is an example of a very bad method.)

No Context

Often I view a tweet which replies to something, but when I click through to details I cannot see what it replies to. Ugh... Twitter please get your act together. (Or better yet, someone make a better website and replace Twitter. Thanks in advance.)

Here is an example. Ann Coulter replies to someone talking about a school censoring "this word", and suggests "raghead" instead. But what word was it? I don't know and there's no reasonable way to find out because Twitter's threading is broken: it doesn't show me the prior tweets in the conversation.

The graph @asymco tweeted is normal enough. Focus your attention on the reply from @Spiff.

The reason some of the revenue data looks compressed on the current revenue scale is because it is. Trying to change that is trying to mislead viewers.

What @Spiff proposes is an example of How To Lie With Statistics. He's intentionally trying to make the graph look different than it normally would to meet an agenda of his and give the viewer a different impression.

His tweet may seem innocuous but it's really bad. He's a bad scholar wannabe. Stuff like this is not OK. Bar graphs for typical quantities should use a linear scale, starting at 0, unless there's a damn good reason to do otherwise. If you do something else, the bars are no longer proportional to each other in the intuitive way, so it misleads viewers.

For example, suppose the graph started with revenue at $50,000,000 instead of 0. Then all the bars would be shorter. This would make the smaller bars shorter by a large percentage, and make them look even shorter compared to the big bars. That'd be really bad! When someone looked at it and thought "this bar is twice as big as this other bar", that would no longer be a valid way of reasoning due to not starting at 0.

Or suppose the graph used a log scale like @Spiff proposed. A log scale mean the revenue would be labelled like $0, $1, $10 $100, $1000, instead of like $0, $10 million, $20 million, $30 million, etc... See how misleading that could be? What it means is, again, when someone compares the sizes of the bars he gets the wrong idea. When he thinks, "This bar is about 20% higher than the bar before it", he's being played for a fool.

Don't play people for a fool. Don't try to trick them. Don't have an agenda for how you want your graph to look and then adjust things to achieve it. Just make a simple graph that makes sense and then leave it alone and let it speak for itself. Tinkering with your graph as @Spiff proposes is dishonest (or clueless and still harmful).

Scholarship, please.

Update: @asymco says:

Share prices are frequently graphed using log scales by default. I don't condone the practice.

I'm sad to hear how common bad scholarship is. That's terrible. But I'm glad @asymco understands this and does a better job. Thumbs up to him! Here's @asymco's blog which I read regularly.

Update 2: @Spiff now agrees with my point about log scales (I think).

Update 3: Here is an example of a very bad article advocating bad use of log charts. It looks mainstream and has a tone of sharing uncontroversial knowledge.