Were the Dbacks Really Such
a Surprise?
By Michael Barnes
Many of you may not remember, but back on May 4th,
my colleague and fellow baseball observer, prognosticator, and fantasy guru Dave
Wysocky wrote an article claiming that the Dbacks were going to do much better
than their record of 16-13 indicated at that point in the season. The evidence was provided using our objective
measure of player value that clearly indicated general under-performance. It is always a source of great satisfaction
to point to successful use of statistical models to accurately predict the
future--the Dbacks not only won the division, but
have already advanced to the NLCS.
As Dave so well articulated—“One of the joys of basing
player values on actual statistical models versus gut feel (seemingly the way
most roto sites assess player values), is that you
can hit the pause button at any point in the season to understand who’s doing
what and how they rank among their peers.”
Building on the table from the original article, we can see where the Dbacks came through:
|
Player
|
Value as of 5/3
|
Predicted as of 5/3
|
% Over / Under
|
Actual 2007 Value
|
Predicted 2007 Value
|
% Over / Under
|
|
Stephen Drew
|
$0.95
|
$3.02
|
-68.5%
|
$6.66
|
$16.87
|
-60.5%
|
|
Chris Young
|
$2.01
|
$3.53
|
-43.1%
|
$19.85
|
$19.72
|
0.7%
|
|
Chad Tracy
|
$2.32
|
$3.85
|
-39.7%
|
$4.28
|
$21.52
|
-80.1%
|
|
Conor Jackson
|
-$0.56
|
$3.14
|
-117.8%
|
$11.19
|
$17.55
|
-36.2%
|
|
Orlando Hudson
|
$5.50
|
$3.03
|
81.5%
|
$15.28
|
$16.95
|
-9.9%
|
|
Miguel Montero
|
$0.26
|
$0.90
|
-71.1%
|
$2.79
|
$4.99
|
-44.1%
|
|
Eric Byrnes
|
$5.06
|
$3.57
|
41.7%
|
$34.51
|
$19.92
|
73.2%
|
|
Brandon Webb
|
$4.18
|
$5.46
|
-23.4%
|
$31.76
|
$30.51
|
4.1%
|
|
Randy Johnson
|
-$0.13
|
$3.73
|
-103.5%
|
$8.12
|
$18.28
|
-55.6%
|
|
Doug Davis
|
$2.99
|
$1.79
|
67.0%
|
$14.11
|
$9.86
|
43.1%
|
|
Jose Valverde
|
$5.76
|
$3.33
|
73.0%
|
$29.49
|
$18.61
|
58.5%
|
Predicted Values based on models from
BaseballInsights.com
As predicted, most of the under-performing players
improved: Drew, Young, Jackson, Montero, and Webb. Guys like Byrnes and Valverde
who were already over-performing, continued to do so. Add to that, the ability of our model to detect
the guys who started making valuable contributions during the year, such as
Micah Owings ($13.77), Brandon Lyon ($9.65), and Tony Clark ($6.12), and it
becomes clear that the Dbacks were playing well, and
ended the season strong, despite all the “contrary indicators.” Too many supposed pundits point to their 20
run deficit, lack of a 0.300 hitter, no hitter with more than 100 RBIs, and
other stats, and ask “How did the Dbacks do it?” Our models and player values show at least
part of the answer—many players simply over-performed.
The models do not account for timely hitting, situational
pitching, and other reasons the Dbacks had a great
season, but at least for us, their success was no surprise.
