Hacker News new | past | comments | ask | show | jobs | submit login
An inside look behind Romney's loss: An epic failure of its Orca big-data app (computerworld.com)
142 points by CrankyBear on Nov 9, 2012 | hide | past | favorite | 105 comments



Stripping out the politics here, some good lessons for all of us:

1. Test your damn product before it goes live.

2. User feedback, user feedback, user feedback. Things like the confusion of the use of "app" (vs "website") and the lack of http->https routing scream of things that a developer would think are "obvious" but even the most basic user testing would reveal early on.

3. Stealth mode can be trouble. This may be a controversial opinion but I've learned to believe more and more that "Surprise! We're here!" approach does more harm than good. There are limited circumstances where this is not the case (situations where knowing the existence of a project can damage that project) but I emphasize limited.


Yes! Long experience has taught me that I can either look dumb in front of a small, friendly audience of early testers or a big audience of people with inflated launch-day expectations.

Not looking dumb is not an option. It is a big fucking world, and it will never fit inside the 3 pounds of meat that people somehow expect me to think with. Especially when some of that meat is dedicated to things like the ego and false pride that tell me that my big launch will be magically perfect because I am so smart.


"Not looking dumb is not an option."

I'm going to be quoting you forever, thanks.


Are we really to believe this is plausible?

1: Romney's big data app failed on election day which led to ineffective use of their volunteers.

2: Romney's campaign performed exactly as predicted by the likes of Nate Silver.

Therefore: the failure of their big-data app was inconsequential. Otherwise we have to assume that the hundreds of polls were consistently biased in the direction of Obama and that the failure of the Orca big data app exactly compensates for this failure of the polls so that the outcome of the election is nonetheless exactly as predicted!

Absolutely 100% ridiculous.


Whoa, there nelly!

First things first, Project Orca (you can read two first hand user accounts here: http://www.breitbart.com/Big-Government/2012/11/08/Orca-How-... and here: http://ace.mu.nu/archives/334783.php ) appears to be a Get Out The Vote and volunteer management app.

Second, 1 and 2 are not incompatible results. Nate Silver and others have pointed out that the polls appear to have underestimate democratic performance at the polls... is that due to Republicans failing to turn out? That seems like a likelihood that's at least worth considering.


Nate Silver and others have pointed out that the polls appear to have underestimate democratic performance at the polls

This is what Nate Silver said on election night:

So far, however, Democrats also have an edge in the party identification numbers in the exit polls. In Ohio, 38 percent of voters identified as Democratic in the exit poll as compared with 31 percent of Republicans. And in Virginia, Democrats had a 37-to-33 advantage in party identification. These numbers are similar to what many pre-election polls showed.

http://fivethirtyeight.blogs.nytimes.com/2012/11/06/live-blo...

Can you point to where he adjusted this analysis?


I don't believe Nate Silver pointed out that the polls undercounted democrats, do you have a cite? In fact his analysis was basically that the aggregate polling "got it right", coming very close to the final counts in almost all cases.

In the four victory-critical states I still have in my head, Silver's polling averages (which weren't very much different from Wang's or TPM's or RCP's, by the way) nailed it dead on in Florida, undercounted Democrats in Colorado and Virginia, and overcounted them in Ohio. But in no case were the errors more than 1.5% or so.


He said as much in an interview. I'm pretty sure it was when he was on the Daily Show, but I could be misremembering.

http://www.thedailyshow.com/watch/wed-november-7-2012/nate-s...


This was the interview I was recalling, thanks pflats!


Two things to consider before immediately going to the "yay Nate Silver" default:

1. GOP had a consistent polling advantage on voter engagement, even going as far back as August. The Romney campaign placed a lot of stock on this. (see here: http://www.politico.com/news/stories/0812/79665.html) The polls might break for Obama, but the Romney campaign was banking on something else...

2. Polls attempt to account for the simple error that people are more likely to say that they'll vote than will actually vote. (see here: http://www.gallup.com/poll/111268/how-gallups-likely-voter-m...) The Romney campaign was predicated on the idea that GOP voters were likely to get out and vote, while Obama voters would say they were going to vote but then drop out on Election Day. That is a polling blind side that Nate Silver couldn't account for.

The Romney campaign didn't accurately test their engagement assertion. They relied on journalist polling, but engagement didn't turn into votes. The Obama campaign did the opposite: they targeted voters who might be disengaged and directly went after those with their GOTV efforts.

There is a HUGE debate that always goes on around election time with politicos around who matters more. Is it more important to try to sway independents (i.e., go after the middle), or is it more important to fire up your base? Romney won independents but Obama got the base turnout, indicating that a GOTV effort targeted at the base might have worked, had it been used effectively. (see here: http://abcnews.go.com/Politics/OTUS/exit-polls-obama-loses-i...)


Technically Romney didn't win independents. http://www.cbsnews.com/8301-250_162-57547239/adviser-romney-...

Over the past 3 years the GOP voters switched their identification to independent while their beliefs stayed the same. This shift accounted for what the polls were seeing among independents.

See "Independents. State polls showed Romney winning big among independents. Historically, any candidate polling that well among independents wins. But as it turned out, many of those independents were former Republicans who now self-identify as independents. The state polls weren't oversampling Democrats and undersampling Republicans - there just weren't as many Republicans this time because they were calling themselves independents. "


> 2. Polls attempt to account for the simple error that people are more likely to say that they'll vote than will actually vote. (see here: http://www.gallup.com/poll/111268/how-gallups-likely-voter-m...) The Romney campaign was predicated on the idea that GOP voters were likely to get out and vote, while Obama voters would say they were going to vote but then drop out on Election Day. That is a polling blind side that Nate Silver couldn't account for.

Actually Silver did account for something similar: That polls could be systematically biased against Romney. Not for political reasons, but incorrect LV models. The accurate pollsters of this cycle used an LV model that was shaped by 2008 exit polls. Others used those based on 2004 or 2010 (Rasmussen). Those that banked in this being an '04 or '10 electorate were way off.

He eludes to this here: http://fivethirtyeight.blogs.nytimes.com/2012/11/02/nov-1-th... and: http://fivethirtyeight.blogs.nytimes.com/2012/11/01/oct-31-o...


The Occams Razor answer is that the "Obama supporters are stoned slackers, they won't even turn out to vote" hypothesis was wrong, not that it was counterbalanced by a failure within the Romney campaign.


Could be more relevant in Colorado and Washington. Just kidding.


"Are we really to believe this is plausible?"

To the degree we find any blogspam title plausible, yes.

In other words, no.

The original article was titled "The Unmitigated Disaster Known As Project ORCA" and merely pointed out what a custerfuck it was. It made no claim that it turned the election.

http://ace.mu.nu/archives/334783.php


Fully agree with your point about polling, but it's worth reading the original blog post (by a GOP campaign volunteer, but otherwise apolitical) to see in detail just how badly their system failed: http://ace.mu.nu/archives/334783.php

It's a classic example of 'fail late, fail terminally.' I still think Obama would have won, but it should have been a bit closer than it was. 'Get out the vote' operations are important.


According to Politico (I don't know if this is the best data, but it was high on the list in Google) Obama won Florida by about 60,000 votes (out of almost 10 million cast).

Now, that Romney volunteer says:

the end result was that 30,000+ of the most active and fired-up volunteers were wandering around confused and frustrated when they could have been doing anything else to help. Like driving people to the polls, phone-banking, walking door-to-door, etc

Now, if those volunteers had been engaged, could that have tipped the state to Romney? I don't know how many of those volunteers were in Florida, and how many votes each dedicated volunteer can turn out, but it's entirely plausible.

(Of course, even if Romney had won Florida, Obama still would have won.


A related story (http://www.breitbart.com/Big-Government/2012/11/08/Orca-How-...) observes that Obama won by about 5-700,000 votes in swing states. Given that there were 37k 'Project ORCA' volunteers in swing states, the author estimates that if each had brought only 20 additional voters to the polls the election could have gone to Romney.

Now, I think this is rather unlikely, and reminiscent of the 'Bargaining' stage in the Kubler-Ross model of the grieving process; polls last weekend in critical states like OH and FL showed strong support for Democratic senate candidates and >70% support for Obama's handling of Hurricane Sandy, and the poll momentum was already headed Obama's way. Likewise, Republican voters who can't stand the President were highly motivated to go out and vote against him anyway and likely did so with or without a phone call - indeed, I heard from some Romney supporters that the campaign relied far too heavily on robo-calling in the leadup to the election, to the point that they were sick of getting reminders.

But having 37k field volunteers confused and dispirited on election day is still a disaster. It's very ironic that a campaign that was nominally against 'trickle down government' was run in such a centralized, dictatorial fashion. Psychologically, it seems like a classic case of projection - unconsciously ascribing one's own personality traits to other people in one's life. Romney's reported preference for abundant data and tight control robbed tens of thousands of his most enthusiastic supporters of their autonomy and initiative.


Check out Silver from yesterday:

http://fivethirtyeight.blogs.nytimes.com/2012/11/08/as-natio...

Obama's tipping point state was CO, which he won by almost 5 points, or somewhere in the vicinity of 80-90k votes. People will pay a lot of attention to Ohio, Florida, and Virginia (crazy that OH was closer than VA!), but Obama could have lost all three and still won.


I agree. I think Romney lost mostly because he flip-flopped. I watched the Frontline episode called "The Choice 2012" and it was quite insightful. The two candidates come across as quite similar in many policy/thinking respects. Because Romney had to cater to the right wing fringe, the republicans essentially ate themselves.

Effectively, the country moved right because the republicans effectively shut Obama down and caused him to polarize. Basically negating the "yes we can change" hypothesis. In 2008 people were sick of unilateral Bush action, the wars, and wanted people who would work together. By the midterm elections fringe groups got super worked up about Obamacare and death panels. Tea party hyperbole basically, a very effective and vocal minority. In addition the vote for Obamacare was drawn down party lines which I believe contributed to the democratic losses. It was a sign of not working together even though the republicans basically stonewalled the entire time.

Essentially, Obama, the great listener who could bring people together was shut out and shut down by politics. He was shutdown as an outsider essentially. We had hoped that an outsider could bring change, but the system itself refused to. Romney himself was an outsider too but would have had the initial advantage of a republican congress.


Romney's flip-flopping was only one reason for his loss. Obama's ground game played a much bigger role, IMO. Rather than trying to woo mercurial independents, Obama simply used the changing demographics to his advantage to turn out a lot of new voters who simply hadn't voted in previous elections. Romney's ground game was much weaker because Obama had been developing his over the last 4 years and didn't have to waste time and money on a long primary campaign.

In swing states like FA and CO, he got many US-born Hispanics who hadn't previously voted (many of whom recently became adults) registered and got them to vote for him by telling them that Romney would deport their foreign-born family members who are in the country illegally. As explained by the NYT[0]:

> In Chicago, the campaign recruited a team of behavioral scientists to build an extraordinarily sophisticated database packed with names of millions of undecided voters and potential supporters. The ever-expanding list let the campaign find and register new voters who fit the demographic pattern of Obama backers and methodically track their views through thousands of telephone calls every night.

> That allowed the Obama campaign not only to alter the very nature of the electorate, making it younger and less white, but also to create a portrait of shifting voter allegiances. The power of this operation stunned Mr. Romney’s aides on election night, as they saw voters they never even knew existed turn out in places like Osceola County, Fla.

In Ohio and Pennsylvania, Obama partially relied on the black vote. These are people who are guaranteed to vote for him, so the main issue was getting them to vote in the first place. He set up neighborhood election stations in black community gathering places like barbershops and beauty salons, getting people to register their friends and family and encourage them to go to the polling stations on the day of the election. In fact, the voting rate among blacks was higher than among whites in Ohio - an amazing coup.

Another key constituency in Ohio and PA was the auto workers and other blue collar voters, mostly whites. This one was pretty easy - Obama bombarded the airwaves with TV ads portraying Romney as a corporate asset-stripper who cackled merrily all the way to the bank as he shipped blue collar jobs to Mexico and China. Obama also reminded them about the auto bailout he engineered, which saved many blue collar jobs in those parts of the country.

In Ohio and PA, as well as places with low minority populations, like Wisconsin and Iowa, Obama relied on the female vote. He associated Romney with Senate candidates like Richard Mourdock and Todd Akin, who made controversial remarks about abortion and rape. I believe Paul Ryan made some similar comments as well. He played up the angle that Republicans want to tell women what they can and can't do with their bodies.

0: http://www.nytimes.com/2012/11/08/us/politics/obama-campaign...


In startup terms they spent all of their money on marketing and not enough on the product itself.


great listener who could bring people together

Cite needed. When has he ever done this?


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

69 million people voted for him in 2008, and 61 million in 2012. This is the most objective metric available. These more subjective ratings show his potential for bringing people together: http://www.thedailybeast.com/articles/2012/07/13/obama-does-...

"According to Pew, 92 percent considered him a “good communicator,” 87 percent deemed him “warm and friendly,” 81 percent said he “cares about people like me,” 79 percent thought him “well-informed,” and 76 percent judged him “trustworthy.” Since then, each of those numbers has declined between 10 and 20 points. But they began at such stratospherically high levels that even with the drop, the public’s perception of Obama as a person remains remarkably cheery."

And that was during an election with a flood of attack ads from both sides. The cause of the double digit drops will attenuate with time.


Nate Silver predicted a range of outcomes including closer losses and victories for Romney. Republican "Get Out The Vote" outperforming history was probably a necessary condition of that 10% shot at winning. Instead it looks like the Romney campaign dropped the ball on it. No conspiracy theory needed.


Well, it's probably better to say that the 10% included all known error estimations. Given the skew between final polls and actual results seen over history, Silver's model figured it was 1 in 10 that some "unknown" factor would pollute the numbers enough to produce a Romney victory. In this he was actually much more conservative than Sam Wang, who AFAICT looked only at the poll sampling error and arrived at a muc higher (e.g. 99%) chance of victory.

And actually, I think you can make the case that Silver was too conservative. Looking at the results, the poll averages were well within sampling error in all battleground states. And sorting the states by margin, Romney would have had to pick up FL, OH, VA and CO to win. Obama won Colorado by 4.7% (not too far from the predicted ~3.5%, I believe), which is comparatively huge and absolutely not explainable by polling error (a Romney victory there looks like about three sigma to me).

In hindsight, the known state of the election on Tuesday morning was never winnable for Romney. Wang was right (even though his averages came down on the wrong side in FL), Silver was too timid.


And actually, I think you can make the case that Silver was too conservative.

Based on what I know about methodology of polls, Silver's approach is better. If your possible errors are all independent, then Wang's approach is correct. But when you introduce methodology dependencies between polls, it is going to be much more confident than it should be.

Ironically if Silver is right, then he'd also predict that with some high probability, at a guess somewhere in the 70-80% range, Wang's numbers are likely to look better than his. But in those outliers there are some shocking results.

For the record, a fundamentally similar methodological error to Wang's helped Wall Street think that bonds backed by subprime mortgages were safe.


Exactly, there's some epistomolgy at work here. There are no first-principle models of polling bias that fit the data. Silver took a history of poll-vs-election skew as a proxy, while Wang ignored the issue and looked only at sampling error (I think -- honestly I don't know for sure).

As it happens, in this election (and across hundreds of polls) there was no poll bias. The polls were right.

Now, is that because we're lucky or because polls have gotten better, or both? I don't know that this is answerable. My intuition (but that's all it is) agrees with you: I wouldn't have put down a bet for Obama given Wang's 100:1 odds, though I might have at Silver's 10:1. Arbitrage vs. Intrade's 2:1 looks a lot like a sure thing in hindsight...


An inside look behind Romney's loss: Bill Clinton


I don't think that analysis is correct. Silver's analysis did not include the effect of Romney's app in the input, and so the expected result of Romney's app failing is to make the result more in line with Silver's prediction.

The whole point of Romney's app was to make Silver's prediction turn out wrong, by changing one of the conditions contributing toward Silver predicting an Obama win.


Just to add to that: Obama would have still won even if he had lost Florida, Ohio and Virginia, combined. See: http://nyti.ms/VJ5rDb


No. Obama won by 313 to 225 electoral votes. Florida had 29, Ohio 18, and Virginia 13. If you switch all three states to Romney then Romney would have won 285 to 260.


Your 313 figure is incorrect. Obama won with 332 votes to Romney's 206. See the link I posted.


Sorry, I was looking at the wrong page while I wrote that. But the greater point stands -- swinging Florida, Ohio, and Virginia would not have swung the election.


313 was Nate Silver's average outcome prediction. The mode of 332 is what actually happened.


No, the parent was correct. 313 to 225 was Nate Silver's prediction (note "Forecast"). The actual electoral vote was (assuming Obama won Florida) 332 to 206.


The main point is that the 'app' was not ready and the system failed to perform. The speculation that it could have made a difference in some states was secondary.


The key mantra about Get Out The Vote is that the battle is mostly won in the long campaign.

The key technique is to get a voter to make an emotional commitment to voting by talking personally to an actual human being (think Asda/Walmart having someone greet you on entrance - it works).

Weakly identifying voters (switchers, intermittent voters) are preferentially contacted by the candidate, other more grounded ones get ordinary party members on their doorsteps, plus personal letters etc.

So this failure is likely symptomatic of a general failure of GOTV planning - which means the polls are likely to be correct.

There is no magic polling day bullet.


You sure tore that strawman a new one, but did anybody say that Orca would have facilitated a win, or even a significant vote gain for Romney?

The closest I could find: "We lost by fairly small margins in Florida, Virginia, Ohio and Colorado. If this had worked could it have closed the gap? I sure hope not for my sanity's sake."


It's right there in the article's title: "An inside look behind Romney's loss". I take that to mean that the article attempted to explain (partially) why Romney lost.


Except it drew that conclusion from an article which made no such claim. The headline is blogspam.

The preblogspam article is titled "The Unmitigated Disaster Known As Project ORCA."

It is found here: http://ace.mu.nu/archives/334783.php


Exactly. "We actually won, but (x) failed us so OBAMA was able to steal the election." Totally expected from Bullshit Mountain. I'd like to see two parties that have some barest grasp on reality in America rather than just one, but I can't say that watching the Republican party disappear even further up its own asshole isn't going to be a hell of good time, in the coming years.


The OP borders on blogspam, read the more detailed account where it was original posted:

http://ace.mu.nu/archives/334783.php

Among the more amusing errors: A checklist for poll-watchers left out a vital item -- "poll watcher certificate" -- in a packet that went out to tens of thousands of volunteers.

This should also be amusing to most web devs: > Next, and this part I find mind-boggingly absurd, the web address was located at "https://www.whateveritwas.com/orca. Notice the "s" after http. This denotes it's a secure connection, something that's used for e-commerce and web-based email. So far, so good. The problem is that they didn't auto-forward the regular "http" to "https" and as a result, many people got a blank page and thought the system was down.


"Project ORCA is a massive undertaking – the Republican Party’s newest, unprecedented and most technologically advanced plan to win the 2012 presidential election."

I hope this puts the last nail in the coffin holding the remains of the hope that Romney would have been a better business and technical strategist for America than Obama.


The presidential election is a game, quite complicated, and the winner gets to be president. Voters and voting is just one aspect of the game.

Romney lost this match, probably partly due to an IT and analysis advantage the Obama team had.


Romney himself has a considerable track record of success in the private sector. Hard to think he would know less about what works than Obama, who has never worked there.


Because no good business/technical strategist has ever overseen a complex software project that had a bug in it?


The article doesn't really describe a "bug" so much as a complete lack of planning, ignoring warning signs, brushing off questions and constructive criticisms, and in general demonstrates the exact wrong way to implement any system.

No, no good strategist would ever release a "big data" system without stress testing it. You'd be hard pressed to find a job doing stuff like that.


That was an unmitigated disaster, possibly contributing to the total destruction of their employer?

Well, I don't label people based on single data points; they might be very nice people and perfectly competent. But I certainly wouldn't hire them after this.


Romney lost because he was a terrible candidate, simple as that. Even the Republican base hated him.

Right now the GOP leadership is busy flailing around trying to figure out why they didn't win. The ones who have gotten closest to the truth are the ones who have zeroed in on this year's lower-than-normal turnout of traditionally Republican voters (as this article touches upon). However, the underlying reason for that wasn't a failure of software or a lack of last-minute "get out the vote" campaigning, it's because people just plain didn't like Romney.

Remember the primaries? For months the Republican primaries were a game of "anybody but Romney". The primary voters at times even gave consideration to total crackpots like Santorum and Cain. Think about it -- they sent the message that they'd rather have some unelectable whackjob like Rick Santorum than be stuck with Romney. That's how much they didn't like him. But when it was all said and done, the Republicans finally gritted their teeth and accepted Romney as the nominee, despite the fact that they really, really didn't like the idea. Then on election day, many Republicans simply stayed home rather than hold their nose and vote for the guy.

Even if this "Orca" thing had worked perfectly it's doubtful it would have changed much. Republican voters' dislike of Romney just ran too deep.


unelectable whackjob

Hard to say that, given how the so-called "electable" guy did.

My objection always was, the republicans nominated the guy who could not beat the last guy who could not beat Obama (i.e. McCain).


Santorum is "unelectable" in the same way as Sarah Palin. Since they are so extreme, there will always be a small, dedicated swath of people who like them, and they can even get enough local support to get elected to an office, however once they hit the national stage they're toast.

Also, I completely agree it is ludicrous how the Republicans chose to nominate someone less popular than McCain and are now acting surprised that he got fewer votes. Gee, no one could have predicted that!


FWIW the same thing happened to Obama in 2008, at least in the key swing state of Virginia. I did some campaigning there for Obama the weekend before the election and the system that managed their get-out-the-vote efforts had crashed statewide. The precinct captains were handing volunteers old printouts of where to knock on doors and then using a highlighter to mark your route on a Google Maps printout. Given the limitations it was still an effective system for organizing a lot of volunteers.


Yeah, strike lists are The Way It Has Been Done for over a century, it works pretty well. What amazes me about this story is not that the software crashed, but that they did not have precinct captains and strike lists at the ready to step in. And that they were only rolling out their GOTV operation on the day of the election, and not sooner? Pretty bizarre.


The RNC establishment absolutely HATES Ron Paul's guts. Given the slap in the face to all the RP volunteers, a large percentage of them voted for the Libertarian candidate in protest, voted for Obama as a protest, or did not even vote.

If you look at Ohio, Florida, and Virginia (I am going from memory here) - the votes gotten by the libertarian candidate represented a large portion of votes needed to swing the state to Romney. And those 3 states' EC votes would have put Romney a lot closer, giving him 266 EC votes. Add in Colorado (numbers below) and he would have won.

Florida: 29EC votes; difference was 61,000 votes, L. got 44,000 votes

Ohio: 18E votes: difference was 104,000 votes, L got 47,000

Virginia: 13EC votes: difference was 116,000, L and Constitution parties got 44,000 .

Colorado: 9EC votes: difference was 113,000, L and Constitution got 37,000 .

Clausewitz, in his treatise "On War" dedicates an entire chapter to "Moral Forces" , what we would term "an organization's morale" .

Did crappy software contribute? Perhaps.

But wars and elections are not won by software, but by people.

That critical Libertarian slice or whatever you want to call it, of the traditional Republican base, was demoralized over Ron Paul and that resulted in an Obama victory.

Description of rule changes at RNC convention: http://articles.latimes.com/2012/aug/28/news/la-pn-ron-paul-...


> The RNC establishment absolutely HATES Ron Paul's guts

It's not just the RNC - it's also most of the voters who don't like Ron Paul. Reading about politics mainly on forums gives you a myopia, but in the actual US most people don't agree with his ideas.


Regardless, there's no way all of those Libertarian voters would have voted Republican. Some who voted Libertarian did so out of concern for civil liberties or drug legalization or avoiding war, and many of those would be more likely to vote Democratic than Republican if the option of voting Libertarian were not present.

You can't just put voters on a left-right axis.


So even making the incorrect assumption that 100% of the Gary Johnson votes were coming from would-be Republicans, none of the outcomes would have changed.


That is not what I said... key sentence is "That critical Libertarian slice or whatever you want to call it, of the traditional Republican base, was demoralized over Ron Paul and that resulted in an Obama victory."

Meaning the libertarian section of the Republican base, not meaning the people who necessarily voted for Gary Johnson.


From an Ask HN post a month ago:

http://news.ycombinator.com/item?id=4583307

> > Fairly often we see stories that link to a piece of useless blogspam rather than the original article. The guidelines say that these should never have been submitted. But are silent on what we should do if we notice the guideline being violated.

> > Is it helpful for us to flag those as we notice them so that editors can more easily find and kill them? Or would those flags just an annoyance for editors, and we should just ignore them in the belief that they will get deleted if they are a problem?

> Yes, flag them.

The story kindly submitted here was nothing but one blogger quoting some earlier publications, with no new added analysis. There have been better submissions to HN on the same underlying story today, from better sources. This is one of the hot stories today.


This is pretty usual when government and software development are mixed together. All of the bad developers I met at college are now working for the government or are contracted by the government. Most hackers I know wouldn't go anywhere close to such a contract.

My guess as to why it happens this way is that they check your scores rather than your actual skills and anyone without a diploma is automatically disqualified. Their culture is the opposite of hacker culture.


It wasn't a problem the Obama campaign had. Nor that the current government websites had. This was a failure of management by someone who, it appears, was good at exploiting tax loopholes rather than being good at managing people doing something productive.

I would argue that this is the result of mixing business people into software development. MBAs focus on ultimate results instead of iterative improvement, believe in secrecy over learning and generally don't understand the human factor of software development, much less the infrastructure involved in running it once it's written. After all, programmers are cost centers to be outsourced as quickly as possible.

The only person who admits to being a technical manager for Romney on LinkedIn is an old-school web developer, not a government worker. He's been a consultant for a decade and a half. His company's website uses Flash, loads six different totally unminimized js files and displays gradient images for the background. He is exactly the person I would expect a somewhat-older business dude to hire, if that business dude knew nothing about web development and was uncomfortable with excessively-technical things. His resume looks impressive, he's older than nearly everyone else with web experience, he can probably speak CEO-jargon-ese and he isn't going to make them feel dumb.


The government wasn't involved at all. The Republican Party (like the Democratic Party) is just a large dumb organization.


* All of the bad developers I met at college are now working for the government or are contracted by the government. *

Interestingly, all of the government developers I know would run circles around almost any startup programmer. These guys design and implement systems that are designed to last decades without even minor bugs. Frequently, they manage to keep systems designed decades ago from collapsing under their own weight.

My guess as to why it happens this way is that they check your scores rather than your actual skills and anyone without a diploma is automatically disqualified. Their culture is the opposite of hacker culture.

Government hiring is credentials and experience based. Don't be bitter that you can't get a job because you lack the credentials or experience to make the cut.

Most hackers I know wouldn't go anywhere close to such a contract.

A certain level of professionalism is expected for a real job. Nobody cares what you wear, or where you work, or how you work when you're making yet another Instagram clone. When you are designing and implementing critical systems, being a professional matters. Most hackers simply don't have the self-discipline to get such a contract.


> A certain level of professionalism is expected for a real job. Nobody cares what you wear, or where you work, or how you work when you're making yet another Instagram clone. When you are designing and implementing critical systems, being a professional matters. Most hackers simply don't have the self-discipline to get such a contract.

Disagree.

I work in critical system infrastructure. Dressing up nice to talk the computer doesn't make the code better; aggressively making the code better does. It's my opinion that the better dressed the guy writing code, the more suspect his code is...


  It's my opinion that the better dressed the guy 
  writing code, the more suspect his code is...
So do you think others make judgements about your ability to do a job based on how you're dressed?


Everyone is making judgements, are you wearing black, are you wearing a suit, are you wearing fashionable clothing, etc.

My opinion/observation is that the more attention that is paid to your clothing, the more your engineering/code quality suffers. It like the IQ test: wearing suits impresses people who think highly of people wearing suits. Maybe it's better to impress people who think highly of quality work.

edit: It's foolishness to dress finely to make people think you're awesome. You're not more of a professional because you wear a tie and suit. You're just a guy in funny-looking clothing who looks like important people. If you want to be professional, do the job well, do it kindly, do it so it can be used for its intended purpose for a long time, do it with respect for other people. Dressing nice has so very little to do with professionalism.


I find it curious that most hackers associate dressing nicely with professionalism. While dressing up can be a marker of professionalism, it is only one factor of many that goes into acting "professionally."


"A certain level of professionalism is expected for a real job. Nobody cares what you wear, or where you work, or how you work when you're making yet another Instagram clone."

Let's see... you can wear shorts, work from home, and move fast and break things, or you can wear a tie to go to an office park and build DMV software. I'm not surprised that the people with a lot of options choose the startups.


Well, I suppose the difference between the two is that in the former, you do nothing of relevance for a product that doesn't matter (which is why it can break repeatedly without incident). In the latter, you are building a program/system that people are actually using to do things.

DMV software may not be sexy, but it is infinitely more valuable than yet another Instragram clone.


A certain level of professionalism is expected for a real job.

Yeah. All those slackers at Apple and Google must be constantly wishing they could be writing COBOL for the IRS and really making a difference in the world.


Since when is Mitt Romney's campaign part of the government? And even if we give you that leap, Obama's campaign was fantastic on this particular front.


It is pretty interesting that they didn't have the system up and running well before early voting started--do campaigns not do get out the vote volunteer work during early voting?


My hunch is that this project was started way too late. At the start of the Republican primary in 2011, Romney's fundraising wasn't aggressive, and it's unlikely his team prioritized a "Big Data" initiative from day 1. In the ensuing primary slugfest, Romney probably needed to put a ton of resources into advertising to win. He also had a very large campaign payroll: his people were well compensated, but that didn't leave a lot money left-over for things like Orca.

Based on this conjecture, Orca probably got started sometime in the Spring, essentially giving the campaign 5-7 months to ship a bullet-proof GotV app.... alongside all the other stuff the team had to build (Tons of miscellaneous internal tools, the public mobile apps, etc). A lack of time and resources probably doomed it.

Moral of the story? Software is hard. Don't take it lightly.


Lack of resources was probably not a problem. The amount of soft money around was tremendous.

But too much resources leads to the temptation to throw lots of bodies at it. Lots and lots of bodies is simply not a good way to hit an aggressive schedule.


But most of that money couldn't be put towards a program like ORCA. The one saving grace of Citizen's United is the limitation on PAC-campaign coordination.


Romney could have lent the money to his own campaign, couldn’t he?


I don't believe the app is the reason for the loss. It does, however, point to a glaring weakness of the campaign: organization.


An interesting takeaway from this article, if one believes this may have helped lead to a Obama victory, is Romney's campaign had a single point of failure: that poor little web server.

One of the classic blunders!


I'm skeptical of blaming the tools so much as the people responsible for buying them, building them, etc.

Also, no amount of data driven malarky changed the fact that a lot of people weren't so excited about Romney or so upset about Obama that they changed the current status quo.

President, house, and senate remain largely the same, so really, people weren't convinced that a dramatically different direction was needed. At the end of the day, Romney lost because he didn't convince enough people to change. Some of that is money, some is tech, some is just he didn't do a good enough job leading his campaign to victory.


I agree about blaming the tools, but I think that last bit is false inference. An awful lot of people were convinced that a dramatically different direction was needed. It's just that we have roughly similar numbers of those people.


Why does everyone blame/praise the data?

It's the voters that made the difference.

We are talking 1% of people changed by campaigning, everyone has already made up their minds. They spend a billion dollars on this nonsense and probably a million "man hours" annoying people door to door. Just imagine what kind of actual good that much money and effort could have done for schools, etc.

The population is just becoming more progressive and "values" from the 1950s are less and less popular.


Remember that voting is not mandatory in the US and in fact voter turnouts are relatively small compared to the rest of the world, at less then 50% (IIRC). Their objective was not trying to flip democratic or undecided voters - it was trying to remind known republican voters to get out of their house to actually vote.


Your point stands but it's a little higher than that in recent elections; low 50s to low 60s depending on the election.

http://www.presidency.ucsb.edu/data/turnout.php (historical figures)

http://www.oregonlive.com/today/index.ssf/2012/11/2012_elect... (A guess at 2012 data)


Certainly I wouldn't argue your point that the money could be used in more productive ways, but the purpose of knocking on doors isn't to change peoples' minds, but rather to find the people who DO support your guy, so that come election day you can make sure they get out to vote.

For example, senior citizens are VERY important electorally, and many of them have mobility trouble. You want to know where they are, so your 'Republicrat Election Funbus' can come pick them up and take them to the polling station.


Well then uniform voting laws needs to be enforced on all the states.

Anyone with a mobility problem should be voting by mail.

And no-one should have to wait seven hours in line to vote, that's literally a crime being perpetrated on them.


US elections are an adversarial system run almost entirely by volunteers. If you don't like how it works in your area or think it disadvantages your guy, you don't have to wait for anyone else's permission to do something about it, you can just create a polling place or volunteer at one, or go help people vote.

There is no shortage of money, they can support all the volunteers they can get.


What? Each state board of elections is usually partisan. The number of voting places and their resources is designated by someone who is most certainly partisan. Virtually all places that had huge lines, take a wild guess if it was the left or right that was responsible.


Never thought of it that way! I was watching on the news and wondering where the hell those places were though. I've never even sniffed a legitimate line when I've voted. I have always voted in highly Democratic areas.


I live in a reliably Democratic area of California and I've both seen and been in long poll lines.

I'm not saying there isn't somewhere where local politicians are conspiring to suppress turnout, but often long lines are just a matter of understaffed or under-hosted polling places, which presumably means that the various campaigns would rather have their election-day volunteers calling people or driving them around than running a polling place.


Are you in the US? The reason I ask is that your comment seemed reasonable to me (non-US) and yet you seem to be getting down voted so was wondering if this was a US/non-US divide showing.


Agreed. Making someone wait a significant amount of time to vote is not that different from a poll tax.


It's not data in this case. It's logistics. Volunteers on foot really do stir up the vote.


I'm a progressive but I don't know if I'm convinced that the populate is moving to the left. I hope so but....

Romney's loss may just be that he just wasn't a candidate that excited the right-wing enough for them to really get out and vote. Romney wasn't conservative enough for them and he made a number of bone-headed blunders along the campaign trail.

That said, I hope you're right and I'm wrong : )


Romney wasn't conservative enough for them

I hear this a lot, usually from social conservatives trying to salvage their seat in the Big Tent after costing the party the Presidency in what should have been a slam-dunk election. It doesn't make any sense. Obama has governed as a centrist, not the wild-eyed leftist that he's been portayed as. People who voted Republican in this election did so because they shared the personal goal of getting rid of Obama by any means necessary.

More conservatism in Romney's platform could not possibly have helped.


I don't define "progressive" as right or left. I pretty much define it as "solution-driven" rather than "status-quo" driven.

The GOP has been "status-quo" since Reagan. Say what you like about Paul Ryan, he's at least looking toward the future. The rest of the GOP establishment is still livings the past. Problem is the country isn't.


The right would have voted for a lump of coal vs Obama after foaming at the mouth for the past four years at Obama. Romney is kind of proof of that since he was the pick of the primaries.

It's not a get out the vote problem, it's just that the old population is slowly dying off and they can only brainwash their young to a certain level when information flows so freely today.

I am disappointed there is only a 2.5% difference in the popular vote when Bush was only 4 years ago but people have tiny memories and change happens slowly. Values like anti-gay, anti-birthcontrol, etc. etc. etc. aren't going to fly with most voters under 25 unless they are raised in a very conservative environment and are never exposed to the rest of the world outside their home.


the old population is slowly dying off

As if the young population won't age and replace them? Believe me, age, responsibility, and having kids changes your outlook. When I was 25 and didn't know anything, I was pretty liberal too. I voted for Jesse Jackson once. And I supported Romney this year, even though he wasn't conservative enough for me, he was better than the alternative.


As if the young population won't age and replace them?

No, I don't think young people with openly gay friends and coworkers are going to decide 10 years from now that they really should be shunned by society. And if you're thinking at all rationally about marijuana, you'll recognize that the greatest danger it represents to your kids is them being arrested over it.

The GOP is seemingly going out of their way to make sure that nobody under 30 even considers voting for them, except for the strongly religious. Which as a fiscal conservative I find quite frustrating.


As I like to explain my downvotes, I don't think your post contributes to Hacker News because:

a) It belongs more in a political blog or something like Reddit. It isn't particularly intellectually insightful or thought-provoking.

b) You're assuming that everyone on HN agrees with your stance (the other half must be brainwashed and ignorant, right?)

c) It's unnecessarily vitriolic.


There are more 1st-person details at the Breitbart account:

http://www.breitbart.com/Big-Government/2012/11/08/Orca-How-...

I find this volunteer account notable:

I do not know if the system was totally broken, or if I just saw the worst of it. But I wonder, because they told us all day that most volunteers were submitting just fine, yet admitted at the end that all of Colorado had the wrong PIN's. They also said the system projected every swing state as pink or red.

Putting aside Hanlon's Razor ("never ascribe to malics what can be explained by incompetence") for a moment: what if some of these failures were due to partisan/hacker tampering?

If you owned the server(s), you could scramble a state's PINs. You could break redirects. You could bias the projections for overconfidence, or mark records as 'voted' or 'followed-up' that hadn't been.

With a little phone-hacking, you could make sure attempts to call-in with problems get no answers.

A single skilled amoral hacker could more easily suppress thousands of actual votes this way, compared to voting-machine/voting-tabulation tampering. And the details could disappear more easily into the secrecy and cover-your-ass bullshitting that is common in a partisan, temporary campaign organization.


That wasn't much of a look inside.


Romney never played the game to win, but to show there are even worse alternatives than the black Bush.


What a vulgar sentiment! President Obama is hardly a black Bush, whatever your thoughts about his policies. If you really believe they are the same--yet utterly fail to articulate your rationale--there probably isn't much reason to engage in a discussion with you.


Disagree w/ your specifics, strongly agree w/ the perception of a narrative being present. Agree that Romney was never in real danger of winning, disagree that he himself threw the contest. He wanted to win, it just wasn't in the cards.




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

Search: