Video Of The Week: Nate Silver at Google

I am currently reading Nate Silver’s The Signal and The Noise which I am enjoying very much. So I’ve got Nate on my brain right now. Here’s a great hour long talk that Nate did at Google in late November.

#Statistics and Probability

Comments (Archived):

  1. JimHirshfield

    There’s a high probability that I won’t find an available hour to watch this. ;-)But I do find his methods fascinating. I worry that too much attention on these things during elections leads people to believe there’s no point in voting; as if decision is already been made.

    1. fredwilson

      he talks about that about half way through

    2. David Petersen

      Question:Are you concerned that during future elections, the accuracy of your predictions will lull readers into a mindset of “it has been foretold, therefore I needn’t bother to vote”?Nate Silver:It worries me a bit. There is probably a danger zone in which a candidate’s supporters take for granted that he’ll win the election and so don’t turn out to vote, but the election is nevertheless close enough for him to lose. That may have happened in the Democratic primary in New Hampshire in 2008, for example. There were a lot of reasons why Hillary beat her polls, but one contributing factor may have been that a lot of independent voters who would otherwise have voted for Barack chose to vote in the GOP primary instead since it seemed more competitive.

    3. Kirsten Lambertsen

      Off topic: could I hit you up, outside AVC comments, with a very fast advice/referral request? I promise, it’s brief.

      1. JimHirshfield

        Sure. jim at disqus

  2. JamesHRH

    I have seen him in a talk where he says that journalists know more than pundits, that the pundits are just entertainment.They should have to file on their taxes: professional noise maker.

      1. pointsnfigures

        outside the 95% confidence interval and more than 3 standard deviations from the mean.

      2. Modernist

        also, David Heinemeier Hansson

        1. jason wright

          google ‘zonda hh’ – is he on the down slope?

    1. pointsnfigures

      I have found the opposite to be true in many cases. Journalists don’t do a good job fact checking. They also aren’t very steeped in the things they are covering. I find them sort of vapid-and trying to make news rather than report news.In 2004, I was at a debate party. A left wing (is there any other kind) Northwestern journalism professor said to me, “Wouldn’t it be better if just people like you and me could vote?”. I reminded him of Jim Crow in the south and that his attitude was no different than theirs.Bloggers have more information, and better stuff than any journalist. No need to read the papers anymore except for the weather and sports scores.

    2. Modernist

      journalism needs an “intrade” layer

      1. Matt A. Myers

        Might have to kill off mainstream media channels first — many people are drawn into ‘bold’ statements.

        1. Modernist

          i agree, we need table stakes to have a viable negotiating position; such as, the threat of extinction for “no skin in the game” journalism.that said, i think hacks will always be drawn into making bold statements, just as hackers are drawn to make money. it is just one part of giving leverage to our ideascan a social layer grow on top of an antisocial medium before infecting the antisocial medium with sociability?(Modernist, founded late 2011, is a deliberate experiment with the negotiation of pageview journalism and intrade-style gambling)

  3. PJSweeney

    It’s really a shame our schools don’t give better basic training on stats, it’s so important to decision making.It’s crazy to me when you ask someone: “if you flipped a coin ten times and got heads each time would bet me $1,000 the next flip would be tails?” the answer would be very different than if you asked them “Want to bet $1,000 on the flip of a coin?” Or the decisions people make anecdotally – “One of our users said we should have Login with Pinterest so let’s put that on the product roadmap”I really enjoyed quant courses in B-School and like the fact you can get to the kind of depth that helped Billy Beane create a world-class team on a budget, I just wish more people (especially in government) would learn the basics!

    1. Richard

      It depends on whether it is a fair coin. (not sure i understand the pinterest comment?)

    2. Luke Chamberlin

      I wonder if there is a correlation between math ability in a population and the efficiency of their government.

    3. fredwilson

      i agree. i loved the statistics courses i took in college

  4. kirklove

    I’d like to borrow that when you’re done.

    1. fredwilson

      i will give it to you. it was a gift to me. i will regift it to you.

  5. William Mougayar

    Nate Silver’s criticism of Big Data is a good one: “Why in an era of “Big Data,” with “2.5 quintillion bytes” generated daily do so many predictions so badly miss the mark?”Nassim Taleb in Antifragility provides the other half of the argument which is – “the higher the frequency of the data, the more difficult it is to interpret it.” Too much real-time, in too many small pieces lets you miss the forest for the trees.

    1. Richard

      You are right. Once you adjust the data (making it stationary), the best forecast is often the most recent state. Thus, a lot of forecasting has to do, not so much with forecasting the mean, but forecasting the dispersion/volatility. While forecasting (time series analysis) and prediction (correlation modeling) are used interchangeably, they differ in some ways.

      1. JLM

        .The most recent state combined with the infinite slice trend IS the forecast — the right forecast — about 99% of the time.It is the exhaustion of the trend — which is not viewable until the first goose falls from the sky rather than flying in formation — or the pivot point which is obscured by the trend which is difficult to see.Believe it or not, the intelligence community does a very, very good job of this by ensuring that intelligence “appreciations” are made by analysts who are compartmentalized and unable to compare notes with each other thereby, in effect, preventing everyone from regressing to the popular position.In the case of OBL the simple but ultimately very big questions were —How many 6’5″ Arabs are out there? <<< they had measured and triangulated his shadow to get his heightHow many compounds do NOT have phone or Internet access today?These two facts coming from two different sources likely were determinant of the entire issue.Pity the poor consumer of five different appreciations who is forced to pick and choose what is really going to happen. Oddly enough, they almost always get it perfectly correct because the real constraint is people and national action.A good analyst could likely have predicted that Iraq and A’stan would end with the US leaving with the job either undone or poorly done. Based purely upon our national tendencies on several fronts..

        1. JamesHRH

          @wmoug:disqus this is ridiculously bang on, even for @JLM:disqus .’Exhaustion of the trend’ is the phrase of the year (admittedly, its early).

    2. Matt A. Myers

      Most big data is useless at the whole. It’s asking the right questions that matters, and even then I’m not sure the mass of data available publicly would really have the nuanced information you’re looking for.

      1. William Mougayar

        Are there good examples of using big data?

        1. Modernist

          since you mention taleb, i think his counterparty kurzweil is a good example.

        2. Matt A. Myers

          I haven’t needed to look into the domain yet in depth to determine what genuinely are the best / good examples – I’ll get there / need to eventually though.

  6. jmcaddell

    Fred, I love this book – one of my favorite reads from last year. I especially like the lessons for leaders. It’s very easy to get overconfident in your own judgment, especially when you are a decisive, opinionated person, as many entrepreneurs are. You need to strike a fine balance between self-confidence and humility – seek out dissenting data and views, use as much data as you can, but still make decisions.

  7. leigh

    Don’t have time for the whole video (and i prefer books and have a vacation coming up) but whenever i hear prediction talk, signal, noise, always takes me to McLuhan. Figure vs. ground. Basic premise — You have to understand the medium and the context together. Most often pple ignore the context. Great insights come from context.

  8. Richard

    These book are fun but there is “no way” to develop real intuition in forecasting either as a frequentist or a bayesian without a good understanding of statistics/probability and and applying it in R or SAS.

    1. fredwilson

      ooh. i will check that out.

  9. Luke Chamberlin

    I like Nate Silver but even he has profited greatly from people’s misunderstanding of statistics.If I say that someone has a 60/40 chance of winning and that person wins, it does not mean that I was right. That is not how statistics work.

    1. pointsnfigures

      Nate Silver also benefitted from having inside information in 2008 to make predictions.

      1. Luke Chamberlin

        What inside information?

        1. pointsnfigures

          he was getting the internal polls of the obama campaign, and then made public predictions. no one else in the media had that kind of access

          1. Matt A. Myers

            That’s not really a prediction though then is it … it’s a calculated guess.

    2. David Petersen

      Nate is the first one to point this out, however.

      1. Luke Chamberlin

        Yes you are right. Wasn’t a comment against him.

    3. thinkdisruptive

      But, there was a 60/40 chance that you were.

  10. Paul Sanwald

    I just finished this book, it’s a fantastic read. Like me, nate has a background in mid-high stakes poker, and was a very active member of the poker strategy community for many years. He was always one of the best strategists, and I loved in his book how much he highlighted what he drew from his poker experience. There’s a lot of skill in poker, and a lot of luck, too, and most of us who played professionally or semi-professionally have a completely different outlook on variance as a result of those experiences.It’s been fantastic to see his success, it’s obviously well deserved.

    1. ShanaC

      ah. What if you don’t like poker?

      1. Paul Sanwald

        Shana, it’s just one chapter in the book. I think you’d still enjoy the book as a whole a great deal, a variety of topics are covered.

    2. fredwilson

      he talks a bit about that in the video.

    3. JamesHRH

      I have never seen a big time poker personality say ‘It the math that matters.’Math helps, but lots of high profile pro play totally on reads.

      1. Paul Sanwald

        James, this isn’t really the case in my experience. I’m assuming by “high profile pros” you mean tournament players, as there are legions of extremely successful cash game players that keep a low profile. But even in the tournament world, even very old school players like Doyle Brunson use pot odds. There aren’t many (if any) successful players who don’t use basic pot odds and understand the probabilities of draws. For example, Daniel Negreauneau is a fantastic hand-reader, probably wouldn’t describe himself as using much math at all, yet he uses bayes’ theorem, thinks very probabilistically even if he wouldn’t use those words to describe his thought process.it is true that there isn’t a ton of math involved in playing a hand of poker, other than understanding odds, bayes’ theorem, etc. A more important consideration for a pro player is managing their bankroll, understanding variance, kelly criterion, and that side of things.There are 3 books I highly recommend:Doyle Brunson: super system. The classic old school poker text; and great as a historical document. There’s a ton of different perspectives here, as each chapter is written by a different high-stakes (for the time) player.Barry Greenstein- ace on the river. Barry is a consummate pro who was a phd in math. he talks a lot about the role of math in poker, and also psychological aspects of playing professionally: running bad, etc. Barry also worked at Symantec for years.Bill Chen – The mathematics of poker. Great, detailed read on the mathematics of hold ’em.

  11. laurie kalmanson

    if something happens to you, it’s 100 percent #math

  12. Modernist

    the (latest episode in) silver psuedo-controversy also furthered the contention that professionals (journalists, academics) advancing strong variant opinions be willing to put skin in the game ie. make a wager for charity. http://marginalrevolution.c…unfortunately, physicists have not figured this out yet, and are more interested in developing unique theories than building a meaningful consensus on the nature of reality as best observable. it’s not just refinement; there is an utter lack of agreement. MIT review:”For example, in answer to the question “Do you believe that physical objects have their properties well defined prior to and independent of measurement?”, 48 per cent replied “no”, while 52 per cent replied “yes, in some cases”. A further 3 per cent said “yes in all cases” and 9 per cent were undecided (respondents were able to select more than one answer).”if we could find a way to force all professionals to make bets on multiple time-frames (with token currencies even), this could build a more significant and less noisy human understanding graph. it could also help make Theory more engaging.mentos started out selling gum-like candy, and now also sells candy-like gum. they are straddling the disconnect. that is what intellectuals and gamblers need to do.*by “force” i mean introduce a betting layer that changes the cultural moresexperts in any given field have very deep understanding sets which are considered impossible to quantify. however, we could gain a more textured set of quantification if certainties in relevant predictions were tradeable as odds. the odds (as perceived to be true) would trend towards equilibrium. then professions could look back and critique themselves quantitatively as well.could gambling ability in a domain of expertise become a modernized (And Standardized) credential?

  13. David Petersen

    It’s too bad more data on aspiring/just-funded startups isn’t in the public domain, I bet Nate could come up with some surprising predictions.Fred, maybe you should share your USV data set with him and see what he comes up with.I like Nate and his predictions have provided me with great entertainment over the years, but not as much as watching his style of dress progress from standard yuppie to straight hipster as he has spent more time living in Brooklyn.Also, he did an AMA the other day: http://www.reddit.com/r/IAm

    1. Dave Pinsen

      Or maybe he might miss something crucial (in hindsight, at least), as he did with the impact of steroids during his baseball forecasting career?

      1. David Petersen

        Did the impact of steroids strongly affect the year-to-year prediction of player performance?

        1. Dave Pinsen

          I don’t know. Did you read the link though? As it notes, cracking down on steroid use had a huge impact on offensive stats, one Silver failed to predict, because he assumed marginal players would be more likely to be users.

          1. JamesHRH

            Predictions that involve numbers always turn on assumptions.

          2. Dave Pinsen

            Sure, but as the blogger notes, Silver said this in 2009, long after everyone knew that some of the biggest names on the sport were juicing.

          3. David Petersen

            I read the link and it’s very interesting. Still, they haven’t yet tested for HGH, and as Victor Conte pointed out, the steroid patches that players are using now are almost undetectable; it’s hard to imagine that the majority of players are clean, making the impact of PEDs harder to analyze. Offensive stats may have cratered for other reasons.Nate Silver’s blindness to this issue is still hard to understand, since it was obvious to me as long as I can remember that the best players were using steroids. (Probably due to the brief period when I took creatine and saw my athletic career go places it never had before, quickly realizing the potential of true PEDs).I always thought the single hardest part of scouting athletes was this: how do you tell if the athlete has used PEDs yet. If he hasn’t, he will possibly start at some point in the future, making this athlete a better prospect than one who has used PEDs to get to this point in his career.

    2. fredwilson

      what data would we share with him?

      1. David Petersen

        Traffic data, revenue models, revenue data, profit data, information about the founders, etc. Although I assume you have already done pretty comprehensive analysis on most of that stuff. Sample size might be an issue, too.

        1. fredwilson

          we’ve tried. it is hard to find the signal in that noise.

  14. george

    Good to see a statistician with a bit more personality! This interview could have been extremely boring…

  15. kidmercury

    off topic but since silicon valley hot shots have decide to take a break from their core competence of bubble blowing to fund a campaign against the 2nd amendment i thought it would be worth noting congressman israel is working on legislation to ban 3D printing of guns: http://www.forbes.com/sites…i hope we can all support this, we demanded a plan so let’s make sure we demand a plan on 3D printers and strict enforcement of them as well. israel says he doesn’t want to hamper innovation or impose pervasive monitoring but i don’t see how 3D printers can be stopped otherwise, especially with organizations like defense distributed taunting the anti-2nd amendment crowd. let’s all support congressman israel here and let’s demand strict controls on 3D printing of weapons, even if that means some restrictions on innovation. our rights aren’t absolute and i’m sure government will do a good job making the right distinction, as they always do.

    1. Dave Pinsen

      The guys behind Defense Distributed have some serious stones. Much easier, politically and popularly, to Demand a Plan (NSFW, if your W is a place where you have a boss who scrutinizes what you watch on YouTube).

      1. PhilipSugar

        That is so great. I love it. Guns for me but not for thee.

      2. kidmercury

        hahahhaha that is a great video!!!

    2. Modernist

      i hate to sound like a conservative, but we should consider their slippery slope argument when it comes to building infastructure to restrict 3d printing.

    3. JLM

      .Without speaking to the issue of whether 3D printing should be part of the debate, the issue of the number of bullets that can fit into a single magazine is one of those nonsensical arguments which keeps real problems from being solved.Everyone is arguing about the box in which the product arrived rather than the contents. Arguing on the margin.A murderous bastard is not going to be dissuaded from committing murder — which by the way is “against the law” — because of the size of the magazine. Or the fact that his particular magazine is “against the law”.”Against the law” is not working.Most recent murderers have carried a number of weapons. The murderous bastard at the Colorado theater had a drum — a BF magazine with over 100 rounds in it — IT JAMMED. He then killed most of those folks with a shotgun which provides multiple shells within the body of the gun itself and does not use a magazine at all.The key to decreasing murders is making a list of criminals and crazy people FIRST. And, then, it is about handguns not “assault rifles”.Rifles figure into less than 3% of all crime..

        1. JLM

          .Yes, Philip, I read that article some time ago. My heart melts at the challenge such people are forced to encounter. It is beyond belief and seemingly beyond help right now.It does however identify the real nexus of the problem — when the mentally ill person’s hand touches the weapon. That is where the barrier has to be erected — not in my gun safe.That is the FEBA — forward edge of the battle area. Where the fight has to start.The gun does not pull its own trigger and thus if the problem is going to be solved we have to focus on the mentally ill trigger finger.I agree more with you than you do with yourself.JLM.

          1. PhilipSugar

            I have spent my time, treasure, and talent on this one.We got to a point where my wife didn’t have to “work”That is in huge quotes because raising a family, and specifically dealing with me is more than “work”, its something you probably have to think there is a reward in the afterlife to tolerate.But I have sponsored and she runs a compassionate care center for veterans and serviceman. (she is a nurse practioner)She does this because they only will deal with veterans because even though they have issues, the general population is a huge mess, and really wants to abuse the system.So many of my Ivy League friends like to helicopter (like parents) into a situation and give their opinion on the problem instead of living, dealing, and working on one piece of the world.

          2. JLM

            .God bless you and your wife..

          3. PhilipSugar

            Before that she headed up the medical center for this place as a favor to the Governor:http://en.wikipedia.org/wik…Specifically the SHU.That I could not take, and has influenced my belief of the “goodness of man” and why I believe strongly in the second amendment for me. Tom Capano was her best patient.

          4. JamesHRH

            I don’t get how people like your wife find the energy for that work.I have been in a SHU (as a visitor / legal representative). It was at a max secure institution & the acronym was not as pleasant when expanded – Special Handling Unit.

          5. PhilipSugar

            I do not either.Suffice to say that even though she was abandoned as a child she managed to put herself through an Ivy League advanced degree by sheer grit (and a full time bartending job)I can tell you the calls in the middle of the night (with the problems) and the calls from threatening relatives has steeled my resolve for my 2nd amendment rights.

          6. Dave W Baldwin

            Love the “helipcoptor in” analogy.

          7. Modernist

            disengagement is the main problem. we shouldn’t think about mental illness as an unsolveable

          8. Matt A. Myers

            Very good observation. Engagement is a very important factor — we are social beings afterall. I have a theory on health relating to this – where everyone is on a spectrum of health with depression being on one end, and the ability to be social being on the other end; If you’re able to be social you’re more likely to be engaging socially, more likely to be active, more likely to have multiple interests, to be learning, more productive, more likely to be self-aware, be connected to other people / the community, more likely to be healthy and health improving, etc.. If you’re on the other end then your health is more likely stagnant or getting worse, you’ll be less likely to be engaging with others, etc.. This is why I am starting a not-for-profit charity to help test a preventative / proactive healthcare system that incentivizes social activities – to get people being socially engaged, to lead them towards this. If people are connected to others, to the community, engaged – they have more they care about, more to live for, and of course more situations / opportunities to learn (no matter what stage of their own evolution and self-awareness and world-awareness they are at). At the foundation you have to support people being able to get healthy and then maintain their health – and then you’ll have to facilitate / support the other issues that come up and are unique to each individual as they start the healing process (whether it’s from childhood situations or physical injury that lead them into a stagnant disengaged lifestyle which then lead to depression, etc..)

          9. Modernist

            thanks matt, yours appears the appropriate approach. people forget zuck was a psychology major

          10. Matt A. Myers

            Not sure that really characterizes him.

          11. Modernist

            haha fair enough

          12. Matt A. Myers

            Though I do know some people who aren’t grounded, not deeply connected to emotion, who can thrive in the psychological field because they’re capable of memorizing and have strong logical minds; This leaves the nuanced aspect and influence of emotion on awareness and thinking to be weak.

          13. Modernist

            personally, i think zuck built facebook as a therapy for himself and everyone else, and its wild success shows how sick we are as a culture

          14. PhilipSugar

            That was an upvote by me.

          15. Modernist

            that’s what hackers do. we solve problems for ourselves and the world along the way. this community has helped me comprehend that

          16. Matt A. Myers

            It’s all the “selfish gene” at work, relating to survival of course, and our innate experimenter / problem solver.

          17. Matt A. Myers

            Well, it’s a bandaid if that’s true. And he’s an abusive therapy provider then too..

          18. PhilipSugar

            I really want to live in your world. As I have said come live in mine for a while and you will change your view.

          19. Modernist

            philip, i think what matt is pointing out is that in many cases, engagement or disengagement is a self-reinforcing phenomenon. while your world is probably far more intense, that shouldn’t stop us from considering the full spectrum. this problem of disengagement is real for all of us, to varying extents

          20. Matt A. Myers

            I can only assume what your surroundings include, though all I can say assuming the worse case scenario is that transitions take time – and once systems are implemented it might take a generation or two before amplification effects really start to work in big ways; It could even mean waiting for those very worst who were somehow let through the cracks in a bad way to die before their negative impact on us all ends. There will always be edge cases too that have to be individually dealt with.

        2. Dave Pinsen

          Careful though: the kid in that article is mentally ill by anyone’s definition, but definitions of mental illness can be malleable and subject to political influence. In ~40 years we’ve gone from homosexuality being considered a mental illness to gay marriage being considered a civil right. Who knows? In the future, maybe conservatism will be classified as a mental illness, as a pretext to disarm conservatives.

          1. Modernist

            indeed dave, the will to exist meaningfully is itself something of an “addiction”. perhaps aaron’s prosecutors wanted all of his drug for themselves

          2. Mark Essel

            Yeah, the problem with classification is that it’s fallible, we’re fallible, and social systems we construct are even more error prone. Doesn’t mean we shouldn’t try though.

          3. Dave Pinsen

            Fallibility isn’t the only concern — pathologizing politically incorrect behavior is another, and the one I had in mind with that comment.

          4. kidmercury

            Guaranteed kooks will be classified as mentally ill

      1. Matt A. Myers

        People who commit crimes aren’t in the mindset to care or be influenced by laws, nor are there really any accurate signals that would predict early enough that they would actually commit a heinous crime.The answer is helping people not become “crazy” people, and likewise to help guide people away from crime and violence. The systems that would be put in place to prevent this would actually be of benefit to all of society, so it would be resources well spent – as productivity, happiness, etc. for all of society would increase.

        1. JLM

          .I agree with you on the broad brush strokes but I completely disagree with you on the detail work.This is exactly where gun regulation should be headed.When the Colorado assassin acquired 4 weapons and 10,000 rounds of ammo we should have had the benefit of the following MEDIUM SIZE DATA solutions:1. There should have been NO ability to buy a gun on the Internet without a background investigation. It would have turned up the fact that this guy was on medication and that his psychiatrist thought he was capable of violence. She had alerted the authorities to that fact.He does not get a gun under these circumstances. He does get mental health assistance.2. There should have been a 60 day waiting period and when it came up he was being treated by a psychiatrist the waiting period should have been indefinite until he was released, if ever, to own a gun.He does not get a gun under these circumstances. He does get mental health assistance.3. Medium size data should have sent out an alert — a kid has bought 4 guns and 10K rounds — that got someone to go see him.Amazon can tell me exactly where each and every one of my shipments is by day, hour, minute — these gun purchases can be tracked similarly.Detectives investigate murders. A detective should have gone to see this kid — verdict? No guns or ammo for this specific kid.The detective should have checked on this kid from time to time.When a murder occurs, a detective goes to visit his snitches to see what the word on the street is and he typically hears something. Detectives are locally mobile.4. This kid should have been on a list. A nationwide list. Medicine, therapist should have won him a spot on The List.If you can list up perverts, we can list up potential mass murderers.None of these common sense notions assails the Second Amendment and today they can all work.None of them is about taking my guns. They are smarter than that. They are preventing the mentally ill persons hands and trigger finger from coming in contact with a deadly weapon and society is being alerted to his existence..

          1. Matt A. Myers

            I agree – those checks, at minimum, should be in place – though where there’s a will, there’s a way – so as this does limit immediate ‘satisfaction’ and success with obtaining weapons, it won’t prevent it as much as if you went for a two-pronged solution which included cultivating a society that kept / lead people away from crime and mental illness.

          2. Mark Essel

            You’re a step ahead of me- just read this after my comment above. The people who can put others on the “watch list” would require variation/screening to avoid abuse. No need to kick up a witch hunt, that wouldn’t help at all.

      2. Mark Essel

        Root cause, I’m a fan of this practice. Treat the symptoms or the cause, the choice is entirely up to us.What social situations, environment, genetic disposition, upbringing leads to murderous rampages? Without going after the real problem legislation is ghost chasing.Minority ReportWe have potent data mining tools at our disposal, can they help identify and target pre-criminals and at the very least minimize their access to lethal weapons/keep an eye on them. Isn’t this how “terrorists” are discovered? Does this type of analysis infringe on the right to privacy of US citizens – yes. Would this system be fallible, of course. It’s a statistics game where your law enforcement agency struggles to minimize the presence of mass murders, while doing it’s best to preserve individual rights. Such an agency would need strict oversight.

    4. JLM

      ,OK as a real world guy who was once upon a time an expert on foreign military weapons, let me tell you what really happens:I get two magazines.I cut off the bottom of one and the top of the other.I true them up and tack weld the bottom half to the top half.I finish up the weld, give it a bit of a grind for looks. Maybe a quick bit of flame from a torch and some gun blue.I put two springs (from the two magazines) in the bottom of the new magazine.Voila! I now have a high capacity magazine.This nibbling around the margins is not going to dissuade even a minimally talented gunsmith. It is all a waste of time.The challenge of gun regulation is not the size of the magazine.JLM.

      1. ShanaC

        how many people are minimally talented gunsmiths/would take the effort out to do this?

        1. JLM

          .It only takes one, right?Would take less than hour to do.Need one for your AK?.

          1. pointsnfigures

            Besides, aren’t that what “maker labs” for?

        2. Dave W Baldwin

          @JLM:disqus is making a good point and joining that to the video above, you have the issue of, “What is the intent of the gun user?”. If he/she thinks they have to fire as many shots possible in short time period, they will figure it out placing the odds way above the 80/20 (won’t/will) as mentioned in the video re poker. At the same time, throwing around political formulas doesn’t take into account that the gun user maybe just happened to use that gun, so even outlawing that gun wouldn’t result in he/she using another.

      2. MalloryBerman

        …..—–goo.gl/uR4nF (Click on Home)

    5. ShanaC

      meh, I still think my put it into the printer solution works best.

    6. fredwilson

      we should ban the 3d printing of guns that are banned. but not the guns that aren’t banned.

      1. kidmercury

        I agree I hope you are interested in ensuring such laws can be sufficiently enforced. Right now there is no real plans for enforcement. If enforcement is to be successful it needs to occur at the printer level, otherwise the pro 2nd amendment people like defense distributed will get around and ignore such legislation as they are already suggesting they will do.

      2. Modernist

        okay, but to what extent is your ban enforced and how?

        1. fredwilson

          you go to jail if you are caught doing it?

          1. thinkdisruptive

            And, when that doesn’t work, we just ban 3D printers. Problem solved.

          2. kidmercury

            That will be about as effective as media Piracy laws. Piracy has skyrocketed thanks to the Internet. Unless regulation is stronger, the same can be expected regarding gun production.

    7. thinkdisruptive

      It’s worth noting that it would be hard to find two congressmen with an IQ higher than cooked spinach.

  16. jason wright

    how many times does Nate say “right”?

    1. fredwilson

      every three seconds. he should work on that.

      1. JamesHRH

        Classic highly analytical verbal tick.Essentially translates to – ‘I see it, don’t you?’

        1. jason wright

          it was starting to give me a neck twitch tick 🙂

      2. jason wright

        he should. it was giving me classroom flashbacks.

  17. jason wright

    is it still illegal to publish opinion polls prior to election day in France? i believe it is, and that polling is corrupt.

  18. John Revay

    “I am currently reading Nate Silver’s The Signal and The Noise”Curious – what is your e reader of choice – kindle or nexus 7

    1. fredwilson

      kindle app on the nexus 7

  19. ShanaC

    How is “The Signal and the Noise” from our favorite stat witch 🙂

  20. Kirsten Lambertsen

    I made – and lost – my first Super Bowl bet ever this year, thanks to Nate Silver. True story. Curse him! Oh, all right. I’m still a fan.

  21. Guest

    It’s all about sample size, right?

  22. johndefi

    The 4th down question is great. I would love to see that data segmented, though. Not sure it makes sense at all times for all teams.

  23. Ryan Jenkins

    Most people don’t know this, but very high-end printers today will contact authorities if you try to photocopy currency. This was put into effect over a decade ago when printers became capable of printing in extremely high resolution. If you try to disable the feature, it also notifies authorities. A similar solution could work here.

  24. jason wright

    publishing polls is also banned in Israel prior to an election;http://www.guardian.co.uk/w…i think the US (and other countries) has something to learn here about the technical enactment of democracy. the media industries favor power rather than the people and distort the process. much to do.

  25. jason wright

    still no votes