Why "head-to-head" battles?

Why use the two-at-a-time, head-to-head method of voting for who is the hottest? Well, when it comes to ranking 1st, 2nd, 3rd, 4th, etc., I found other voting systems lacking:

The Bracket System

As virtually all of you know, this the kind where 64 entities (college basketball playoffs being the most famous example) face off two at a time until there are 32 left, then 16, then 8, then 4, and finally the final confrontation between the remaining two to create an overall winner. This is great when your objective is to produce a single champion.

But since we are trying to create a numerical ranking, this is far from being a desirable and accurate method. Here's why. It's possible that the first and second seeds - the two who are actually the best - can face off in one of the first round matches. Each party can be highly skilled and talented, the battle can be fierce and the end results can be close, but ultimately the loser of this match wouldn't even be considered to be in the top 32 of all contending entities. In reality the loser of that match may have actually been victorious when pitted against most of or all other opponents; yet since this second seed fought the very best right off the bat they were eliminated and had no chance at proving themselves further and reaching the top slots of the bracket.

Again, this may be ideal for sports where your objective is finding the single best. But here we want to show who is 17th overall in addition to who is in first place.

Various Polling Methods

We've all seen where groups of individual choices are presented. They can be lists of three, five, eight, twelve, or whatever:

Good, simple solution, but again you're only voting for whom you think should be in first place. It's not forcing you to pick who's better between who you would consider to be in second and third places, and so on. This can make it inaccurate when it comes to having the full preferences of each voter expressed.

This has the inadequacies of the picking a single winner, and doesn't allow non-winners from previous polls to battle against contenders in new polls. Plus it also takes quite a bit of time (sometimes weeks) to find the overall winner.

This is nice method, and works terrifically at laying out the ranking of the contenders 1, 2, 3, 4, 5, etc. A small drawback is that the voter must at times rack his or her brains out trying to order two or three items that they have no preference over each other (perhaps they'll have three contenders all "tied" in 4th place). This can be poor in usability terms, annoying to to voter, and time consuming.

Rating Methods

Rating a picture on a scale on a scale of 1 to 5 or 1 to 10 may seem like a logical method, but in reality it's piss poor. People will tend to rate everyone they like a 9 or 10, and rate everyone they hate a 0 or 1. This can leave a very skewed (for or against) ranking for each contender. Plus you can get enough people manipulating the voting system (see below), and you end up with results that are highly inaccurate, and quite frankly, completely useless.

Gaming the System

Another disadvantage to the aforementioned polling methods is the ability to game the system. Whether it be by spamming, voting by bots or automated scripts, deleting cookies, voting from multiple IP addresses, signing up with multiple accounts, there are ways to vote numerous times and seriously fix the results in your favor.

Advantages of One-on-One, Head-to-Head Battles