Learned League, Mid-Season Review

It's day thirteen of Learned League Season 55. With 12 matches behind me, twelve before me, and one currently pending, I thought I'd take a look at what's been going on.

Learned League is an online trivia contest that features head-to-head competition: everybody gets the same questions on a given day, but you are matched up with a single player, and your success is measured solely against theirs. To make it more interesting, you decide on what points your opponent will receive for correct answers (and vice versa), so even if you answer fewer questions, you might still win on points, if the ones you got right were the ones your opponent thought were the hardest.

(More info: Learned League FAQ)

Normally, players are grouped into "rundles" based on performance in previous seasons, so you can expect that the folks you are facing are at about your same level of triviality. Rookies, however, get lumped together in a big groups, resulting in battles of widely different levels. And this year, I'm a rookie.

Numbers Racket

Here are the stats for my rundle (and here are my stats, registration required). I'm currently in 11th place of 30 with a record of 7 wins, 4 losses, and 1 tie; but at my height a few days ago, I was in 4th. In the two matches since then, I've been really unlucky (on Friday, I could only get one answer right, and received zero points for it), but it looks like I'll be up against softer opponents for a little bit, so I should be able to make up more ground.

One thing that's been bugging me is that I've got a slightly-harder-than-average draw, especially in this first half. There are 25 matches, but 30 players in my rundle, leaving five people I won't face. And all five of those are currently ranked in the bottom half. And I've mostly been facing harder opponents thus far. (I am currently tied at #1 for "Correct Answers Allowed" which is a general indicator of how strong one's opponents have been.) You may notice I've been entranced by statistics, particular. Every day, I import the updated stats into an Excel sheet, so I can see my past and future matches color-coded against the median.

Similarly Erratic Results

After two days of competition, I tweeted:

My first LearnedLeague 6 pack comes after a painful loss on day 1. I expect similarly erratic results from now on.

It's turned out to be a good prediction. A 6-pack is, of course, getting all the questions correct on a given day, a relatively rare occurrence in all but the higher echelons of the league. But beyond that, well, see above.

A lot of my success comes from managing to craft good guesses based on the context of clues, rather than being certain of particular knowledge. It can be frustrating, especially when a guess (or two! (or three!!)) goes slightly off. When I first played a live version of this game, it was against a group of pretty serious folks (the NPL), and I left feeling like nothing was in my control, which put me off of the league for a while. But the prospect of settling into a nice matched group is pretty appealing, so I'm eager to finish this season.

Play Along at Home

The LL website has been slowly developing, and now it's really well designed for playing along even if you're not registered. After the day's match is over, the website is updated with the questions and how well all the players did, with the answers concealed by a script. Here are the first day's questions (you can reach other days by using the "Match Day 2" arrow near the top of the page, or by choosing from the calendar on the main page). Individual questions have their own pages with detailed breakdowns of accuracy, the most common wrong answers, and the "best wrong answers." For example, here is Question 3 of Day 9:

The work of what 19th c. English engineer and mathematician on what he called a Difference Engine and Analytical Engine, which are considered today among the first mechanical computers, has earned him the moniker "Father of the Computer"?

Forty-eight percent of players answered correctly; the most common wrong answer was "Charles Turing" at 11%, and the best wrong answers were Charles Widmore, Dr. Emmett Brown, Sir William Computer, and Sir Nigel Speakandspell.

So far, some of my favorite questions have been:

Hey girl, who is the only former MMC (Mickey Mouse Club) Mouseketeer to be nominated for an Academy Award for acting? (Match Day 2, Question 6)

Identify this musical group. [Image] (Match Day 10, Question 2; be sure to look at the best wrong answers)

In the mid to late 1960s, author Arthur Hailey published two simply named novels, which each explored the operation and politics of a single specific location/establishment (different in each novel), and both of which inspired film adaptations (and one a television series). Name both novels. (Match Day 12, Question 6)

Previous seasons are available for review, too, though as you go back further, the display interface gets rougher. There are also themed "minileagues" and one-day competitions that go on between main seasons. One of the things that really excited me about joining the League was kibitzing on thedan's hilarious collection of Before & After trivia, where each question contains two parts that merge together. (Ex: "Name the 1960s comedian who was famously convicted for obscenity based on live performances in which he demonstrated his original martial art, Jeet Kune Do." Answer: "Lenny Bruce Lee.") And if you know the answer to this question, you are officially an awesome person:

Name the fictional game show on which the host (played by Bill Murray) asked contestants to determine which of Lorenzo Lamas and Ricardo Montalban is more like WNYX station owner Jimmy James (as portrayed in his poorly translated autobiography).