CS% stats over time

[dominik]

Here are the CS% values for every year since MLB started (3rd collumn from right)
http://www.baseball-reference.com/le...LB/field.shtml

While it always stays in a certain range I see some quite significant fluktuations. What is the reason for this? for example in the last years it went down a little, it's now under 30% for some years.

however the most striking thing I see is the 30s-50s. It's consistently well over 40%. do you have an explanation for this? is run environment and the number of steals the reason? (although that doesn't make much sense because that was a high run time which doesn't favor high steal rates).

why are the rates so much lower today when pitchers and catchers work so hard to shorten the time to the plate and pop times?

[ipitch]

why are the rates so much lower today when pitchers and catchers work so hard to shorten the time to the plate and pop times?

These days, with easy access to game videos, the offensive team has the pitcher's time to the plate, the catcher's time to 2nd base, and the runner's time. That gives the advantage to the base stealer, since they won't steal (generally) unless the times are in their favor. That's my theory, anyway.

[brett]

Iteresting. Runs are down from '93-'04 levels, but why are R/G in the 2's? Anyway what is most significant is that steals are UP, and success rates are up. I would expect steals go up when offense goes down, but to have them rise until the CS rate hits some critical level. That doesn't seem to really be the case. Look at 1967 which was a low offensive time. CS rates were up to 41% even though we only had .84 steals per game. Steals then rose over the next 20 years with CS rates dropping almost lockstep down to 30% and double the number of steals per game.

Come to think of it though, a team or league is not going to run until they get caught at a break even level overall, they are going to run more as long as the next marginal attempt is likely to be valuable. At 26% though with a decreasing offensive level it would seem to predict that we could go back to 1980s steal numbers with guys challenging 100+.

[pheasant]

I believe the CS% was higher 70-100 years ago for a few additional reasons not mentioned: The higher strike zone forced catchers not to crotch as much, which made throwing much easier(particularly in the Dead Ball era). And today, I believe that the catchers arms in general are too muscle-bound due to weight training. Speaking from personal experience, I think that weight-training hurts the throwing arm. Granted, proper stretching or steroids should be able to offset the negative affects of weight training on one's throwing arm. I believe the CS% has dropped recently due to a drop in steroid use.

[brett]

I wonder how much the removal of the high strike helped or hurt CS rates.

also in the deadball period, a lot of CS were hit an runs that failed.

[dominik]

I wonder how much the removal of the high strike helped or hurt CS rates.

also in the deadball period, a lot of CS were hit an runs that failed.

that might have a role. however the 30s-50s were a high scoring environment and still rates were high. maybe the increased use of breaking balls also plays a role. but I don't really see a pattern yet regarding the CS rates.

[leewileyfan]

In the deadball era and through my early teens, catchers were seen as the de facto team "captains" in the field. With opponents on base, they played the dual pronged role of traffic cop [eyes and arm] and benificent "Svengali" intent on keeping pitchers aware of in-game dynamics. There was pride in "arms" that could snap throws to any base for pick-off attempts and a very special pride in being able to nail would-be base stealers. In fact, one of the common "drawing" contests between games of a doubleheader was the one in which the participants were a pitcher, a catcher, and an empty beer barrel parked at second base, with the opened "top" facing home plate. Opposing catchers would have "X" numbers of throws to fire the ball into the mouth of the beer barrel.

Pride in that throwing arm made preservation of that talent extremely important ... "If it ain't broke, don't fix it." Weight lifting and rolling muscles were NOT seen as throwing arm enhancements; they were invitations to injury and loss of range of motion. You kept a throwing arms strong and accurate by THROWING ... but not to the point of fatigue or strain.

Long ball has drawn fans to the park in ever increasing numbers; but the more emphasis is placed on long ball, the less attention is paid to the finer points and nuances of the game, especially those that are defense-oriented. Pitchers are more lax in holding runners on; and any catcher who has followed a regamen of weight traing, has probably done no benefit to the natural abilities of his throwing arm.

The math isn't rocket science. A pitched ball will travel from the point of release by the pitcher, to the catcher, in about .43 of a second. The drama for the base runner, in anticipating the rhythm of the pitcher, starts a second or two earlier ... a critical time in which a pitch to the plate might be traded in favor of a throw to first base. There is that .5 second instant between the pitcher commitment to the plate and the release that determines the base runner's "jump."

If the runner has a 15' lead when he heads for second base, allowing for the transition for sliding, the runner will take about 3.2 seconds to hook onto the bag in some fashion. If we give the catcher .5 second for receiving, transfer and release of throw, we have:

[.42] : released pitch to plate + [.5] pitcher movement that "tips" his commitment to come to the plate = .92 seconds;

[.50]: catcher receipt, transfer + release = [.50] seconds

If the runner has 3.3 seconds to execute his steal attempt; and if the 1.42 seconds already taken BEFORE the catcher's release, the catcher has +/- 1.7 seconds to make a throw that nails the runner. IF the catcher can make a throw that nets out at 85 mph over 130+ feet, he can complete the entire play in about 2.6 seconds, having half-a-second to spare. Most plays at second base on attempted steals are closer than that; so a "rifle-armed" catcher is probably releasing a throw that averages 75 mph to 2B.

That's a heckuva arm. If we consider pitch speed studies, we are told that a pitch released at 100 mph will reach the plate at 92 mph over less than half the distance a catcher's throw must cover ... without leg kick; without rubber to drive off; and laden with catcher's gear.

It is a skill that the modern power game has been discounting for some time.

[brett]

But keep in mind that the break even mark for value on steals has ranged from just under 60% to just over 70%. In a 60% setting a guy who goes 70-30 has netted +25 steals above break even. In a 70% setting, of course he's netted nothing, and used a lot of energy to do so.

I think steals also occur in a bimodal fashion. Let's say you'r facing a staff and catcher who throw out near 40% in a league with a 70% break even level. When do you steal? a) when it matters a lot; b) when it matters very little. So we get more surprise steals that might not be meaningful, and we get steals against good pitchers in low scoring games when nobody is hitting. Is small ball back? Maybe we are seeing a lot of the less meaningful surprise steals.

I was wondering if perhaps the high K rates today also contribute somehow. Do you want to steal and give your batter a strike? Also, how have non-sac fly sacs changed? Do we see fewer sac bunts today (as sabermetrics has suggested as optimal?)

[leewileyfan]

But keep in mind that the break even mark for value on steals has ranged from just under 60% to just over 70%. In a 60% setting a guy who goes 70-30 has netted +25 steals above break even. In a 70% setting, of course he's netted nothing, and used a lot of energy to do so.

I'd suggest that the sabermetric values attached to the risk-reward of base stealing comports pretyy well with the speed-time elements I suggested. If you take a calculator, a camera, a stop watch and a scout and study miles and miles of MLB tapes of stolen base attempts, you'll come up with a model of time and motion that gets broken into 3 basic elements:

1. pitcher stretch, leg kick and stride, push off mechanics and resume of "stuff" [and how fast it gets to the plate;
2. catcher mechanics and catching methods [like one or two-handed]; ball transfer and mechanics of release;
3. team baserunners, relative speeds, lead know-how for optimized "jump."

It boils down to a 3 second drama. + or -; and [as I see it] sabermetrics SEEMS to come down on the side of the pitcher-catcher element, seeing stolen base attempts as a losing proposition UNLESS you can predict a success rate of +/ 65%.

If an overpowering fastball leaves the pitcher's hand at 100 mph and arrives in the catcher's mitt at 92 mph, how much mustard must be on the point of release to travel almost 2.15 X further than from the mound to the plate? The catcher is behind the plate by a few feet; and the pitcher's point of release is in front of the rubber by several feet, so the difference becomes exaggerated. An 85 mph [net] throw to 2B has to be released with approxzimately the same speed as that overpowering fastball; and it has a very resticted "strike" zone at the 2B end [or the person covering the bag will lose precious tenths of a second jumping, stretching, scooping and shifting to make a tag. Even so, the odds say that steal attempts are a losing proposition.

For me, that magnifies the values of the catcher who can keep runners honest [and keep battery-mates attentive]. Given the style of the game today, I'm a bit surprised that there isn't more stealing, because the skills to combat the steal are eroding.

[stevebogus]

I think that from the early 1920s to the late 1940s the straight steal fell into disfavor. Before 1920 in the deadball era teams were aggressive on the basepaths, trying to create offense by challenging the defense. There were many more errors in that period, so while there were more runners caught stealing than today there were also more instances where an error would result in additional bases gained. So, even at a 50%-60% success rate the steal may have been a good play.

The beginning of the liveball era in 1920 immediately made teams more cautious on the bases and began a downward trend of steal attempts. However, the hit and run was probably still a part of the game. In the 1920s the strikeout rate collapsed and batters stood a much better chance of making contact. With fewer straight steals you might expect that runners would pick their spots better, but that may not have been the case. If the steal remained a "challenge" play rather than a strategic play then runners may have been going in situations when the defense was expecting it. The downward trend in stolen bases did not reverse until 1948.

In 1948 steals began to increase. I think this is primarily due to the integration of baseball. The Negro League style of play included aggressive baserunning, and this was reintroduced to the major leagues. Within about ten years the emphasis was placed on basestealing specialists who had high success rates, rather than players who were merely fast. So, beginning in the late 1950s we had an increasing number of excellent high percentage basestealers. The introduction of Astroturf in the late 1960s and through the 1970s also emphasized speed. Much of the change in the success rate came from 1955-1987, when teams were combining increased attempts with increased success. After 1987 we have seen further increases in the success rate with decreasing attempts, as baserunners are picking their spots and emphasising success rate over raw steal totals.

[Tyrus4189Cobb]

http://www.youtube.com/watch?v=xgz5-XToJIw

That is the science behind stealing a base. Many minute factors can change the outcome. A lot of the theories posted are viable here because the window of time between stealing the base and being caught is ever so slim.

I've also always thought that dead-ball era stars were caught a lot because of the prevalent running game. Perhaps catchers, recruited more for defense and less for hitting, were constantly wary of even the slowest of runners? Games lasted less than two hours, so I believe that the men behind the plate were always on the prowl in these high-octane games.

[reviewboy1]

Interesting. I didn't realize that CS stats were available for early seasons, like Dead Ball I and late 1800s. In fact, B-R.com does not seem to list CS stats for individual players for some of those seasons, for example 1908.

Are early era CS stats truly accurate?

Just wondering.....