This election proved you’re only as good as your data
As you’ve probably heard, while Hillary Clinton won the majority of the popular vote, Donald Trump was awarded more than 270 votes in the Electoral College. Many, particularly on social media, were incredulous, partly at the candidates but also at the pollsters — in particular, famous polling analysts like Nate Silver from Disney’s FiveThirtyEight blog. Silver rose to fame as the guy who successfully predicted the past few presidential elections. In 2008, he correctly predicted 49 of 50 states, and in 2012 he nailed all 50. With that, plus an impressive showing in the midterms, a legend was born. The 2016 presidential election was not so kind to FiveThirtyEight, with misses in the battlegrounds of Florida, North Carolina, Pennsylvania, Michigan and Wisconsin turning the odds quickly in favor of Trump. But throughout nearly the entire general election campaign, Mrs. Clinton was an overwhelming favorite.
So the question is: What went wrong? There are explanations all over the internet this week, but there’s one that should not be overlooked: Polling …
FiveThirtyEight is one polling aggregator, but the Princeton Election Consortium, the Huffington Post, Daily Kos, RealClearPolitics and the New York Times’ Upshot all do some form of polling aggregation. The premise behind a polling aggregator is simple elementary statistics. Each poll is a sample with a certain sample size, but if you aggregate the polls, you can increase the sample size, decrease the variance and make a better estimate of the true values.
Everyone has a slightly different way of doing this to draw their conclusions. Princeton Election Consortium treats each state as an individual probability and then calculates an overall probability. FiveThirtyEight runs a series of corrections, then simulates what the results might be given the data numbers they’re left with and a series of assumptions (like states flock together). RealClearPolitics just runs a raw average. You can quibble with some of the assumptions and operations of the models, and a few do, but there is a diversity of approaches that should, given good polling, reflect the true state of the race. FiveThirtyEight, in particular, was criticized quite a bit during the election season for wild swings in the model — most notably by Chris Arnade and Nassim Taleb. (The criticism goes that wild swings in the model increase uncertainty, which means the odds should be 50/50.) That being said, FiveThirtyEight’s model was consistently one of the most conservative, yet it still went into Election Night with Hillary Clinton favored by a 2:1 margin. Many were shocked, understandably, to see things change so quickly.
From month before election to 9 days before it, we had 80 live interview polls in 2012 in 10 states closest to national vote. In 2016? 36.
— (((Harry Enten))) (@ForecasterEnten) November 1, 2016
Possible explanation 1: Quantity of the data
Let’s look at the polls from three Upper Midwest states that had large polling errors in 2016 and compare them to polls from the previous two presidential elections.
| NUMBER OF POLLS CONDUCTED | ||||
| 2008 | 11 | 25 | 16 | 52 |
| 2012 | 10 | 13 | 12 | 35 |
| 2016 | 9 | 18 | 10 | 37 |
| Percent decrease 2008-2016 | 18% | 28% | 37.5% | 28.8% |
There were significant decreases from 2008 to 2012 and 2016. Polls were still happening in these states, so you could still make an assessment of the race, but every decrease in the number of polls causes an increase in the likelihood of a polling error.
Let’s consider how much the RealClearPolitics polling average differed from the final results in these same three Rust Belt states.
| POINT DIFFERENCE BETWEEN ACTUAL RESULT AND ESTIMATE |
Michigan
Ohio
Wisconsin
Total
2008
2.9
2.1
2.9
2.63
2012
5.5
0.1
2.7
2.76
2016
3.7
5.1
7.5
5.43
Considered in aggregate, it does appear that the polling misses have increased in these Rust Belt states as poll volume has decreased.
The decrease could be the result of budget cuts across the newspaper industry. If newspapers and journalism have less funding, polling could be one of the first things to go. Live interview polls, which are regarded to be the most accurate, require paying someone to call participants, likely a third-party company, and can become quite expensive. If places like FiveThirtyEight are simply using these polls to aggregate results, and polls are driving clicks to FiveThirtyEight and not the newspaper, it’s hard to justify this use of time and money.
Possible explanation 2: Data quality
A poll requires a simple random sample to be able to make inference about the true mean of the population. If everyone in the population is not equally likely to get selected, it’s not a random sample of the whole population, and if you can’t acquire a random sample, then any conclusions you make about the state of the race are flawed.
The 1991 Telephone Consumer Protection Act banned the use of random dialers on cell phones, so pollsters have had to rely on people calling individuals on cell phones. Many of us don’t pick up calls from unknown numbers. Surveyors are spending more time and money calling more people and having a harder time getting responses, and thus getting appropriate samples.
We could certainly look into online polling instead, but random sampling there is not foolproof either. Not everyone is online, and people are reluctant to respond to internet surveys.
Pollsters have an ideal representative sample, based on the electorate they’re trying to poll. If they aren’t able to get a representative random sample, they will count the responses of the individuals they do get more times (called weighting) to reach their desired sample. If a demographic that is harder to reach via cell phone or internet goes heavily towards one candidate, the chances of a polling error increase.
These are methodologically sound ways of dealing with response bias, but they are not preferred to actual simple random samples. The much-maligned LA Times poll was weighing one black Trump supporter 30 times more than the average participant in the poll, and 300 times more than the least weighted.
Possible explanation 3: Misleading data
Trump surrogates had been claiming for months that there was a “hidden” Trump vote. Social desirability bias, a term coined by social scientists, is a type of polling-response bias where individuals don’t share their true preferences in a poll because they desire to be viewed favorably by the person polling them. This type of bias is more common in live interview polls than in online polls. If Trump supporters were misrepresenting their preferences, it would have skewed the polls toward Clinton and resulted in a polling miss.
Possible explanation 4: Uncertainty
Nate Silver had been arguing for weeks that while it looked like Clinton would win, this was far from certain, mostly due to a high number of undecided and third-party voters. The results showed that a huge number of late-breaking voters in the deciding states went for Trump. If uncertainty was present all along, it’s entirely possible that the polls were right, but people weren’t accounting enough for the uncertainty.
Polling has become both easier than ever because of our increased ability to process large amounts of data quickly and harder than ever because of our increased reliance on our phones, among other things. The polling industry has not caught up with these changes. We simultaneously know more than ever and less than ever. As we look toward the 2018 midterms and the 2020 general election, expect a lot of discussion about how to make online polls more accurate to decrease the uncertainty and reduce our reliance on live interview polls for accurate results.
Soccer may use video replays to help with concussions
Soccer may not be full-contact like football, but the risks of getting a concussion from a collision on the pitch are still a huge concern. At its annual meeting in March, the International Football Association Board (IFAB) will decide whether a doctor on the touchline will be able to review video replays. After doing so, medics will be able to determine if a blow to the head was severe enough to require a substitution. If approved, the system could be in place as soon as next season and ahead of the next World Cup in 2018.
Right now, goal-line technology is used during games to review whether or not the ball crossed into the goal. The IFAB announced in June that testing would begin on in-game video replays to review goals, penalties, fouls and more in several leagues around the world. The governing body isn’t expected to decide whether or not the Video Assistant Referees (VARs) system will be officially approved for use until 2018 or 2019. However, the Associated Press reports that mangers will have access to video replays for tactical reasons, but they won’t be able to use them to argue calls. I’m sure they’ll abide by that guideline with no issues.
FIFA enacted protocols two years ago that allow referees to stop a match for three minutes while a potential head injury is assessed. IFAB says that even with that change, there have been cases where a player shouldn’t have been allowed to continue. Until the VARs system is approved, officials won’t be able to consult replays that are available for medical reasons.
“We have seen enough examples where potentially if the doctor was able to use video he might have made a decision not to put a player back on the field,” IFAB secretary Lukas Brud explained to the AP.
Via: The Verge
Source: Associated Press
Antelope Club is a Slendertone for full-body workouts
German startup Antelope Club is looking to make fitness training more effective with its new electrical muscle stimulation suit. It’s essentially a giant Slendertone that will work on more than just your abs, ideally getting you fitter and happier much faster. The company has built a t-shirt with stimulators on the stomach, pectorals and biceps, while a pair of accompanying shorts will help your quads. Additionally, calf simulators can be hooked up to a separate battery to help tone your lower legs.
The headline claim is that every 20 minutes that you spend working out with this kit is the same as an hour in the gym. For best results, users are encouraged to wear it while running, cycling or weightlifting, adjusting the intensity to best suit their needs. Power is sent to the unit via a charging module that’s re-juiced over micro-USB and connects to the suit with a series of magnetic pins.
One of the downsides to most EMS devices is that you need to periodically replace the adhesive pads to maintain the connection. Antelope, however, is using a stretchy silicon mix that, it claims, won’t become less effective over time. It’s also said that you won’t need to use a gel or water to make the pads work properly, which will be useful during those cold winter runs.
Unfortunately, full body EMS suits don’t come cheap, or within the budget of most casual exercisers. To purchase the men’s version of the full outfit will set you back $1,349, although a bolero-style top for your arms and pecs will set you back $275. In addition, a tank top that just offers abs and pecs training for $299, so there are savings to be had if you’re prepared to compromise.
Source: Antelope Club
Google’s defense against anti-trust claims: ‘we’re open’
Google has a response for the European Commission’s anti-trust allegations. In a lengthy blog post, the tech juggernaut addressed the EC’s concerns point by point. That starts with the EC’s stance that Android isn’t in competition with Apple’s iOS mobile operating system, and Google citing the Commission’s own research that 89 percent of survey respondents feel that the two are competitors. That last bit is a recurring theme, with Google pointing toward the survey responses for the EC’s stance on Android’s “stable and consistent framework” across devices as well.
In perhaps the most poignant response, Google made a GIF that illustrates how many apps are typically pre-installed/bundled on Android devices versus the competition — something the EC directly called out. By Mountain View’s count, of the Samsung Galaxy S7 with Android 6.0.1’s 38 pre-installed apps, only 11 were from Google. Contrast that with 39 out of 47 on the Lumia 550 from Microsoft and 39 out of 39 from Apple on the iPhone 7 running iOS 10.0.2.
“Android hasn’t hurt competition, it’s expanded it,” Google’s Senior Vice President and General Counsel Kent Walker said in a statement. “Android is the most flexibe mobile platform out there, balancing the needs of thousands of manufacturers and operators, millions of app developers and more than a billion consumers.
“Upsetting this balance would raise prices and hamper innovation, choice and competition. That wouldn’t just be a bad outcome for us. It would be a bad outcome for the entire ecosystem, and — most critically — for consumers.”
And with that, the battle moves onward. Maybe the EC’s stance won’t leak ahead of the next round. Maybe.
Source: Google
Apple Wins ‘Gold’ Customer Service Award From Twitter
Twitter today announced the results of its first ever #Awards initiative, naming Apple as the top customer service provider on the social networking platform.
Apple was awarded with the “Gold” #Customer award for the @AppleSupport Twitter account, which Apple uses to quickly respond to customer service requests. According to Twitter, the @AppleSupport account ranks #1 among support accounts for average engagements and average engagements for followers, racking up 150,000 likes and retweets since its launch.
The average online customer service response time is a long, grueling 17 hours. Hours! @AppleSupport decided to do something about that sad stat and use Twitter to respond to customer inquiries and comments. Because of the live nature of Twitter they are also able to pick up conversations with customers at any time, and incorporate rich media into responses to provide even more help to people when and where they need it most.
Apple launched its Apple Support Twitter account in March of 2016, marking its first support account on a social networking site. Apple uses the Apple Support account to answer customer questions, offer support, and provide Apple customers with tips and tricks on using Apple products.
Illustrate your point. Draw, zoom, and add text to images before sending. pic.twitter.com/WKXfjlac5j
— Apple Support (@AppleSupport) November 9, 2016
Apple interacts directly with customers through the Apple Support Twitter account and provides quick solutions to problems, often responding to requests in a matter of minutes during the hours of 5:00 a.m. to 8:00 p.m. Pacific Time.
Since its March launch, Apple Support has gained more than 524,000 followers and has shared dozens of useful product tips that often garner hundreds of likes and retweets.
Tags: Twitter, Apple Support
Discuss this article in our forums
T-Rex, Hedgehog, Sandwich, Vampire and Zombie Among Emoji Proposed for 2017’s Unicode 10
The Unicode Consortium today published the list of proposed emoji that could potentially be included in Unicode 10, set to be released in 2017, giving us a first glimpse at the emojis we might see included in future iOS updates.
Today’s proposal includes 51 emoji candidates [PDF], including flying saucer, shocked face with exploding head, face with open mouth vomiting, grinning face with star eyes, coconut, broccoli, pie, pretzel, sandwich, t-rex, giraffe face, sauropod (basically a brontosaurus), hedgehog, mage, vampire, fairy, elf, genie, zombie, and tons more.
Image via Emojipedia
Emoji candidates are often chosen for multiple reasons, ranging from expected high frequency of use to requests from online communities. Others fill in gaps in existing emoji sets or are compatible with current emoji characters.
The Unicode 10 emoji candidates are not yet finalized, so it is possible some of the characters on the list could be eliminated. We may also see new emoji additions before Unicode 10 is finalized in 2017.
With emoji growing in popularity, Apple has started adopting new emojis more frequently. iOS 10.2 includes support for Unicode 9 emoji, introducing characters like clown face, drooling face, selfie, face palm, fox face, avocado, bacon, and croissant, along with dozens of profession emoji.
Unicode 9 was first approved in June of 2016, so we can perhaps expect to see Apple adopt Unicode 10 within six months of its release.
Tags: emoji, Unicode 10
Discuss this article in our forums
First Touch Bar MacBook Pro Orders Now Preparing for Shipment
The first 13 and 15-inch MacBook Pro models equipped with a Touch Bar will soon be making it into the hands of customers, as initial MacBook Pro orders have shifted into the “Preparing for Shipment” stage.
Customers who ordered a MacBook Pro right when it became available for purchase on October 27 received shipping estimates ranging from November 17 to November 25, and with orders ready to go, it appears customers will begin receiving their machines next week.
According to reports from MacRumors readers, it looks like 15-inch MacBook Pro models and custom ordered machines will be shipping out in the near future. We aren’t yet seeing reports of order changes for 13-inch machines, but those should also be going out soon.
New MacBook Pro models sold out quickly, with shipping estimates slipping from 2-3 weeks to 3-4 weeks within a matter of hours after they went on sale. It is not yet known when stores will begin receiving supplies of the new MacBook Pro, but right around November 17 is a good guess.
Last week, Apple’s Phil Schiller said Apple has received more online orders for the new MacBook Pro than for any previous-generation MacBook Pro. Sales estimates from Slice Intelligence have also suggested the new MacBook Pro has already outsold all competing laptops in 2016, despite its high price tag.
Pricing on the 13-inch MacBook Pro with a Touch Bar begins at $1,799, while pricing on the 15-inch Touch Bar model begins at $2,399.
Related Roundup: MacBook Pro
Buyer’s Guide: MacBook Pro (Buy Now)
Discuss this article in our forums
WhatsApp beta finally adds support for two-step verification
You can now add a passcode to your WhatsApp account.
The latest update to WhatsApp beta lets you secure your account with a six-digit passcode. WhatsApp verifies your phone number via an SMS confirmation when you’re setting up the service on a new phone, and with two-step verification enabled, you’ll have to enter the passcode as well to gain access to your account.
You also get the option to add your email address, which can be used to disable two-step verification should you forget your passcode. If you’re looking to go ahead and enable two-step verification, you’ll have to consider the following:
If you have two-step verification enabled, your number will not be permitted to reverify on WhatsApp within 7 days of last using WhatsApp without your passcode. Thus, if you forget your own passcode, but did not provide an email to disable two-step verification, even you will not be permitted to reverify on WhatsApp within 7 days of last using WhatsApp.
After these 7 days, your number will be permitted to reverify on WhatsApp without your passcode, but you will lose all pending messages upon reverifying – they will be deleted. If your number is reverified on WhatsApp after 30 days of last using WhatsApp, and without your passcode, your account will be deleted and a new one will be created upon successfully reverifying.
WhatsApp doesn’t verify your email address, so it asks that you “provide an accurate email address.” The service will, however, periodically remind you to enter your passcode.
The feature is now live in the beta channel (2.16.346), and should make its way to the stable release soon. WhatsApp is picking up features at a steady pace, with the ability to make video calls going live on the platform in recent weeks.
Sony is offering Canadians $180 off a PS4 bundle with a Xperia X Performance purchase
Sony is getting in on the deals game early this year with the offer of $180 a PlayStation 4 Uncharted 4 bundle with the purchase of an Xperia X Performance from Bell.
The deal is fairly simple, but a bit arduous, especially if you’re in a rush to get the new console. You need to buy an Xperia X Performance from Bell and submit the IMEI for verification, which the company says could take between two and three weeks, before receiving an email with a discount code for $180 off the PS4 purchase.
Valid between now and December 31, the deal is pretty good if you’re not lusting after the more-expensive PS4 Pro — this is the standard, non-slim PS4 with Uncharted 4.
Who’s buying?
More: PS4 vs PS4 Slim vs PS4 Pro: Which should you buy?
Google needs to stop shipping ideas and start shipping products
Nearly all of Google’s new products are now shipping, but none of them feel complete.
What a month, right?! In the last 30 days Google has announced and shipped a solid answer to Apple’s iPhone, competition for Amazon Echo, and a serious message that Samsung’s Gear VR isn’t the only game in town for great portable VR. As an Android nerd, a connected home nerd, and possibly the biggest VR nerd on the planet (by weight) I should be in some kind of joyous coma with commentary about kids at Christmas flying around.
But I’m not. In fact, I’m a little disappointed in all three of these products and I’m more than a little hesitant about the soon to be released Google Wifi after using these other #madebygoogle products. I’m frustrated because the only thing stopping each of these products from being the best things in their respective categories is how entirely unfinished their biggest features are.
Hashtag Eventually
Google’s new Pixel phones are incredible. That camera! The UI is so snappy even after I’ve actively tried to bog it down! It just feels like a great phone when you hold it, and I’m finally over missing my Note 7 now that Daydream is here. Only, it’s not quite finished, is it? Google Assistant, the big thing that is supposed to take these phones to the next level, is still very much a beta. Some of that is to be expected, Google is at their best when they have lots of real people to use as digital lab rats. Some of the things that got overlooked before shipping I have a real problem with, like not being able to tell me the name of a song when I ask. Google Assistant is supposed to be the future, but it isn’t yet as capable as Google Now? Come on now, that’s just bad form.
SERIOUSLY?! pic.twitter.com/cH4B4bPfKW
— Russell Holly (@russellholly) November 4, 2016
I want a Google Home in every room. The speaker is great, better than any other Google Cast speaker you could get at that price, but half of the features Google showed us at the launch event don’t exist yet. You can Chromecast YouTube with your voice, which is kinda neat, but Netflix and Google Play Movies support are nowhere to be found. Communicating with Nest and Hue works on a basic level, but none of the Google Assistant threaded conversation stuff works in that context. I can’t say “can you turn them back on” after asking Home to turn out the lights, which is supposed to be the whole point of this platform. It’s supposed to learn from me and feel like a conversation. When does that start?
Stepping up from Google Cardboard to Daydream is a revelation. A mobile VR headset with lots of streaming video choices means I can lose myself in a movie on a flight, or play a game while waiting in the car for my kids to get out of school. The headset looks nice, feels great, and the Pixel XL is superb as a VR display. But ever since Google I/O we’ve been hearing about how Daydream will support many different phones and be a high-end democratized platform. So far not even LG’s V20, the other Nougat phone, actually has Daydream support on the way. As long as you have a Pixel and are cool with only 19 apps while others trickle in, it’s awesome.
Seeing a pattern here? It’s not a good look.
Shipping is hard, but it’s time to step up
The truth is, Google’s relationship with shipping hardware has always been a little dicey. How many Nexus launches were totally shut down because the biggest internet company in the world wouldn’t figure out a shopping cart solution that could handle a couple hundred thousand people at once? Almost all of them is the answer, just in case you were wondering.
This year is a little different. Google’s strategy has changed, and getting the hardware out there has been a lot less painful. It’s great hardware, too. There’s no denying the design team has been putting out quality gear, but the software for some reason doesn’t match and that’s a real shame. Right now Google has a collection of really great ideas inside of very attractive packaging, but precious few complete thoughts. A good bit of this is just how Google software works. This is a company that thrives on millions of data points, and putting your software out there in the world lets you quickly collect those data points. That data collection lets Google iterate quickly and as long as the product isn’t totally abandoned Nexus Q style, things improve pretty quickly.
I’m excited by the new direction Google is taking, but it couldn’t be more clear the company has an execution problem.
This strategy gets highlighted this year because Google’s releasing multiple products that are very clearly being sold as competition to existing heavyweights. The first real iPhone competitor directly from Google. The Google take on Amazon Echo. Mobile VR that one-ups the Gear VR. These are all category champions that have the benefit of time and have gained quite a bit of polish along the way. Google’s offerings in each category this year could easily one day be clearly the best thing available, but people aren’t asked to buy them later. People are being asked to buy something that has the ability to get better over time, with a frankly spotty record of that actually happening.
I’m excited by the new direction Google is taking, but it couldn’t be clearer the company has an execution problem. If you’re releasing something as direct competition to something dominating a category, telling people to buy it now and watch it grow and improve isn’t going to work outside of Google’s early adopter bubble. These aren’t early products only available in the Google Store, though. They’re on the shelf at Best Buy, with big friendly boxes telling you how great they’ll be once you get to know them.
AIVAnet 