Author: JP

Soon, AI will be used as the scapegoat for anything that doesn’t work

Following on from my last two posts, I’ve continued to ruminate on instances where inflexible rules (often compounded by unnecessary complexity) created problematic circumstances that could only be solved by humans with empathy and empowerment.

I loved the @mildlyamused tweet that Stefan referred to in his post, where Donna talked about “arguing with a motion sensor about whether or not your hands are in the sink”. As we continue to implement inflexibility and poor design in software, we’re going to see more and more of this.

It’s easy to blame the software or firmware or hardware. All designed and implemented by us. Soon we’re going to blame “AI” for everything that doesn’t work, in an ironic variant of Douglas Hofstadter‘s definition of AI, sometimes referred to as Tesler’s Law or Effect, sometimes attributed to others; Hofstadter described AI as “anything that hasn’t been done yet”.

The Cheap Day Return saga

Many years ago, around 1986 or 1987, I’d planned a day out playing golf with a bunch of colleagues. We’d booked a tee time at Gatton Manor, near Ockley, which at the time prided itself as the course with the most holes with water hazards; it also had the longest Championship par 5 in the country, I think it was the 16th, 625 yards. (The course has been shut for development for many years; I hear it’s reopened, and I will plan a visit soon).

Anyway, back to my story. I’ve never driven in my life, as in driven a motor vehicle. (After seeing me play off the tee, the unkind amongst you would say that the distinction is a tautology).

Since I didn’t drive, I’d arranged to meet the rest of the four ball at Camberley station. Why Camberley? Because they lived there. So I wandered down to Windsor and Eton Riverside station at the crack of dawn, asked for a return to Camberley, changing at Staines. The man at the ticket counter asked me when I’d be returning, I told him, and having heard that, he issued me a Cheap Day Return and I went to the platform in a state of anticipatory bliss. (I’m like that, many small things fill me with joy).

Now you can’t get to Camberley without passing through Ascot and Bagshot first. When I got to Bagshot, I was shaken out of my early-morning reverie by the raucous sounds emanating from the car park there. My colleagues and golf partners. Trying, urgently, to attract my attention. Which they succeeded at. So I got off the train, one stop “early”, accompanied by my clubs and bag.

Even at that unearthly time, there was a ticket inspector at the exit. I gave him my ticket.

“This ticket’s not valid. You can’t get off here”. “Why ever not? I’m paid up to the next stop, and this is a valid “alighting” stop, isn’t it?” “Yes, but not on this ticket. Cheap Day Returns from Windsor cannot be used at this time to get to Bagshot, but are valid from Camberley onwards”.

Heaven help us.

(Incidentally, in the Wikipedia entry for Cheap Day Return, the second reference is to the Jethro Tull song from Aqualung. Well worth listening to. Here’s the link.)

Hyperlinks subvert hierarchies

I know I’ve written about this before, probably over a decade ago. I gave him my card, told him he can write to fine me or whatever he had to do, and barged past him to meet my friends, reverie now no longer as blissful as before. (British Rail actually wrote to me to explain the rules and to “let me off” grudgingly just this once).

Now all this was dreamt up by people and imposed on people in a very analogue way; as more and more of what we do goes digital, we’re going to find more things that go wrong. That doesn’t mean we shouldn’t be digitising processes. What it does mean is that process design in a digital world really needs to understand the concept of leeway.

David Weinberger, writing in The Cluetrain Manifesto, introduced me to the lovely phrase “Hyperlinks subvert hierarchies“. In this particular context, I think of complex rules and processes as the hierarchy, and trust-based human interactions as the hyperlink.

A classic example for me is something I’ve had to face a few times:

I’m heading for the airport. I get a notification that the plane I’m booked on is delayed by a few hours. So I make use of the time doing something else and then arrive at the airport well in time for the now-delayed flight. Hand baggage only as usual. (Paper) boarding pass in my hand. (Why paper? I will come to that later). Walk up to the turnstiles that welcome me to security. Place pass on scanner. Red light. You can’t pass Go or collect 200 or anything else.

Why was I stopped? Because the system thinks I’ve missed my flight, or at least that I’ve missed the deadline by which I should have passed through security and headed for the gate. It doesn’t matter that the flight is late, and that the plane hasn’t even arrived at the gate to let people board. All that matters is that I’ve missed the preset deadline.

This has happened to me more times than I would have liked. Each time, I was late-on-purpose, to catch a flight that was later-still. Each time, I had to find someone with the hey-presto ability to magic me past the turnstile. At least once, in a mainland Europe airport, I’ve had to vault the turnstile, inelegantly and clumsily, to make it through security on time; it was the last flight to depart that night, and there was no one left manning the check-in desks.

Stuff like this happens all the time, not just tied to the underlying process. We’ve moved a lot of functionality to the cloud, and made our smart devices carry critical workloads.

But.

It doesn’t all work the way it’s meant to. Bar codes and QR codes don’t scan. Image recognition doesn’t work. Phones run out of battery. Taps, towel dispensers and soap servers don’t recognise the presence of your hands. Signals are weak. Surfaces overheat. Touch screens become unresponsive. Sunlight makes the screen unreadable; ambient noise makes the instruction unhearable.

Stuff happens.

I was at a phone shop, waiting patiently to be served, finally at the front of the queue, when a woman came running in to the shop and beseeched me to let her jump the queue and be served before me. All she wanted was a charged battery, or the chance to get a little bit of charge on her phone. She’d come from the shop across the road, groceries piled up in her cart, went to pay, her phone had “randomly” run out of battery, even though she was sure it was over 50% ten minutes earlier. Her phone was her only way to pay. She had a parking slot that was about to run out and a child that needed picking up from school.

It all worked out. But it needed shop assistants in both shops to understand, to make exceptions; it needed the others involved to be human and not mechanical, social and not selfish.

The good thing is that most of the time, humans are human. Social. Empathetic. Machines are machines.

Trust is the hyperlink that can subvert hierarchies.

Kletskassa

Herb Kim, who runs the wonderful Thinking Digital conference, shared a story on one of the social channels a few days ago. He’s one of those people who regularly shares stories that will enrich and encourage others, rather than criticise them or cut them down. One such story was about a Dutch supermarket operator who’d opened slow-checkout queues to help combat loneliness amongst the elderly. (An aside. Legend has it that when Mahatma Gandhi first visited somewhere in the West, he was asked what he thought of Western civilisation. He considered the question and then replied, I think it would be a good idea).

I love the very idea. People don’t necessarily do something for the “obvious” or “visible” reason. When we design stuff or automate stuff, we need to consider that.

As I write this, my local train operator plans to shut a number of ticket offices down. No more ticket counters. Instead, machines that are platform-mounted, exposed to the elements, with screens that are hard to read in what passes for sunlight here. Machines. With hard-to-follow instructions based on complex rules. With error conditions that are often neither displayed nor guessable. Offering choices that are very context dependent, missing the nuances that a trained and experienced ticket clerk provides.

My local station has a wonderful ticket clerk. She’s part of the community, part of what makes the village tick. She knows everyone, greets everyone, knows what they need. Knows enough about them to give them advice that is precision-tailored for them. Asks the right questions. And does all this in an environment she works to make friendly and welcoming. Christmas decorations and Easter bunnies and everything. Warm when needed, cool when needed.

Her personality and attitude and service are probably key reasons why some amongst the elderly feel less lonely. They come for a chat, not just to buy tickets to go somewhere. They come for the warmth they are guaranteed before they go wherever they’re going. These things are really important in villages with the demographics we have, an ageing population with many people living on their own.

My village isn’t unique. All over the West, the demographics are heading that way. All over the West, there are instances of misunderstandings as to what a process is really about: the human contact, the ability to have conversations, the camaraderie, the shelter, the dignity and purpose that come from all that.

We can’t blame devices for not working. Hierarchies doing what they do best.

What we can do is be the hyperlinks. Subvert the hierarchies.

Planning for offline behaviours

When I was at the Euro final a couple of years ago, the ticket technology just failed to keep up, and people stormed the gates. Ugly scenes. I’ve seen variants all over the place, where the “design” of a process is focused on objectives like preventing fraud and controlling secondary markets, while paying limited attention to the things that will fail and the leeways needed. I think David Birch, when writing about digital currencies, often makes the point that no digital currency will work unless it caters explicitly for offline behaviours.

Devices will fail. Lost, stolen, out of range, battery dead, screen shattered, microphone or camera not working, Bluetooth disabled, whatever.

Processes will not work as intended. The connection with the mother ship may be down. The mother ship may itself have problems. Stuff happens.

The conditions in which the failure happens are usually the least helpful. Inclement weather; a time when no one is around; a time when everyone is around and everyone is frustrated; long queues here, deserted halls there.

Stuff happens. Stuff that can be fixed because people design for human intervention. Empowered and empathetic.

Many years ago, I gave lectures on the topic of Designing for Loss of Control. Many years ago, I spoke here and there on The Future of Lurk. Both topics continue to intrigue me. Not everyone is digital-savvy, not everyone has the wherewithal to work around the obstacles formed by failing systems, processes and connections. And yet we continue to entrust much that is critical to that world.

It can work.

It will work.

Only if we make the allowances we need to. Only if we treat people with respect, with tolerance, with dignity. Only if we work on the basis of mutual trust.

Only if we continue to be human.

The background

Last week, I wrote about cricket and codes, shaken 0ut of my self-imposed Rip Van Winkle reverie away from public writing. One of the comments I received was from Stefan Czerniawski, reminding me about David Weinberger’s talks and writings about leeway.

(Incidentally, if you haven’t read David’s work, please do so, starting with The Cluetrain Manifesto. I cannot believe that soon we will be celebrating twenty-five years since he, Chris Locke (sadly no longer with us), Doc Searls and Rick Levine helped prepare us for what lay ahead.)

On his blog, in 2002, David led us into thinking about the importance of leeway in a digital world. As he says, computers don’t do leeway. He then chose to devote an entire section to the concept of leeway:

The Need for Leeway

Let’s say you a sign a lease for an apartment. It stipulates that you are not to paint without explicit permission. But your dog scratches the bottom of the door, so you buy a pint of matching paint and touch up the dog’s damage. You are technically in violation of the lease but no one cares. 

Let’s say you’re a client of the Gartner Group. Their latest report says “Do not photocopy” at the bottom of every page. But it’d be really helpful if at an internal meeting you could distribute copies of page 212 because there’s a complex chart on it. So you print up 12 copies and hand them out, warning the marketing guy that he’s not to send it out to the press. If Gartner were to haul you into court, the judge would lecture the Gartner lawyer for wasting the court’s time. In fact, by violating your license, you helped ensconce Gartner more firmly in your company.

You are standing on a street corner when a father takes his young daughter by the hand and jaywalks. You don’t call the cops. You don’t even lecture him about why jaywalking is bad. You don’t do nothin’.

Leeway is the only way we manage to live together: We ignore what isn’t our business. We cut one another some slack. We forgive one another when we transgress.

By bending the rules we’re not violating fairness. The equal and blind application of rules is a bureaucracy’s idea of fairness. Judiciously granting leeway is what fairness is all about. Fairness comes in dealing with the exceptions.

And there will always be exceptions because rules are imposed on an unruly reality. The analog world is continuous. It has no edges and barely has corners. Rules at best work pretty well. That’s why in the analog world we have a variety of judges, arbiters, and referees to settle issues fairly when smudgy reality outstrips clear rules.

Matters are different in the digital world. Bits are all edges. Nothing is continuous. Everything is precise. Bits are uniform so no exceptions are required, no leeway is permitted.

Which brings us to “digital rights management” which implements in code a digital view of rights. Yes, vendors and users should have a wide variety of agreements possible, but the nature of those agreements is necessarily digital. If I agree to buy the report from Gartner with no right to print, the software won’t be able to look the other way when I need print out page 212. The equivalent is not having a landlord install video cameras everywhere in your apartment. It’s having him physically remove your mom when she takes ill because your lease says you can’t have overnight guests. 

If we build software that enables us to “negotiate” usage rules with content providers, the rules can be as favorable as we’d like but their enforcement will necessarily be strict, literal and unforgiving. Binary, not human. 

Leeway with rights is how we live together. Leeway with ideas is how we progress our thinking. No leeway when it comes to rights about ideas is a bad, bad idea.

Stefan’s post builds on David’s, and is itself well worth the read.

“Spirit” versus “letter”

In the post, Stefan tells us:

Leeway doesn’t mean that there are no rules or that some people are entitled to ignore the rules, it means that at the margin it may be more important to respect the spirit of a rule than the letter.

There have been a number of comments about the Bairstow incident suggesting that, when it comes to cricket, “spirit” is invoked whenever the outcome of following the letter of the law is not to the satisfaction of the spirit-invoker. I think that does a disservice both to the spirit of cricket as well as to the letter of the law of cricket.

Take Mankading for example, the act of a bowler running out a non-striker for backing up too enthusiastically. I’ve lost count of the number of times I’ve been involved in discussions about Mankading being unsportsmanlike behaviour, somehow bruising the spirit of the game.

That’s nonsense.

Fairness

The non-striker, by taking off before the ball has been delivered, is stealing distance. He’s giving himself a shorter distance to run. Illegally.

The bowler, by running him out, is acting legally. And the non-striker is out. Legally.

This is where spirit comes in. It’s all to do with intent. Did the non-striker actually mean to steal the yard or two? It could have been a heat-of-the-moment aberration, like a false start in a sprint. And that is how, and why, the idea of a friendly warning probably came into common usage.

Leeway. Judgment. In context.

So it became normal for a bowler to warn the non-striker about his transgression, and to indicate that the next time, he may be run out. Spirit. And in time, letter.

Don Bradman, who I believe was at the crease when his partner was run out by Vinoo Mankad, felt that the non-striker was taking an unfair advantage.

It’s all to do with fairness.

Why do I write this now?

It is possible that we are going to have another right royal argument about spirit and letter and fairness very soon.

Why do I think so? As I inferred at the end of my last post, at Lords in the recent Second Test between England and Australia, the two teams pretty much won the Nobel Prize for missing the daily overs target. This, despite adjustments for wickets and weather and weirdnesses like the Just Stop Oil pitch invasion.

The two teams are up in arms about the potential penalties, since it means that both could land up with negative points from the match.

So the matter is being debated at some ICC meeting somewhere this week, and a ruling is expected shortly.

It’s a hard one. I can understand the sense of unfairness players may feel at having worked hard, creating a wonderful spectacle for all of us to enjoy, and then losing not just their match fees but also earning negative match points.

But.

The rules were clear. And established well in advance of the match. And in place for over two years, including an entire WTC cycle, 2019-2021.

The deficit was reported daily on the scoreboard, for the bowling team, and updated regularly. Neither side can say they didn’t know.

In fact, the same rules were used to pose penalties on both teams for the first Test at Edgbaston. Nobody complained then, probably because the penalties were lighter, which was because the transgressions were smaller.

I’ve seen neither the provisional judgment nor the appeal, so I’m surmising. How can the teams justify an appeal? They’re likely to argue that the modern game has changed, there’s an increased focus on winning, audiences are being treated to more entertaining fare as a result, run rates are running rampant, whatever. Yadda yadda.

The rules were clear. The rules were known. And the rules were broken.

The opportunity

However hard it seems, I think it could be a mistake to change the rules retrospectively. India, England and Australia are probably the three teams that have accrued the most penalty points since the introduction of the rules. Those three teams also probably have the majority of voting power at the ICC. Changing the rules retrospectively could have a very negative impact on the perception of fairness in the game by the other countries. That’s not something that is desirable.

There is, however, an argument that the new WTC cycle has just begun, no team has completed a series, so the retrospection only covers two completed Tests and will therefore be uniform across the whole cycle. A valid argument. But one that relies on how the teams outside the Big Three perceive the action and its fairness.

Whatever happens, there is an opportunity to improve the situation. Let’s step back for a bit. Why was the rule introduced? Because there was a worrying trend in the number of overs bowled in a day, across the whole of Test cricket. Some teams were guiltier than others of this particular transgression, and the penalty approach was seen as a way to improve performance on this front.

Now there is an argument that what matters is the result. I’m not so sure of that. I detest Pink Ball cricket, given how short those day-night Tests tend to be. (There is anecdotal evidence that it has to do with the lacquer used for making the ball pink; for sure something has to be done here).

I’ve also heard rumours and debates about reducing Tests from five to four days, and I’m not a fan of that change either. (I think the Ireland Test early in June was actually advertised as a four-day Test, and I was shocked).

Anyway, this is how I’m thinking. Leave the penalty process for unbowled overs untouched. Instead of changing that, add a few incentives. Add a positive number if the Test ends in a win for either team, with both teams getting that benefit. Add another positive number if the match only completes in the final session of the final day, win, lose or draw, again given to both teams. And add a third, larger, positive number if the match ends in a fifteenth-session tie.

A few other frills may be needed. I think it’s time for a limit on the number of times equipment can be changed, be it the ball, the glove or the bat, per session. A limit on the number of times the run-up and crease ground can be repaired, again per session. I’ve even considered the possibility of bringing in a rule that says once a team is more than x below the over rate (say 3.5), only spinners can be used until the “technical debt” is paid off.

Just some thoughts. Cricket is a wonderful game, and it’s all about fairness. That fairness is brought about by leeways that have been developed over centuries. Sometimes it becomes necessary to convert some aspect of leeway into law. Sometimes it will also become necessary to adjust, amend or even remove some laws.

As long as fairness is seen to be upheld, it will work. Otherwise it’s just not cricket.

The four types of code

Most of my working life, I’ve worked with information in one form or the other. As more and more things became “digital”, I spent time trying to understand how all this would work in society, in the real world, in a physical context. What was different and why it mattered. What I could do about it.

Some of that time was spent reading; some experimenting, seeing what worked; and some talking to people. Learning all the time.

One of the lessons I found most valuable was to think of there being four types of “code” operating in any given space: the law; software; market conventions; social norms. This “lesson” was heavily influenced by reading, listening to, and chatting with, Larry Lessig, and underpinned by the works of and conversations with Howard Rheingold, Cory Doctorow and Amy Jo Kim, amongst others.

Applying the framework to cricket

If you’re reading this, then you know I love cricket. I’ve been travelling to Test matches for over 56 years. This past month alone, I’ve been at Lords for the Ireland match; then at the Oval for India v Australia; then at Edgbaston for the first Ashes Test, and finally at Lords again for the second Ashes Test, which finished last Sunday.

We’ve had two great Ashes Tests: thrilling, packed with action and heroics, speckled with incident and controversy.

I wish I could get away with just “speckled”. My fear is that one of the incidents — the Bairstow stumping — will discolour and overshadow most of the memories people have of the Test. More worryingly, I fear that the incident could have negative ramifications that last for some time, perhaps years.

Some part of me then immediately goes into “root cause analysis” and “prevention of recurrence” mode. Why did this happen? And what can be done to prevent it happening again?

The “facts”

Let’s look at what happened.

1. Bairstow ducked under a bouncer, which was then collected by the keeper. He then went walkabout.
2. Carey, the keeper, having collected the ball, lobbed it underhand at the stumps, and hit them.
3. The Australian team appealed.
4. The on-field umpires, who had both been distracted, sent the appeal upstairs.
5. The off-field umpire decided to uphold the appeal.
6. Bairstow was given out.

Those were the “facts”.

The root causes, as seen using the four codes framework

So we’ve seen the “what”. Let’s now look at the “why”. Why did this happen?

Code as Law: The “Laws of Cricket ” view

What happened could not have happened unless Bairstow left the crease. This leads to one of the commonest observations from experts everywhere. He should not have left the crease until after he had confirmed that the ball was dead. But. Let’s pause there for a moment.

How does a batsman confirm that the ball is dead? To answer that we need to understand how a ball dies.

According to the Laws, a ball “becomes dead” when it is “finally settled in the hands of the wicketkeeper or of the bowler”, or when a boundary is scored, or the batter dismissed, or becomes trapped in clothing or equipment, or if a penalty is due related to improper re-entry or fielding, or if it hits a protective helmet on the ground, or the match ended. (Law 20.1)

In addition, a ball is “considered to be dead” “when it is clear to the bowler’s end umpire that the fielding side and both batters at the wicket have ceased to regard it as in play”. (Law 20.1.2)

To compound matters further, only the umpire can decide whether a ball is “finally settled”. (Law 20.2).

To summarise, the ball can either become dead or be considered to be dead by only one person, the umpire at the bowler’s end.

That’s what the Laws say.

So, if Bairstow had waited until the bowler’s end umpire had signalled Dead ball before meandering beyond the crease, would it all have ended happily ever after?

Possibly. Except. Except that he might have waited a very long time. Godot may have arrived first.

It turns out that the umpire does not have to signal Dead ball. To quote the relevant Law, “When the ball has become dead under 20.1, the umpire may call and signal Dead ball if it is necessary to inform the players”. (Law 20.4.1).

May. If it is necessary.

Now umpires don’t call or signal Dead Ball every ball. For good reason. It would be tiresome. So it only tends to happen when there is doubt. This is not a Laws thing, it is a market convention thing. Informal agreements that make the world work, probably an arcane example of Erik Brynjolfsson’s Theory of Incomplete Contracts.

Was there doubt? Hmmm. The bowler’s end umpire was busy completing his cloakroom-attendant duties, preparing to return Cameron Green’s cap to him. At the point when the ball hit the stumps, his mind appeared to be otherwise occupied. The leg umpire was also busy doing something else, as he prepared to start his next stint at the stumps; he was walking over from square leg. So neither appeared to be watching the action. They were going about their business preparing for the next over. Soft signals that suggest the two of them had already considered the ball to be dead. Yet neither of them said so. Perhaps they weren’t asked. Perhaps what they were asked was a different question, whether “over” had been called.

The argument of whether “over” was called is neither here nor there, other than the fact that it would have proved conclusively whether the ball was dead or not.

The calling of “over” cannot be done until after the ball has been deemed dead. (Law 20.3). “Nether the call of Over (see Law 17.4) nor the call of Time (see Law 12.2) is to be made until the ball is dead, either under 20.1 or 20.4.”

It is theoretically possible that Bairstow had assumed the ball was dead due to the actions of one or both umpires, even though Dead ball had not been signalled and Over not called. But that’s a stretch. Bairstow had been seen going walkabout earlier, he appeared to have a habit of wandering, particularly after ducking. So let’s leave that as a stretch.

It’s not important. Because what Bairstow believed does not have any impact on the umpiring decision, which only had to do with two things. Was the ball dead? And if not, was Bairstow in his crease or not when the ball hit the stumps?

That’s where I hit my first problem. I can’t find evidence that the bowler’s end umpire was actually asked whether he thought the ball was dead. Since it was technically a stumping attempt, the referring umpire should be the square leg umpire, not the bowler’s end umpire. The only person who could have deemed the ball dead may not have been asked the question. I believe Ben Stokes actually raised this point later, but I haven’t seen a formal response.

That’s enough as far as the Laws are concerned. If the ball wasn’t dead (a point that is still in question) and if the batsman wasn’t in his crease when the ball hit the stumps, then the batsman is out. That’s what the off-field umpire ruled. To most people, that’s the end of the argument. And maybe it is. Maybe the off-field umpire did ask the bowler’s end umpire whether he deemed the ball dead and was told he hadn’t. I just haven’t got the info.

Code in the form of software: The replay of the event

There’s nothing much to comment on here. The frames showed what they showed. At the point where the ball hit the stumps, Bairstow was out of his crease. The TV umpire decided that the ball was not dead, and that was that.

A few nuggets of interest remain and are captured in the frames. Did Carey launch the ball at the stumps before Bairstow went wandering? Not an issue from a Laws viewpoint, but potentially of interest from a Market Conventions viewpoint. Was the Bowler’s End umpire in the act of preparing to hand over the bowler’s cap before the stumps were shattered? Could it be surmised that the ball had been considered dead as a consequence. Possibly of esoteric interest from a Laws viewpoint, as and when people work on improving the Dead ball Laws. Until then, it’s interesting but not of much use. The result cannot be changed.

Code in the form of Market Conventions

Let’s assume that issues of Law have been settled, even though they haven’t. Let’s now move to the Letter-versus-Spirit argument.

Most of the debates I’ve seen that go down this route descend into pure whataboutery: what about what Bairstow did earlier in the match, what about Dhoni and Bell, what about Stuart Broad’s snick, and so on. I’m going to avoid those rabbit holes, won’t solve anything.

There are already Laws in place about unfair play and player’s conduct, as detailed in Laws 41 and 42. They deal with unfair actions, threatening behaviour damage to pitch or ball, etc. Some of them lead to outright penalties, and others lead to warnings that precede penalties if the offending behaviour continues.

Where the Law was found to be inadequate, informal conventions arose. For example, it was expected that a bowler would warn a backing-up batsman, one who was too enthusiastic in his backing-up, before “Mankading” him. (I dislike the use of the term Mankad to describe this but have to use it nevertheless). Now, with recent changes to the relevant Law, this may no longer be necessary. Onus has shifted to the umpire, if I’ve interpreted the change correctly. (Law 41.16, Batters Stealing a Run).

This makes sense. Sometimes the Law needs work. And convention evolves to fill the gap until the Law changes.

The Dead ball Laws need a serious upgrade. Conventions have evolved: the informal eye-check between batsman and wicketkeeper; the tap of the bat in the crease before the walkabout; soft signals that evolve to fill gaps. But these signals are still informal conventions, and need validation and formalisation.

The warning mechanisms have hitherto worked on the basis of “mate, you’re taking advantage, and it’s not fair. It goes against the spirit of the game. Please stop or we will take the action available to us, however unsporting it may feel to you”.

It has been reported that the Australians had noticed Bairstow’s behaviour. If, as has been suggested, Carey took a shy at the stumps before Bairstow had left his ground, on the off chance that he repeats his prior behaviour, it suggests premeditation. Premeditation without a warning. That becomes a market convention issue. Are we now to see a Mankading in one of the remaining Tests, one carried out without warning? No! Why? Because the Laws have caught up.

Code in the form of Social Norms

We’ve now had a situation where:

A bowler has bowled a ball which had no possibility of hitting the stumps

The ball sailed over the batsman’s head and was easily ducked

The batsman facing the ball made no attempt to play it

The two primary actions of cricket are missing from the episode

And yet a wicket has fallen.

To make matters worse, there is no argument to be made that Bairstow somehow gained an unfair advantage by doing what he did. (At least with Mankading, or with Running On The Pitch, the perpetrator was seen to seek unfair advantage. No such accusation can be made against Bairstow).

Not cricket.

That’s why there was uproar. No one is arguing about the decision per se, even though some elements of doubt remain. The umpire ruled him out. And that is that. Game over. The argument is whether the appeal should have been rescinded. And that’s really about market conventions and social norms. Spirit rather than letter.

It is possible that the event led to some extra fire in Stokes’s belly, and we could all enjoy the fireworks.

It is possible that even if the appeal was rescinded, the Australians would have won the match. There was still work to do, and England’s performance so far has been patchy.

It is also possible that the event motivates the whole England team, and that we have an even more exciting Ashes series in prospect.

Time will tell.

What is important is that the rancour does not spread, that we don’t see the scenes we saw in the Long Room or in parts of the ground outside the stadium.

Prevention of Recurrence

As John Mayall famously sang all those years ago, The Laws Must Change. (Incidentally, if you’ve never heard this before, I envy you. What a treat you have in store for you. I can say that about the whole album. The Turning Point is one of the finest albums ever made. California. Thoughts About Roxanne. Room To Move. A triplet as worthy as the Deadhead favourite of Help Slip Franks).

Time to improve the Laws related to Dead ball. The umpire’s job is hard enough without all this ambiguity.

When doing the improving, it’s important to come up with mechanisms that don’t slow the game down. Over rates are already appalling, and only made worse by the new entrants such as ball-change shenanigans. I don’t think that the penalties for the second Test have been issued yet, but I was at the game. I think England closed at -8 and Australia at -8 or 9. There were only 12 points at stake in the first place, so the net take from that Test between the two teams might turn out to be negative. (Unless there is some miraculous way to zeroise the under bowled overs since there was a result).

Introduction

This is a post about cricket. A post about Test cricket. A boring, mind-numbing post about some of the numbers in Test cricket. Some very particular and arcane numbers. You have been warned.

15 March 1877. That’s the day when the first officially recognised Test match began, between England and Australia, at the Melbourne Cricket Ground (MCG).

3 January 2022. 52,889 days later. Today. Tests #2444 and #2445 are in flight, the first between New Zealand and Bangladesh at Mount Maunganui, the other between South Africa and India at Johannesburg.

Each Test consists of up to four team innings, up to two per team. Each team innings consists of up to 11 completed batsman’s innings (although at least one batsman per innings would remain “not out” at the end of the innings. So each Test match is capable of generating up to 44 completed batsman’s innings.

I haven’t counted all the completed innings so far; my guess would be that there have been over 90,000 such innings so far, perhaps as many as 100,000. (Here, I speak solely about men’s cricket, and solely about the Test variety. While more attention is being paid to the women’s game, I have yet to find good sources of in-depth statistical information that I can use. I’m sure this will be rectified in the next few years. It’s been a long time coming).

The bingo board

Over those 2445 Tests (*and counting), the highest score ever made by a batsman in a single completed innings has been adjusted regularly, as shown below:

During the first Test, at the MCG in 1877, Charles Bannerman, an English-born Australian, scored 165 and set a high bar for everyone else to follow. That record would stand for another seven years, until Billy Murdoch, another Australian, scored 211 in the 16th Test, at the Oval, in August 1884.

Murdoch’s 211 remained the target to beat till 1903, when RE “Tip” Foster became the first Englishman to hold the title. He scored 287 that year, away to the Australians at the SCG. (Incidentally, Tip Foster apparently remains the only person to have captained England at both cricket as well as football).

Foster held the record for another 27 years, till Andy Sandham, another Englishman, went past the magical 300 mark at the 4th Test versus the West Indies in 1930, in Kingston. He scored 325.

Sandham wasn’t to wear the crown for long; barely three months later, Don Bradman whipped it off him, returning control to the Aussies, with his 334 at Leeds. Less than three years later, Wally Hammond reclaimed it for England, scoring 336 versus New Zealand in Auckland.

It stayed with England for a while after that. Len Hutton’s 364 versus Australia at the Oval in 1938 raised the bar again. It would stay that way till well after the Second World War. Hutton’s 364 was only the second time the high score was set by a batsman playing at home. The only previous instance was Bannerman’s 165, English-born, playing in his adopted home.

The first seven holders of the “highest Test score” record were either English (4) or Australian (3). The eighth holder, West Indian maestro Gary Sobers, became the first outside those two countries, with his 365 extending Hutton’s record, which had stood nearly 20 years, by just one run. While it was another “home” achievement, what stood out was that it was Sobers’ maiden century: the 21-year old had never scored a century before, much less a triple.

36 years later, the title moved again, but stayed with the West Indies. Brian Lara became the ninth holder of the High Score record, scoring 375 versus England, at home, at St John’s, Antigua in 1994. Nearly a decade later, Matthew Hayden briefly usurped Lara with a 380 versus Zimbabwe in Perth, to become the 10th holder of the title. Barely six months later, in 2004, the crown settled again on Brian Lara’s head, as he scored 400, again at home, at St John’s, Antigua.

And that’s the story of the men’s cricket single-completed-innings high score trail. Since Bannerman’s 165 in the first ever Test, the sequence has been 211-287-325-334-336-364-365-375-380-400. Eleven stops in all. Just ten holders, including the incumbent, who has held it twice. And no movement for nearly 18 years.

So why 229?

I mentioned a bingo board. If you’ve read this far, you will know that the highest score was slowly yet inexorably taken from 0 (the lowest possible score) to 400, over the 2445 intervening Tests, two of which are still in play as I write.

Now think of all the individual scores that have been achieved in that time. Thousands of Tests, probably a hundred thousand completed innings. We can create a “bingo card” with all those scores: we know that the smallest number is 0 and that the largest is 400. Every time a batsman completes an innings, we scratch that number off. What’s the lowest unscratched number? As of today, that’s 229. And it’s been 229 since Thursday 2nd January 2003, nineteen years yesterday, when Herschelle Gibbs scored 228 versus Pakistan at Cape Town.

Development of the lowest unscratched number

I’m nuts about cricket; those of you who know me would have expected me to do this, to plot the movement of the lowest unscratched number, the history and future of 229. So here it is.

Surprisingly, after just one Test, the lowest unscratched number was 14. By the end of the second Test, it had moved to 16, after Harry Charlwood removed 14 from play. The fourth Test took it further to 25, as George Bonnor scored 16; the next Test, the fifth, moved the needle to 29, courtesy Jack Blackham’s 25.

We then had a brief gap until the 9th Test, which set the bar at 44, when Billy Murdoch (Mr 211 himself), rubbed out 29. That wouldn’t budge till the 24th Test, when 44 was chalked off by Arthur Shrewsbury and replaced by 46. Seven Tests later, 46 became 60, with Bobby Abel’s help. That would last another 10 Tests: the 41st Test saw the record shift to 65: Alec Bannerman saw to that.

Some frenetic activity ensued. Seven Tests later, 65 became 71, thanks to Arthur Hill; after another six Tests, 71 became 76, with Syd Gregory’s help; eight matches later, 76 became 78, in the 62nd Test, courtesy Jack Worrall.

The glacial movements that characterised the later years then become visible. 78 became 110 forty-four Tests later, with Aubrey Faulkner’s assistance; moved to 125 fifty-two matches later, with Bill Ponsford scratching 110; swept to 139 one hundred and thirteen Tests later, with Pieter van der Bijl contributing the 125; and then jumped to 171, taking two hundred and forty-six Tests to make that move, as a result of Everton Weekes’ classic 139.

If you thought that was slow, you’re not prepared for what followed. The bar moved to 186 during the 675th Test, thanks to Ian Redpath, a whole two hundred and seventy-one Tests later; then on to 199 with a gap of two hundred and sixty-seven Tests, with considerable help from Zaheer Abbas.

That brings us to relatively modern times. 218 was set by Sanjay Manjrekar in the 1130th Test, between Pakistan and India, at Lahore in 1989, moving the target to 224. That was then taken by Vinod Kambli in 1993, versus England in Mumbai. The spotlight then moves to Herschelle Gibbs, whose 228 I mentioned earlier.

And we arrived at 229 being the number to beat.

That was Test 1637. Over eight hundred Tests ago. Over nineteen years ago.

A gentle walk from 14 to 16, then on to 25, 29, 44, 46, 60, 65, 71, 76, 78, 110, 125, 139, 171, 186, 199, 218, 224, 228 to bring us to today and 229.

So what’s next after 229?

There are still 18 unscratched numbers in the 200s, starting with 229. They are: 229 252 265 272 273 276 279 282 283 284 286 288 289 292 295 296 297 and 298.

7 numbers in the 200s were taken out in the last decade alone, so things are still moving along. Alastair Cook’s 294; Ross Taylor’s 290; Adam Voges’ 269; Tom Latham’s 264, Alastair Cook (again) with his 263; Shoaib Malik’s 245; and, most recently, Kane Williamson’s 238.

Pickings in the 300s are somewhat richer, with 76 numbers remaining untouched.

There’s some progress even there: David Warner, with his 335, and Karan Nair, with his 303, took two numbers off the table.

So we now have 94 numbers left in the range 0-400. The lower end of that range will not change unless someone comes up with the daft idea of negative numbers for batsmen. (Given the shenanigans I see in so much of sport, I wouldn’t rule it out, but I live in hope that it doesn’t happen in my lifetime).

The upper end of that range has remained steady for eighteen years, since Lara’s 400.

229 has been around longer.

The last number to be scratched on the bingo card was Kane Williamson’s 238. Exactly a year ago.

This year, for sure I expect to see some of the 94 numbers taken out. But will it be a bumper year? Will we see the 229 moved all the way to 252? Will 252 itself still be around by then? And will 400 remain the upper limit? All with or without the madness of Umpire’s Call, DRS and whatever passes for the technologies in use at the time.

We shall see.

Introduction

I can’t believe it’s been nearly forty years since I first watched one of my all-time favourite films, Local Hero. if you haven’t watched it as yet, I envy you. You’re in for a treat. (If you don’t feel like watching the film, just listen to the soundtrack).

I’m not going to say anything about the plot just in case. The excerpt from the script is pretty innocuous as excerpts go. One line stands out for me. It consists of just three words.

It’s never locked.

Memory’s a strange thing. I could have sworn that the actual phrase used in the film was “We don’t lock doors here”. But this is from the official script so I’ll live with it.

I was born in what was then the family home in Lower Circular Road, Sealdah, Calcutta, over 64 years ago. When I was a little over two, we moved to Hindustan Park, and spent nearly a decade there. Shortly before my twelfth birthday, we moved to Moira Street, where I spent the rest of my life in Calcutta, until I left for the UK in November 1980.

I can still remember the names of at least of the immediate neighbouring families in Lower Circular Road. We visited that house many times in the years since we left there, it was in the family till the early 1980s.

I think I can still remember the names of all my neighbours in the apartment block in Hindustan Park, and have visited the building at least once since leaving India.

When it comes to Moira Street, that was a whole different ball game. Ten flats. One of them permanently rented by a business during the time we were there. The rest of the building? All families.

The centre of my universe for many years

In the four decades since I left there, I must have visited that building half a dozen times: a couple of times in the 80s, once in the 90s, three times this century. Pretty much every time I went to Calcutta.

Why am I telling you this? Not because I’ve decided to go on yet another nostalgia trip, something I’m quite happy to do regularly.

But because it was a place where we didn’t really lock doors. At least three families there grew up almost as one, with a few others almost-as-integrated. (You know who you are, and I remain forever grateful for the times we had together). We flitted from apartment to apartment at will, through open front doors, up the fire escape and through the back doors, occasionally climbing in through open balconies, and, very very rarely, using well-honed techniques to barge through doors that were apparently shut.

In the main, the doors weren’t shut. We had liberty withal, as large a charter as the wind, to blow on whom [we] [pleased].

Covenant relationships and mutualisation

Even in those days, Calcutta was a very crowded city, and there was physical and pragmatic security-through-transparency. Village-like, everyone knew everyone within their ????? (para) or neighbourhood. People looked out for each other, strangers were noticed straightaway, and usually challenged, in a typically blunt-curious way. The community looked after its own.

Neighbours, particularly children, growing up together, thick as thieves, fast friends forever. Until they grow up a little more. And flap their wings.

That’s the normal plot line.

But it’s not what happened here. The extended family I speak of is in at least four continents, and are still in touch with each other, fifty years after we came together. Social media does have its occasional uses, and has simplified access and connectivity.

The experiences we had (and continue to have) are good examples of covenant relationships, rather than the ersatz contract versions. In a contract relationship, when something goes wrong, when there’s a “breach”, the key question is “who pays?”, closely followed by the how much and the when. In a covenant relationship, the key question is fundamentally different. “How shall we fix it?”. And the critical word is “we”.

The society I grew up in was founded on those principles of mutualisation. The relationships we had were based on a bedrock of shared experiences and an immutable trust. It permeated my family, my friends, my schoolmates, my community-at-large.

The communities were relatively small, Dunbar-like in their compactness, a tight core of 20 or 30 in an extended network of around 150. We had address books, the telephone numbers were there as well, but we didn’t really have to use the address books or directories as a rule. We knew where everyone lived, and, where relevant, knew the phone numbers off by heart. Directories and address books were for “outstation” data.

Phone calls were for people who lived far away, not for people who lived in the same building. And if someone couldn’t be found, the wireless powerless megaphone would come into use: a yell from the balcony, which would then get relayed in some form or the other to wherever the child or children in question were playing, hiding, fighting, whatever. Audible smoke signals transmitted at speed without using up any scarce natural resources.

During my time there, the “building” saw the usual coming-of-age and rites of passage: births, marriages, deaths, “sacred thread” ceremonies, everything. The community came together as one, an integrated family in action. My father died while we still lived there, just over 40 years ago, and I saw that community-in-action “up close and personal”, covering my family in covenant protection in ways I can never forget, ways I will never forget.

Covenant relationships in an environment of mutual trust and respect.

Satisficers and maximisers

Ever since Herbert Simon coined the term “to satisfice”, there’s been a lot of good work done by great people on satisficers, and I’m not even going to try to summarise any of it here. Suffice it to say that a satisficer looks for “good enough” while a maximiser or even an optimiser looks for “the best”. (In doing that, I can be accused of choosing a “good enough” definition rather than “the best” one. So be it).

Years later, in the early 2000s, I was reading a book by Barry Schwartz, called The Paradox of Choice. One of the most interesting hypotheses put up in that book was the suggestion that satisficers were somehow happier than maximisers or optimisers. Over the years, I’ve also seen research and arguments that suggest there is a genetic predisposition to be one or the other, but I’m not yet convinced of that argument. I do like the “happier” suggestion though.

I’ve also come across papers that put forward an intriguing argument that looks at the “transaction costs” of satisficing and maximising and avers that satisficers are happy because they trade some cognitive workload in making their choices. And that gives me a feel for “happier because….” More to think about.

I like that idea. A part of me really likes that idea, because it gives me the chance to link to Janis Joplin singing Me and Bobby McGee. Amazing singer, amazing song.

Freedom’s just another word for nothing left to lose

While reading Schwartz’s book, I couldn’t help having the Kristofferson-penned words ear worming their way though my head. Except they were ever-so-slightly malformed in the process. What I keep hearing was “freedom’s just another word for nothing left to choose”, a real Marmite statement. Everyone wants to have choice. Until they don’t. And only in places they don’t. In the wrong context, choice can be a tyranny.

When I came to the UK, I had never seen a supermarket, much less been in one. I had no idea what to do when presented with such immense choice for a commodity. (My first visit to a supermarket, I needed toothpaste and came back with some sort of denture cream, grabbing the first thing I saw, dashing for the till and then hurrying out. I wasn’t comfortable there).

In Calcutta, if we ran out of toothpaste in one of the bathrooms at home, one of us would go round the corner to S Stores (which I’m delighted to see is still there, exactly where it was in November 1969). We’d go there and ask for toothpaste. And be given toothpaste. Not a snake. Nor a stone. Toothpaste.

Levels of sophistication did emerge, sometimes one of us would want the “new” go-faster stripes Signal. Or maybe a tip towards the swadeshi, Vajradanti. Most of the time, we stayed uncomplicated. Toothpaste. Thanks.

The conversation was quick and simple, across a single counter, with someone you knew and who knew you. A relationship of trust, a relationship built over time, a covenant relationship. If there were qualifying questions to be had, they would get asked simply and unthreateningly. The shopkeeper was avuncular, and the uncle-ness was part of the relationship. Advice and counsel was part of the deal, and there was no question of being ripped off.

The customers who shopped there were proper customers: they gave the shop their custom. As part of a relationship founded on trust and enriched over time.

Yes, I’ve been a satisficer all my life. I like giving people my custom rather than continually shopping around; maybe I’m trading some cognitive workload in the process, maybe I’m genetically predisposed to do so. I don’t know. What I do know is that I feel good about it.

There is something about the Cheers mentality that I like:

sometimes you want to go

where everybody knows your name

and they’re always glad you came

As the saying goes, people buy from people, people sell to people. It’s all about people and relationships in relatively small communities. Global villages are good, but let’s remember the village part and not just the global part.

Good neighbours

A few days ago, a neighbour noticed something untoward in his garden, told me about it, and helped me answer something that had been bugging me. Why had a particular garden chair been moved from its usual place?

Today, as I walked back from the newsagent with my Sunday paper, I bumped into three different neighbours, and had the chance to speak with each of them. Yes, of course we exchanged pleasantries, wished each other top of the morning and the season and the year (as you do), and then did what everyone in the UK does nowadays. Instead of talking about the weather, we spoke of Covid-19 and masks and boosters. Social objects by another name, I guess.

We knew each other. We knew where everyone lived. We’ve covered for each other, taking parcels in, joining neighbourhood protests, noticing when alarms go off, seeing how we could help if the other(s) needed help. A loose, gentle relationship, but a relationship all the same, more covenant than contract. In no way transactional.

The Calcuttan in me was used to knowing my neighbours. When I left India and migrated to England, I lived in a part of Liverpool for a while. And everyone knew everyone there. And said hello. And smiled. And asked after each other. Civil. Polite. Helpful.

Then I came to London. A very big place, and I have no wish to generalise. But the part of London I was in, everyone was in a hurry, all the time. Disappearing from home while it’s still dark, returning only when it’s dark again. Surfacing occasionally at weekends. And in between, never meeting eyes, never speaking, never engaging. Heaven forfend.

I moved around until I found somewhere in London that still felt like a village, where people did speak to each other, where it wasn’t a crime to be on first-name terms with neighbours. Where everyone wasn’t always on a Danny Kaye schedule.

Phenomena like Nextdoor intrigue me, though I still don’t actually use it. I’m still wrestling with centralised-versus-decentralised, and, in general, tend to find that people conflate distributed with decentralised, and that isn’t necessarily a good thing. I love the neighbour-to-neighbour connect, I love many of the functions Nextdoor appears to have, but I still hum and haw about connections to businesses beyond the neighbourhood; I need to know more before I can truly engage.

Startups like Tredish fascinate me, enough for me to invest in them and to get involved in whatever way I can.

The neighbourhood is a unit of community that makes a lot of sense to me; and while I can see immense value in the ability to discover and share patterns between neighbourhoods, I still think of each neighbourhood as unique and vibrant and alive and deserving of its uniqueness.

So where is all this leading?

If you’ve visited here before you know what to expect.

I’m largely retired now, and have the privilege and luxury of time to observe life around. Time to reflect on what I observe. Time to research whatever I find interesting or at least intriguing.

One of those things is growth, particularly unfettered growth. I believe less and less in the constant need for growth. I spend some of my time learning about what the right minima and maxima should be for a given measure in a given context. Wherever possible, I look to nature as a teacher in this quest, looking particularly at complex adaptive systems.

One of those things is trust. I find the term “trustless” odd, odd enough to discomfort me. So I pull a thread here, remove a scab there, gnaw away at it, trying to understand why such a term should exist.

I have this hunch that these two intriguing things, the quest for unfettered growth and the development of “trustless” systems, that these two are related. So I pull and tear, gnaw and chew. Until I learn more. And then start all over again, refining my start position.

Conclusion

I’m increasingly drawn to a world where the “conquest” of distance and time means less to me than it did when I first heard it. I’m similarly keener on worlds where people know their neighbours and operate in a climate of mutual trust and respect, where communities are compact and coherent. I think ideas of continuous growth and scale need to be re-examined constantly, particularly as we learn to price social and environmental costs in ways we haven’t been able to do before. In much of this, technology can and will play a role; but there are bound to be missteps on the way.

The missteps that intrigue, nay concern, me the most are those where communal trust is weakened. I think the neighbourhood is a fascinating level of aggregation to look at; maybe I should look further into Geoffrey West‘s work on cities and try and tease out the neighbourhood implications, going wherever that leads me.

I want to live in a world where it’s not necessary to lock doors. I want to live in a world where we learn how to do this — again —in a way that we can preserve the valuable bits of technological advancement while filtering out the dross, the detritus. Keeping baby. Throwing out bath water.

I write to elicit and excite conversations about these things. Not with thousands of people. Not even hundreds. Just enough to help me learn that little bit more.