Tag Archives: voting

“like” problems: social ‘voting’ is a bad idea

The news story making the rounds about Facebook the past few days indicates they’re working on a kind of “dislike” button.

The problem with the Facebook “like” button is the same problem Google has with Google+ and their “+1” button: it doesn’t tell you anything meaningful.

Voting on Reddit doesn’t really convey much meaning, either.

Stack Overflow tries to address this with its up/down voting and being able to see the gestalt votes as a ratio (if your rep is high enough (an admittedly low bar, but still a bar, and an aspect of the gamification of Stack Oveflow)). But that doesn’t really cut it, either.

The problem with online “voting” (or “liking”, or “plussing”, etc) is that it is a dimensionless data point.

Does getting 300 “likes” on a post make it “good”? Does it reflect on its quality in any way? How about getting nearly 400 upvotes (and only a handful of downvotes) on a question about MySQL (along with 100+ “favorites”) mean the question is good? Does it show something is popular? Are people clicking the vote mechanism out of peer pressure, because they actually agree, or because they think it needs more visibility? Or something else entirely?

Dimensionless data that gets used as if it has meaning is a problem – one of many problems of social media and web sites in general.

Of course, you will object, quality is a potentially-subjective term – what does “quality” mean, exactly, when talking about a post, website, question, etc? Is it how well-written it is? Is it how long? How funny? How sad?

Take this question I asked on Stack Overflow, “CSS – how to trim text output?” It’s clearly-written, was answered excellently in 2 minutes, and is a “real” problem I had. Yet in the 4.5 years since asking, it’s only gotten 2 votes total (both “up”, but still only two).

Reddit has upvotes and downvotes – and your comment/post score is merely the sum of the ups and downs; below a certain [relative] threshold, you won’t see content unless you ask for it.

One of the biggest problems with all of these systems is that the “score” doesn’t actually tell you anything. An atheist subreddit, for example, will tend to downvote-into-oblivion comments that are theistic in nature (especially from Christians). Quora‘s voting system is highly untransparent – downvotes don’t really seem to mean much, and upvotes are pretty much just for show.

This derives from the fact that these sites use dimensionless data and try to give it a value or meaning outside of what it really is – a number.

What should be shown is the total number of “votes” a given post has gotten – positive negative, reshare, etc – but never combined. A ratio could be displayed, but the sum of the votes is a poor plan.

Facebook, Google+, and others should offer various voting options – “up”, “down”, “disagree”, “agree”, “share”, and possibly others – some of which may be mutually-exclusive (you cannot upvote and downvote the same thing), but you might downvote something you agree with (or upvote something you disagree with) just because of how it is written/presented, etc.

And the total of each type of click should be shown – show me 10,000 people disagreed with what I said, 15,000 agreed; 20,000 upvoted, and 30,000 downvoted; 12,000 reshared it (with, or without, comment).

Using voting as a means of hiding things (and trying to prevent others from seeing them) can be somewhat akin to online bullying – revenge voting has its problems; as does blindly upvoting anything a particular person says/does. Which is why assigning (and then displaying) dimensionless data anything more than a count is dangerous.

a simple restructuring of elections

In close follow-up with my desire to see political parties abolished, we also need to rethink how voting is done.

In the United States, you can only vote for a single candidate for most positions (town councils are an exception).

You do not have the opportunity to say anything more than a binary yes|no to a given person for a given office.

You can vote for Bob for mayor. But not voting for Mary, Quentin, and Zoe doesn’t really say anything about what you think of them – just that you liked Bob the best.

And there is the problem. There is an explicit elimination of relative preference when voting: all you can do is vote “yes” for a candidate.

That is very different from voting “no” against a candidate.

What should happen instead is you should vote for your favorite candidates in order of preference, so Bob is number 1, Zoe number 2, Quentin number 3, and Mary number 4.

Then when I vote, and rank them Mary 1, Zoe 2, Quentin 3, and Bob 4, we can get a picture of the relative preference of any given candidate running for the office.

Do this across all voters in a given election, and assign the winner to the person with the lowest score (in the numbering shown above – flip the values to assign the winner to the person with the highest score).

Perhaps even look at the top 3 or 4 after gestalt ranking, then vote again to determine the winner (this would be ideal for a Primary-then-General Election method).

What research shows is that while you and I may wildly disagree on “best” and “worst”, we’ll probably be pretty close on who we think is “good enough”.

In the Bob-Mary-Quentin-Zoe example with two voters, Mary & Bob both got 5 points. Quentin received 6, but Zoe earned 4.

The two voters, therefore, think Zoe is “good enough”, even though they part ways on “best” and “worst” (Bob & Mary).

Combine such a ranking system with a fully-open Primary election (ie you go rank every candidate regardless of “party”), and we would see much more representative-of-the-citizenry candidates appear at final Election.

why the electoral college matters

This year’s election results seem to – again – be confusing a LOT of people.

The incumbent presidential candidate, Mr Obama, won ~51% of the popular vote. His main opponent, Mr Romney, won ~48% of the  popular vote.

However, when you look at the electoral votes (the only ones that really matter), you see a different picture: 332 vs 206, which puts Mr Obama’s electoral victory at 61% of the Electoral College, and Mr Romney at 39%.

For some reason, and I have my personal theories on this, civics and American History is no longer actually taught in schools. No one today knows what the Connecticut Compromise was about. Let’s do a little history lesson to bring everyone up to speed.

In 1787 there was no “United States of America” – folks were still trying to figure out what to do with the nascent country that just won its independence from the British Empire. Virginia representatives proposed having a two-house structure for Congress (the Senate and House). However, they wanted both houses of Congress to be apportioned based on population – at the time, that would’ve meant a disproportionate level of influence from the more populous states over lower-populated ones (irony: New Jersey in 1787 was one of the smallest states by population while Virginia was one of the largest: NJ has almost a million more people today than does VA). For obvious reasons, the smaller states felt this was a Bad Idea™ – their voice would never be heard.

The Compromise brought the ideas that New Jersey wanted (a unicameral representation based on the concept of one vote per state) and the one Virginia was lobbying for (bicameral, but both houses based on population) into the system we have today: a bicameral Congress with one house based [loosely] on population*, and the second a flat number per state (ie, our House of Representatives and Senate).

With Congress out of the way, let’s look at how the President is actually elected. Article II of the Constitution covers this (along with Amendment 12). This is where things get interesting: to help mitigate the disproportionate effect of large states on small ones, each state votes for Electors who will then “really” vote later for the President (and Vice President).

Why is this important?

First, it is an evidence of the fact that we do not live in democracy – we live in a representative republic.

Second, it allows every state to have at least minimum voice in an election – which means that it views every state as important.

Third, it means that pure favoritism shouldn’t be the exclusive basis for why any given candidate becomes President. Being President isn’t supposed to be a popularity contest in the way a beauty pageant is, it is supposed to be a race to determine the best leader for the country (of course, “best” is subjective, and few actually seem to campaign because they want to ‘lead’ – they seem more to run for the thrill of being “in charge” .. but that’s another post entirely).

How are electors apportioned? Most states distribute electors in a winner-take-all form: if a candidate receives a simple majority of the popular vote in the state, they get all the electors of the state (eg a 51% win in CA gets all 55 electors even though 18.4 million of the state’s population of the state may disagree with the 18.6 that elected a given candidate). Hypothetically this shows that the States are joining together to vote for the President rather than merely the populace.

Not all states follow that model, however – Nebraska is a notable exception which awards Electors based on the vote percentages of its population.

Some argue that this system inherently creates “swing states” which lead to disproportionate campaign expenditures and focus instead of spending approximately-equal time in every state.

Personally, I think this is a fantastic system because pure democracies devolve into anarchy and/or split into multiple groups upon reaching a given size.

The Founders of our country were a lot smarter and forward-thinking than most are willing to give credit for. Were they perfect? No. Did they have flaws in the initial proposals? Absolutely. But this is one artifact of our founding that needs to stay.


*“The Number of Representatives shall not exceed one for every thirty Thousand” – Article I, Section 2. If we followed this minimum today, we would have >10,000 representatives in Congress (2012 US population ~310,000,000)