“When I got hurt this year I feel that really affected me, took a lot away from me. I wasn’t able to steal bases. Every time I hit the ball in the hole I wasn’t able to run. I just couldn’t run. It’s kind of frustrating because you feel like you would have done a better job. To me this year has been a learning experience about a lot of things, about knowing where I’m playing, knowing the city of New York, knowing myself. I feel good with it.”
—Carlos Beltran, on his 2005 season
So we’ve finally reached a turning point in our series on quantifying baserunning. Since mid-July we’ve developed a methodology and framework for crediting baserunners for advancing on ground outs (Equivalent Ground Advancement Runs, or EqGAR), advancing on outs in the air (Equivalent Air Advancement Runs, or EqAAR), and attempted stolen bases as well as pick offs (Equivalent Stolen Base Runs, or EqSBR). This week we’ll look at total picture and evaluate which players got the most and least from their legs over the past six years.
Adding to the Toolbox
To get those readers up to speed who may not have seen this framework before, the final piece of the puzzle is crediting runners with advancing on hits. To that end we can use the same basic methodology as we did when creating the other metrics by relying on the Run Expectancy matrix. Simply put, we’ll credit or debit runners for changes in Run Expectancy in the following scenarios:
- Runner on first, second not occupied, and the batter singles
- Runner on second, third not occupied, and the batter singles
- Runner on first, second not occupied, and the batter doubles
In each scenario we create a table that shows how often runners typically advance to each subsequent base or get thrown out; this will be broken down by the number of outs, handedness of the batter, and the position of the fielder who fielded the ball.
As we saw with advancing on ground and air outs, the probabilities of advancing and the number of bases a runner can advance change dramatically as the number of outs and the position of the fielder change. For example, when a batter singles with nobody out and a runner on second, the runner reaches third around 55% of the time and scores 45% of the time. With two outs, however, those probabilities change to 6% and 89%, respectively, with an increasing percentage of the runners getting thrown out (less than 1% up to 5%).
Totaling the credit assigned to each opportunity (and not crediting the runner for advancing the minimum number of bases) for players and teams allows us to assign a number of theoretical runs above and beyond what a typical player or team would have contributed given the same opportunities. Finally, as we did when looking at advancing on outs in the air, we apply a park factor in order to take into account the fact that at some parks it is easier or more difficult to advance given the dimensions or configuration (wall height for example).
Some readers will recall that this is essentially the framework I published last fall in an essay in The Hardball Times Baseball Annual 2006 with the metric called Incremental Runs (IR), and its associated rate statistic Incremental Run Percentage (IRP). Since that time, I’ve made a couple small refinements to the framework and have recalculated the park factor to equally weight all of the previous six years worth of data in order to smooth it out a bit. Along with those changes, we’ll also take this opportunity to rechristen IR, Equivalent Hit Advancement Runs (EqHAR) so that it fits nicely into our toolbox.
It should be mentioned that EqHAR is essentially the same in both methodology and in what it attempts to quantify as James Click’s metric discussed in the 2005 Baseball Prospectus. Both systems were being developed at the same time.
So then let’s take a look at the leaders and trailers in EqHAR for 2005 as well as the individual leader and trailer for each of the six seasons in our study. The table below lists the number of opportunities, the number of times the runner was thrown out advancing, the number of runs we credited them with, and a rate statistic that removes the bias associated with a greater number of opportunities.
2005 Leaders and Trailers in EqHAR Name Opp OA EqHAR Rate Carlos Beltran 45 1 3.53 1.66 Robinson Cano 54 1 3.36 1.49 Edgar Renteria 49 0 3.21 1.37 Grady Sizemore 59 0 3.14 1.31 Darin Erstad 59 0 2.99 1.38 David DeJesus 50 0 2.99 1.47 Scott Podsednik 45 0 2.98 1.43 Rafael Furcal 52 0 2.94 1.38 Julio Lugo 58 0 2.85 1.40 David Wright 41 0 2.82 1.39 ------------------------------------------------ Pat Burrell 34 4 -5.60 0.17 Luis Gonzalez 44 4 -5.33 -0.05 Lance Berkman 37 3 -4.56 0.00 David Ortiz 41 1 -4.12 0.25 Bengie Molina 34 0 -4.06 0.22 Mark Loretta 46 4 -4.02 0.58 Matt Lawton 49 3 -3.70 0.40 Kevin Millar 47 1 -3.59 0.34 Aramis Ramirez 45 2 -2.82 0.57 Rickie Weeks 30 5 -2.81 0.46
Yearly Leaders and Trailers in EqHAR Year Name Opp OA EqHAR Rate 2000 Luis Castillo 57 0 4.78 1.52 2001 Juan Pierre 41 0 4.04 1.81 2002 David Eckstein 60 0 3.66 1.52 2003 Raul Ibanez 55 0 4.75 1.54 2004 Vernon Wells 34 0 4.90 1.73 2005 Carlos Beltran 45 1 3.53 1.66 -------------------------------------------------------- 2000 Joe Randa 50 2 -3.91 0.42 2001 Adrian Beltre 25 4 -4.59 -0.14 2002 Frank Thomas 40 4 -5.17 0.08 2003 Jim Thome 51 3 -5.15 0.49 2004 Bill Mueller 47 3 -5.17 0.29 2005 Pat Burrell 34 4 -5.60 0.17
From these lists it’s apparent that at the extremes EqHAR falls roughly in the +5 to -5 range, or the equivalent of about one win, putting it in the same category as EqSBR in terms of magnitude, as shown below.
Magnitude of Baserunning Metrics / Single Player, Single YearMetric Min Max EqGAR -1.50 4.00 EqAAR -3.00 2.00 EqSBR -6.00 5.00 EqHAR -5.00 5.00
Interestingly, EqSBR actually has a slightly larger range on the negative side since there are more potential opportunities for stolen base attempts coupled with the fact that some players don’t know when to quit (see Guerrero, Vladimir). Both EqSBR and EqHAR, however, have larger ranges than EqGAR and EqAAR. This reflects two primary factors. First, there are a greater number of opportunities available to runners in terms of stolen base attempts and advancing on hits. The leaders in EqHAR and EqSBR typically had more than 50 opportunities in the 2000-2005 time period while those for EqGAR had slightly fewer; for EqAAR it’s in the 30s.
Second, the success rates for EqHAR and EqSBR are such that good (or bad) baserunners have a bit more room to distance themselves from the pack. For example, with EqAAR almost all runners score from third on fly balls and so the difference between a good runner and one that is merely average is compressed to less than 5%. On the other hand, with EqHAR a good runner may take 16 to 20% more bases than an average one in a given scenario (for example when advancing from first to third on a single).
But as we mentioned earlier in this series, those seasonal ranges don’t mean that the best players and worst baserunners over the six year span will be credited with -30 to +30 runs. As with any metric, part of the derived value simply reflects random variation and so that variation–combined with the fact that the same players don’t necessarily end up at the top and bottom each season–means that over the entire period that span between the leaders and trailers is on the order of 25 to 30 runs, or three wins:
Leaders and Trailers in EqHAR for 2000-2005 Name Opp OA EqHAR Rate Juan Pierre 315 4 15.22 1.41 Luis Castillo 331 6 14.64 1.32 Rafael Furcal 272 3 12.09 1.33 Ray Durham 249 0 11.87 1.37 Mike Cameron 188 2 11.59 1.37 Darin Erstad 288 6 11.43 1.23 Jay Payton 201 3 11.35 1.36 Carlos Beltran 253 1 11.23 1.27 David Eckstein 280 4 10.98 1.30 Cristian Guzman 226 3 10.36 1.37 ------------------------------------------------ Edgar Martinez 178 3 -12.50 0.58 Rafael Palmeiro 231 9 -11.50 0.63 Dmitri Young 181 10 -11.04 0.60 Richie Sexson 153 6 -10.83 0.53 Juan Encarnacion 201 12 -10.73 0.66 Carlos Delgado 237 8 -10.71 0.73 Rich Aurilia 185 9 -10.48 0.62 David Ortiz 172 5 -10.36 0.63 Bill Mueller 168 9 -9.91 0.58 Luis Gonzalez 266 10 -9.58 0.77
Finally, and as mentioned previously, EqHAR also has a rate stat (calculated as the ratio of actual runs to expected runs); if we want to see which runners performed the best regardless of the quantity of opportunities (and remember we’ve already controlled for the quality of those opportunities by comparing what they did against the league average for each situation they found themselves in and then park adjusting the results) we can rank them according to rate:
Leaders and Trailers in EqHAR rate for 2000-2005 (100 or more opportunities) Name Opp OA EqHAR Rate Timo Perez 100 1 4.86 1.42 Juan Pierre 315 4 15.22 1.41 Raul Mondesi 126 1 6.25 1.38 Scott Podsednik 132 1 7.03 1.37 Mike Cameron 188 2 11.59 1.37 Cristian Guzman 226 3 10.36 1.37 Ray Durham 249 0 11.87 1.37 Jay Payton 201 3 11.35 1.36 Miguel Cairo 125 1 5.33 1.36 Felipe Lopez 110 1 5.35 1.34 ------------------------------------------------ Richie Sexson 153 6 -10.83 0.53 Kevin Millar 163 3 -9.38 0.56 Edgar Martinez 178 3 -12.50 0.58 Bill Mueller 168 9 -9.91 0.58 Dmitri Young 181 10 -11.04 0.60 Bengie Molina 140 1 -9.44 0.61 John Olerud 130 5 -6.86 0.62 Javy Lopez 104 4 -5.56 0.62 Rich Aurilia 185 9 -10.48 0.62 Rafael Palmeiro 231 9 -11.50 0.63
Of active players with fewer than 100 opportunities Robinson Cano (1.49), David DeJesus (1.46), and Ryan Freel (1.46) also all come out very well.
This list once again highlights the situation where a player like Cano does well in one metric but poorly in others. In Cano’s case, his EqHAR was among the leaders in 2005 at 3.36, while his EqGAR, EqAAR, and EqSBR values were at -1.13, -0.93, and -1.5 respectively, putting him on the negative side (-0.21) when you add it all up. This may reflect the fact that different skills are required to do well in the different metrics (for example, judgment may be more important in EqGAR and EqSBR than in EqHAR where sheer speed is what counts most), or simply that random variation and small sample size is at work–after all, Cano had just 4 stolen base attempts in 2005. I would bet on a mix of the two, although it will be interesting to see how Cano stacks up this season.
And because I know I’ll be asked, Bobby Abreu and Juan Encarnacion lead all players in getting thrown out on the bases in these scenarios at 12, with Lance Berkman and Matt Lawton close behind at 11.
Contributing with their Legs
Finally, we can now provide a more complete picture of baserunning. Today we’ll focus on individuals, and next week we’ll take it to the team level. So first, here are the leaders and trailers for 2005:
2005 Leaders and Trailers in Total Baserunning Name Opp EqGAR Opp EqAAR Opp EqSBR Opp EqHAR Total Chone Figgins 53 4.52 39 1.79 80 -0.30 58 2.27 8.29 Jose Reyes 52 1.76 33 1.34 77 1.73 50 1.98 6.81 Juan Pierre 54 3.52 30 -0.07 75 0.82 56 2.25 6.52 Alfonso Soriano 31 -0.08 45 -0.07 32 4.92 42 1.10 5.86 Jason Bay 20 0.82 29 1.02 22 2.39 54 1.39 5.63 Marcus Giles 27 -0.59 41 1.75 18 2.04 42 2.29 5.49 Johnny Damon 42 0.66 54 1.66 19 2.84 66 0.04 5.20 Carlos Beltran 19 1.16 22 0.24 22 0.14 45 3.53 5.07 Rafael Furcal 41 0.32 38 -0.71 57 2.36 52 2.94 4.90 Ichiro Suzuki 47 0.55 38 1.49 42 1.23 63 1.34 4.61 ---------------------------------------------------------------------------------------- Pat Burrell 19 -0.48 25 -0.69 1 -0.46 34 -5.60 -7.23 Brad Wilkerson 48 -0.80 26 0.09 19 -6.23 43 0.31 -6.64 Matt Lawton 35 -0.12 38 -0.30 28 -1.91 49 -3.70 -6.03 David Ortiz 22 -1.16 34 -0.73 1 0.09 41 -4.12 -5.92 Mark Loretta 20 -0.33 29 0.02 12 -1.29 46 -4.02 -5.62 Bengie Molina 14 0.51 16 -0.80 2 -1.07 34 -4.06 -5.41 Oscar Robles 20 -0.59 19 0.46 8 -4.26 26 -0.23 -4.63 Luis Gonzalez 16 -0.47 32 1.44 4 -0.03 44 -5.33 -4.39 Jeromy Burnitz 17 -0.28 35 0.41 12 -4.02 61 -0.40 -4.29 Carlos Lee 18 -0.78 21 -1.27 18 0.61 29 -2.64 -4.07
Again, you can see that although some players do well in all categories (Jose Reyes and Jason Bay, for example) there are others who excel in just one, like Alfonso Soriano. You’ll also notice that in total the player who comes out on top contributes about 7 runs and those who do poorly cost their teams about 7 runs; historically, as shown in the next table, the leaders and trailers have a span of more like -8 to +8.
Yearly Leaders and Trailers in Total Baserunning Year Name Opp EqGAR Opp EqAAR Opp EqSBR Opp EqHAR Total 2000 Tom Goodwin 41 0.94 34 1.38 66 4.66 40 3.45 10.43 2001 David Eckstein 44 1.17 41 0.35 33 2.68 51 3.75 7.95 2002 Derek Jeter 34 -0.61 39 1.55 35 4.20 56 2.70 7.84 2003 Scott Podsednik 38 1.66 29 1.21 54 3.14 44 3.16 9.16 2004 Tony Womack 46 1.59 33 0.68 33 2.12 55 2.55 6.94 2005 Chone Figgins 53 4.52 39 1.79 80 -0.30 58 2.27 8.29 ------------------------------------------------------------------------------------------ 2000 Vladimir Guerrero 27 -0.07 40 -1.00 22 -4.87 30 -1.77 -7.72 2001 Doug Mientkiewicz 31 -0.75 25 0.62 9 -3.23 43 -3.50 -6.86 2002 Deivi Cruz 27 -0.97 18 -2.08 4 -1.56 31 -0.75 -5.35 2003 D'Angelo Jimenez 30 -1.11 24 -3.51 19 -1.97 52 -1.04 -7.62 2004 Jim Thome 19 -0.41 28 -3.04 2 -1.41 60 -3.38 -8.24 2005 Pat Burrell 19 -0.48 25 -0.69 1 -0.46 34 -5.60 -7.23
To wrap up, we want to provide a first order answer to the question of just who may be the best and worst total baserunners in baseball since 2000. And so without further ado, the following table lists the top and bottom 10 in total runs from the combination of all four metrics:
2000-2005 Leaders and Trailers in Total Baserunning Name Opp EqGAR Opp EqAAR Opp EqSBR Opp EqHAR Total Carlos Beltran 126 0.66 169 1.80 204 11.75 253 11.23 25.44 Derek Jeter 218 -0.51 252 6.93 149 10.59 317 8.09 25.09 Johnny Damon 258 1.78 265 3.15 217 9.03 322 9.40 23.36 Rafael Furcal 224 3.17 192 2.58 246 2.76 272 12.09 20.60 Tom Goodwin 90 2.59 83 3.74 147 5.15 97 7.43 18.91 Juan Pierre 279 9.37 204 2.91 378 -8.85 315 15.22 18.64 Tony Womack 201 4.26 178 3.18 219 2.68 224 7.65 17.77 Jimmy Rollins 212 3.49 180 3.28 223 1.69 243 8.62 17.08 Scott Podsednik 126 2.43 103 1.95 227 4.52 132 7.03 15.94 Darin Erstad 176 1.04 195 -2.23 136 4.10 288 11.43 14.35 ---------------------------------------------------------------------------------------- Jorge Posada 158 -2.67 149 -1.46 26 -8.08 199 -8.18 -20.39 Jim Thome 139 -3.12 129 -4.07 10 -3.74 222 -8.22 -19.16 Carlos Delgado 141 -1.06 177 -3.55 8 -3.65 237 -10.71 -18.95 Richie Sexson 104 -1.60 100 -1.73 14 -3.78 153 -10.83 -17.94 Paul Lo Duca 154 -4.16 133 -2.91 30 -6.09 190 -4.07 -17.23 Luis Gonzalez 133 -3.41 170 0.02 36 -4.00 266 -9.58 -16.97 Edgar Martinez 113 -2.97 142 -0.56 12 -0.55 178 -12.50 -16.58 Dmitri Young 132 -3.14 119 2.62 22 -4.95 181 -11.04 -16.51 Matt Lawton 185 1.65 198 -3.48 164 -9.32 258 -5.13 -16.28 Rafael Palmeiro 129 -3.92 159 0.22 15 -0.90 231 -11.50 -16.10 Juan Encarnacion 145 -1.61 145 1.60 113 -5.16 201 -10.73 -15.90
Carlos Beltran came out on top at +25.44 by virtue of his ability to steal bases at a high percentage and advance on hits. Meanwhile, Jorge Posada had trouble in all departments and ended up at -20.39. You’ll notice that had Juan Pierre at least come out neutral in EqSBR he would have taken the top spot. But still, the difference between the top and bottom is on the order of four to five wins, a difference far smaller than that between the best and worst offensive players and best and worst fielders. The upshot is that while good baserunning pays off and is repeatable (more on that next week) to a certain extent, its primary uses from an offensive perspective are probably more strategic than general–even the best baserunners seemingly cannot contribute a significant number of wins in the long haul. Not a surprising conclusion, to be sure, but one that is often forgotten.
Interestingly, although Beltran’s injury last season affected his stolen bases and despite his comments to the contrary, it didn’t appear to have quite the same effect on other parts of his running game as his totals from each season show:
Year Opp EqGAR Opp EqAAR Opp EqSBR Opp EqHAR Total 2000 7 -0.11 12 0.10 14 1.29 33 0.30 1.57 2001 29 0.20 31 -0.09 34 3.20 43 0.75 4.06 2002 18 0.38 26 -0.88 45 -0.33 39 2.47 1.64 2003 28 -0.80 31 1.80 44 3.69 42 2.40 7.08 2004 25 -0.17 47 0.63 45 3.76 51 1.79 6.01 2005 19 1.16 22 0.24 22 0.14 45 3.53 5.07
Alas, by adding up the metrics as we’ve done here, those on this list are heavily influenced by the number of their opportunities. To correct for this, next week we’ll develop a rate statistic that encompasses all four metrics, and we’ll discuss the relative importance of each, and perhaps even delve into aging patterns, how teams stack up, and any other interesting avenue we happen to wander down.
Thank you for reading
This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly proprietary environment.
Subscribe now