Defense Independent Pitching Statistic (DIPS) is a baseball statistic invented in 1999 by a man named Voros McCracken. DIPS provides a lens through which we might be able to understand why murder in New Orleans is rising so far in 2015.

McCracken wrote an article in Baseball Prospectus in 2001 to present his new research, which found that while a pitcher can control certain outcomes (mainly strikeouts, walks and home runs), he has little control over what happens when the ball is put into play and one of his fielders must try to make an out.

The idea that a pitcher lacks control over what happens once a ball is put in play was hugely controversial, a fact that McCracken himself noted from the outset. What his research often found, however, was that pitchers who were the best at getting outs on balls put in play one year would be the worst the following year, and vice-versa. Studying the results of many pitchers over many years showed McCracken that pitchers allow hits on balls in play at a nearly identical rate to the vast majority of other pitchers over the course of their careers.

McCracken concluded his 2001 article by saying:

The critical thing to understand is that major-league pitchers don’t appear to have the ability to prevent hits on balls in play. There are many possible reasons why this is the case, and I don’t really have a concrete idea as to why it is.

But the one thing I do know is that it is the case.

The Fatal Shooting Percentage

How does all of this relate to gun violence? With a new statistic called the Fatal Shooting Percentage.

The Fatal Shooting Percentage is a number which shows that the percentage of shooting incidents ending in a fatality has remained relatively steady at around 36 percent in New Orleans over the last few years. In other words, roughly 36 percent of all shooting incidents in New Orleans over time should be expected to end in at least one fatality. While the city clearly has an ability to reduce gun violence incidents, vacillations in the percentage of shooting incidents that end in death appear to be random.

There were 1,893 shooting incidents in New Orleans between 2011 and the end of June 2015 with 694 (36.66%) of them producing a fatality. The breakdown of incidents since 2011 is provided in the table below. In some years, the proportion of shootings that are fatal have been above this threshold; in other years, the proportion has been below it.

The Fatal Shooting Percentage in New Orleans, 2011 to June 2015. Source: NOPD Calls for Service data.
The Fatal Shooting Percentage in New Orleans, 2011 to June 2015. Source: NOPD Calls for Service data.

The data suggest that the percentage of shootings ending in a fatality will vacillate above and below 36 percent, but over time it will remain relatively steady. Improvements in medical techniques over the 1990s and early 2000s may have had a positive impact on survival rates of gunshot victims, but those techniques do not appear to be affecting survival rates over the short/medium term.

Estimating 2010

Proof of a Fatal Shooting Percentage average of around 36 percent is made even stronger by estimating how many shooting incidents happened in 2010.

The first step in estimating 2010 shooting incidents is this chart showing non-fatal shooting victims for New Orleans from 2009-2013. The chart indicates the number of non-fatal shooting victims from 2010 to 2013 was: 415, 400, 378 and 321 respectively. A separate article shows that there were 396 non-fatal shooting victims in 2014.

Comparing the number of non-fatal shooting victims from 2011-2014 with the number of total shooting incidents over those years shows fairly good stability in terms of, on average, the number of non-fatal shooting victims per shooting incident. Shooting incidents between 2011 and 2014 produced between 0.853 and 0.917 non-fatal victims per incident, with 1,696 incidents producing an average of 0.881 non-fatal victims per shooting.

As noted above, 2010 is said to have produced 415 non-fatal victims. Based on the rates of non-fatal victimization between 2011 and 2014, there would have been between 452 and 486 shooting incidents in 2010. A previous public records request fulfilled by NOPD showed there were 147 fatal shooting incidents in 2010, which means the Fatal Shooting Percentage for 2010 was most likely around 31.3 percent (147 fatal incidents, with about 469 total incidents).

What adding 2010’s shooting total and Fatal Shooting Percentage suggest is that, rather than being on a steady downward trend, the proportion of shootings in New Orleans that will result in death tends to concentrate around 36 percent over the long term. With this added confidence in the Fatal Shooting Percentage, we can turn our attention to gun violence in 2014 and 2015.

Fatal Shooting Percentage, 2010 to June 2015 with 2010 estimated. Source: NOPD Calls for Service data.
Fatal Shooting Percentage, 2010 to June 2015 with 2010 estimated. Source: NOPD Calls for Service data.

2014: Shootings Rise, Murder Does Not

Much like the pitcher who cannot control hits on balls in play, comparing shooting incidents and the Fatal Shooting Percentage from 2014 to 2015 can help explain why murder is on pace to rise. Reviewing 2013 showed how the city can make an impact on how many shooting incidents occur from year to year. But the Fatal Shooting Percentage in 2014 and 2015 highlights the truism that whether a New Orleans shooting ends in a fatality appears to be largely random.

Shooting incidents in New Orleans dropped from around 460 in 2011 to about 440 in 2012 and 360 in 2013 with corresponding declines in murder over those years. But oddly, murder fell again in 2014 even though shootings increased for the year to approximately 430, nearly reaching 2012 levels.

Although city policies appear to have help bring about 2013’s reductions in both shooting incidents and murders, analyzing gun violence for 2014 suggests randomness played a significant role in the slide in murder for that year. Other factors in 2014 unrelated to gun violence also played a role in murder falling, but the greatest difference between New Orleans murder in 2013 and 2014 was that a smaller percentage of shooting incidents ended in a fatality in 2014.

While 2013’s murder decline was a very positive sign for gun violence reduction strategies, shooting incidents rose substantially in 2014. Murder stayed relatively low in 2014, however, because only about 32 percent of the shootings that year ended in a fatality. As previously shown, 36 percent of shooting incidents should be expected to end in a fatality. That only 32 percent of shootings in 2014 did means that 17 shootings were randomly non-fatal rather than fatal in 2014 over the course of 430 shooting incidents. In other words, had the usual averages applied, New Orleans would have had around 167 murders in 2014, rather than 150.

2015: Fewer Shootings, Regression to the Mean

In this light, a rise of murder in New Orleans in 2015 without a substantial drop in shooting incidents shouldn’t be shocking; according to the Fatal Shooting Percentage, it is wholly predictable. You wouldn’t know it from reading the local newspapers, but New Orleans had a stellar January through April in terms of overall gun violence. The city’s 112 shootings through April 30 were 23 fewer than the city experienced through that point in 2014, 36 fewer than 2013 and 2012, and 55 fewer than 2011.

The positive start to 2015 is illustrated by the below chart showing the pace of shooting incidents in New Orleans from 2011. Shootings definitely started to pick up in mid-May, but the city has still had a positive year in 2015 in terms of gun violence relative to 2014. There were 27 fewer shooting incidents in New Orleans over the first six months of 2015 compared to the first six months of 2014. Murder is up, however, because the randomness that benefited New Orleans in 2014 has caught up with the city.

Annualized Pace of Shootings by day, January 2011 - June 2015. Source: NOPD Calls for Service data.
Annualized Pace of Shootings by day, January 2011 – 2015 (UPDATED TO AUGUST 2015). Source: NOPD Calls for Service data.

Although total shooting incidents are lower through June 2015, fatal shooting incidents are up because over 40 percent of shooting incidents in the first half of the year ended in a fatality. Randomness played a positive role in reducing murder in New Orleans in 2014, when only 32 percent of shootings ended in a fatality, but it is playing a negative role in murder in New Orleans so far in 2015.

The Fatal Shooting Percentage is slowly falling in 2015 but it remains well above the expected annual average of 36 percent.

Fatal Shooting Percentage by day, 2011 - June 2015. Source: NOPD Calls for Service data.
Fatal Shooting Percentage by day, 2011 – June 2015 (UPDATED TO AUGUST 2015). Source: NOPD Calls for Service data.

Explaining Randomness

Voros McCracken was kind enough to chat with me briefly about randomness and gun violence in December 2014. He talked about how he attempted to convince a highly skeptical audience about a potentially controversial concept such as the role of luck and randomness in baseball and what the term randomness truly means.

Unfortunately you have stumbled on one of life’s truisms when it comes to not just studying baseball or crime or whatever, but any time probability is involved: human beings, as a rule, do not like to ponder the extent to which luck (or random chance or whatever) affects outcomes in the world around them, and I’m guessing they particularly don’t like to ponder it when it comes to things like life and death.

As for tips in dealing with this normal reaction, here’s one that I have absolutely no guarantee will work: one of the keys to explaining probability and chance to a skeptical audience is to emphasize the vast amount of factors that could affect the outcomes you’re studying and that while you can control for some of these, most are so hidden or hard to know, their effects on the outcome seem random.

Random chance and ‘luck’ isn’t about things which have a random number generator affecting them, it’s about our own lack of knowledge on the factors determining the outcome. When we don’t know and at least temporarily can’t know about those factors, we use random numbers and probability as a way to still be able to move forward to find out whatever information we can. Otherwise our only option would be to throw up our hands and not bother to try and learn anything.

The reality is that the reason 36 percent of New Orleans shooting incidents over time end in a fatality is unknown right now. To quote McCracken’s words from 2001, “there are many possible reasons why this is the case, and I don’t really have a concrete idea as to why it is. But the one thing I do know is that it is the case.”

We can’t explain why the Fatal Shooting Percentage appears to exist in New Orleans, but it does appear to exist and is influencing murder levels in the city.

Further study of this concept is absolutely needed, because that study may be able to find a way to reduce fatal shootings. As McCracken noted, there could be countless explanations for why 36 percent of shootings end in death, and understanding why could help save lives.

What is clear is that this phenomenon is occurring and it can be used to help explain what New Orleans is currently experiencing and to predict what might happen in the future.

New Orleans will probably finish with more murders in 2015 than it has seen in the last couple of years, and, as difficult as it may be to believe, randomness appears to be the cause.