Changing the score

“Change does not roll in on the wheels of inevitability”;[1]  an axiom, in the context of data, you’d think would either drive or inhibit progress. And you’d be right. This well-rehearsed paradox is playing out relentlessly in the context of the global pandemic and endeavours to disagree on which ‘science to follow’ in optimising our chances of survival.

It goes without saying that our innate instinct and resolve as a species to survive (although not recently always obvious), is exacerbated and compounded where the ‘threat’ is at scale –  in this case global. However, one level down from united humanity, any threat affects everyone differently and this creates factions.

Without being unnecessarily philosophical re-imagining a ‘better way’ routinely stems from these factional and alternative opinions, combining thinking and data to move us forward.  This has certainly played out in the COvid-19 narrative and now in another global and fundamental question of race equality: most rational people agree that things need to change if we are to survive.

Both Covid-19 and the mobilisation of the Black Lives Matter movement have exposed and re-exposed possibly two of the most fundamental challenges conceivable in modern times and with it our collective will to come together towards a solution e.g. optimising our chances against both a pandemic and changing the perpetual pattern of inequality and the drivers of violence.  So far, so good.

Hopefully, the sentiment and intention that follows is self-evidently about finding a solution. There are no balances or comparisons to be struck with Covid-19 or Race equality but in both cases we have the data, we have the will, we want to change but there remains an avoidable stalemate, a reductive reading of the data. To be clear, our point here is neither about a pandemic nor about race equality;  it is about understanding why, in both cases, despite will and information and data, the bodies keep piling up. It is either ironic or predictable but certainly unintentional, that these two chosen examples (Covid-19 and American police) both work out disproportionally worse for Black people. 

Getting to the point: if the objective is to change, survive, redress, coexist, and be free, why are the statistics – the outputs, the numbers of deaths – unchanging? Why, in the context of COvid-19, if we are really ‘following the science’, are the data presented as immutable and absolute, and yet are contradictory? (I’m sure there is a quote about that?).  Why, if we know that Black Americans have been disproportionally killed by police year after year and despite the will and anger and imperative to change this output, does this not happen?


There must be a reason.  In an attempt to get closer to any reason and optimistically any solution, we invented two contrived data effigies which attempted to artificially level two sets of facts derived from the two examples highlighted above.  Our panel and contributors were presented with these scenarios and asked to respond.  We learned two things.  Firstly, that it’s pretty easy to simulate with a few people and some data, the wider societal reaction to what are very different but emotive and fundamental issues. Second, that no number of disclaimers or caveats are sufficient to disaggregate data from issue from emotion. Even where the objective is to objectively look at data (follow the science), emotive and polemic issues as fundamental as these simultaneously generate and evoke the commitment and passion necessary to achieve change, and the polarity that becomes quickly embedded in each faction.  This leads to stalemate or repeated outputs.  The catch is that it is the determination and will to change the output, to make things different, that both mobilises calls for progress, and eclipses the input data, leaving it unchecked. These are the inputs that sustain the outputs (the number of deaths), and their eclipse disguises the foundations that perpetuate them.  The more emotive, personal and fundamental the question or cause the more polarised and determined the struggle and by extension, the allegiance to partisan data, becomes.

Our panel followed the following steps:

1. Find belief or political position or outcome you want. 

2. Find the nearest Google data results to support your position

3. Wave this data (and only this data) about 

4. Suppress or attempt to discredit any data that doesn’t support your argument. 

5. Cement polarisation and in/out group stalemate by personalising ‘your’ data story and experience.

For clarity, focussing on the outputs – describing and quantifying the problem and not the input drivers – is a risk not exclusive to these two examples. However much this post and thinking requires refinement, if we can’t successfully make the basic argument that we are looking at the wrong data in fundamentals like equality and survival then it’s a busted flush. It is worth calling out that we understand that there are any number of criticisms about the validity or point here but hopefully you get our intent.

Here are some other less controversial, yet doomsday type examples to illustrate what we’re on about…

Human extinction is pretty fundamental, and according to the Global Challenges Foundation, is five-times more likely to kill the average American than a car crash.[2] Although we have survived volcanoes and pandemics for a lot longer than the threat of nuclear war and other anthropogenic threats, the fact that we have not ‘completed’ an extinction in our past is not mathematical evidence about its likelihood in the future.  Not to mention that any world that has made itself extinct has de facto no observers left to report back.

Simpsons creator Matt Groening told The Guardian that placing Trump in the White House was the most ‘absurd’ cameo they could think of. The cartoon is routinely cited as having predicted everything from 9/11, to Covid-19, to the deaths of both Kobe Bryant and supposedly George Floyd;  the serious point being that the sheer number of predications made in each of the 640 plus episodes, per Mosteller and Diaconis ‘law of large numbers’, mean that some inevitably come true and receive focus relative to all the others.  

The Guardian newspaper also reported in 2015 that there was a total of 55 fatal police shootings in England and Wales between 1990 and 2014. Only 15 people were shot fatally by German police in 2010 and 2011 combined. The kind of output data that is cited to prove less guns, less dead people.

Output data: In America, the police killed on average three people per day in 2019.  24% of these were Black, despite only representing 13% of the population.  Output data: The very next link down showed the American police consistently killed more white people (in volume) than black.[3] More output data: The very same article then presents the data per population million.[4] The conclusions from this output, the data describing the problem, shows simultaneously that the police in America kill more white people a year in total but of everyone killed, Black people are more disproportionality killed and wider conclusion that the police in American kill too many people. These figures, the output data, describe the problem year after year.

What about input data then?

According to a 2020 Reuters report, in 1967, the U.S. Supreme Court introduced the legal doctrine of ‘qualified immunity’, originally with the rationale of protecting law enforcement officials from frivolous lawsuits. By 2005, this doctrine was described as a “failsafe tool to let police brutality go unpunished and deny victims their constitutional rights”.

This fact, combined with the local-federal spilt decision making process to militarise the police and war modelling has led to a “perceived enemy” culture. Input data: COvid-19 tests return a false negative 1 in three times. Only 0.01% of the population is a risk of becoming seriously ill or dying from Covid-19 yet the economy is collapsing, education is on hold etc.  The list goes on.

There is a ‘wilful blindness’ and bias.  Myopic, selective data only polarises and creates stalemate – not progress or change.  With all these examples, the data we can see and that is pushed into our conscious, is distracting us.  We design it ourselves and choose to let it present alternatives where there really are none – it’s like your mum asking if you want a bath before or after your dinner.  Either way, you’re getting wet, but you feel like you have a choice.

The solutions for change come quickly from exposing all data to understand its inputs, its drivers, and then to dismantle and redress its outputs. Map all our ways forward. Let’s look at the inputs and change the score.

[1] Martin Luther King, Jr., “The Death of Evil upon the Seashore,’ sermon given at the Cathedral of St. John the Divine, New York City, 17 May 1956

[2] The 2016 annual report by the Global Challenges Foundation estimates that an average American is more than five times more likely to die during a human-extinction event than in a car crash.[18][19]



Photograph (c) Nicola Denley

Leave a Comment

This site uses Akismet to reduce spam. Learn how your comment data is processed.