Batting Stats 9: wRAA

Weighted Runs Above Average (wRAA) takes Weighted Runs Created (wRC) and puts it in the context of comparing it to the average player.  Below, we see one of our test subjects, Jorge Acosta.

In 2015, Acosta accumulated 56.9 wRC.  Is that good?  As with all counting stats, it depends.  Like other counting stats, the number of plate appearances matters.  (I know I have AB shown here, but bear with me.) A player who racks up 50 RBIs over a full season isn’t terribly impressive.  One who does it in 200 plate appearances is amazing.  Same with wRC.  Fangraphs gives us some general guidelines on what’s good, great, average, etc. over the course of a full season.  Still, that’s a number that is devoid of any context.  How does that stack up against other players in the league?

That’s where wRAA comes in.  League average wRAA is always 0.  So, looking at Acosta with his -7.3 wRAA, we see that his 56.9 wRC is not impressive at all: 7% below league average.  In 2015, Acosta played like a scrub.

We derive wRAA thusly:
(wOBA-lg_woba)/wOBAscale)*PA

It does, and should, look very similar to wRC.

My testing paramaters are based on the idea that there’s generally around 10 points separating the middle categories of poor to above average.  In order that my stats stay close to the game’s, I started off with a threshold of 3 points.  All of my results were within that range, so I tightened up a bit.  I changed my “good” zone to 1.0.  Here are my results:

Again, I’m really happy with how this turned out.  I don’t think I need to do any tweaking or further investigation.  Ready to move on!

Our script with the addition of wRAA:

#Calculated batting stats for OOTP
    DROP TABLE IF EXISTS CalcBatting;
    CREATE TABLE IF NOT EXISTS CalcBatting AS

    SELECT b.year
    , b.league_id
    , b.player_id
    , b.stint #We can eventually move this down the list
    , b.split_id #We can eventually remove
    , b.team_id #We can eventually move this down the list
    , l.abbr as Lg
    , t.abbr as Team
    , b.g
    , b.ab
    , @PA := b.ab+b.bb+b.sh+b.sf+b.hp AS PA
    , b.r 
    , b.h
    , b.d
    , b.t
    , b.hr
    , b.rbi
    , b.sb
    , b.cs
    , b.bb
    , b.k
    , b.ibb
    , b.hp
    , b.sh
    , b.sf
    , b.gdp
    , b.ci
    , @BA := round(b.h/b.ab,3) AS ba
    , round(b.k/@PA,3) as krate
    , round((b.bb)/@PA,3) as bbrate
    , @OBP := round((b.h + b.bb + b.hp)/(@PA-b.sh-b.ci),3) AS obp
    , round(100*(@OBP/r.woba),0) as OBPplus
    , @SLG := round((b.h+b.d+2*b.t+3*b.hr)/b.ab,3) as slg
    , round(@OBP+@SLG,3) as ops
    , round(@SLG-@BA,3) as iso
    , round((b.h-b.hr)/(b.ab-b.k-b.hr+b.sf),3) as babip
    , @woba := round((r.wobaBB*(b.bb-b.ibb) + r.wobaHB*b.hp + r.woba1B*(b.h-b.d-b.t-b.hr) +
       r.woba2B*b.d + r.woba3B*b.t + r.wobaHR*b.hr)
       /(b.ab+b.bb-b.ibb+b.sf+b.hp),3) as woba
    , round(((@woba-r.woba)/r.wOBAscale)*@PA,1) as wRAA
    , round((((@woba-r.woba)/r.wOBAscale)+(lro.totr/lro.totpa))*@PA,1) as wRC
    /* NOT yet modified for OOTP and MySQL
    , ((([wRAA]/[PA] + l.RperPA) + (l.RperPA - t.bpf*l.RperPA))/(lb.wRC/lb.PA))*100 AS [wRC+]
    */
    FROM 
      players_career_batting_stats b 
      INNER JOIN leagues l ON b.league_id=l.league_id 
      INNER JOIN teams t ON b.team_id=t.team_id
      INNER JOIN tblRunValues2 r ON b.year=r.year AND b.league_id=r.league_id
      INNER JOIN vLeagueRunsPerOut lro ON b.year=lro.year AND b.league_id=lro.league_id
    WHERE b.ab<>0 AND b.split_id=1
    ORDER BY b.player_id, b.year