Close to three decades ago, an Australian computer programmer pioneered online text commentary for cricket matches
This article was originally published to mark ESPNcricinfo's 20th anniversary, in 2013
It's Christmas day in 1991. Australia and India are gearing up for the Boxing Day Test in Melbourne, the second in a five-match series.
Kevin Robert Elz, a senior systems programmer at the University of Melbourne, who has recently returned from a trip to Thailand, types out a long email to the newsgroup rec.sport.cricket.
"Yes, I'm back…" he writes. "I was able to get pretty good reception from Radio Aust for the games they broadcast… and the Thai English language papers carry reasonable cricket reports for the few that weren't broadcast. However, I won't be sending reports from Melbourne match, I'll be at it watching… its no use sending me mail asking what the score is, or anything else like that…"
Cricket is a huge part of Elz's life. "When there's cricket on to watch, there is never anything better to do," says Elz, a member of the Melbourne Cricket Club since around 1980. He claims to have watched every minute of Test cricket played in Melbourne from the ground since the early '70s, and to almost always be near a television set when Australia are playing a Test.
"I commonly work while watching the cricket," he wrote in his Christmas email to the newsgroup. "I have a TV on my desk positioned so I can more or less watch both screens at the same time. I also have a terminal at home… and it's placed so I can use it and watch cricket at the same time."
The history of the internet can be traced back to the late 1950s and the Advanced Research Projects Agency Network (ARPANET), funded by the United States Department of Defense and developed by researchers at American universities and labs.
Australian researchers began harnessing ARPANET's potential in the '70s, and while several key developments occurred through the '70s and '80s - like the setting up of the Australian Computer Science Network (ACSnet), and advancements in email and file-transfer mechanisms - the utility of resources was limited to computer science researchers in universities. The rest of the country was blissfully offline.
In 1989 a 56Kbps satellite circuit was set up to connect Australia and the US. The historic moment arrived on the night of June 23. Torben Nielsen of the University of Hawaii pinged a signal with the succinct message "Link Up". It traversed the Pacific Ocean to reach a laboratory at Melbourne University where a researcher effected the connection. That researcher was Robert Elz.
Elz had a rock-star reputation among Australia's tech community. A book titled AARNet: 20 years of the Internet in Australia 1989-2009 tells the story of a workshop in Adelaide where Elz, sitting in a corner during dinner, was surrounded by hordes of techs. "Some of them were kneeling. They were hanging off his every word."
Being intimately linked to the country's push for the internet gave Elz a headstart and he was part of a select band of Australians with internet access in 1991.
Melbourne-based writer and historian Gideon Haigh puts things in perspective: "End of 1991, I joined the Independent Monthly, a small current-affairs journal that published long-form reportage," says Haigh. "I used a Mac Classic with Word and Claris Works. I did not own a computer myself - few people did - so I would take it home on weekends - terminal, keyboard and mouse under my arm. I wrote The Cricket War on it."
In contrast, by the end of 1991, Elz had set up permanent internet connections at work and home. On the second day of the Boxing Day Test, he fired off a 1200-word report - sprinkled with finely observed detail - to rec.sport.cricket. Here's a sample:
"Today at ground level there was a noticeable east to west breeze - noticeable because of the amount of flying trash continually needing to be captured and removed from the batsman's line of sight."
Seventy-three-year-old K Sankara Rao, who bears a striking resemblance to the Indian author RK Narayan, remembers his childhood in Eluru, a small town in the West Godavari district of Andhra Pradesh.
"I used to wake up early every morning and run two miles to my friend's house to listen to Radio Australia commentary, to hear Michael Charlton, Alan McGilvray and Johnny Moyes. There was no other way to find out what was happening in the matches."
Little did Rao know then that in 1972, when he moved from IIT Madras to teach at North Dakota State University in Fargo, he would be cut off from cricket altogether.
"Even if I had run 200 miles from Fargo, I would have not got any cricket scores," says Rao, now retired and based in the state of Washington. "For close to 15 years, my only resource for cricket [was] Indian newspapers in our university library - and even those used to reach a month after the matches had finished."
Around 1988, Rao decided to invest US$300 on a Sony digital shortwave radio - "a huge decision for me" - so that he could listen to the BBC's Sports Roundup. "If I remember right, we used to get it at 7:45AM and 4:45PM Fargo time."
A few months after buying the radio, he began posting scores on his profile page on his university network, using his Unix machine. "There was a 'finger' function which people could use to access user profiles. So anyone who accessed my profile could see the updated scores.
"But soon I had another idea: to set up a mailing list and help everyone else. I asked a few of my friends - mostly professors at different US universities - for their email ids and began sending cricket scores every morning and evening. I still remember the email id I sent it from - firstname.lastname@example.org. Gradually the list grew."
In 1979, two Duke University students conceived Usenet, a discussion forum where users could read and post messages. Elz, currently a lecturer at the Department of Computer Engineering at Prince of Songkla University in Hat Yai, Thailand, was a user from the "very early days" - way before the internet appeared in Australia.
"Because I was running one of the major Usenet relay sites, and for some other reasons, I was included in the 'Usenet cabal', which was the set of people that the software would allow to create new groups."
A request for creating a newsgroup dedicated to cricket, rec.sport.cricket, was posted in March 1990, according to a 2007 paper by Sanjay Joshi in the International Journal of the History of Sport. "There were 181 different postings to the new newsgroup in April 1990, its first month of existence, and this more than doubled in the following month. The newsgroup averaged more than 300 posts for the first six months of 1991."
Elz does not remember who came up with the idea of the group, but he says it is possible that he sent the message to activate it.
About 10,000 miles away, in Fargo, North Dakota, Rao trawled through the list of groups and stumbled upon rec.sport.cricket.
"I noticed that cricket scores were getting lost amid a lot of discussion. So I decided that we needed a separate sub-group for scores alone. Which is when I created rec.sport.cricket.scores."
Rao's original mailing group - where he posted cricket scores from the BBC round-up - was now a thriving community with many users posting scores from around the world.
"I decided to take those updates from the mailing groups and post it on rec.sport.cricket.scores. That way many more people could have access to scores."
Initially Rao was the only moderator for the group but he gradually began adding contributors as moderators. One of moderators "sending regular updates" was Elz.
Elz, or kre as he was known in internet circles, developed an enduring love for scoring cricket matches from an early age.
"When I was still at school I had developed my own system of scoring, with my own formats, on my own forms that I designed. I used to score all the games I attended or watched on TV from some time in the late-60s probably until perhaps the mid-70s. I [also] used to fill in the summaries in the ABC cricket guide pages that were available for people to record the scores of the matches."
Elz used to a play a bit of tennis but he didn't gravitate towards scoring tennis matches.
"Cricket's short brief periods of action with nice little gaps (while the bowler walks back to get ready to bowl again) make it ideal for this kind of coverage. There's time to describe what happened with each ball before it is necessary to watch the next one… And then when anything really important happens there's an even longer break to make sure the description is just right."
The third Australia-India Test - Shane Warne's debut - began on January 2. Elz doesn't remember if he was at home in front of his Teletype 5620 "aka Blit" terminal or in his office, equipped with a Sun 3/50, on the first morning of the Test.
rec.sport.cricket archives show that he sent a message with the subject line "3rd test, Aust v India, start of play 1st day" listing the teams and the toss result for the match.
His next message to the group, after the first 16 overs, is titled: "3rd test, Aust v India, 1st hour, Scores". The scorecard isn't tabbed but it resembles something you would see in a newspaper. There are even columns for no-balls and wides.
The most significant message, sent a few minutes later, carried the subject line "3rd test, Aust v India, 1st hour, commentary" and was 650 words long. It began thus:
"With 4 seam bowlers, and only Shastri to bowl spin, it wasn't surprising that Azharuddin decided to send Australia in. Border didn't seem very disappointed that he will get to bowl last."
"Others had been providing summaries of games," Elz says of why he might have chosen to send updates every hour rather than one long summary at stumps. "As I remember it, I just thought that I could do a bit better and provide more frequent reports, and make information available to people who obviously really wanted it.
"Doing ball-by-ball, while more actual typing work, is much easier to do than summaries - it is much easier to say 'nice glance fine for four' or 'top edge to wild swing over slips and down to third man on the bounce, just a single' than to attempt to summarise an innings in a paragraph or two."
That first-hour update opened the sluice gates. Next came a 470-word missive at lunch, a 336-word bulletin at the end of the third hour, a 482-word tea report, a 469-word fifth-hour summary, and a 655-word wrap at the end of the day. Elz preceded each of these descriptive postings with scorecards, all immaculately written up.
"I would just type my impressions of what was happening," he says, "and keep a rough score summary… My impression is that the earliest [updates] were more paragraphs of summaries with just the notable events recorded, and [they] gradually got more detailed, and into a more line-by-line, ball-by-ball format, as that seemed both more useful to people and actually easier to produce."
Reading the archive, it's clear that Elz was bent on getting all the details right. He apologised for factual mess-ups and appended each scorecard with a parenthesised note: "these numbers are approx correct, but watch later socrecards [sic] for revisions, there may be the occasional error... "
What is also clear is that Elz did his utmost to ensure his biases didn't creep into his descriptions. This was particularly crucial since the whole tour - like so many of India's visits to Australia - was dogged by umpiring controversies.
The second day's play began with a short one-paragraph report, but come the end of the first hour, Elz was back in full swing, hammering out a 720-word update. The lunch update began with a description of Warne's first contributions to Test cricket, playing out a maiden off Manoj Prabhakar, before a post-script at the end of the report that said: "I'm glad that I got to see it [the session] - aided by a car that wouldn't start, preventing me doing what I should have been doing…"
Elz flooded the newsgroup with live updates over the next few months. He tracked the Benson & Hedges World Series (between Australia, West Indies and India) and went on to provide ball-by-ball updates for the World Cup.
His commentary style evolved with every match, and on January 12, 1992, during a one-dayer between Australia and West Indies in Brisbane, he settled on the ball-by-ball format. Dave Liverman's "Cricinfo History" online has Elz's updates:
WSC, WI v Aust, Jan 12, Aust after 20 overs
Patterson to Boon
-. B no ball, pushed forward of square 1/72
1. B down leg side, inside edge, single 1/73
2. J dropped at his feet, Patterson fields
3. J well up, yorker, dug out, to close mid wicket
4. J driven to cover, no run
5. J late cut, just to point, no run
6. J well picthed up betwen bat and pad, leg stump gone
Jones Bowled Patterson 0 2/73
After 19 2/73 (Boon 36, Extras 8)
Neeran Karnik, then a graduate student in computer science at the University of Minnesota in Minneapolis who would later go on to play a big role in the early days of ESPNcricinfo, calls Elz a "godsend".
"Not only did he post live updates (well, once every five overs typically, but with almost bbb commentary), he also made cricketing sense in his comments. For example, in a thread on r.s.c. in the early 1990s, he responded to someone's prediction that [Craig] McDermott would surpass Kapil/Hadlee's wickets record by predicting that Warne was more likely to.
"But frankly, our emotion was more of gratitude, since we were so cut off from cricket coverage. I wouldn't have cared too much if he had been a rabid, biased Aussie supporter as long as be kept posting the updated scores."
Progress was rapid. Over the next few months, cricket commentary moved from rec.sport.cricket to internet relay chat (IRC) - a live, interactive text-messaging system - and a number of university students picked up the baton.
Then came an audio hook-up: a transistor radio with BBC TMS commentary, left in front of a Sun workstation at Oxford University, relaying the audio using a software utility called 'Vat' to the University of Minnesota, as Karnik remembers it. There Simon King, now known as the founder of Cricinfo, and Karnik (and a couple of volunteers at Stanford, Rohan Chandran and Vallury Prabhakar), would listen to the broadcast and frantically type ball-by-ball commentary on IRC.
The big breakthrough arrived on March 15, 1993 when King, a post-doctoral student at the University of Minnesota in Minneapolis, placed a bot on the IRC channel #cricket to provide scores to anyone who typed in request commands. The bot was called Cricinfo.
Elz didn't send many updates after the World Cup, partly because others were filling the breach and because he was busy with work, which included overseeing .au domain names, a task that he was assigned from 1986 to 2001.
But he pioneered online text commentary for cricket. As Elz pounded the keys in Melbourne, ESPN was still four years away from creating its own website. And the earliest accounts of play-by-play commentary in the National Football League - when "laptop quarterback" Bob Sansevere tracked Minnesota Vikings' games - were still seven years away. Even US politics didn't latch on to the internet's potential until the 1996 elections.
"Everything was happening so fast that we didn't realise he was doing something that was never done before," says Rao about those heady days in 1991-92. "But go back to the archive and you will see how grateful everyone was for his updates."
Rao never met Elz - "our only interaction was on email and in the newsgroups" - but he is not in the least surprised that Cricinfo was birthed in university labs around the world.
"You needed four things to get something like this going: access to cricket, an internet connection, time, and passion. We in American universities had three of these but there was no access to cricket. Elz had all four. So he was the chosen one. You could call it destiny."