Monday, November 4, 2013

Numbers



    In the realm of Azeroth we seem to be surrounded by numbers.  They are everywhere.  Item level, character level, DPS, there is no escaping numbers.  Numbers can be a good thing, but they can also be abused.

   Let’s take the pet score.  This may be a number you are not familiar with.  Wowprogress is a site that ranks all kinds of things, realms, and characters by item level, guilds by progression and more.  One of the rankings on this site is something called a pet score.  Without knowing anything about how the site makes a ranking determination you could make all kinds of assumptions.  You might assume that the number of pet battles or the number of PvP battles you have one might be factored into computing this score.  One assumption I made was that the number of pet related achievement points was factored into this score.  The other day I found out I was very wrong in my assumption.

   In my guild, is a shaman, Wildsong by name.  She is (or was) the undisputed pet champion.  If it’s a collectible pet, she has it.  She has the Ethereal Soul Trader; she has the various rare drops from throne of thunder.  She has over 500 unique pets.  In checking her ranking, she is first in the guild and fifth on the server for pet score.  This, of course makes perfect sense.  It did until the other day.

   I am online the other day, plodding about and an achievement toast comes up in guild chat.  “Wildsong has earned the achievement Taming the World”.  Along with that toast was another one announcing she had defeated Grand Master Aki.  I was stunned.  I got the tamer achievement a long time ago.  I would have figured Wildsong did too.  I congratulated her on her achievement and took the time to ask how the Celestial Tournament was going for her.  Her response was that she had not yet beaten it, but now that she had beaten the tamers in Pandaria, she had hope and confidence.

   I thought it would have gone without saying that she would have beat not only the Pandaria tamers, but the Tournament as well, based on her pet score and ranking.  I have managed to beat the Tournament five times now.  I have two of the reward pets to prove it and am working on the third.  My ranking is second within the guild and ninth on the server.  Apparently these scores and rankings followed some logic I was not familiar with.

   I looked into exactly how WoWprogress computes these numbers.  They do it based solely on the number of pets you have, the level of said pets, and the quality (rare, uncommon, etc…) of said pets.  For instance a level 25 rare pet is worth is worth 62.10 points.  The same pet as an uncommon is worth 57.32 points.  A rare level 23 pet is worth 51.96 points.  It is entirely feasible that a player could never fight any tamers and simply accumulate a huge collection of wild pets from Pandaria and have a huge score.  Fill up all 1000 pet slots with bandicoons and marsh fiddlers and level them to 25 and you too could have a high pet score and ranking.  Let’s do the math.  If I decided to capture every wild pet in Pandaria with a quality of uncommon or rare, and then level them to 25 I am quite sure I would be first on my server.  Let’s assume that I wind up with 75% uncommon (750 pets) and 25% rare (250 pets).  My pet score would then be 750*57.32 (this equals 42,990) plus 250*62.10 (this equals 15,525.  My total pet score would be 58,515.  By way of comparison, the number one pet score on my server is 17,088.  The number one slot in the entire US has a score of 36204.

    Clearly we can only draw limited assumptions from Wowprogress’s pet score.  My hypothetical collector of bandicoons and marsh fiddlers would probably be ranked number one in the world.  This same hypothetical character would probably never win a PvP pet battle, would not be able to defeat most tamers and would probably have very few pet achievements.

   All numbers are the same in this regards.  Let’s take iLevel.  The common wisdom seems to be that higher iLevel generally equates to higher performance.  This of course is simply not true.  One thing I seem to have plenty of is trinkets.  I have a nice mix of tanking and DPS trinkets.  Let’s say that my highest iLevel trinkets are my DPS trinkets (these will be the 553 pieces from SoO normal) and that all my tanking trinkets are from LFR (so my tanking trinkets are iLevel 528).  A flex raid is forming and they need a tank.  I decide to cheese iLevel and equip my 553 DPS trinkets.  Sure this will give me a boost in iLevel, but probably won’t really do wonders for my performance as a tank.

   How about DPS?  The higher the DPS, the better the player, right?  People seem to forget that not all characters are created equal.  Let’s take a shadow priest and a frost mage ad let’s say they are geared to LFR level (they both have an iLevel of exactly 528).  At the end of a boss fight, the meters announce the frost mage did 219K DPS, and the shadow priest did 165K DPS.  Must be the mage is some kind of hero and that the shadow priest might need to learn his class.  Thing is, this is the expected performance for both classes.  Check out the site Noxxic.  This site tells you the projected DPS for a character based on their class/specialization and gear level.  In the example I just used, each character is doing their expected amount of DPS.

   I could go on and on.   Numbers have their uses, but players have a very bad tendency to dogmatically make assumptions about numbers and take a raw number as some kind of gospel truth about a character’s performance or worth.  Playing with numbers requires just a little bit of thought.  It’s not as simple as “character x has higher DPS and therefore is the better player”.
   This seems like a good place to conclude.
   Peace.