This website is a repository for information collected by the Society of Learned league Obscure and Byzantine Research, from this point on referred to as SLOBR. It consists of mostly web pages and download-able csv files that contain statistics compiled by SLOBR. These statistics will be updated at the end of every season. As of release 1.0.14, the following statistics are available:
The following is a csv file of all players and locations: click here.
The following is a csv file of all players and schools that they attended: click here.
A Llama-Cycle is a pattern of repeated win-loss-tie numbers repeated for the entire season. The cycle-size is the length of this pattern. For example, if a llama's score is 1-1-1 after match 3, 2-2-2 after match 6, 3-3-3 after match 9, all the way to 8-8-8 after match 24, then this llama would have achieved a llama cycle of size 3. So far, the shortest cycles found have been 5 matches.
To see the shortest llama cycles, click here.
To download a csv file containing all cycles for all llamas, click here. The data in the csv files contains fields of the form X-Y where X is the season number and Y is the cycle length. For example 76-8 would mean that in season 76, there was a cycle of length 8 by that player. All the numbers in the player's won-loss-tie record for match 16 should be twice the numbers in the player's match 8 record, and the won-lost-tie record for match 24 should be three times the numbers in match 8.
WONDER stands for Warren's Overtly Narcissistic Defensive Efficiency Rating. This metric measures how many (or fewer) match points a player would have if only the number of questions answered were used for scoring. If a llama ties a match where their opponent answered more questions correctly, then their WONDER number is incremented by one. If that llama wins a match where their oppenent answered more questions correctly, then their WONDER number increases by two. If a llama wins a match where the players tie in total number of questions correct, then their WONDER number increase by one. Similarly, a player loses the WONDER number value if their opponent gains a WONDER number value. This statistic was developed by Your's Truly during a season where his WONDER number was high.
To see the highest and lowest WONDER numbers, and to see the highest and lowest WONDER numbers per match played, click here. This data looks to be skewed by how many matches a llama has played. The highest and lowest totals contain players who generally have played many matches (makes sense), and the highest and lowest averages contain players who have played few matches (makes sense because a smaller sample space is available).
To download a csv file containing all WONDER numbers for all llamas, click here. The fields in this csv file are the WONDER number, and the average WONDER value (WONDER number divided by matches played).
In Learned League matches, it is possible for a player to have 26 different
scores. 0(0), 0(1), 1(1), 2(1), 3(1) ... 9(6).
We are not counting 0(F) because we do not want to encourage forfeits.
Some llamas have commented that they have achieved all possible scores and
wondered how often this happens and what the minimum number of matches
were needed for a player to achieve this feat. It turns out that roughly
10% of all llamas have done this. To see a list of lamas who managed
to achieve this feat in their first 4 seasons
click here.
To download a csv file for all llamas click here. The fields in this csv file are the number of matches needed to get all possible scores, or the matches played so far if the player has not accomplished this. The next number is the number of different scores needed to get all possible scores. This is followed by a list of scores that the player needs.
MOPS stands for My Own Private Scorigami. This concept was developed by Jon Bois for NFL scores and has been tracked by llama MattinglyD (see Scorigami). MattinglyD has graciously allowed me to appropriation this name.
HUN stands for Hamill-Usui Number, a concept named after HamillB, who first proposed it, and UsuiW, who first implemented it. Every Llama has a Hamill-Usui Number relative to every other Llama. To compute your Hamill-Usui Number versus another Llama, check every question that you played in common with the other llama. Add one to the numerator for every question that you both got right or both missed. Divide this by the total number of questions that you have answered in common. The closer you are to 100%, the more that you have in common with this llama.
Because of the nature of this value, every player has a different HUN versus every other player. A csv file of HUN data for every user would be unique to that user. Because of this, the amount of time and storage needed for every HUN for every user would be too large for SLOBR to want to do. Instead, we have provided a tar file and some simple instructions to calculate your own hun tables.
First, on a Linux system with python 3 installed, download this file to a local directory. In the example below, that directory is /tmp. Then mkdir a fresh directory and cd there, and then execute the following commands (where xxx is the name of the llama).
tar xvf /tmp/hun_values.tar . python find_hun.py xxx
This takes a little over one minute to execute on my machine. When finished, cd to generated_files. You will see the following files there:
Special thanks to Llama JamesK for providing these instructions for installing the Hun calculator on Windows.
Scores of 9(5), 8(4), 7(3), 5(2), and 3(1) are the most one can score with the given number of correct answers (poorest defense possible). The highest scoring llamas in each category, the sum of all categories, and the corresponding average values are displayed here. Like WONDER numbers, the raw scores do skew toward those who have played a lot of matches, and the average values skew toward rookies.
To download a csv file for all llamas click here. Each line in this csv file contains in order a player's name, the number of 9(5), 8(4), 7(3), 5(2), and 3(1) matches they have scored, the sum of the previous numbers, and the total number of matched played.
Llama_slobber is also a python package available through pypi.org that contains methods to help scrape Learned League data and to compute these statistics. It is open source python 3 code that can be cloned from github. See here for more details.