Eng v NZ, Edgbaston 2021: Soft signals, hard questions

The second in a series of first-person reflections on experiencing cricket in 2021.

The difference between in-person and on-screen viewing was vividly brought home by England’s second outing of the summer.

At the most basic level, this applied to attending or watching from home. After a year of enforced watching from home, tempting emails had been sent and received. “We are the Fortress. We are Edgbaston. And Test cricket is back in Birmingham next week…Around 18,000 spectators can attend each day of the England v New Zealand Test Match as part of the government’s Events Research Programme (ERP) – and you can join them. […] Prices starting from just £20 [for Days 4 and 5].” £20 for the chance to mingle again – no social distancing! – with 18000 others, drinking beer and singing – of a sort – in the Hollies stand. As the chants rang out, it reinforced the conviction that this was how live sport was meant to be experienced.

It was the incident on the second day, however, that threw up another old difference in viewing: the contrast between the on-field umpires’ live view and the TV umpire’s screen-mediated view.

The scenario was similar and not unfamiliar. Devon Conway nicked Stuart Broad to third slip. Zak Crawley appeared to take a diving catch, but it was not clear whether it was clean. The umpires conferred and decided to send it upstairs, with a soft signal of “not out”. The TV umpire, on replay, decided the ball had clearly touched the ground and returned a decision of not out.

Broad, it may come as a surprise to those only accustomed to his on-field enthusiasm, often speaks well, thoughtfully, and perceptively off the field. In a Sky interview, he declared later: “if the umpires are unsure, let’s go through the amazing technology we’ve got and get the right decision.” He also claimed that “when you calmly look at the positives, the pros of the soft signal and the cons of the soft signal, the cons completely outweigh the pros.”

Unfortunately for his case, a calm consideration reveals exactly the opposite to be true. Nick Friend’s article for thecricketer.com explains well why scrapping the soft signal would not work in the way Broad presumably imagines. Put simply, it would do nothing to help progress a situation where the footage is unclear: such scenarios would still require a (potentially controversial) judgment call having to be made. To that, this point could also be added: the soft signal increases the chance of the “right”* decision for the reason that the soft signal ensures that the input of the officials closest to the action is factored in.

To remove the soft signal would be to reduce the amount of relevant information considered. Thanks to perspective compression, that will almost certainly reduce the number of catches that are given. I doubt this is what Broad would be hoping for.

Jason Holder’s claim that the soft signal is being allowed to “cloud” the game is also misdirected in its ire. The cloudiness comes from situations arising that are difficult to judge, whether in real-time or when replayed. The soft signal does not increase the cloudiness, or more formally the uncertainty, but reduces it. The umpires are making their best judgment and providing their opinion. They shouldn’t be criticised for doing so but neither is there any need to pity them or suggest that they need to be relieved of this responsibility.

Though it was welcome to hear sympathetic expressions for the umpires from Broad, it was hard to escape the impression that his criticism of soft signals was directly linked to the decision not falling his way on this occasion. It seems that he has, incorrectly, equated “what TV pictures seem to show” with “what actually happened” (in his view). (Ironically, judging by the vocalised deliberations of the TV umpire, the verdict would have been not out, even if no soft signal had been given.) The broader issue (no pun intended) is that we tend to give too much weight to technology and insufficient weight to the judgment of human umpires.

Yet technology has tended to confirm that human umpires are extremely good at their jobs. Recall that by the time they have graduated to Test level, they will likely have adjudicated on hundreds of catches. Yes, their ground level view may be 40 yards away from the action, but the nearest camera may be 70 yards away. We shouldn’t underestimate the power of the human eye and the human brain to make a reasonable judgment.

You could also view this in a very reductive sense, if one were to classify the human brain as a computer: removing the umpires’ opinions would then be a case of removing two highly trained neural networks from the task. (However, if one were to assemble a silicon-based neural network making similar judgment, I predict that a high number of people would happily accept it on being told that it was “AI” and therefore supposedly more accurate/impartial/perceptive.)

Broad may or may not know that an alteration to the soft signal protocol has already been mooted by the MCC World Cricket Committee, designed to prevent the umpires from having to offer an opinion where they are both genuinely unsighted, for instance for a boundary catch.

In addition, there is one modification to the soft-signal protocol that would be worth considering, in the light of this incident. Broad’s claim that the TV umpire’s hands are “tied” by the soft signal is a gross exaggeration: ICC playing conditions state that the soft signal is only used if the TV umpire advises that “the replay evidence is inconclusive”. However, there is a grain of sense in that claim: as was suggested on social media, might the TV umpire be subsconciously influenced by his colleagues’ soft signal? To counteract that possibility, why not keep the on-field umpires’ decision initially hidden? The protocol would then look like this:

  1. On-field umpires agree that they are not sure.
  2. On-field umpires note their “soft” decision but do not reveal it.
  3. On-field umpires ask the TV umpire to review the catch.
  4. TV umpire reviews the catch and returns one of three verdicts:
    1. Conclusively Out.
    2. Conclusively Not out.
    3. Inconclusive.
  5. If the TV umpire returns a verdict of Conclusively Out or Conclusively Not Out, that decision is used.
  6. If the TV umpire returns Inconclusive, the on-field umpires’ “soft” decision is revealed and used.

This would maintain the primacy of the on-field umpires while making full use of the TV umpire’s facilities.

We should be grateful that in Test cricket, we don’t have to rely solely on the on-field umpires’ snap reaction, nor solely on the TV umpire’s protracted deliberations – both of which are shaped, for better and for worse, by the manner in which they perceive the action, and the method in which it is presented to them. Rather, we have a protocol that makes use of the technology when neeeded (which also, let’s not forget, requires a judgment call by a human umpire).

As is the case with simply viewing a match as an onlooker: sometimes TV is best; sometimes in-person is best. The game’s richer for both options.

*”Right decision” is left in quotation marks because, arguably, there is no such thing as a definitively “right” decision; at some level a degree of opinion will always come into play, even with a fully-automated technological solution – but that’s a subject for another time.

Cricket World Cup 2019: Week 1, update #2—Tamper tantrums

The second of an informal weekly (p)review of the 2019 Cricket World Cup – or, as I like to call it, the #Champ10nsTrophy.

Truth be told, #CWC19 needed the England vs Pakistan result. After a couple of one-sided shellackings (expect me to return to that at a future date as part of yet another criticism of the 10-team format) a high-scoring game that went down to the wire and, most importantly, saw the favourites finish in second place, was needed to kickstart the competition.

England shouldn’t worry too much. Neither of the previous two hosts to win the World Cup managed a clean sweep: India lost to South Africa in 2011, and Australia lost to New Zealand in 2015. Neither had too much trouble in lifting the trophy a few matches later.

What it didn’t need was a ball-tampering controversy. Maybe “ball-tampering” and “controversy” are too strong a pair of terms: a couple of English batsmen were noticeably reticent – albeit also noticeably exercised – when given the opportunity to comment on the subject, thereby damping down the flames.

At Trent Bridge the umpires were quick to jump on the signs of proto-tampering and nip them in the bud. To be specific, both sides were warned against throwing the ball in on more than one bounce to the keeper, a well-worn tactic (geddit?) for roughing up one side. Whether they were consistent or not, or whether both teams complied with the direction equally, is immaterial to this discussion.

It strikes me as excessive to prevent teams from engaging in legitimate activity that has an incidental and beneficial side effect. The key point is that throwing the ball in on the bounce is justifiable for non-tampering reasons: the flatter trajectory is faster. Perhaps two or three bounces – which is what the umpires were taking exception to – is pushing it; is there a valid reason, other than an attempt to change the condition of the ball, for such throwing? Could one not argue that it saves time and reduces the sting on the bowler’s receiving hands?

Obviously this all seems to be overlooked when a run out is on the cards: I’ve yet to hear of umpires pulling up fielders who’ve effected outfield dismissals via a bounced throw. Yet even when not aiming for a run out, the the fielding side has a right, nay, a responsibility to save time where possible – as Bumble would say, to get on with the game. (There’s another #CWC19 disappointment: why’s Bumble missing from the commentary box?)

That’s not to say the umpires acted incorrectly. Law 41.3.2, the Law that prohibits ball-tampering, reveals a double standard: one for batsmen and one for everyone else.

41.3.2 It is an offence for any player to take any action which changes the condition of the ball.
Except in carrying out his/her normal duties, a batsman is not allowed to wilfully damage the ball. […]


How, pray, can one interact with the ball in any way that does not change its condition? Bowling anything other than full tosses will cause it to land and rough up. Catching it will transfer sweat or other substances onto its surface. Batsmen, however, are given special dispensation to wilfully damage the ball in carrying out their normal duties. Why are bowlers and fielders not similarly allowed to effect damage in the course of their normal duties?

Wasn’t the game so much more interesting when Hassan Ali found a bit of inswing? In fact, forget England losing: so far, the biggest surprise of the World Cup has been the lesser spotted Kookaburra swing. Even up the playing field, I say, and let bowlers damage the ball incidentally, if not accidentally.

Square leg scorer

3.1 Appointment of scorers

Two scorers shall be appointed to record all runs scored, all wickets taken and, where appropriate, number of overs bowled.

3.2 Correctness of scores

The scorers shall frequently check to ensure that their records agree.  They shall agree with the umpires, at least at every interval, other than drinks intervals, and at the conclusion of the match, the runs scored, the wickets that have fallen and, where appropriate, the number of overs bowled.  See Law 2.15 (Correctness of scores).


Marylebone Cricket Club, 2017, Law 3 – The Scorers

Law 3 is not a controversial Law. It makes sense. Two scorers are better than one: they can check with each other, watch out for each other, and even cover for each other (in extremis).

Law 3 is also, probably, the most flouted law in the book, particularly in the recreational game. How often do captains agree to “just copy the book” at the end of the game and get away with one scorer, drawn from whoever isn’t batting at the time? Chances are that scorer hasn’t had official training, either. It’s not in the least surprising that scoring errors plague lower (and sometimes the not such lower) levels of cricket.

One of the more unexpected discoveries from my excursion to play cricket in Quebec (more on that another time) was the system that a social cricket club had developed to address the frequent scoring discrepancies that threatened to derail games. They had, in effect, re-invented linear scoring. The umpire was given a clipboard and blank bowling record. Provided he/she filled it in, the teams would then have two records – one by the umpire, one by the book scorer, and errors could be detected. Most importantly, the final score could be agreed without undue wrangling.

It’s such a simple idea that it must already be in use elsewhere. Nevertheless, I liked this idea so much that I’ve adapted the core idea for use in my club, Haymakers CC, during its social fixtures. As a strategy in casual cricket, it has a lot going for it. Appoint one bowler’s end umpire (i.e. somebody who knows the LBW law), and one striker’s end umpire. Give the latter the job of maintaining this A6 size bowling record. The advantages are myriad:

  • It allows every over, and indeed the entire innings, more or less, to be reconstructed.
  • It gives the fielding side access to the score on the field when the scoreboard inevitably falls behind.
  • It makes the role of the striker’s end umpire a bit more interesting.
  • It means the striker’s end umpire can’t avoid keeping count of the number of balls in the over.
  • It starts to introduce players who fulfill this role to the scoring symbols.
  • It finally satisfies Law 3.1 (two scorers are thereby appointed) and Law 3.2 (runs, wickets and overs are recorded and can be meaningfully cross-checked).
  • It relieves the bowler’s end umpire from keeping score (some do, some don’t).

The downside is, of course, that it does make the job of striker’s end umpire more involved. Yet it will be worth it for the time and aggravation it’ll save down the line. Whole overs won’t go missing (not that that’s ever happened – ahem).

Here is version 1 of my Square Leg Scoring record. It includes:

  • A row for each over (maximum of 30)
  • A box for each delivery
  • A space to record an identifier for the bowler (whether name, number, initials, hairstyle, or something else)
  • Columns for the over total and the running total
  • Two features for those new to the job:
    • A quick reference guide to scoring symbols
    • A couple of example overs

The underlying format is an Excel spreadsheet. Each page covers 30 overs. You can print it at A4 if you wish, but my strategy is to print 4 to the page, as shown in the PDF version. This can then be folded twice to leave an A6 size sheet. Add pencil and backing board (time to source an A6 clipboard), and your Square Leg Scrumpire / Striker’s End Umporer is ready to sally forth and tally.

I intend to hone this in light of user feedback, and to that end I’d be interested to hear from anyone who uses this, or a similar system, as to how it might be improved.

Special thanks to Angus Bell of the Pirates of the St Lawrence CC for the inspiration, and the welcome to cricket in Montreal.