Fantasy Baseball Calculator

Subtitle

Using More Appropriate Data

If you just got done reading how to determine replacement level players you have got to be asking yourself what is wrong with this ranking? Well, the problem can be pinpointed all the way back to the original z-score calculations. When calculating the z-scores for the statistical categories we used every player that had at least one plate appearance. This means that the average HR per player is getting reduced by the late season call ups and the early season injuries. These incomplete data sets will have a negative impact on the accuracy of their distribution. 


When looking at the 2012 hitters, the average number of games played for all hitters was 79.2 games. This means that we are evaluating top level talent off of streaky performances and pretty much nonsense averages dragged down by players ranked 1000+. To get more reflective data from the player pool that will actually be used, the Fantasy Baseball Calculator spreadsheet does something fairly clever.


Because all of the players have aleady been identified that will theoretically serve a purpose in the league, we can compare how much their stats will deviate from one another by stricly looking at their stats alone!


For instance, in the yahoo default settings, for all the players that have been identified to have given value the average number of home runs for each of those players is 20.43. This is a significant change over the league wide 2012 average of 7.76. The standard deviation of this data comes in at a slightly more 9.90.


Remember our Players A, B, and C?


Well just for fun we will bring them back, only this time calculate their HR z-scores with these new values in mind:

Player
 HR
z-score: mean = 20.43 ; SD = 9.90
Player A
35
1.47
Player B
20
-0.04
Player C
12
-0.85

These look a little differently don't they? Whereas Player A was 1.631 higher than Player B with every player taken in account, he is only 1.51 higher now. For comparison, I will calculate the rest of the 5x5 hitter calcs with the new means and standard deviations:

Player 
 R
z-score: mean = 77.83 ; SD = 16.18
Player A
100
1.37
Player B
105
1.68
Player C
130
3.22
Player 
RBI 
z-score: mean = 75.93 ; SD = 21.23
Player A
100
1.13
Player B
89
0.62
Player C
75
-0.04
Player
SB 
z-score: mean = 13.68 ; SD = 12.16
Player A
10
-0.30
Player B
5
-0.71
Player C
65
4.22
Player
 xBA ; sum hit = 18621 sum ab = 66093
z-score: mean = -5.89* ; SD = 9.94
Player A
1.78
0.77
Player B
4.47
1.04
Player C
15.78
2.18
*Note = you will see that the mean xBA does not equal zero in this case - this is because this is the average of all the xBA for all hitters with the filtered hit and ab sums of those qualifying. This makes sense that the average would be negative, because in this setup there are more players that do not have value than those that do have value, and the players that have value have better stats in comparison. All values are compared against the new hit and ab sums, thus the xBA are different. The xBA for all qualifying players is still zero (rate stats can really make things confusing sometimes)
Player
 z HR
 z R
 z RBI
 z SB
 z xBA
z total
Player A
1.47
1.37
1.13
-0.30
0.77
4.44
Player B
-0.04
1.68
0.62
-0.71
1.04
2.59
Player C
-0.85
3.22
-0.04
4.22
2.18
8.73
From this, we can see that through the use of appropriate, relevant data, we have gone from Player C being valued at 1.58 x Player A ; to Player C being valued at 1.97 x Player A. 

All of the players will have a z-score measured with the averages and standard deviations of the qualifying players. The Fantasy Baseball Calculator spreadsheet will then do the sorting and replacement player analyzing with these new values all over again to determine the final rankings and values for the specified criteria.

And yes, these are the values that will be used in the final rankings! That doesn't mean there isn't more we can do though. We talked about earlier that the question we should be asking is 'How much MORE a stat is worth than another'. Well know that we have put a value on this, as well as put a value on all the selected scoring categories, we can assign an easy to compare Dollar Value to each player so we can now directly see how much more of your leagues budget is one player worth in comparison to another.

To see how the dollar values and auction prices are determined, Click Here