1. Welcome to Smogon Forums! Please take a minute to read the rules.
  2. Click here to ensure that you never miss a new SmogonU video upload!

Data Types and Tier Dominance: Some Cool Visualizations

Discussion in 'Smogon Metagames' started by Antar, May 21, 2016.

  1. Antar

    Antar
    is a Battle Server Administratoris a Programmeris a Super Moderatoris a Community Contributor
    Official Data Miner

    Joined:
    Feb 17, 2010
    Messages:
    3,874
    A few weeks ago a user by the name of smooth came to me with some really cool charts showing the type breakdowns of each tier and was thus able to do some analysis related to which types dominated which tiers.

    His analyses for April can be found here: https://onedrive.live.com/view.aspx?resid=B45EAEE46513E2B1!852&ithint=file,xlsx&app=Excel&authkey=!AOFvCSuP-_eZIcw

    and some charts for June here: http://imgur.com/a/WicgM

    Here's an example:
    [​IMG]

    Beyond that, I'll let him explain them more fully!

    Great work, smooth! Can't wait to see more analyses like this.
    Last edited: May 21, 2016
  2. smooth

    smooth

    Joined:
    Jan 25, 2015
    Messages:
    4
  3. smooth

    smooth

    Joined:
    Jan 25, 2015
    Messages:
    4
    Alright, here are what some of the charts mean:

    The Borda Count Charts:
    Imagine if every tier had a point value from 0 to 10. The value of each tier goes up by 2 the higher the level it is. From Ubers down the numbers are: 10 for Ubers, 8 for OU, 6 for UU, 4 RU, 2 NU, and 0 for PU. PU gets 0 because it is the lowest tier. Borderlines also get a point relative to the tiers they are in between. BL gets 7, BL2 gets 5, BL3 gets 3, and BL4 gets 1.

    If you take the mean of all the points that a type would receive based on what tiers its pokemon are in, you would get these numbers (represented by the Borda Count graphs). I made three graphs, one of them with PU taken into account, one without PU, and the last showing the difference between the two. Notice the massive changes in ratings that the Bug, Ice, and Normal types undergo when PU is included.

    The Tier Pie Graphs:
    The pie graphs simply show the percentage of dominance of each type in each tier plus the borderline group below that tier. One interesting thing to do is to look at one type specifically from Ubers and to then scroll down, while still looking specifically at one type, you may see how it changes as it goes from higher tiers to lower tiers. An interesting example of this is the steel type, where it plateaus at Ubers and OU, but then gradually decreases in dominance from there. Another example is the Dark type. The Dark type starts out a 6%, then drops to 4% in OU + BL, then reaches its apex at UU + BL2, and then proceeds to gradually drop in percentage as the tiers go down.

    The next graph with all of the colorful bars:
    This is probably the most complex graph of them. Essentially, this graph is the collection of the tier pie graphs, and then are split up based on typing rather than tiers themselves. Each colored bar represents a tier/borderline, and all of them added together make 100% just like how a full circle represents 100% in a pie graph.

    The Privilege Chart:
    This chart is mostly a gag chart based on the "Check your privilege" slogan said by some feminists. This chart however, is a nice overall representation of the level of performance of pokemon of their respective types. This is calculated with a summation formula that is cumulative over all the tiers AND borderlines. The fomulas are as follows:


    Privilage Rating=100×ROUND((Numerator Value)/(Average numerator value of all types),2)

    Numerator Value=(∑_(n=PU)^Ubers▒((Average # of Pokemon per type at nth tier & below )/(Average # of Pokemon per type in All tiers)-(# Pokemon of type at nth tier & below)/(# Pokemon of type in All tiers)) )+2
    GMars, Swaggy-G, talha123 and 5 others like this.

Users Viewing Thread (Users: 0, Guests: 0)