The X-Wing community has already built some amazing tools to study and analyze tournament data. Examples of these are sites like MetaWing, PBM, and Advanced Targeting Computer.
These are great tools for gaining deep insight into a tournament after it has finished, but they require the data to be exported from the tournament page and uploaded to Listfortress. However, with the sunsetting of TTO and the switch to Longshanks, there is currently no way to export the tournament data from Longshanks. This means a chunk of the tournament data is missing from Listfortress.
Pattern Analyzer tries to fill this gap by obtaining data directly from Longshanks (and other vendors) and allowing people to discover what squads are played and gain some additional insights into an X-Wing tournament while it is still in progress. And since Pattern Analyzer acquires all this data anyway, it can also provide an export for Listforstress, so that it stays up to date and other tools like MetaWing and PBM have information to consume.
Some squad lists may have been entered in a format that can not be analyzed by Pattern Analyzer and will be not considered in most of the statistics. In other cases they will show up as “unknown” or “???” to indicate how much of the information could be used to generate the statistic.
Below you will find explanations of some commonly used terms:
Percentile: The percentile indicates how well an entity (pilot, upgrade, ...) performs in comparison to other entities in the same set (single or multiple tournaments). It is calculated by averaging the results, and expressed as a percentage.
For example, if a pilot makes 1st and 3rd place in a tournament with 10 participants, her percentile will be 89% since there is one occurrence of the the pilot that is better than 100% of the field (1st place) and the occurrence (3rd place) is better then 78% of the field. Thus on average the pilot performed better than 89% of the field.
Std. Deviation: The standard deviation measures how dispersed the performance (percentile) of an entity (pilot, upgrade, ...) is in relation to its mean. Low standard deviation means the results are clustered around the mean performance. On the other hand, high standard deviation indicates that the results are more spread out. Basically, a high standard deviation means a high variance in results and vice versa.
If there is only one occurance of an entity, a "-" is displayed rather than the standard deviation of 0% to indicate that there can not be any deviation.
Winrate: The winrate is expressed as a percentage and represents how much games an entity (pilot, upgrade,…) has won in comparison to the total games it appeared in. Sometimes round data is missing which causes an entity to have 0 games recorded. In this case a "-" is displayed to indicate the missing information.
Frequency: The frequency is expressed as a percentage and represents how much games an entity (pilot, upgrade,…) appeared in. This is usually in relation to the faction (in case of pilots) or to the slot (in case of upgrades).
Count: Since frequency is a relative value, the absolute number of occurrences is also presented.
Score: The score value attempts to balance small sample sizes and their percentile. However, be aware that there is no elaborate science behind this number, even though it tries to quantify the percentile, and count of an entity. The goal of the score is to indicate whether a result is just a happy accident or if it is somewhat reliable because multiple people achieved a similar result. Essentially, it"s a feel-good metric!