Archive for the ‘Stuff’ Category

Visual acuity redux

Tuesday, May 21st, 2013

A visual acuity test (or something like it)” was prepared in conjunction with an essay on differences (I’d call it “discrimination” but that word’s been damaged) for the August Cites & Insights–and, sigh, I think I did the image wrong. There were supposed to be six rectangles on a pure-white background, with each rectangle differing by having one or two of the primary colors (RGB) reduced from #FF to #FE.

Somehow (still haven’t figured out how, unless JPEG is to blame) I wound up with only five colors–three of the six rectangles had the same color.

I did get some useful comments, including at least one person who apparently has extraordinary color acuity.

This time, I’ll give you the image and tell up front that there should be ten colors on it–nine off-white shades with two narrow bands of pure white between them. This time, I closed the file and reopened the JPEG version; Paint.NET’s color picker does show nine different values. If you’re like me, you may be able to see the bottom three rectangles if you look really hard. Some of you may see some of the center row.

This time, the top row should be one FD and two FFs; the center row was intended to be FCFFFF, FCFCFF, and FCFCFC respectively; the bottom row goes all the way down to FB for one of the three primaries in each box. I say “should be” and “intended to be” because reopening the JPEG version shows some shifts in the top two rows–still nine colors, but not precisely the nine I started out with.

Don’t think you can actually perceive 16 million colors? You may be right.

On the other hand, there are cases where–at least to my eyes–a one-digit change in one of the three primaries is visible if you put the two colors side by side. I have that example in the essay (which is mostly not about color acuity but about differentiation in general).

Anyway, here’s the ten-color image, for what it’s worth:


A visual acuity test (or something like it)

Monday, May 20th, 2013


How many colors are in that image?

“What image, and why does the text start so far down?”

That’s a reasonable reaction.

If you see anything above my question other than an undifferentiated field of white, I invite you to answer the question (the “How many” one) in comments.

If you don’t, that’s OK. (Do I? That would be telling…)


A very silly post of consequence whatsoever

Tuesday, May 14th, 2013

That’s right: Another post about video poker (which means comments are off because most of them wouldn’t get through filters anyway).

Not completely about video poker, though: Also about a silly statement in a pretty good book about statistics, one I read when I was first pondering the possibility of writing a book about coping with everyday statistics.

Still pondering, as a recent post made clear. No decision yet.

Anyway: The book said–accurately–that neither (honest) dice nor (honest) cards (nor, for that matter, honest slot machines) have memories, any more than an honest coin does. Therefore, getting, say, fifty heads in a row when flipping a coin does not mean that the next flip is “nearly certain to be tails.” Assuming the coin’s legit, it means the odds are even–50% chance you’ll get tails, 50% chance heads.

Similarly, if you’ve played 40,000 hands of draw video poker and haven’t received a royal flush (which should come up roughly one out of every 40,000 hands), the chances of getting one on the next play are, well, 1 in about 40,000. And if you get two royal flushes in a row, the chances of getting a third one are 1 in about 40,000–even though the chances of that three-royal-flush streak are truly small.

All of that’s accurate. Here’s the part that’s not: The author says that there are no such things as streaks.

That’s nonsense. In fact, the nature of random play means that there almost have to be streaks–periods during which the randomness doesn’t seem random. What might have been said honestly is that a streak can end at any time, because it’s just a series of random events.

But there are streaks. The last time I played video poker for money (which was quite a while back, the last time ALA Annual was in New Orleans), I had a hot streak: Although I was playing machines with roughly 96% payback (because I wasn’t betting five coins at a time, so wouldn’t get the Big Payout for a royal flush), I was probably averaging 105% payback. For two days (maybe three hours total).

It works both ways, to be sure. And therein hangs the tale of the current silly post, a long way of saying “Aarrggh…”, the non-wagering video poker site run by the company that produces most video poker slot machines, just added a third contest to its other two (a daily $50 contest and a monthly contest with several winners running up to $500): A monthly contest to reward those of us who win a fair number of rounds (a round is 100 deals; a deal is usually three hands but sometimes only one and sometimes 25-100) but rarely get lucky enough to get the day’s top score. The new contest gives you a point each time you win a round at all–and gives you another round when you want it (you’re normally limited to either 5 or 8 rounds a day, depending). Whoever has the most points at the end of the month wins.

I’ve played when I felt like it, and did so-so, and saved my extra points for games that I really like and have usually done well at. Normally, I can figure to win 25%-40% of rounds; that’s reasonable for games with 98.5% payback [full Nevada odds, varying slightly depending on the game]. These two variants have actually been somewhat better, typically in the 40%-50% range.

So yesterday was one of the “good” ones. I figured to win 2 or 3 of the eight standard rounds and keep playing those extra credits until they were gone. Played eight rounds. Three of the eight had >90% payback; two of those three had >97% payback. But not one round was a winner. I used two existing extra credits to play ten rounds. Not one winner. That’s exceptionally bad for this particular variant–for any variant, actually.

Today? Another good one. So far, I’ve done five rounds. Every round >90%. Two rounds at 99% or higher.

Not one winner.

So, yes, there are streaks. Hot streaks and cold streaks. Right now, if you suggested going to a casino, all housing and travel expenses paid, I’d probably laugh: At least with, there’s no possibility of losing money. (And no second-hand smoke, and no music unless I want it…) My cold streak could snap any time–possibly the very next time I play–but right now, it’s making it remarkably easy to write, read, weed (I hate weeding!), watch TV…

The only lesson here: Of course there are streaks. If there are no streaks, you’re playing a crooked game–one that has memory*. But, also of course, streaks are neither predictable nor meaningful in the long run.

*Real-world blackjack and table poker do, of course, have a form of memory–cards that have already been played can’t be played again until the deck’s reshuffled. Different issue.

Update 7:30 p.m.: It is, of course, entirely coincidental that after posting this lament I managed to win four out of the next six rounds, after having lost 16 in a row. Doesn’t mean I’m suddenly on a hot streak, only that the cold streak seems to have ended. For now.

What’s going on

Tuesday, April 23rd, 2013

A little randomness that may help explain why posting’s been even lighter than usual the past couple of weeks.

And if you get Marvin Gaye as an earbug–you won’t get an apology from me, as there are few better songs/singers available.

Oregon and Washington Librarians

I hope to see a few of you over the next few days. Later today, I’ll fly to Vancouver, Washington for the 2013 joint Oregon/Washington Library Association(s) conference, where I’ll do a three-hour preconference on open access and talks on “Give Us a Dollar…” and micropublishing.

OK, so I won’t actually fly to Vancouver. I’ll fly to Portland. And will probably spend longer on BART getting to SFO than I spend in the air. I love Livermore, but for flights–much rarer now than in the past–it’s less than ideal.

Preparing for Oregon and Washington

I’ve been busy preparing for those talks. Other than one talk last year, I really haven’t done much of any speaking, and I’d like to do a good job.

Actually, I started preparing quite some time ago–and preparations included a special Oregon/Washington version of Give Us a Dollar…, which I think is a good model of what I could do for other states/regions if anybody wanted it.

The link is to the free, $0, no-cost PDF version–did I mention you don’t have to pay anything for it?–that can be turned into a neat little booklet using a duplexing color printer and quick instructions here. There’s also a color hardcover edition–the book has multiline graphs that require color–but so far my own copy is the only copy, and that’s OK.

Preparatory work for the OA precon has resulted in two offshoots…

Hot Times for Open Access: The June Cites & Insights

When I devoted 60,000 words to OA in the January and February Cites & Insights, I thought I was done with it for the year. But the last few months have been unusually active–so much so that the June issue will be almost entirely devoted to another OA roundup.

Look for it some time next week, probably right around May 1 or 2.

But there’s also…

The Big Deal and the Damage Done

Working on the precon and the issue, I ran into another of those statements asserting that the Big Deal was a wonderful thing all around and that it essentially solved the serials crisis back in 2004.

Wayne Bivens-Tatum also did a nice blog post, “Politics, Economics, and Screwing the Humanities,” which reminded me of what was getting damaged if serials prices were still rising too quickly: Namely, the humanities (which still depend on books) and the flexibility of academic libraries to be anything other than licensing agencies. I suspected that the damage was complex, and decided to investigate just a little.

The result will be an ebook (and paperback book–while it has lots of graphs, I’ve designed them so they’re workable without color if need be), probably out in the first half of May. Several of the graphs prepared in the process will show up in the OA precon PowerPoints.

Portions of the book will probably appear in the July 2013 C&I.

It will be interesting to see whether the response is as overwhelming as it has been for this year’s two previous library-analysis pieces, the March issue and the May issue, which so far have substantially fewer downloads than other issues.

Oh, as to the apparent random use of indented quotations: Think of them as asides. Or not.

Reason I may yet write that stats/numbers book, #346

Wednesday, April 3rd, 2013

It’s just a little item in the April 2013 Fast Company, over in the bottom-left corner of a page, about the sales of steel wallets—mostly (I think) sold to people who are worried about RFID chips in their smart credit cards or passports being read.

Here’s a rough version of the graph that appeared with the piece:

When you see that, you’d tend to agree with the text of the article: “But apparently, the freak-out has already faded.”

But the graph is showing percentage increase in sales, year-over-year.

So, just for fun, let’s assume that 100,000 wallets were sold in 2006 (the actual number doesn’t matter) and draw a new graph using exactly the same data:


Now, does the assertion in the text seem quite as valid?

No? What? You say “Damn—those are pretty fast sales increases since 2009”? Well, you’re not Fast enough for FastCo.

Of course, there’s another aspect of this: We don’t know what the baseline for 2006 was–and percentages don’t mean much without a baseline. For all I know, maybe steel wallets were such a novelty item in 2006 that only 100 of them sold…which would make 2012 sales 1,905 and that Ginormous Jump from 2008 to 2009 an increase of…well, 675 wallets. Whoopdedo.


A quick reality check on the economics of Netflix

Saturday, February 9th, 2013

Netflix has done a great job of adding streaming-only subscribers and losing disc subscribers, to the point that, at the end of 2012, it had more than three times as many paying domestic streaming subscribers as paying domestic disc subscribers.

And, of course, that has to be treat for Netflix financially, right? It gets rid of all that postage?

So here’s the quick reality check–come up with your answer without actually, you know, looking it up.

How much more profitable is Netflix’ domestic streaming operation than its domestic DVD operation?

[I mention domestic because, if you include international streaming…well, let’s not go there just yet.]

For those disinclined to actually check the spreadsheet, I’ll add the answer on Monday.

Schwab, on the other hand…

Monday, January 28th, 2013

Given the grumpy tone of this post about our travails with a Panasonic plasma HDTV, I thought a little karmic balance might be appropriate. To wit, a recent encounter (begun last Thursday, completed today) with Charles Schwab, with which we have a modest account.

We have little CDs maturing. We find the current rates for CDs, thanks to the Fed’s “Punish the Savers” initiatives, absurd.

Not sure why the Fed has determined that savers are required to shove all our money into the stock market, but that appears to be the case. Maybe gambling interests have taken over behind the scenes and want to demonstrate that everything’s a gamble? Is getting 3%–or, say, 1% above inflation–on guaranteed savings such an outlandish notion?

So we were looking at possible safe havens–low-risk investments that might yield, say, 3% or 4% without losing capital and with flexibility if interest rates are ever allowed to return to levels that encourage actual savings.

One mutual fund–part of Schwab’s Select List, with no purchase or redemption fees and very low overhead–looked promising. But there was one aspect of the information card & prospectus that troubled us.

So I called last Thursday and talked to a representative about it. He didn’t have an answer, and wasn’t sure he could get one, but said he’d look into it (emailing the fund manager, since it’s a Schwab fund) and call me back. I think he forgot that he wasn’t working on Friday, since he said he’d call me back the next day.

This morning (Monday), he did, in fact, call. He read me the response from the fund management, which was in English (not Managementese or Legalese), was understandable, directly addressed our concern, and was actually a satisfactory answer. He also forwarded the response to my email, so I’d have it in writing.

Did I mention that this was free? (OK, so I do have a Schwab account…)

I think we already felt positively about Schwab. After this interaction, we feel more positively about it. Kudos for responsive customer support on a difficult question.


The year in posts & readers

Tuesday, January 1st, 2013

It’s the time of the year when some bloggers recount how many thousands of visitors they had in the previous year and various other astounding statistics (maybe even as an infographic!)

So, here goes:

Some people read Walt at Random last year. A very few people even commented (plus thousands of spammers, to be sure).

The actual numbers are neither especially trustworthy (how many were people and how many were spambots and spiders?) nor particularly interesting.

Happy New Year and, if it works for you, keep on blogging.

Wrapping up 2012

Monday, December 31st, 2012

It’s been an interesting year–both in the apocryphal “Chinese curse” sense and in a more positive sense. So, what the heck, here are a few idle end-of-year musings, of no special import.

Better than 2011 (at least in professional terms)

Some comparisons with 2011 seem reasonable…and encouraging.

  • Speaking: My already-slowed pace of speaking invitations disappeared entirely in 2011, with not a single speech requested or given (after one in 2010 and two in 2009). In 2012, I spoke once (at Internet Librarian), in an odd sort of invited/requested situation. I’m pleased to say that at least one two-state group (or one committed person and groups he was able to convince) think[s] I still have something to say: I’ll be speaking at least three times in 2013, albeit all at the same conference.
  • Writing 1–Paid print columns and articles: This one’s less encouraging. After 27 years of having at least one (and at one point four) ongoing paid print magazine columns, the end of 2011 also saw the end of the last such column…with a single paid article in 2012.
  • Writing 2–Cites & Insights: For various reasons, some personal, some otherwise, I came very close to giving up C&I altogether in 2011, which saw the fewest issues ever (nine), the fewest words since 2002 (214,521), and the fewest pages since 2002 (274). I thought turning all my energy to book writing–for real publishers–definitely made financial sense and might make sense otherwise. That might have been the sensible decision, but C&I started creeping back in. 2012 was back to 12 issues (if not quite monthly, since a combined January-February issue was balanced by a special Fall issue), and relatively long ones at that (312,023 words in 394 pages–actually the second-highest word count, with 2009 higher, and the third-highest page count, with 2009 and 2010 higher). My goal for 2013? Do issues that I find fun to do and somewhat worthwhile; no specific goal for length (which has never worked out) or number. (Since the first issue of 2013 is a long 40 pages and the second is sure to be at least roughly that long, saying “16 to 24 pages” seems a bit ludicrous.)
  • Writing 3–Books: 2011 saw my first professionally-published book in eight years, Open Access: What You Need to Know Now. 2012 saw my second, The Librarian’s Guide to Micropublishing, and my third–Successful Social Networking in Public Libraries–should be out right about now. Having finally given up on books about liblogs and library blogs, I tried something more directly relevant to public libraries, Give Us a Dollar and We’ll Give You Back Four (2012-13). The jury’s still out on that one, and I wish I had a way to market it to library trustees, who are possibly a natural audience for it. Here, 2013 is currently a void: I have no specific projects in mind.
  • Writing 4–Blogging: You can see the numbers in the right sidebar. 2011 was down from 2010; 2012 is up (slightly) from 2011. Admittedly, a lot of the 2012 posts are related to one particular project–but fewer than a third of the total for the year (49 of 188). I have a little list of posts I should write. One of these days, I will. (No, this ramble isn’t on the list.)
  • Social networking: I haven’t yet been asked to leave by LSW on Friendfeed, and I don’t think I’ve quite alienated everybody there. I continue to check in on Facebook, Google+ (where I started an LSW group as a placeholder) and Twitter, with relatively little actual activity on any of them…and, once in a great while, on LinkedIn, with even less activity. I don’t think I’m ever going to be an SEO-worthy Social Networker and Personal Brand Builder, and that’s OK.

Anniversaries and other oddities

In September, my wife and I attended my high school graduating class’s 50th reunion. Fifty years. (Thanks, in part, to George Lucas, who I never knew in high school and still haven’t actually talked to, the Class of ’62 continues to have large, lively reunions every five years.)

In October (I think) my wife and I attended her brother’s 80th birthday party.

And tomorrow (January 1, 2013), my wife and I will celebrate our 35th anniversary at New Year’s Day brunch–with a good friend of ours, at (oddly enough) a golf course restaurant in Livermore. Thirty-five years…and looking forward to quite a few more.

After more than 35 years of subscribing to the San Francisco Chronicle as a daily print newspaper, we became digital converts and tablet owners, largely because the print paper had just gotten too expensive ($559/year as compared to $71.88/year on the Kindle or $60/year on the iPad) and its arrival too uncertain in the morning. We’ve now owned a Kindle Fire HD 8.9 for…lessee…eleven days. We had two weeks to try out the Chronicle Kindle version before deciding whether to drop that and send back the Kindle (and probably get an iPad); it took us four days to decide, cancelling the print subscription. The Kindle’s staying; next up is to get a case and stand. (My wife crafted a temporary cardboard stand that works beautifully, but it’s not a long-term solution and won’t do squat to protect the Kindle when we eventually start traveling again.) I’ve blogged about that once and probably will again. We’re still not ebook people; I downloaded the free Sherlock Holmes collection on the day it was free (which mostly means a free introduction + nicely-formatted Project Gutenberg texts of public domain fiction), but haven’t really looked at it. When we start traveling again, we might add some actual ebooks, but that’s not why we got the Fire HD–we got it as a newspaper, and so far that’s what it is. I miss the print paper…but not all that much.

The downside, at the moment: Our ^#T(* Panasonic HDTV still isn’t repaired, thanks partly to Panasonic’s apparent policy of delaying everything as much as possible (seemingly even replacement parts). Since October 7, 2012, we’ve had a working TV for 33 days and a big ugly sculpture for 51 days (from October 7 until November 1, and from December 4 until…well, it’s still not fixed). We are sick and tired of the whole situation. (Thanks to my brother, we have a 19″ Polaroid with which we can watch DVDs on the itty-bitty screen; it lacks a QAM tuner, so broadcast TV via cable isn’t happening.) The Panasonic had a great picture: I remember that. I also remember everything else that’s happened since it failed the first time… We will not be watching any NYE countdowns, but that’s OK.

We’re both reasonably healthy, we’re not starving, we live in a town and neighborhood and setting so nice that it really is like being on vacation all the time (one of several reasons we haven’t really done much traveling in the past 2-3 years)…and I hope that all continues.

I’ll see some of you in Vancouver, Washington in April; I’ll chat with others of you on Friendfeed and possibly elsewhere; I wish you all a good new year.

Losing credibility with one quick answer

Sunday, December 30th, 2012

When you’re an Expert, it helps to know your facts. Especially if they’re Facts people might not want to hear.

And it doesn’t help to make up Facts on the spot.

This is just a little rant. Sorry. You can skip it if you want.

On my way to and from the library today (the public library, and yes, it was busy, as it always is), I was listening to small chunks of a conversation involving some doctor (didn’t catch the name) who’s clearly an Expert on food and health, with lots of books and lots of advice–eat lots of soy, don’t eat anything processed, avoid fructose, lots of cooked Asian mushrooms but no raw or American mushrooms, do this, don’t do that. [Possibly worth noting: this was on a PBS station.]

And the interviewer mentioned a local attempt to tax soft drinks at one cent per ounce, an attempt that went down to defeat.

At which point, the Expert made a pat, convenient statement:

If we taxed soft drinks at one cent per ounce, we’d raise enough money to pay for health care for everybody.

Fortunately, I held on to the wheel tightly as I said “That can’t be true.” Now, I’ll admit, my wife and I are outliers: Neither of us drinks soda at all (except that my wife occasionally uses a little ginger ale to try to clear out her throat–maybe two ounces in a day).

But still…

So I came home and did a little investigation.

I wasn’t aware that American consumption of soft drinks has been declining for some years, but that’s a different issue. I took the current figure (2011 consumption) that I could find repeated by more than one apparently authoritative body and a much higher figure (date not given) reported by an Alarmist You’re All Going to Die Fat organization.

Here’s what I found:

  • If you take the Alarmist number, a one cent tax per ounce would raise $72 per person per year.
  • If you take the apparent 2011 figure, that tax would raise $57 per person per year.

I don’t know about your health care costs, but I’m guessing the general average is just a wee bit higher than $57 to $72 per year.

Just a wee bit as in two orders of magnitude.

Those per capita numbers add up to between $18 billion and $23 billion. That’s billion with a B: 315 million Americans times $57 to $72.

Health care expenditures in the US appear to be on the order of $2.5 trillion. That’s trillion with a T.

Yes, health care expenditures in the U.S. are way too high–but even if we managed to get them down to $1.8 trillion, a 28% cut, that would still be 100 times as much as the more probable figure for “one cent per ounce.”

Here’s the thing…

The Expert was saying provocative things. I found some of them annoying and more than a little preachy, but hey, he’s an Expert: Maybe I should look into them further.

And then he popped out that line.

And I thought: Why should I believe anything else he says, given that he’ll make something up on the spot that’s so wildly wrong? (He’s a doctor. Either he really believes that most people drink, what, ten gallons of soda a day–which would still only be $12.80 per capita per day in soda tax, not close to enough to pay for health care–or he’s terrible with numbers.)

Credibility: Lost. Just that easy.