Monday, June 09, 2008

Stealer's Wheels (Offensive Edition)

In Dayn Perry's exploratory book "Winners: How Good Baseball Teams Become Great Ones", Perry attempts to quantify and explain what elements playoff teams in the modern era all shared in common. The chapters are divided into areas such as "the Slugger", "the Ace", "the Middle Reliever" and so on. In chapter six Perry tackles the value of having a base stealer on the roster. Baseball researchers dating back to George Lindsey to Pete Palmer to John Thorn have applied the run value of the stolen base and also the runs lost when a would-be stealer is thrown out. At Tangotiger.net, the researchers there have updated the formula and released their findings that a stolen base results in 0.185 runs. Conversely, being caught stealing costs the team 0.46 runs as the out recorded is much worse than the added base. In order to reap dividends, a player or team needs to have a success rate of greater than 72% in order to have a gain in runs, anything less than 72% is detrimental to a team. Likewise, a catcher (and the pitching staff as well) needs to be successful 29% of the time in throwing runners out to avoid having opponents score additional runs.

Based on that model, Perry identified the top ten all-time teams based on their net runs via stolen bases. The 1985 St Louis Cardinals, who had Vince Coleman (110), Willie McGee (56), Andy Van Slyke (34), Ozzie Smith (31) and Tommy Herr (31) had what is regarded as one of the speediest teams on the basepaths. Thanks to the quick astroturf at the old Busch Stadium, the Cardinals were able to swipe 314 bases and being caught just 96 times leading to 17.8 net runs. The 1985 Cardinals finished 101-61, just three games ahead of the New York Mets. Those 17.8 runs could have meant the difference of three games to the Cardinals resulting in clinching the NL East championship title. Oddly enough, the Cardinals who had used their speed to project themselves into the playoffs, only attempted to steal just five times in the seven game World Series against cross-state rival Kansas City. Royals catcher Jim Sundberg threw out two Cardinals as well. (During the regular season, Sandberg had thrown out 29.2% of potential base runners - a significant deduction from his 1984 season with the Texas Rangers when he threw out 50% of would-be stealers). The 2002 Minnesota Twins, on the other hand, were the second worst base-stealers ever. The team that led the organization back into the playoffs for the first time since 1991 were not particularly adept at pilfering the extra base. Torii Hunter led the team with 23 steals (8 times caught) while Cristian Guzman, Corey Koskie and Jacque Jones were all thrown out more times than they were successful. In total, the team stole 79 bases and were caught 62 times costing them 14.7 runs. Since the Twins won the Central by 13 games, the net loss was moot.

This is all relevant to the 2008 Twins because one of the team's biggest attributes, speed, has also been to this point, a hindrance. In 2007 during his short tenure with the New York Mets Carlos Gomez stole 12 bases in 15 attempts (an 80% success rate). This 90-foot gamble netted Gomez and the Mets an additional 0.84 runs. In the offseason when the Twins targeted the speedy outfielder there were optimistic estimates that Gomez could steal 50-75 bases as soon as this year - a threshold that hasn't been crossed since former Twins Chuck Knoblauch stole 62 bases in 1997. Those 62 bases were a Twins record (fourth best in the franchise dating back to Washington). Knoblauch, who would go on to steal 407 career bases (a 77% success rate), gave the Twins an additional 6.87 runs that season. To be sure, stealing 62 bases added just 7 runs. Through his first 57 games in 1997, Knoblauch had obtained 27 bases by means of thievery while being erased just twice (93% success rate). Gomez hasn't been nearly as fortunate. In his first 57 games over a decade later, Gomez has swiped just 17 bases and has been thrown out seven times (70% success rate). While Gomez still have over 100 games to steal 45 more bases, the rate of which he is being caught is alarming. Over the past 15 games, Gomez has been successful once in his five attempts. At times it doesn't appear that Gomez is analyzing his opportunity at all. Nevertheless, his 17 total bases stolen places him in the top ten of base stealers:

Player Team SB CS SB%
Ellsbury BOS 28 3 90.3%
Suzuki SEA 26 2 92.9%
Upton TB 20 6 76.9%
Roberts BAL 18 5 78.3%
Gomez MIN 17 7 70.8%
Kinsler TEX 17 0 100%
Crawford TB 16 4 80.0%
Gathright KC 16 2 88.9%
Sizemore CLE 15 2 88.2%
Rios TOR 14 4 77.8%

However, if we were to rank the players on the runs which the stolen bases/caught stealing led to we see a different picture all together:

Player Team SB CS SB% Runs +/-
Suzuki SEA 26 2 92.9% 3.9
Ellsbury BOS 28 3 90.3% 3.8
Kinsler TEX 17 0 100% 3.1
Gathright KC 16 2 88.9% 2.1
Sizemore CLE 15 2 88.2% 1.9
Crawford TB 16 4 80.0% 1.2
Roberts BAL 18 5 78.3% 1.0
Upton TB 20 6 76.9% 0.9
Rios TOR 14 4 88.2% 0.8
Gomez MIN 17 7 70.8% -0.1
Minnesota Twins' Carlos Gomez, left, is tagged out at second by Kansas City Royals shortstop Mike Aviles while attempting to steal a base in the first inning of a baseball game Thursday, May 29, 2008, in Kansas City, Mo.

Gomez, thanks to his league leading seven caught stealing, has actually costing the Twins (even the same fraction that it is) runs. When we look further into the numbers, is evident that the Dome's surface aids Gomez. Part of Gomez's problem is his inability to accept that he is not a good of a base stealer on grass as he is on turf. While playing at the Dome, Gomez had stole 11 bases in 13 attempts, a 84% success rate. On the road, Gomez has been successful only five times out of ten attempts, a 50% success rate. This is a stark contrast to the remainder of the team. At home the Twins have stolen 25 bases and have been caught 10 times (71% success rate). On the road, the Twins have 20 stolen bases and been caught stealing eight times (71% success rate). As a whole, the Twins are currently ranked fourth among the American League franchises with 45 stolen bases:

Team SB CS SB%
TB 69 23 75%
BOS 58 12 82%
LAA 49 17 74%
MIN 45 18 71%
SEA 44 11 80%
BAL 43 20 68%
TOR 43 21 67%
KC 36 20 64%
CLE 33 10 76%
TEX 33 8 80%

Like the analysis above, the application of the stolen base formula shows a different picture. The efficient and sabermetrically influenced Boston Red Sox have used their speed at the appropriate times resulting in a 82% success rate. This basepath prowess has provided the Red Sox with 5.2 additional runs (a drop in the bucket considering their 332 runs scored). Seattle, which is having a hard time finding runs with an average of 4.03 per game (second lowest in the AL), has benefited from running at opportune moments adding 3.0 runs. Tampa Bay, who were originally topping the list with 69 stolen bases, have added runs (2.1) through the sheer volume of attempts, stealing at 75% success rate while running 92 times, but fell to fourth on the list. The Twins have dropped from fourth to seventh as their 71% stolen base rate has led to zero runs.

Team SB CS SB% Runs +/-
BOS 58 12 82% 5.2
SEA 44 11 80% 3.0
TEX 33 8 80% 2.4
TB 69 23 75% 2.1
CLE 33 10 76% 1.5
LAA 49 17 74% 1.3
MIN 45 18 71% 0.0
BAL 43 20 68% -1.2
TOR 43 21 67% -1.2
KC 36 20 64% -2.5

As this bizarre American League Central season wears on, it is apparent that the Twins are going to need to retain every run that they could possibly obtain. Like the 1985 Cardinals who had overcome the powerful New York Mets by three games in the NL East, the Twins may be in a dogfight with the White Sox come fall. It is one thing to have the necessary components such as speed, but it is another thing to use it in a way that hurts the bottom-line. The White Sox, by the way, are debted -1.2 runs due to their basepath adventures. Limiting Gomez on the basepaths to unleash his speed more frequently at home rather than on the road would provide a positive net runs gained.