Stat targets & value calculation

theodog  3/13/2013 11:32 PM

Is there a way to calculate stat targets based on custom league settings?  

Alternatively, could I use the team-level "value" ratings in place of targets?  I play in a very shallow league (9 hitters, 6 pitchers, 12 teams).  Is the value calculation for each category based on the average given my league settings, or is it just based on the average for Major League Baseball?

mbuser   3/13/2013 11:55 PM

Every value you see is specific to the League Settings that you have entered and saved. And the concepts of the z-scores are the same across all non-points leagues, which means a team score in any single category of 0.0 is league average, +1.0 or greater is good, and +2.0 or greater is excellent, +3.0 or greater is dominant.

theodog  3/14/2013 5:49 PM

On the draft screen, and I'm guessing it's the same on the team analysis screen, it looks like the values are being summed to give team-level category values.  Does this make sense?  Aside from square root terms not being additive, after just 4 rounds of a mock draft, my team-level HR value score is a 4, which would suggest I've drafted 99.997% more homeruns the rest of my league, but I'm pretty sure I could construct another team or two that met or exceeded a total value score of 4 for HRs.

It seems like a more appropriate (but probably not perfect) calculation would be to add all of the squared z-scores (subtract if the z-score was negative) and then take the square root of the (absolute value) sum to come up with an aggregate category value (and then slap the negative term back on, if necessary).  That would sort of be like adding up all the HRs above/below the mean your team has drafted, and then converting back to a z-score to show you at the aggregate level how you compare to other hypothetical teams.

Also, are the z-scores for individual players based on my "relevant player pool size" or the entire pool?  For example, we only start 9 hitters.  So if I draft Albert Pujols, is his 2.23 z-score based on a population of the top 108 players (or let's say somewhere between 108 and 144 to account for bench spots), or the entire population of hitters in the hitter universe?

mbuser   3/14/2013 7:43 PM

The z-scores are not relative to other teams or rated versus players drafted to that point. Z-scores are calculated on a player average basis (the average of your drafted players) and scored relative to expected available values for all teams, based on calculations taken from your player pool.

So what happened was that you drafted a lot of home runs early and got well ahead of the pace needed to place well in the category. The kicker here is that most other teams had high z-scores across the board, as well, because it's the elite players coming off the board. As you draft, and the less-than-elite players influence those z-scores, they start to normalize.