2004 Baserunning Framework

Using the same kind of approach as that taken by David Pinto when he calculates his Probabalistic Model of Range (PMR), I created a baserunning framework (Probabalistic Model of Baserunning?) last year that I explained here. Basically, this methodology uses standard advancement tables to calculate how many extra bases (which I called Incremental Bases or IB) a baserunner gained in the following scenarios:

1) Runner on first batter singles
2) Runner on first batter doubles
3) Runner on second batter singles

The advancement tables then take into account the number of outs, the handedness of the batter, and which fielder fielded the ball. I also calculate the ratio of actual bases to Expected Bases (EB) to create an Incremental Base Percentage (IBP).

In any case, thanks to SABR's Mike Emeigh I have now obtained the 2004 play-by-play data and rerun my framework. Drum roll please....the leaders for 2004 (20 or more opportunities are):

                  Opp Bases     EB     IB   IBP OA
Jorge Cantu        21   36   29.70   6.30  1.21  0
Rafael Furcal      73  127  106.15  20.85  1.20  0
Matt Holliday      50   95   79.69  15.31  1.19  0
Jack Wilson        43   76   64.00  12.00  1.19  1
Alfonso Soriano    36   64   54.07   9.93  1.18  0
Chase Utley        22   38   32.16   5.84  1.18  0
Vernon Wells       47   83   70.56  12.44  1.18  0
Jose Valentin      43   70   59.74  10.23  1.17  0
Ryan Freel         49   83   71.30  11.70  1.16  0
Tony Womack        61  103   88.88  14.13  1.16  1
Torii Hunter       43   74   63.86  10.14  1.16  1
Gabe Kapler        38   63   55.03   7.97  1.14  0
David DeJesus      46   75   66.12   8.88  1.13  0
Laynce Nix         42   75   66.19   8.81  1.13  1
Rocco Baldelli     44   74   65.36   8.64  1.13  0
Reed Johnson       59  101   89.34  11.66  1.13  0
Johnny Damon       70  125  110.71  14.29  1.13  0
Joe Crede          38   66   58.47   7.53  1.13  0
Carlos Beltran    134  226  200.74  25.26  1.13  0
Willie Bloomquist  31   47   41.79   5.21  1.12  0
Sean Burroughs     67  109   96.92  12.08  1.12  1
Mike Cameron       41   68   60.53   7.47  1.12  0
Cesar Izturis      68  113  100.72  12.28  1.12  0
Juan Pierre        83  131  116.87  14.13  1.12  1
Robert Fick        30   50   44.63   5.37  1.12  0

Just as in 2003 the top performers came out about 21% above average in terms of gaining extra bases. And as you'd imagine most of the top performers seem to be guys you would think might be decent baserunners (especially those with over 50 opportunities). On the bottom of the list of 333 qualifer were:

                  Opp Bases     EB     IB   IBP OA
Randall Simon      34   32   51.61 -19.61  0.62  4
Ken Griffey Jr.    24   30   39.84  -9.84  0.75  3
Ross Gload         32   37   48.83 -11.83  0.76  5
Ben Molina         31   35   46.08 -11.08  0.76  1
Mike Piazza        45   49   63.36 -14.36  0.77  2
Bill Mueller       64   79  102.01 -23.01  0.77  3
Jason LaRue        44   54   68.30 -14.30  0.79  2
Kevin Youkilis     30   37   46.72  -9.72  0.79  2
Chad Moeller       20   24   29.88  -5.88  0.80  2
Gary Bennett       25   32   39.12  -7.12  0.82  1
Jacob Cruz         24   33   40.15  -7.15  0.82  2
Karim Garcia       44   56   68.02 -12.02  0.82  2
Ben Davis          40   46   55.63  -9.63  0.83  2
A.J. Pierzynski    49   59   71.25 -12.25  0.83  3
Doug Mirabelli     22   27   32.44  -5.44  0.83  0

Once again, nothing too surprising. The range of the best to worst baserunners using this measure is on the order of 30 to 40 bases.

When I ran the 2003 results against 1992 data I did a quick study of the players who were in both data sets to see if there was any correlation and to see if as players age their IBP would decrease as expected. Indeed I found that 9 of the 13 players had higher IBPs in 1992 than in 2003 and that their cumulative IBP was just a tad higher in 1992.

Now having the 2004 data gives me an opportunity to do a side by side comparison. In all there were 261 players with 20 or more opportunities that played in both seasons. Their cumulative IBP for 2003 was 1.00205 whereas for 2004 it was 1.00099. This equates to 20 more bases gained in 2003 than in 2004. While not a large difference it is in the right direction on the assumption that baserunning skills decline with age as a player slows down.

I also ran a regression on the data and calculated a correlation coefficient of .298 for the two years. When I increased the threshold to 50 opportunities in both season the correlation rose to .320. Not a particularly stong correlation but a positive one that indicates there is some predictive power here.

Up next I plan on creating a cumulative set of advancement tables for 2003-2004 and then recalculate the leaders for each season as well as the leaders across both seasons.

From a team perspective it broke down like this:

    Opp Bases     EB     IB   IBP  OA
COL 616 1008  958.07  49.93  1.05  16
SLN 626  982  948.87  33.13  1.03  11
MON 584  884  857.88  26.12  1.03  12
LAN 651  981  952.85  28.15  1.03  10
TEX 576  899  873.32  25.68  1.03   9
CHA 607  925  902.94  22.06  1.02  14
MIN 587  914  892.81  21.19  1.02  10
KCA 646  976  959.17  16.83  1.02  11
ARI 620  942  927.44  14.56  1.02   7
DET 576  874  865.26   8.74  1.01  13
SDN 717 1069 1061.14   7.86  1.01  13
CLE 654 1025 1021.44   3.56  1.00  15
FLO 636  961  957.72   3.28  1.00  18
ATL 597  903  901.92   1.08  1.00   9
NYA 631  952  954.75  -2.75  1.00  13
CHN 582  874  878.57  -4.57  0.99  15
TOR 663  986  991.94  -5.94  0.99  12
ANA 668  990  996.91  -6.91  0.99  19
BAL 708 1050 1062.57 -12.57  0.99  20
TBA 554  815  827.36 -12.36  0.99  10
SFN 698 1032 1051.49 -19.49  0.98  14
HOU 670  994 1013.12 -19.12  0.98  18
SEA 740 1073 1095.85 -22.85  0.98  12
PHI 610  919  939.74 -20.74  0.98  19
MIL 533  793  814.62 -21.62  0.97  10
OAK 628  925  953.57 -28.57  0.97  13
NYN 579  863  893.98 -30.98  0.97  15
PIT 605  868  905.84 -37.84  0.96  17
CIN 549  809  847.42 -38.42  0.95  16
BOS 778 1166 1226.45 -60.45  0.95  19

Once again the range here is around 75 to 100 bases. It's interesting that Colorado led in both 2003 and 2004 with IBPs of 1.04 and 1.05. This immediately cries out for a ballpark explanation. Off the top of my head the larger outfield and fast surface probably contribute to the ability of baserunners to take extra bases.

A next step here would be to run these numbers for the various parks.