The Uncertainty Principle

Here is a brief explanation of why the "Heisenberg uncertainty principle" is implied by quantum theory. It's not nearly as mysterious as people think. Yes, if you don't have any matrix math background you won't be able to follow this. But you still might see it's just a little bit of math, there isn't a lot of stuff to it. The Uncertainty Principle is not a Principle, it's not a law of physics, it's just one of many results you can work out about quantum theory with a small amount of math:

-- In quantum theory observables can be represented by Hermitian matrices.

-- If an observable of a system can be represented by a particular matrix at a particular instant, then all matrices of the same dimension represent observables of that system.

-- In a state specified by the vector |psi>, an observable X is sharp if and only if X|psi> = x|psi> for some real number x. In which case x is an eigenvalue of X and |psi> is an eigenvector of X.

Now let Y be any matrix that does not have |psi> among its eigenvectors. (For any vector, there exists an infinity of such matrices.)

If the actual state is |psi>, the observable Y cannot be sharp. (Because of the 'if and only if' above.)

Elliot Temple | Permalink | Messages (4)

Elliot Temple | Permalink | Messages (7)

Letter to SENS

I sent the below letter to SENS, which is a medical research non-profit seeking to solve human aging. I like them because they have a good plan for how to do this which makes sense. Aubrey de Grey is their leader, I had a long discussion with him which you can read here.

SENS claims to be basically the most important thing in the world. SENS' web presence is inconsistent with this claim. SENS' web presence communicates low-prestige, low-intellectual-seriousness amateur hour. I offer criticism for several issues, partly on Aubrey's direct invitation, in hopes of helping.

Concrete Examples:

The SENS website LOOKS like a very standard generic format that doesn't stand out at all or get attention.

The SENS website has many basic web design errors such as:

- requires giving your country and even US State to sign up for newsletter. email should be the ONLY required field, period. and don't even ask for stuff like people's zip code. it's not OK to add friction to newsletter signups.

- SENS front page should be aimed at the public. that means you don't put things like "jobs" and "terms of use" there. you put all the stuff the public doesn't care about on an About page or other internal page.

- the February newsletter webpage does not link to the previous newsletter, or the archives, at the bottom.

- SENS has 3 blogs instead of 1 blog with categories. this splits up viewer attention. and since all 3 are very inactive, it just makes them look even more inactive – even with triple content in one place it'd still look bad and like SENS is inactive.

- It just plain looks like a cheap generic site in terms of layout and design. It's hard to explicitly explain why it does, but lots of people can tell because they've seen many other websites that look similar. The look of the site doesn't stand out and doesn't DIFFERENTIATE SENS. It doesn't communicate that this is something special or important.

- The images used look generic and unimpressive too. They don't stand out.

- It's not a .com site. That's bad because lots of people don't understand other TLDs besides com. (People given the website URL in person will literally do things like try to go to sens.org.com or just forget and go to sens.com. This especially applies to older people who I'm guessing are a larger part of the SENS audience. This issue is well known and makes a substantial difference.)

- The site doesn't have a bunch of awesome impressive essays (or other content) with amazing ideas. Or if it does they aren't prominent and I managed to miss them.

The SENS newsletter isn't even consistently once per month (which would be the bare minimum frequency to not look bad and have people forget about you).

The SENS newsletter looks like a normal newsletter, it doesn't stand out, it doesn't communicate SENS is SUPER FUCKING IMPORTANT.

The SENS contact form looks like a generic "we have to put up a contact form to pretend we listen to feedback" black hole. I don't know whether it is or not, but it looks that way. It looks generic and boring, and like you won't get a reply just like you don't from many other organizations. And it even adds annoying friction like making you categorize your inquiry – which is asking people, if they want to contact SENS at all, to do extra work which they aren't good at and don't want to do.

The SENS website homepage links to the SENS subreddit. This is not OK because that subreddit is very inactive (the 15th highest submission is 3 months old!). Do not send homepage visitors to a dead site, only link them places they should actually go and will be glad they went.

When you claim SENS is super duper important, but lots of the stuff you do implicitly contradicts, you destroy your own credibility and drive away most people.

Here's an example of acting inconsistently with your claims from Facebook:

Jonathan Weaver That's $10,000 in 2-3 days. Nice booster.
Like · Reply · December 5, 2014 at 6:45am

SENS Foundation Jonathan Weaver That's right! We're very thankful.
Like · Reply · December 5, 2014 at 8:31am

SENS claims to need something like $100,000,000/yr for the RMR project to go full speed and save everyone's lives. 10k/2.5 days would be too little by a factor of 68 if you got it constantly all year. 10k fundraising also just looks bad for being a small amount of money, all kinds of unimportant projects get more than 10k on kickstarter in 2-3 days. By being happy with a small amount, you accept it as appropriate to SENS, and accept a status below all sorts of stuff that can raise more.

If you really think you need 100mil/yr or MILLIONS OF PEOPLE DIE (which is what even a few year delay for SENS means), then sound the alarm instead of saying you're happy with an amount of fundraising that kills millions. When you act happy with pennies, you are telling people SENS isn't really that big a deal.

You may doubt the importance of these things. Keep in mind the cultural context. People don't expect to be listened to. If SENS is any different (which I'm unclear on), you have to shout it from the rooftops before anyone will notice. You have to make the difference extremely clear.

When Joe Random has what he thinks is a good idea, he knows he'll have a hell of a time getting anyone to listen, be it a big company, a small company, a scientist, a politician, etc. It's true that the majority of Joe Randoms have bad ideas, but some have good ideas and some others could learn to have good ideas with some pointers in the right directions. If you want Joe to communicate with SENS, you have to get his attention, not blend in with every other organization that he expects to ignore him.

I posted at the subreddit per Aubrey's recommendation and got replies which said, basically:

1) Leave and email Aubrey personally (or Michael Rae or SENS) instead.

2) Leave and go to the longevity subreddit which is more active. [Note: the longevity subreddit isn't really active either.]

3) I like SENS but got discouraged from the SENS subreddit because my posts kept getting downvoted.

4) You could try posting here and hope that somehow things will work out, contrary to your reasonable expectation.

I was not impressed. And the subreddit does nothing to stand out and communicate SENS IS IMPORTANT.

I think the talk to Aubrey/Michael personally plan is problematic because they are busy. For SENS to succeed on a big scale, there needs to be division of labor rather than expecting Aubrey/Michael to do most stuff personally. It also communicates that SENS is small time and un-prestigious if it doesn't have anyone below the top people to answer questions and have discussions with the public – there should be tiers with only a few things being escalated to the top people.

I checked the SENS Facebook page that Aubrey mentioned. It, again, does nothing to stand out and communicate that SENS is something different that's really important. It's more active than the subreddit. I dislike Facebook so I'm not familiar enough with Facebook pages to say if the activity level is OK or not, but it's definitely not GREAT.

I'd like to differentiate between three different styles of promoting SENS. Three categories of how to approach this. SENS is not doing well for any of them.

Style 1) Prestige

Impress people and say how SENS is smarter than you, and works with prestigious people and has a fancy reputation, etc, etc

This is irrational and will alienate the best and smartest people, but will impress the second tier people. It could work I guess (I'm not a fan of this style and don't recommend it).

SENS does some stuff clearly in this direction, but overall isn't good at this. An example in this style is writing, "Extramural research at PRESTIGIOUS universities and other state-of-the-art laboratory facilities throughout the world". Which isn't even well done, it's crude and blatant. Achieving prestige works better with more subtlety.

Style 2) Generic

You can just be yet another charity organization for yet another undifferentiated cause and try to get somewhere anyway. Some organizations have success with this. They aren't super important, they aren't super prestigious, but they put in the work and get somewhere.

SENS does some stuff in this direction (e.g. runs yet another small stakes matching fundraising), but isn't by any means great at it. For example the website isn't very well done, nor the subreddit, blog or newsletter.

Note, btw, that matching donation drives are bad and should not be done. See: http://blog.givewell.org/2011/12/15/why-you-shouldnt-let-donation-matching-affect-your-giving/

I tried explaining the problems with matching donations to "Reason" (the Fight Aging guy) at more length at the GRG email group but he was unwilling to address/discuss the problem.

Style 3) Reason

The third style is to focus on ideas and the intellect. Really seriously, not in the token way that's common. Here is one way to do this to give you the flavor:

Have high quality public discussions and challenge the entire public to offer any criticism of SENS, and answer every single criticism so you can honestly say there are literally no unanswered criticisms of SENS.

Saying that properly requires not just answering all the criticisms you know of, but also making a serious effort to seek them out in the first place, which involves, for example, having discussion forum of some kind for people to post criticisms at where they expect to be heard and taken seriously. For criticism to be fully possible, you also have to answer questions so people can get you to take stances on every issue and potentially criticize your answers to the questions. They have to be able to draw out more claims from you and get things clarified.

This approach isn't just about telling people SENS is super important and intellectually correct, and acting the part. It also means SENS will get all kinds of ideas, suggestions, comments, feedback and criticism from the public. And some of it will be correct and SENS will learn something too. And it also means one member of the public can answer the question of another member of the public – there can be an interested group of people being helpful.

Broadly, I would say if people are too damn stupid and irrational and have no interest in thinking, SENS is pretty screwed anyway. But I don't think they all are, and I think you ought to try and give people the benefit of the doubt and stop treating them like they are beneath you. I think SENS ought to take the position that people really do have minds, and they matter – if they don't there honestly isn't much point in saving their lives anyway. Don't just ask for monetary donations, show you care about ideas by seeking them out too.

Note these 3 styles are incompatible. The prestige approach appeals to the irrational side of people. Focusing on reason isn't generic, it would stand out. Being generic isn't prestigious. So it's important to pick something and focus, rather than do a little of everything badly.

I recommend the Reason style because it's the only one where SENS is at an advantage. SENS does not have the most expertise at impressing fools with prestige, or at grassroots hard work and community building and running charities. And SENS has no inherent advantage at those activities. That SENS could save millions of lives, and has some good arguments for its importance, is only a major advantage intellectually. In the prestige and generic games, people with much worse causes will say they are important too or whatever else, and since there isn't an intellectual atmosphere they can get away with those claims.

I think SENS should focus on where it has a large advantage over almost all rivals. (I am not personally convinced SENS is the most important cause in the world. But I agree it's a top cause, much better than the vast majority of causes.)

As a separate topic, consider that SENS would like a LOT of money. Like $100,000,000/yr for a decade. SENS, therefore, could use knowledge about money and economics. This kind of knowledge is necessary to use the money well. Consider that you wouldn't want an economically illiterate person deciding how to spend a million dollars. Well, at the billion dollar level, you wouldn't want a person with, say, "above average" economics knowledge either, you'd want world class knowledge to be involved. And it really helps to know how to deal with this money before asking for it, instead of telling people to trust that you'll figure it out correctly after getting it. And understanding these things is important for speaking intelligently to potential donors about these subjects.

This means, for example, familiarity with economics books such as _Capitalism: A Treatise on Economics_ and _Human Action_ (the best two major economics books). Preferably much more.

This does NOT mean that Aubrey should read those books. Understanding economics (not just reading a few books but studying it enough to really understand the material) is HARD and TIME CONSUMING. Therefore, it is an appropriate area for specialization and division of labor. SENS should have access to SOMEONE who knows this stuff, and who can relay important points to Aubrey and others when they are relevant.

Economics is not something everyone should learn, but it is important to basically everyone, and certainly to SENS which wants to deal with huge quantities of wealth. This is just like science: not everyone should be a scientist, division of labor is good, but science is important to everyone (and many organizations ought to have science advisors of some sort).

Similar lines of reasoning apply to quite a few other areas besides economics, such as epistemology (an understanding of the best methods of reasoning, and of philosophy of science, are two things that could aid SENS), moral philosophy (some of the objections to SENS involve moral issues), political philosophy (some actual and potential SENS projects involve the government), and computer science (maybe instead of preserving our bodies, we should upload our minds into computers. if we could accomplish that faster and cheaper than SENS, it could be the better option).

For each area, there are ongoing debates about which ideas in the field are right, which specialist experts are actually fools in disguise, which books are good, and so on. How is SENS to deal with this?

There is no way other than open rational public discussion. It leads back into the issue of discussion. Get a SENS economics expert who will address all public criticism, address all questions and issues about his economics claims, and so on. Open-ended rational discussion addressing all the issues is the only way to sort out the messes in all the various fields full of disagreement. I know this is hard and not SENS' expertise, but there is no way around it. This is what reason, truth-seeking and getting stuff right requires. The truth isn't easy to come by, too bad, suck it up and deal with it; there are no shortcuts.

SENS should not BET ITS FUTURE on the proposition that economics is irrelevant and ignorance of it won't lead to any major mistakes. Nor should SENS bet its future on siding with any particular side in the economics debates and not have that stance fully open to criticism and revision in case it's mistaken. And the same goes for other fields besides economics too.

SENS is struggling. It's badly underfunded. This stuff is URGENT and LIFE OR DEATH. SAY SO. CLEARLY. EVERYWHERE. Don't tell people everything is fine, tell the truth, it's NOT. Most current SENS communications act like these ideas about SENS' urgency are FALSE and actually everything is fine and not too urgent.

I think the most important thing is consistency. Have a consistent message and act commensurate with it. Have a consistent plan instead of a little from several styles.


I have more to say (lots), and more details for these points, but I think this is enough to get started. Please do not say "good points, you're very smart" and then proceed to do your (inevitable) initial misunderstandings of what I meant, without further discussion, in private (as is typical with this kind of thing).


PS Why didn't I write this sooner? Partly because of the contact form, as addressed above, and also the lack of any good SENS discussion place. Another major reason is b/c even now I don't really expect much to change, I don't expect this to have much effect. One reason is because I don't expect you guys to agree with everything I say INITIALLY (which is completely fine and reasonable). And I don't expect you to discuss all this to resolution (which is problematic, it blocks Paths Forward, which is irrational). One reason for these low expectations is SENS does little to differentiate itself from all the other non-profits out there, and I certainly wouldn't expect most orgs to really listen to comments like these and make big changes.

But Aubrey asked me to write (some of) this, and anyway I think it's interesting. And SENS is important – as far as medical science, it impresses me more than anything else I've seen – so I hope this helps.

Update: I received a bad reply from Michael Rae and wrote some comments on it.

Update 2: I wrote SENS Against Specialization and Division of Labor.

Elliot Temple | Permalink | Messages (0)

Don't Talk To The Cops

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.

Elliot Temple | Permalink | Messages (0)

SSBM Training 2: Reverse Dolphin Slash

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.

Elliot Temple | Permalink | Messages (0)

SSBM Training 1: Marth's SH Double Fair

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.

Elliot Temple | Permalink | Message (1)

Israel and Iran

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.

Elliot Temple | Permalink | Messages (0)

Anti-Deviance Strategy

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.

Elliot Temple | Permalink | Messages (0)