2011_Grey_compare
 Share
The version of the browser you are using is no longer supported. Please upgrade to a supported browser.Dismiss

 
View only
 
 
Still loading...
ABCDEFGHIJKL
1
1Downloaded all 2011 pre-season hitters' projections that were available on FanGraphs, plus Grey's.
2
2Sorted to only those players that had projections from all 6 sources.
3
3Eliminated Kendrys Morales and and Ryan Kalish (0 AB) and Manny Ramirez (17AB)
4
4That left 177 hitters that were projected by all.
5
5Preferred comparison method: compare projected player 'value' to end of year actual 'value'.
6
Created a 'total player's value' by:
7
R x 1
8
+HR x 3.5
9
+RBI x 1
10
+SB x 2
11
+AVG x 290 (factor which brings it to relative value)
12
=total player value
13
prct
14
Ex.Ichiro Suzuki, pre-season:end of year value:diff
15
FANGRAPHS374=110+(8x3.5)+69+(36x2)+(.329x290)(see 525AB tab)3031.23worst
16
ROTOCHAMPS3423031.13
17
JAMES3103031.02
18
MARCEL2903030.96
19
ZIPS3033031.00best
20
GREY2993030.99
21
22
Averaged the prct diff for all players for each pundit. INITIAL results (see 'compare' tab bottom page):
23
MARCEL1.17
24
ZIPS1.32
25
ROTOCHAMPS1.34
26
JAMES1.36
27
GREY1.36
28
FANGRAPHS1.41
29
On average, Marcel was 17% over initial projections, James and Grey 36% over.
30
Not exactly sure what goes into Marcel, but believe past 3-yr average plays heavily.
31
32
6Theorized that injuries were what made Marcel's so much better. So…
33
I first eliminated those players with fewer than 500 AB. Results were quite different.
34
Stepped up to those with 525 AB or greater. These the results (see 525AB tab):
35
Comparison of projections to end of year:
36
GREY1.00
37
ROTOCHAMPS1.01
38
JAMES1.02
39
ZIPS0.97
40
FANGRAPHS1.05
41
MARCEL0.88
42
43
My evaluation is that if one is projecting as if a player will be healthy all season, Grey is tops.
44
45
7Another method for comparing can be seen on the 'compare tab' at the top. I simply give a '+' to the closest prediction and a '-' to the furthest away for each cat.
46
These could be summed for all players, but I doubt that the results would be more definitive.
47
48
8notes:
49
The 'all' tab has all of the raw data from the downloads (which includes cats that are not counted, but might be useful for further analyzing).
50
I left Grey's position identification to the far right to allow for analysis by position. Haven't done it yet, but projections might differ slightly in that regard.
51
52
9Precision: R-squared Analysis
53
54
Correlation Coefficients, R-squared, 177 playersRHRRBISBAVGAve of R-sq
55
FANGRAPHS0.230.440.230.230.230.28
56
ROTOCHAMPS0.240.410.240.670.260.36
57
JAMES0.290.450.290.630.200.37
58
MARCEL0.270.410.270.620.210.36
59
ZIPS0.210.420.210.650.300.36
60
GREY0.280.400.280.610.200.35
61
62
63
64
Correlation Coefficient, R-squared, 67 Players, 525+ ABRHRRBISBAVGAve of R-sq
65
FANGRAPHS0.230.620.430.760.280.47
66
ROTOCHAMPS0.110.580.450.740.270.43
67
JAMES0.400.620.470.740.270.50
68
MARCEL0.100.500.330.710.180.36
69
ZIPS0.320.630.480.770.360.51
70
GREY0.260.630.540.720.240.48
Loading...
 
 
 
SUMMARY
byPOS
525AB
Rate Stats
last_name
compare
all
fans
james
marcel
zips
grey