This post is meant to serve as a brief guide to understanding Pink Brain Matter. (PBM)
About the data
All data in PBM come from ListFortress and is automatically loaded daily. The data quality is, for most things, exactly as on ListFortress, so duplicate tournaments might exist, lists may be wrong etc.
Almost everything you see in PBM is based on “Seasons”. A season is a dataset either defined automatically or manually curated by the PBM administrator. New seasons are added whenever a new wave releases or point costs are changed. Seasons are named after the wave they cover, plus an optional letter. For instance, Season 2a is from the release of First Order & Resistance in December 2018, while Season 2b is from the January 28th 2019 points change.
Several sub-divisions exist for each of these seasons that further divides the data into formats as well as some (hand-picked) curated variants.
For instance, for Season 2a, the following can be selected:
- Wave 2a – Hyperspace: All tournaments between Wave 2 release and the January 2019 points change, where every list recorded is Hyperspace legal
- Wave 2a – Extended (Curated, Large Only): A hand-picked set of larger Extended tournaments in the period.
- Wave 2a – All data: All the data in Wave 2a, across formats. (Also includes custom formats)
Note that this means that a given tournament may be included in more than one season.
You can select which season you want to see data for in the leftmost dropdown menu at the top of the screen. The numbers shown in parenthesis after the season name is the number of tournaments in that season.
A little terminology
A few terms are important to understand in order to get the most out of the data:
- Pilot: A specific named pilot, for instance “Whisper” or “Academy Pilot”
- Ship: A type of chassis, for instance “T-65 X-Wing”
- List: A combination of pilots form a list
- Archetype: Slightly more general than a List, for instance “Resistance Aces” is an archetype. Archetypes are assigned using a complicated rules system and is still very much under development, especially for 4+ ship lists.
- Loadout: A specific upgrade-combination on a single pilot.
- Percentile: An indicator of how well the pilot/ship/list/archetype listed performs. This is calculated by averaging the results, and is expressed as a percentage. If the pilot/ship/list/archetype in question has won every tournament it has appeared in, the score will be 100%. If it, on average ends up in the middle, the score will be 50%. Be aware that in general popular “netlists” might score lower than their potential, as they are played by a wider skillrange of players.
- Standard Deviation: Related to Percentile. A statistical term used to indicate how far the results deviate from the mean, on average. A high number indicates a large variance and vice versa.
- p.p: Percentage Points. Used to indicate a performance-difference between something and it’s baseline. See “Pilots” for an example.
- Representation: Shows, in percent, how big the share of the meta is. This is usually in relation to the faction. You will often find a comment at the bottom of the page explaining what the representation (%) signifies. In some instances In Lists% or Frequency is used to synonymously.
- Count: In a few cases, Representation (%) is not feasible and a simple count is displayed instead.
By selecting “Season Overview” in the rightmost dropdown menu, you get a high-level overview of the selected season.
At the top you will see a chart showing the top 20 most frequently flown lists. Color is used to indicate the faction. The more popular a list is, the higher it will appear on the chart, and the better it performs the further to the right it will be. You can drag an area using the left mouse-button to zoom, right click to reset. (zoom out)
The second chart shows the top 20 most frequently flow archetypes. You can doubleclick an archetype to open the detailed page for that archetype.
At the bottom is a table with all the tournaments in the season. You can click an individual tournament to get details on that tournament. Here you can click on an individual player to drill-down further, including match-data and other tournaments attended by that player.
The Faction Overview shows data about the performance of each faction in the season as well as the most commonly flown archetypes of that faction. You can click the archetypes to open the detailed page for that archetype.
Selecting “Pilots” in the dropdown shows an overview of every pilot, divided by faction and ship chassis. Clicking the pilot name allows you to drill-down further into detailed statistics for each pilot. Here you can, for instance, see how the pilot performs in lists of varying size, with different upgrades, or when flown together with a specific other pilot.
Key here is to look for the “pp” value – for instance Poe might have an average performance (percentile) of 51.0% – but with Nien Numb it is 57.8% – giving a “pp” of +6.8. So on average, a list with both Poe and Nien Numb will perform 6.8 percentage points better than an average Poe list. Reading the “pp”‘s requires some thought though, as a comparion with a very high frequency will mean that the average is already weighed strongly towards that. (The 51.0% average performance on Poe in the example includes quite a few lists with Nien Numb already in them)
Lists all upgrades, grouped by upgrade slot, and their average performance. Clicking the upgrade name (or an upgrade on the chart) will allow you to drill-down further into detail statistics for each upgrade, similar to the pilot drill-down above.
Bids show how bids are distributed across lists. On the vertical axis is “number of lists” – cumulative, meaning that you see number of lists with that bid OR less.
Each line on the “Average Initiative”-chart indicate an initiative “bracket” – so a list with two i5’s and an i3 (Average: 4,33) will be included in the “4-5” bracket.
The “Highest Initiative”-chart on the other hand will only look at the pilot with highest initiative in the list.
At the bottom there are two tables. The first (“Bidding Chart – Move Last”) shows your chance of moving last at a given initiative and bid-level. The second (“Bidding Chart – Move after specific initiative”) allows you to see your chance of having a bid that matters at any given initiative, so you can, for instance, judge whether you’d be able to move after the “beef” vs a Ace+Beef list.
Timelines are different from the other options, as they ignore the season selection. Instead a series of charts are displayed showing data divided by calendar months. Two sets of timelines can be selected – one for all data and one for Hyperspace only.
In “All” you can see:
- Faction representation: How much is each faction played?
- Faction performance: How well does each faction do?
- Overall attendance: How many players are attending tournaments?
In “Hyperspace” Faction representation and Performance is repeated, but only counting Hyperspace events – but it also shows:
- Average performance of Hyperspace-legal lists in non-hyperspace tournaments: This gives an idea of the relative power-curve between Hyperspace and Extended for each faction
- Format popularity: Shows the popularity of Hyperspace vs other formats
As PBM is still in development, you will, on occasion, run into links that leads nowhere or places where raw XWS pilotnames are used. This will eventually be fixed 🙂
Please comment below, or on the newest post on the blog if you have feedback or ideas for the future development of PBM.
One Reply to “Using PBM”
Kyber Cup Season 5 is in ListFortress now ready to be included in the curated for Wave 8; largest event on TTS yet!