Friday, August 30, 2013

Team Speed

While watching the Cubs/Padres games this past Sunday, the Stat Sunday feature was on team speed. There's a very nice page on Baseball-Reference that puts a wide variety of different proxy measures of team speed in one place. All data is through Thursday, August 28th, 2013.

What's the difference between an error and a bobbled ball that becomes just another out? Speed--if Norichika Aoki hits a smash to 3rd and the fielder mishandles the ball chances are very good he'll reach base via an error--at least it seems that way since he's tied for the league lead (with Andrelton Simmons) in reaching on errors with 12. On the other hand, if the batter is Paul Konerko, the third baseman has time to go into the stands and grab a hot dog and beer, sign a couple of autographs, scope out the "talent" and STILL make the play. Here's how teams rank in 2013:

As a Cubs fan I'll freely admit to being surprised that they lead the league in this category, and it's pretty obvious how far this has gotten them this year. Like any other metric, NOTHING can be taken by itself as an explanation of team success--for example, the Rays, a team that prides itself on aggressive base running and team speed, is last by a healthy margin. 


The stolen base as a strategy has been all over the place in baseball history--very prevalent in the Dead Ball Era, dormant from around 1920-1960, revived in the 1960s and really took off in the 1970s and 1980s, had another period of dormancy from around 1990-2007 and appears to currently be enjoying another period of renaissance today. In NONE of these periods was there any correlation between stealing bases and team success--some teams, like the Reds and Royals of the mid-70s were good teams with high stolen base totals, but those teams certainly had other factors in their favor. Likewise, other teams like the Angels and Cardinals of that era also had speed and had no success to show for it.

I find base stealing to be potentially confusing since I see three different dynamics at work:
1. How many times do teams try to steal? It ranges from 53 attempts for the Cardinals to 157 attempts for the Brewers.
2. How successful were they? The Brewers 114 steals have to be counterbalanced against their 43 caught stealing, but it could be worse--the Astros lead the league in caught stealing with 45 and managed that feat in 20 fewer steal attempts.
3. How many opportunities did teams have to steal a base? Just as RBI are very much a function of opportunity in that a runner that isn't on base can't be driven in, a team needs to have base runners in order to have steal attempts. If aggressiveness on the base paths can be determined by:

(Stolen Bases + Caught Stealing)/Stolen Base Opportunities (SBO)
Then the Brewers (they attempt to steal in 9.3% of their opportunities), Astros (8.4%) and Rangers (8.0%) are the most aggressive and the Tigers (2.3%), Cardinals (2.8%) and Mariners (3.1%) are the least. 

One of the most-read posts on this blog was one of my first where I discussed the Tom Tango Run Expectancy Matrix. It just so happened that one of the readers was...Tom Tango, or at least one part of him/her/them. Here's the table (it's the second one on the web page):

This table shows the odds a team will score a run by base and out situation.  Assume a runner on 2nd with 0 outs--this chart suggests a 63.7% chance of scoring. If the runner steals 3rd, yes, the odds of scoring a run improve dramatically, to 85.3%--anytime a team can improve the odds of scoring by over 20% needs to be taken into consideration.

But that's not the only consideration--what if that runner is thrown out? Now the odds of a run scoring in the inning drops to 17.2%. In other words, stealing 3rd increases the odds of scoring by around 20% but getting thrown out reduces the odds by around 45%. The cost of the out is 2.25x greater than the benefit of the stolen base. I'd explain further but I doubt many will care, so I'll just state that the knowledge of these numbers is widespread throughout baseball and why steals of 3rd are far lower than of 2nd.

Advancing to 3rd on a groundout improves the odds of scoring to 67.4%--granted, it's not 85.3%, but it's far better than with no base runner. The 1993-2010 time period was one of enhanced offense, and when hitting is ascendant stealing has a tendency to play a smaller role--I'll be very curious how the diminished offense of the past couple of years has affected these numbers.

What doesn't change is the difficulty of stealing home, although this year hasn't been bad--8 of 21 (38.9%)--historically it's around 25-30%. No matter what, stealing home is HARD--unless human physiology changes in a dramatic fashion a thrown ball will reach home far faster than a human being running from 3rd.


This table shows the number of times a player was thrown out trying to advance on the bases. It does NOT include force outs since there is very little a runner can do when the ball is hit directly at a fielder. Whether this is a measure of team speed as much as team aggressiveness is a valid point --for example, Detroit appears fairly aggressive on the base paths but doesn't try to steal much.

One thing I've noticed as I enter box score data is that it seemed to me that teams were FAR MORE aggressive, almost reckless, on the base paths in years gone by. When I stumbled upon this page earlier this year I realized I could actually check this:

It's always nice when actual data lines up with a thought and it appears my suspicions were correct. This is yet another result of a movement that began with Earl Weaver (and probably earlier) to not give away outs with risky base running. The Tango table shows very clearly that there IS an incremental advantage to taking an extra base but that the increase is far less than the cost of an out. I'm not suggesting that aggressive base running be eliminated, because it's not even a fine line between aggressive and reckless, it's more of an attitude. I'll explain that further in the last section.

This very busy table has a plethora of information and I'll break it down by section:


The first section (BT, XBT%) are Bases Taken and Extra Bases Taken % on items like fly balls, wild pitches, passed balls, etc. This is an excellent measure of team base running aggressiveness in that it becomes a proxy measure of team speed, since slow runners have a more difficult time taking advantage of these situations. The XBT% helps put these numbers in context--a brief look at the Cubs and their very low number of 81 bases taken in these instances is properly placed when it is seen that they don't have as many opportunities. Some of this is definitional--I'm not sure if it's a passed ball or wild pitch UNLESS the runner advances (5 seconds of Wikipedia research suggests this is the case), so a slower team can't take advantage. There's a mix, but look at the teams that are near the top in these categories--the Rays, White Sox, Reds and Red Sox, all known for both their team speed and aggressive base running.

The next three sets of columns are related but tell a very compelling story--the first set are runners on 1st when a single is hit (1stS), with the next two being how often they advance to 2nd (1stS2) and 3rd (1stS3). The extra base taken by going to 3rd occurs much less often than I suspect people realize, around 25-30% of the time. This is perfectly logical as I bet around two-thirds of singles are hit to left field, making it VERY difficult to advance to 3rd unless the left fielder is a complete stiff, the runner is extremely fast (and rash) and the ball is hit in JUST the right place. Here are the spray charts for the left-hand hitter Joey Votto and right-hand hitter Dustin Pedroia (both available from Brooks, which now has PITCHf/x data for pitchers AND hitters):
 I may not have picked the best examples, particularly with Pedroia as he appears to do a decent job of using the entire field (note how the green dots, which are hits aren't just located in left field) but my general point is made--most of those ground balls that make it into the outfield are singles. Any opposite field single or even one to center gives the runner a chance to advance to 3rd, anything a hitter pulls much less so. I'd be very interested to see if data supports this but can't find it--hit location splits data from Baseball-Reference breaks down hits by pulled, up the middle and opposite field, but I have no clue what those distinctions are.

The next set of columns are doubles with a runner is on 1st (1stD) and how often the runner advances to 3rd (1stD3) or scores (1stDH). The percentages increase dramatically since doubles are usually hit where players ain't, giving the base runner a much greater chance of advancing. In addition, the location of the hit becomes less important as now even the left fielder has to make a strong AND accurate throw, and even with these advantages, runners advance from 1st on a double and score only about 40% of the time. 

The last columns are a runner on 2nd when a single is hit (2ndS) and how often they advance to 3rd (2ndS3) or score (2ndSH). Now the number increases to around 60%, and I think part of explanation is that it's easier for players to advance two bases as opposed to three when a runner is on 1st. Whether this is due to changes in strategy or ability is difficult to tell.

Has this incidence of taking the extra base changed over time?

Reach your own conclusions on whether this is "bad" or if players "don't play the game like we used to <spit>!" Without more segmented data I can't state this conclusively, but I believe this demonstrates greater caution on team's parts vs. a diminishing of skills. Today's players are definitely faster than those in the past and there is NO way they aren't just as baseball-smart, given they've been playing baseball almost year-round from the age of 10. To me this chart suggests teams are much more willing to take what is given as opposed to "making something happen." It is true that anything can happen on any play but just as true that the higher the level, the less the opportunity of those mistakes occurring. Caution doesn't have to mean timid or less-skilled.

I'll finish a fairly lengthy post by putting this all together, adding together all the speed items:

This takes the positive elements of speed (reaching on errors, stolen bases, extra bases taken on outs, WP, PB, advancing to 3rd on a single, or scoring from 2nd on a single or 1st on a double) and subtracts the negative (caught stealing, pickoffs and out on bases) for a simple number. To me, I see two types of teams here:
1. Teams that DO have team speed
2. Teams that are very aggressive on the base paths

While these two notions aren't mutually exclusive, they also don't march in lockstep with each other. Team speed and aggressiveness on the base paths are no guarantee of team success, nor is their lack a kiss of death. If offense in baseball is declining, then smart teams will pay more attention to these items and make the fullest use of them--they're all extra bases at heart. Since there is a very direct relationship between total bases and runs scored, the more bases, no matter how they're acquired, the better. We may be in the process of seeing offensive changes in baseball that may make these items, all of which are part of what many call "small ball" become important again. To be honest, they've never gone out of style, and it's not "small ball" as much as the successful execution of strategy, but as the home run specifically and power in general declines, the importance of the ability to advance on bases in other manners will increase.

No comments :

Post a Comment