At the beginning of this year a friend of mine sent me a job advert – it was the infamous open letter from Dominic Cummings looking to hire “weirdos and geeks”. According to my friend, this was the perfect opportunity for a man such as myself.  I’m not one to shirk a challenge, and sensing that my time had finally come I crafted an email to the great man, attaching my CV.

I never received a reply, but that’s fair enough really. I don’t think I’d have replied to me – I’m 57, semi-retired and years behind the curve on artificial intelligence. I can quite see that from Cummings’ perspective I’d be the wrong type of weirdo. I didn’t fit among the misfits.

I’ve been building mathematical models all my life – I’m a fair mathematician, if only an average programmer, but I’m determined and generally I get the job done. I’ve worked for Government departments, banks and companies in a variety of industries, and I’ve always enjoyed my work. My Magnum Opus is surely Betfair, the internet betting exchange; I conceived it, designed it, laid out the database schema, built the prototype version myself and oversaw the technology and ongoing development for the first few years. It’s been a huge commercial success and it’s made me a rich man – I don’t need to work any more.

So why apply to work for a Government department?  I did have my misgivings.

When I read the job description, all I could see was extreme maths talk – including Artificial Intelligence gobbledygook that would go over the heads of all but the most committed of maths nerds. The real key to working successfully with models is invariably not about understanding how the maths can work for you – it is more often about understanding what the maths will not and should not be doing for you. I was essentially offering Cummings my services as a non-exec handbrake and naysayer.

Fast forward to August, and the A-level “fiasco” that has recently unfolded.  It was, many have observed, a “predictable” outcome. I agree, but I also believe that there were ways of getting the job done that would have survived the court of public opinion. We don’t have enough evidence at this time to allocate blame, but on the basis of past experience I can’t be blaming Ofqual. I think the problems lay not in the answers they gave, but in the questions that were being asked.

No – the blame, in my opinion, almost certainly lies with the education secretary, Gavin Williamson, and his team, though I wouldn’t want to be too harsh as I think they were doomed to failure from the start. To simply distil the available information down to three grades for each student was folly.  The problem is not that it was done badly – for all we know it was done as well as was possible. The problem is that it was done at all.

Much more was required – perhaps stage one should have been to produce a detailed report starting with a projected mark. This would then have worked down through adjustments based on our knowledge of history and ending with a suggested adjusted mark and associated grade, all packaged as sympathetically as possible to ease understanding.

Stage two of any process must surely be consultation – the part where you share your early plans with all related and affected parties, both for the purpose of arriving at a better and more suitable place with your report, but also to obtain their buy-in to the plan. You will need them when things start to get tricky, which in this instance is an inevitability.

In this case, the government should’ve consult both bodies involved with admissions and those in charge of statistics, to make sure that the former could work with what it would be providing and validation from the latter that the algorithm cut the mathematical mustard. Remember of course that a consultation process is about divining and absorbing opinion: there must be no egos, no politics and, most of all, no red lines.

The government might then additionally have consulted the schools, letting them know that the assessment of their impartiality would feature prominently as part of the report. Who knows, the schools might even have amended their predictions in light of this information, which may in turn have helped the government’s cause.   They might also have been able to widen the scope of its intelligence gathering at this point.

I don’t know how much consultation happened or how effective the consultation was in this case, but I don’t see anything to make me believe that it occurred in any great measure.  If the war was to be won, the key battle was the one at the highest level: the universities and colleges would have to have lent their full and vocal support to the process. If that happened in this case, I didn’t notice.

The final piece of the puzzle should have been to facilitate a quick and easy appeals process, and the key ingredient here was time.  The results should have been announced perhaps two weeks earlier than they were to allow for the necessary and inevitable outcry.  It would have been a painful process for sure, but there was an obligation to allow it to happen in a fair and controlled fashion. What actually happened did not resemble fairness or control, and that’s why the U-turn proved to be inevitable.

It’s easy to be a critic and it’s easy to plan with the benefit of hindsight. No doubt there are gaping holes in my thinking. But the whole process raises fundamental questions in my mind about the function and competence of our current government.

To draw an analogy with the private sector, the government needed to clean up not only its process but also its services operation. Within business there are typically commercial and sales departments. Commercial seek to drive shareholder value by maximizing profit – service are incentivised to produce quality product to keep the customers happy. Governments are different – the customer and the shareholder are the same and the emphasis must always be on quality of service.

The management of the A-level process had all the hallmarks of being run in the style of a commercial project. The government sought to drive it through as quickly as possible with the minimum of fuss;  They provided simple information with little working or explanation; they gave no time to appeal; there was a confident rhetoric assuring us all that everything that could have been done has been done.

But where was the empathy for the customer?  Where was the accountability and transparency to safeguard integrity?

Social media today is dominated by a harsh, bullying blame culture, and this type of harsh culture does not, as is popularly believed, drive quality to the top, and this includes our government.  The people who survive and thrive in blame cultures are people who are good at deflecting blame; creative empathetic types, on the other hand, fare very poorly in such environments. Change is needed; our leaders need to be more service orientated than they are currently.  Maybe this disaster can be used as a catalyst to help us find a better way forward in these changing times.

Andrew Black is the founder of Betfair, the world’s largest online betting exchange.