Hacker Newsnew | past | comments | ask | show | jobs | submit | freedom123's commentslogin

This post is a little long. Anyone who is curious, Rodney Mullen invented the 1st flip trick on skateboarding and about 80% of the flip tricks you see today on video or tv. He has an interesting take on skateboarding where he states you have to try and fail over and over with physical pain to land 1 trick. This process of physical failure to achieve, gives skateboarders a good lesson in life.


@freedom123, I feel your TL;DR version misses the crucial bit.

It's not just failure:

It's the combination of accepting failure (despite risk) with imagining solutions and new concepts, and meticulously applying them until you succeed.

That is what many skateboarders do, and I know that first hand. I'm now a professional coder.

You look at the a bit of land, and imagine ways you could skate it, you look at your board and imagine how you could spin it, or stand or slide on it, and you just try and try and fail and try until you succeed. It's hard to express the creativity and drive you need to have to get anywhere on a board.

Personally, I find my iterative creative mindset and my ability to accept failure help me at skateboarding, coding and lead me at one point to be a pro musician too... These skills exist in many areas. Skateboarding is just incredibly hard to attain any real level of skill at.


Absolutely. There's seems to always be this interesting duality to achieving anything difficult.

You have to have faith that you can do well, but you also have to acknowedge the hard facts of reality.

You have to be kind, but you also have to be firm and not a pushover.

You have to help people, but you can't be taken advantage of.

You have to get out of your comfort zone, but you have to avoid burnout.

So on and so forth.


Skateboarders (rollerskaters too), because they have to, embrace failure. I've seen skilled guys fall apparently hard without injuring themselves because they know how to diffuse the energy as smooth as possible.


Not to mention concussions... just saying.


Not really with Rodney Mullen's tricks. A lot of the amazing stuff he does with a skateboard doesn't involve a vert ramp or grinding over stairs - https://www.youtube.com/watch?v=1U-cgn3cEGA


That's insane. He does flip tricks that I still don't see current pro skaters doing some 20 years later.


And the style that he has just puts it over the top. There are tons of skaters now who grew up attempting to directly copy his repertoire, but when they get to the more difficult flip tricks to manuals and chains he puts together they are lucky if they can land it much less make it look that good. That much physical talent and equal creativity are an extremely rare combination.


I recommend checking out his autobiography, he's led an interesting life.

I'm not sure if any professional athlete has ever dominated a sport like Rodney dominated freestyle skateboarding. At the height of his dominance when he'd show up at a competition it was never a question of who would win, it was a question of who would come in 2nd place behind Rodney.


Yeah, I haven't seen anyone do a darkside other than him yet.


Wes Kramer, this years skater of the year, was the first to do it...pretty stylishly too: http://youtu.be/PIY-qD8rUmw?t=4m27s


Mullen could pull it off on command, with a casper slide to top it off: https://www.youtube.com/watch?v=vcjN5LhYmdE#t=259


Toss education into the private sector, problem solved. The market will sort out what education formats perform best - what else you guys want to talk about?


"The market" prioritizes short-term gains, and stratifies the wealthy and the impoverished. Allowing the market to work on products that amplify that stratification is inherently unstable.


I suggest the other top HN article today, https://news.ycombinator.com/item?id=8274677 where the market is clearly rewarding the long term strategy of Amazon.

This applies to any stock with a high P/E ratio.

Also, anyone with a spare $100 can get started investing in securities.


The point of the article is that investment in Amazon is growing because Amazon has a business model built on stable infrastructure and expansion modeled as a large conglomerate of essentially separate business groups which distributes risk, and is thus better able to weather negative environments (see e.g. the stock's behavior during recent economic crises). It's being used as a hedge against volatility... in the short term. Because that's how markets operate. If you believed that you could make more money in the same amount of time elsewhere you would put your money there instead.

anyone with a spare $100

Being impoverished doesn't just mean that you're unlikely to have a spare $100 at any given time, it means that any negative life event is that much more likely to wipe out any savings you may have been able to accumulate, including an investment in securities.


Don't all people have an equal right to education?


You'd have to consult an attorney for a thorough answer, but my understanding is that US federal law requires public school districts (K-12) to provide an education for every eligible child. The ADA applies to children with disabilities, and districts must provide their special educational needs as much as can be done (there are limits of course but expectations are high).

Schools often don't look for kids who need beyond average services, and may defer, even resist doing so unless parents make a point of pushing the issue. I've known parents who have taken schools to court to get cooperation. Too bad that's necessary, schools invariably relent. It is the law.


Basic education is a right, and you have the freedom to pursue advanced education, often subsidized by the government.


Do you feel like you have a right to the same education as the best-educated person? I don't.


Why should they?


Education isn't a right. It is a choice.


Education is a human right. http://www.un.org/en/documents/udhr/index.shtml#a26

Of all of the countries that are top performing (as rated by NCEE), which have privatized school systems?

http://www.ncee.org/programs-affiliates/center-on-internatio...


I can tell with certainty that my kids do not have all that much choice when it comes to education. By the time they will be old enough to get that choice, their education will be mostly done thing.


>Education isn't a right. It is a choice.

That's a poorly thought out platitude. Plenty of rights are also choices such as the right to vote, to remain silent, to bear arms, to a trial by jury.


I just said its not a right. So to equate it to other rights doesn't make sense.


You said "Education isn't a right. It is a choice." By the pragmatics of English, this is reasonably interpreted as "Education isn't a right [because] it is a choice". Pointing out that other rights are choices refutes that. If it wasn't what you were trying to say, you should clarify.


No, it's your duty. Society suffers, in general, when people are not educated. We really should insist on requiring more of people.


You can say the same thing about public sanitation.


It's funny how the people who believe that the Invisible Hand of the Market will fix everything forever generally claim to be opposed to faith-based ideology.


Dunno, how's high school treating you?


Kabam lol


"We have hundreds of A/B experiments running at anytime" which means you have a very big attribution problem.


I work at Pinterest on Growth, so I use our AB framework every day. Most of our experiments have nothing to do with each other – two of our current experiments are "add a button promoting group boards, compare to control" and "test different designs for our typeahead, compare against each other". Pinterest is a pretty big website, so it's unlikely that our experiments affect each other in ways that'd make our tests invalid.


This is an unfortunate over simplification (for both responses). Argument towards invalid data is stating an assumption- "Most of our experiments have nothing to do with each other" < that isn't a scientific method. Its simply an opinion that you are taking no effort to gain clean data - meaning the truth.

If a scenarios exist that can invalidate your findings, then you use every scientific method to avoid them not enable.

"Pinterest is a pretty big website, so it's unlikely that our experiments affect each other in ways that'd make our tests invalid." << This is the very reason why you will have an attribution problem.


Attribution issues aside, there's two scarier potential issues I see, actually I see these with all posts on "here's our company's cool new A/B testing framework", that really scare me:

1. You're running a ton of tests, yet I see no mention of how you're adjusting your tests to account for multiple testing. The more tests you run the higher the chance you have of getting a false positive. Couple this with the majority of things you test probably won't be significantly better, and your chance of encountering a false positive is much much higher than you might think. You're running hundreds of tests and using a p-value of 0.05, but your chances of a false positive in your tests is much much much higher than 5%. Beware multiple tests and especially beware of the base rate fallacy.

2. Your post has no mention of statistical power or the size of the effect you're looking to detect. That makes me think you might not have considered this. If you don't know the effect size you're looking for or your statistical power, your A/B test results can't be trusted -- as you have no idea what your chances are of actually detecting a beneficial result if it exists.


These are both good points. I think it's worth calling out here that this post is really about the infrastructure to perform experiments, not necessarily the means of analysis (although the ui you see in the screenshots performs that analysis).

In terms of these issues, we handle them in a few ways. In particular, experiment review catches many of these issues. Think of it like code review for your experiments.

In order to run an experiment, we require you to have an "experiment helper" sign off on your change. This involves reviewing your group sizes, verifying that you have the statistical power you need to test the magnitude of change your hypothesis expects, verifying you interact with the framework correctly etc.. Training to become an experiment helper is generally not very easy, and involves a combination of shadowing existing reviewers, performing enough reviews across the stack, and taking a test to verify you understand potential errors (the test itself being composed of many experiments where we have made mistakes).

Changes to experiments (increasing group sizes, terminating an experiment, modifying an experiment etc.) all require this review.


1. Beware multiple tests: agreed. We could apply corrections (e.g., Bonferroni-like), but don't underestimate the logistical complication of this. Before a month starts, we don't even know how many experiments will be run (although we could try to predict). A different way to address this problem: use other information. Do the results make sense (good)? Is there corroboratory evidence (good)? Are there crazy outliers or things that smell funny (bad)? Of course the experimenter often has a bias (it worked!). Can they convince others? etc.

2. To compute statistical power, you need an estimate of effect size. For many experiments, we don't know; we could run pilots, but in the naive case that doubles the number of experiments to run, and time to wait. By default, we recommend a sample size that will detect a certain effect size (but not smaller ones). That is, we decide we are not interested in small effects, because it makes our life simpler.


1) it is true that the more tests you run the higher chance you see one FP, however the tests are not rejected/accepted as a whole. each test will show it's own results.

2) the size effect did take into considerations, there's also r scripts to analyze the results. the impact is trivial


+1


I work at Pinterest, and have some involvement in how we learn from experiments.

In each experiment, we assign users randomly based on a hash of experiment name and userid. So, we assume (and have tested) that the groups of one experiment have a fairly uniform mix of users from the groups of any other experiment. This means the primary effect of being in a particular experiment group G should be much larger than the effect of being in G and some other experiment's group. So, we usually analyze the effects of an experiment independent of any other experiments. Millions of users helps here.

I think this is how trials normally work: assume that random assignment into groups will spread all other factors equally across the groups.

There are experiments which depend on each other (change the same UI, one assumes a precondition of another), and that situation is complicated. We try to avoid it when we can, and negotiate carefully when we find ourselves in it.

You might also be referring to the fact that looking at hundreds of things will find "statistically significant" effects simply at random. As my advisor said, "95% confidence means wrong 1 in 20 times." That is a risk to be managed. We always have to ask things like: is this a reasonable outcome for this experiment? Do we have other corroborating evidence? Is the effect consistent? Do we have reason to disbelieve the result? Some fraction of our experiments are run incorrectly or have a broken implementation; we detect some, likely not all.

Also: is the risk higher to run an experiment that may be misinterpreted, or not to experiment at all and just release things? We like to learn from data if we think we can. There is art in the tradeoffs.


Did you guys look into Bayesian Bandits? They seem much more tradeoff-able.


Using Bayesian methods to learn about our experimental effects is very interesting to me. For example http://www.evanmiller.org/bayesian-ab-testing.html, which may be the type of bandit you refer to.

In the best case, it might be more robust to low-data outliers that cause random blips that look significant, as well as addressing the problem of looking at an experiment over and over (which frequentists are uncomfortable with: more chances to succeed).

However, it is non-trivial to understand how this works for us in practice. Examples: we want overall results as well as days-in results (to look for novelty effects); we would have to choose priors with consideration, because they have a huge impact on most results; etc.

This takes time and effort, and there are a lot of other things competing for those resources.


Not necessarily.

Tests and metrics might be page-specific. For instance, "How long did Experiment User stay on Page X vs. How long did Control User stay on Page X?" That wouldn't cause any difficulty, assuming at most one experiment per page.

Even if more than one experiment is running per page, or if metrics are more complex (e.g. "Does showing a more prominent image result in more repins of that image over the next seven days?") there are methods that can help isolate the effects of an individual experiment.


Agreed. The question I'm interested in is, how do we solve that attribution problem?


Can you share a slide within a blog post without using an iFrame


Nope, presentations can only be embedded as iframes.


I will summarize real quick. "Rich" and "Poor" are 2 variable of millions - And no matter how many variables you address you cannot escape one simple fact about "who gets ahead" - "If its important to you, you will find a way, if it is not, you will find an excuse."


> no matter how many variables you address you cannot escape one simple fact about "who gets ahead" - "If its important to you, you will find a way, if it is not, you will find an excuse."

Do you have evidence for that? All the evidence I read, including the research discussed in this HN post, says that it depends heavily on where you start, including factors such as family income.

Are you saying that children born into poor families or who face other such disadvantages achieve less because they have less willpower than children born into wealthier families? If so, perhaps we should look at the relationship between family income and willpower.

EDIT: Minor clarification


This is a good way to think on an individual level: ie. taking control of one's own destiny, but it's not adequate on a statistical/societal level. We know for a fact that there are number of correlating factors to success beyond an individual's ambition/discipline/chutzpah. If we ignore this as a society and the rich people from advantaged backgrounds (ie. the majority) continue to pat themselves on the back for their work ethic, then the growing divide between rich and poor will accelerate and destabilize the very status quo which the self-satisfied rich depend on to maintain their wealth and privilege.


crickets


Bitcoin, no matter how you word it or rationalize it will always have a root problem and inherent risk...it is backed by nothing. Users today think its usage gives it market value and it can but to a limit. Money (USD) originally (the dollar bill) worked because it was back by gold. A metal that worked because everyone on the planet wants it. That dollar bill was a "check" or agreement stating, this dollar bill represents this much gold thus the value of money. The USA today does not have gold backing its money-- so as you can see we already have a bitcoin and the USA will not allow you to complete with its money - enjoy


If gold wasn't used as a store of value its value would be very maybe a hundredth of what it is now: Gold-plated jewelry in our modern era is essentially indistinguishable (even at the level of a physics lab using scales/xrays/etc if need be) from solid gold jewelry.

Gold, for all intents and purposes, is also "backed by nothing".


Aluminum was more expensive than gold until a cheap way to purify it was developed. Now we wrap our food in it. Gold is only valuable until the first metallic asteroid is placed into high Earth orbit. Not tomorrow, but probably sooner than you think. reply << Has nothing to do with Bitcoin being back by nothing. Even if gold did lessen in value other forms of metals/currency can take golds place. Gold is specifically a variable for anything however it is "some thing". (the root point).

If gold wasn't used as a store of value its value would be very maybe a hundredth of what it is now: Gold-plated jewelry in our modern era is essentially indistinguishable (even at the level of a physics lab using scales/xrays/etc if need be) from solid gold jewelry. Gold, for all intents and purposes, is also "backed by nothing". << again has almost nothing to do with Bitcoin being back by nothing. Gold is the item of value it doesn't need to be backed, it is the back'er.


I suggest you return to your economic and history books. Gold has no intrinsic value as you would probably call it. It is only because of some VERY distinctive properties that are not found in other metals that gold was ultimately selected as the better store of value available. Its only value comes from a market agreement.

Bitcoin is essentially the same as it replicates the same properties of gold while having a more predictable supply and features that allow it to be used more commonly as a currency, unlike gold.


Gold has intrinsic value because 1: Its rare 2: Its tangible 3: People want it. Look at history and you will see gold wrapped around everyone's neck and fingers. As mentioned before, a currency has to be back by something. Bitcoin is backed by nothing - nothing cannot be something - no matter how much you exclaim - that is the risk of investing any time with Bitcoin. Just because you like the utility of of bitcoin almost has nothing to do with it being a currency. Example - You can have 50 coins of solid gold or 50 zeros on a computer.


Also, unlike Bitcoin, gold remains gold without the need for a constant supply of electricity or technology. It's an element, and part of the reason it worked as a store of value was because "ooh shiny" and zero upkeep.


Exactly. My point was that "the things people want" is true wealth, gold is just another measuring stick for converting billable hours of work into diapers (or whatever). If Bitcoin is perceived as a superior measuring stick it will be preferred.


Aluminum was more expensive than gold until a cheap way to purify it was developed. Now we wrap our food in it. Gold is only valuable until the first metallic asteroid is placed into high Earth orbit. Not tomorrow, but probably sooner than you think.


You are going in an infinite loop trying to equate something (gold) to nothing (bitcoins). Typing zeros into a screen will not equate to a hand full of gold coins. The value of gold or any other metal is not a relevant argument because in a free market you can have competing currencies. Bitcoin could compete if the USA was a free market (its not) and if it was backed by something (its not). Thus the reason why investing in Bitcoin in any manner has risk. The argument with the asteroid doesn't support simply bc the metals will still be around after it hits earth. Your hard drives with stored zeros will be annihilated (much like your premise).


In the USA


The American Dream isn't about a price of annual income. The American Dream is about the possibility of having an annual income. Originally American was free which brought on the Enlightenment Period, the Industrial Revolution and allow with it, the concept of making money. The dream is you can make money. This is an important concept to understand back then because making your own money was impossible with the enslavement by kings and variants of Statism. Today American isn't free there is no American dream. Today we are more taxed and more regulated than ever. The concept of making money only exists by having 50% of it redistributed - enjoy.


Today we are more taxed and more regulated than ever.

Say what? Tax rates in the U.S. have been trending down for decades, especially for the wealthy.

http://en.wikipedia.org/wiki/Income_tax_in_the_United_States

See the section titled "History of income tax rates adjusted for inflation".


You are talking about 1 element, out of a million, pointing to it and saying this is a falsehood. For example, you are basically looking at 1 tree saying its not growing, while ignoring the forest around you - enjoy.


Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: