# David Letterman and me

In the latest episode of The Incomparable, Jason Snell talks Letterman with Aaron Barnhart, Andy Ihnatko, John Gruber, Tim Goodman, and Philip Michaels. In a departure from his usual panel format, Jason interviewed each of his guests separately and edited them together topic by topic. It’s a nice piece of work, and you should give it a listen if you haven’t already. If your thirst for Lettermania is as unquenchable as mine, you’ll also want to listen to the full interviews in the 3½ hour bonus track episode. You might even want to read on for my disjointed reminiscences. I’ll try not to pile on with things that Jason and his guests already said, but will fill in around the edges.

I’m 10–15 years older than Jason and most of his guests (Aaron, I suspect, is closer to my age). When Late Night premiered on NBC, I wasn’t a little kid who needed permission to stay up late.1 I was a 21-year-old college student—pretty much the show’s target audience. And it hit the target. I loved it from the start, as did all of my friends.

Our love was, unfortunately, not shared by the management of WICD, the local NBC affiliate that served Champaign-Urbana, Illinois. They apparently thought it was too weird to appeal to what they considered their core audience of farmers and small-town shopkeepers, so they soon replaced Late Night with Thicke of the Night, starring the unbelievably bland Alan Thicke (pre-Growing Pains) and his soap opera star wife, Gloria Loring. This left me with a huge grudge against all things Thicke. This year’s $7 million judgement against their horrible son for “Blurred Lines” has only partially balanced the karma scales. At some point, though, the geniuses at WICD realized that farmers and county extension agents don’t watch TV after 11:30, but college students do, so they brought Dave back.2 And there was much rejoicing among fans of toast on a stick. But let’s not get ahead of ourselves. All proper David Letterman retrospectives must start with a discussion of Johnny Carson, and it just so happens that I watched a lot of Carson when I was in junior high school. I’m not sure what was up with my parents, but I spent three years watching Johnny almost every night. And this was back in the early-to-mid-70s, when The Tonight Show still ran for 90 minutes. By this time, Johnny was about a decade into his run and had cleared out what little competition there had been. Wikipedia says there was some limited version of Dick Cavett’s show still running on ABC, but it was effectively dead. Carson was the king of late night TV, and the rituals of the show were well-established, especially show opening, with “Heeeeeere’s Johnny!” followed by Carson coming through the curtain, a cut to Ed McMahon doing his little namaste bow, and then another cut to Doc Severinsen doing his bigger bow with that elaborate hand-rolling motion. The monologue was a set of topical jokes—in those days, mostly about Nixon and Watergate—with a very specific rhythm that impersonators were just starting to pick up on. At this time, despite there being nothing of note on ABC or CBS, Carson was still, I think, basically an NBC employee. He had not yet won the contracts that shortened both his day and his week and gave him control of the time slot that later became Late Night’s. In fact, as I recall, Johnny was doing a lot moonlighting during this time, using his days off from The Tonight Show to book short but very lucrative runs in Las Vegas. I thought of this while listening to Jason and his guests speculate on what Letterman may do after The Late Show. They mentioned that while Carson talked about doing other projects after The Tonight Show, he never did. That’s true, but they either forgot or never knew about the non-Late Night side projects Johnny had when he was younger. Remember his clothing line? One of the most interesting things about Carson’s run is that there was almost never a time when people weren’t talking about when he would be replaced and by whom. This seems odd now, but you have to realize that his predecessors, Steve Allen and Jack Paar, had hosted the show for only three and five years, respectively. In the 60s, the thought was that he’d be unseated by someone on another network. During the 70s, as The Tonight Show maintained and extended its dominance, speculation shifted to who would be his successor. After his success in the first year of Saturday Night Live, Chevy Chase assumed that successor would be Chevy Chase. Johnny had other ideas. He was widely, albeit unofficially, quoted as saying Chase “couldn’t ad lib a fart after a baked bean dinner.”3 As Johnny’s contract improved, he knew he needed someone to sit in on his increasing number of days off, and he knew that person had to be quick and at least appear to be self-deprecating. Eventually, he knew that had to be Dave. There is a great episode of The Tonight Show that unfortunately I can’t find on YouTube. Carson is the host, and his first two guests are Chevy Chase and Dan Aykroyd, promoting a movie. This is some years after “couldn’t ad lib a fart,” and everyone knows about it. But the Chase/Aykroyd segment goes well enough, and they move down the couch for the next guest. It’s Dave, and he’s reached the level where he doesn’t have to do a stand-up bit anymore—he goes straight to the chair next to Johnny’s desk.4 And he kills. Chevy has to sit there knowing full well that Dave is the heir apparent to the throne he had assumed was his. And it seemed clear at the time that Johnny had orchestrated this show just to rub Chevy’s nose in it. Irony is the much-used word to describe Late Night, and I’m afraid I can’t think of a better one. The show appealed to people who’d grown up watching television in general and talk shows in particular, people who’d consumed a lot of what Paul Shaffer called “this business of show.” You knew the conventions of the format, Dave knew the conventions of the format, and you both knew that each other knew the conventions. That was the starting point, and that was why Late Night was able to flip or even ignore those conventions. It wasn’t sarcastic or mean, at least not usually and not when it wasn’t warranted. Mostly it was gentle and silly, with bits like Stupid Pet Tricks that harmed no one. But Dave could be mean when the occasion called for it. One of my favorite interviews was with Andrew Greeley, a priest who wrote bestselling novels in a Jacqueline Susann style with a Roman Catholic twist. He was kind of an asshole, and had spent a lot of time on talk shows acting clever. One of Dave’s questions was how a celibate man could write so much about sex. This was the kind of softball question that Greeley could have answered in many ways. He chose to answer by saying “That’s my affair,” expecting a laugh for his great wit. Dave pretended not to understand the double entendre and slowly, painfully extracted from Greeley an explanation of the answer. When Greeley was done, Dave said, “Oh, I see. It’s kind of like a joke.” In The Incomparable, Jason said he thought Late Night was the product of three people: Dave, head writer Merrill Markoe, and director Hal Gurtner (that’s Gurnee, Dave). I agree and would be remiss if I didn’t link to a few recent articles by and about Markoe, who deserves every accolade she gets. I think almost everything she does is terrific. I remember seeing an episode of Late Night in which she appeared as a guest. This was some time after she’d broken up with Dave and quit the show to write other things. The ostensible reason for her appearance was to promote a book, but I think the real reason was to make Dave uncomfortable. I don’t remember the details, but I got the clear impression that everyone on the staff adored her and was happy to see her again, but that Dave didn’t quite know how to handle the situation. Which made for a great segment, which I’m sure both Dave and Merrill appreciated. I saw one episode of Late Night in person, during a week in 1989 when the show was in Chicago. It was recorded in the Chicago Theatre, which was kind of like the revamped Ed Sullivan Theatre that the show moved a few years later. The guests were Michael Jordan (before his appearance with Stuart Smalley on SNL) and Jay Leno (back when he was still edgy and funny) and it was great. My wife didn’t go with me because she was breastfeeding our baby daughter and didn’t think she could leave her with grandma for a whole day. Kids can really ruin your life. In his interview with Jason, Andy Ihnatko said Late Night was like the iPhone in that after it appeared, everything changed. I kind of agree, but not fully. There’s no question but that Late Night got everyone’s attention, and there were lots of magazine and newspaper articles about the new Age of Irony. But although the talk shows we see now are clearly the progeny of Dave, that really wasn’t the case right away, at least not on the big three networks. Sure, we now have the Dave-infused Conan and Kimmel shows, but I don’t think anyone would argue that Arsenio Hall’s or Joan Rivers’s shows were direct descendants of Late Night. But Dave certainly was getting through elsewhere. My wife went through a phase in the late-80s where she’d stay up late and watch Spanish language TV. And she found a talk show in which the host was clearly doing Dave. The rhythms of his speech, the looks he gave the audience: all Dave. I don’t speak Spanish, but even I could see what the guy was doing, and that was before I saw him throw index cards. So Telemundo—or maybe it was Univision—was clearly ahead of the English language networks. And I’ve read that there were European Daves, as well. Like Jason and his guests, I’ve focused on Late Night rather than The Late Show because it was the innovator. When The Late Show started I watched it and enjoyed it, because it was a really good show. But it wasn’t quite the same. The first thing you noticed if you were a fan of Late Night is that The Late Show’s monologue actually tried to be funny. And it was, but that was never the point of Late Night’s monologue. The second thing you noticed was that Paul Shaffer was put in the background. He eventually came back to the prominence he’d had on the NBC show, but there was obviously an initial decision that he was too weird for 10:30. I assume he was well compensated for the loss of status. One thing The Late Show has that Late Night didn’t is a good interviewer. Dave’s interviews in the early years were entertaining, but only because they were off-center. Now he can do off-center or uncomfortable if he wants to, but those aren’t the only tools in his toolbox. Dave can now do the kind of warm interview that Carson used to do with Jimmy Stewart or with children. Interestingly though, Dave almost never does that kind of interview with kids because kids now are used to edgy and uncomfortable. They’ve grown up in the Age of Dave. 1. I also had the advantage of living in the Central time zone, where the show came on at 11:30 instead of 12:30. 2. Also, Thick of the Night had the plug pulled on it, so they had no choice but to go back to Dave. 3. Phil Michaels says this referred to the entire SNL cast, but I remember it as referring to Chase alone. 4. I think it’s after Late Night has started, but I’m a little fuzzy on that. # Know that you were doing wrong When the deflate-gate story hit back in January, I didn’t write anything about it because I didn’t want to add to the confusion. It wasn’t that I didn’t know how to use the ideal gas law, it was that I understood it well enough to know that there wasn’t enough publicly available information to use it correctly (except by accident). With the release of the Wells report—and more important, the Exponent report that’s included as Appendix I of the Wells report—we now have the information that was missing back in January. Let me start by saying that I know Exponent pretty well. I’ve been reviewing their work for about a quarter of a century, including a few reports by one of the authors of the Wells appendix. I am not in awe of them or their reputation as “one of the leading scientific and engineering consulting firms in the country.” In fact, I’ve often found crippling errors in their work, and it’s possible that if I went through this recent report on football inflation in detail—as I do when I’m being paid to review their work—I’d also find errors. But I don’t need that kind of detailed review to know that Exponent addressed the right issues and went about its testing and analysis the right way. I should also say that I am intensely jealous of them. This must have been an extremely fun project to work on. Not only are the engineering issues relatively simple, the stakes are low. This is not a product design in which people could be injured if the analysis is wrong, nor is it an investigation in a lawsuit in which millions of dollars hinge on the results (although today’s suspension of Tom Brady could change that). And fun is only part of it. At Exponent’s usual rates—and I don’t see any reason they’d give the NFL a discount—this investigation would bring them hundreds of thousands of dollars. But let’s get back to the engineering. In January, when everyone was dusting off their high school chemistry books and running ideal gas law calculations, three fundamental pieces of information necessary to make those calculations valid were missing from every report I read: 1. Does the volume of a football remain essentially constant over the range of pressures of interest and when the ball gets used during a game and changes from dry to wet? 2. How accurate and how consistent were the pressure gauges used by the officials? 3. What was the temperature of the air in the footballs when their pressure was checked at halftime? It’s certainly reasonable to make an initial calculation assuming that the volume of a football remains constant. But the pressure changes of interest in this matter were in the neighborhood of 1 psi, which is only about 4% of the absolute pressure in the ball. If the volume of the ball changes by, say, a couple of percent when it gets wet, the error in a constant-volume ideal gas law calculation will be significant and will invalidate the result. Exponent reviewed this issue by using 3D scans of a ball in various conditions and concluded that any volume changes were low enough to have no significant effect—constant-volume calculations were valid. Figure from Exponent report. As for the accuracy and consistency of the gauges used by the officials, Exponent got the actual gauges used on the day in question (and a bunch of other gauges from the same manufacturer) and tested them. They used these test results to account for gauges’ variability and bias. Figure from Exponent report. By the way, did anyone back in January report that the officials used two gauges? I don’t remember seeing that. Most important, perhaps, was the temperature of the footballs when they were tested at halftime. Everyone defending the Patriots back in January did their calculations assuming that the temperature of the footballs at the time they were tested was the temperature on the field. This leads to the largest possible calculated pressure drop—which is in New England’s favor—but made no sense to me. Is it realistic to assume that the officials did their halftime measurements out on a cold, wet field? No. But if they took the measurements in the locker room, how long were the balls in there before the testing and how rapidly does the air in them rise back up to room temperature? There is no definitive answer to the first question, although apparently the time before testing can be bracketed through witness accounts. Exponent did answer the second question by instrumenting a ball with a pressure gauge and a thermocouple to track the rate of temperature and pressure rise as the ball was moved from a cold environment to a warm one (like I said, this is fun stuff). They used that data along with the bracketed time estimates to get more realistic values for the temperature-related pressure drop. Figure from Exponent report. There was more, including a “vigorous rubbing” test inspired by Bill Belichick’s bizarre press conference (Exponent must love him for adding to their scope of work). But Exponent’s investigation of these three issues show me that they were asking the right questions and using the right methods to answer them. Engineering is not about plugging numbers into formulas. It’s knowing where the numbers and the formulas come from, what their limitations are, and using judgement in their application. Of course, it’s still possible Exponent made mistakes, but did you read those text messages between McNally and Jastremski? And finally, we have the obvious source for this post’s title: # Customer service and security I placed an online order with Office Depot today, using my company’s corporate account, an account we’ve had for 20 years. Hilarity ensued. Late in afternoon, I got a call from someone who said she was with Office Depot (Caller ID agreed). She said the order had been placed on hold subject to confirmation of some details about our account. “I won’t give you any details about the company over the phone. You called me, and I have no way of knowing you’re actually from Office Depot.” Because of Caller ID, I was pretty sure she was who she claimed, but why risk it? “Sir, I need to confirm your account information.” “Yes, I understand that, but I have no way of confirming you’re really from Office Depot. You called me out of the blue. Will I be able to get to your department if I call the Office Depot customer service number?” “I just need to confirm your information for this order to go through.” “I’m hanging up now and calling customer service. Will that get me to your department?” “Yes, but—” “Thank you.” So I called the customer service line. The automated system asked me to punch in my account number, so I did and was put in the queue to “speak with a representative.” The wait was only a minute or so, and the first thing the representative asked me was my account number. This happens with almost every telephone support system I deal with, so I no longer give the poor rep grief about it, but it still bothers me. This is one of those dumbass pseudo-security features that make a manager somewhere feel like he’s doing something valuable. In any event, the rep confirmed that my order was on hold, but she said that it was on hold because an invoice due in mid-April had not been paid—not exactly what the first woman had told me. I got the details on the unpaid bill and asked if she could hold while I checked our system to see if it had been paid. She said yes, so I went off to our accounting office. I fired up QuickBooks and saw that we hadn’t made any payments to Office Depot since late February and none that matched the amount of the unpaid bill. I also learned that we had no pending Office Depot invoices in our accounts payable file. The bill was lost in the mail, most likely. When I got back to my office and picked up the phone, I learned that the Office Depot lady had hung up. I called back, went through the same nonsense of giving our account number twice, and explained the situation to the third customer service rep of the day. “How would you like to settle your account, sir?” “Well, I’d like to pay it, but I need to have a bill.” “The outstanding amount is$139, but you can make a payment of \$58 to get the hold on your order cleared.”

“I’ll pay the amount in full, but I need to have an invoice in hand before I do that. Just send me a copy of the bill.”

“All right, sir. Can I have your fax number?”

“My tax number? Why do you need that?”

“You’re kidding, right?”

He wasn’t. A few minutes later, the fax (which, like the company and our Office Depot account, is 20 years old) rang and printed the bill.

“OK, there’s a late fee and an interest charge on this. You need to take those off.”

“Well, sir, I don’t know…”

“How long have we had this account?”

“It says here the account started in 1995.”

“And have we ever paid a bill late before?”

“Well, I don’t have the full history, but it does look like you have an excellent payment record. I’ll remove the late fee.”

“And the interest charge.”

“I can’t do that, sir.”

“Sure you can.”

“I’ll have talk to a supervisor.”

“I’ll wait.”

The interest charge was, of course, removed, and I made payment arrangements.

I don’t blame Office Depot for wanting to get paid on time, and I understand the foot-dragging on getting the added charges removed. But I’m not inclined sympathize with a company that

1. Starts a conversation by misinforming me about the purpose of the call.
2. Then asks me to do something that’s thoroughly insecure.
3. Forces me to waste my time with a “security feature” that provides no protection.
4. Hangs up on me mid-call. I should point out that I had to wait several minutes during these calls as information was looked up, verified, and approved. But you knew that already, didn’t you?
5. Continues the “that’s not possible” charade after it’s clear that I’m not buying it.

There’s an inertia to our continued use of Office Depot for certain supplies. We’ve switched to Amazon for so much, maybe it’s time to go all in. At least I know I won’t be getting phone calls from them.

# Moving averages and the iPad

On Monday, all the real Apple bloggers read the press release, listened in on the conference call, and rushed to update their graphs of Apple sales figures. Most of the graphs looked like this one by Jason Snell

or this one by Federico Viticci

in that they plotted the raw numbers provided by Apple. The advantage of these plots is that there’s no screwing around—the numbers are what they are, and there’s no interpretation between the figures and you. However…

Apple’s sales figures are strongly seasonal, and because the graphs jump wildly from quarter to quarter, it’s often hard to see trends. When I wrote about the iPad last fall, I included this screenshot from the October 2013 Special Event:

By plotting cumulative sales instead of quarterly sales (and by doing some unspecified smoothing as well), Apple eliminated the jitter. It also, presumably, hoped to hide the downward trend in iPad quarterly sales—cumulative sales always go up. But for those of us who remember our calculus, the downward curvature at the right end of Tim’s graph told the tale.

There are myriad other ways to smooth a graph without switching to cumulative figures. One of the simplest is the moving average. In this technique, instead of plotting the raw data, you plot the average of a few data points in the neighborhood of each time value.

For many data sets, the best size of this neighborhood is not obvious. With Apple’s sales figures, though, I think it’s clear that the best choice is to average over four quarters: the quarter that you’re plotting and the three previous. This smooths over the seasonal jumpiness while not including so much past data as to ignore real trends. As an example, the 2015 Q2 value of a four-quarter moving average of iPad sales is 14.9, which is the average of 13.3, 12.3, 21.4, and 12.6.

Here’s my plot of Mac, iPhone, and iPad sales using a four-quarter moving average for each. I should also mention that the dates along the horizontal axis represent regular calendar dates, not Apple’s annoying fiscal year. In this graph, a year goes from January through December, not October through September.

I think this way of presenting the data makes the iPad’s situation much clearer. Sales are not “flattening”, nor are they “flat.” They were flat in 2013, but now they’re going down, and they have been for a year. What’s most interesting to me is how the upward trend, still very strong in 2012, just stopped dead in 2013. This is something you can’t see—or at least I can’t see it—in the graphs of raw data.

A few other points:

• In this week’s Connected, Federico, who is moving from Stage 1 to Stage 2 in the Kübler-Ross model, said that Apple is “still” selling three times as many iPads as Macs. This is true, except for the “still” part. The iPad is now selling at 3X Mac sales; it used to be 4X or more.
• The notion that iPad sales have been adversely affected by their durability doesn’t hold water. I’m sure it’s true that iPads aren’t being turned over at the rate iPhones are, but neither are Macs, and their sales are holding steady or increasing slightly.
• The sales decline can’t be explained by cannibalization from the iPhone. The problem with the iPad started when iPhone sales were one of its slowest growth periods.

I didn’t write this post to dump on the iPad. Although I know it’s not right for me, my wife has had one for 3–4 years and I doubt she’ll ever want to go back to a more conventional computer. Yes, the iPad’s one-app-at-a-time limitations sometimes grate on her, but its straightforward UI more than makes up for that.

The real questions are when will iPad sales level off again—this time coming down from above instead of up from below—and what are Apple’s plans to reverse the trend. There’s no reason to get hysterical and start thinking the iPad will turn into the iPod, but iPad sales have been in this state for two years, and there are no outward signs of any change from Cupertino. I do wonder if the iPad has been starved of attention because of the iPhone (which must be tended to because it’s 70% of revenue) and the march to bring out the Apple Watch.

Update 4/30/15 8:02 AM
Kieran Healy, holder of the Krzyzewski Chair of Sociological R at a small college near the University of North Carolina, has taken a more professional approach to the smoothing of the sales figures, breaking them down into three components: underlying trend, seasonal, and remainder. To my eye, his iPhone results look a little oversmoothed, but there’s always room to quibble about smoothing algorithms and parameters.

Most interesting, I think, is how clearly Prof. Healy’s results show something we all kind of expect: whereas the seasonal components of both the iPhone and iPad have a one-quarter peak during the holiday quarter, the Mac has a two-quarter peak, encompassing both the back-to-school and holiday quarters.

If you want to do your own smoothing or other manipulation of the sales data, I’ve put it in a zip archive for downloading. The archive holds three files, one for each product, and the lines of each file look like this,

2010-Q3 3.27
2010-Q4 4.19
2011-Q1 7.33
2011-Q2 4.69


where the Apple fiscal quarter and the sales number are separated by a tab.

Ben Packard, in an email, suggests my reason for dismissing the durability explanation for the iPad sales decline is weak. The Mac, he says, has been around long enough for there to be substantial numbers of owners at every stage of ownership, but far more iPad users are still on their first iPad. Whatever the long term replacement cycle of iPads turns out to be, we’re still in the first one, too early for variations in when people bought their most recent iPad to have evened out.

It’s a legitimate criticism. I was too glib in making a direct comparison between the Mac and the iPad and didn’t flesh out the argument as much as I should have. I think the iPad durability argument fails because its lower price and more rapid hardware improvements here in the early stages of the product’s life should make its ownership cycle shorter than the Mac’s. From a functionality standpoint, those who bought iPads during the ramp-up in 2011 and 2012 should be getting new ones. If they aren’t, and Tim Cook says new iPad sales are going substantially to first-time users, that means they’re finding no compelling reason to upgrade despite the huge improvements in the product itself.