CricInfo – How it all began

I don’t think any of us involved in the founding of CricInfo in 1993 envisaged what our baby would grow into today. I’m often asked how it all started, and with this being the 20^th anniversary year, in celebration of which ESPN is producing a documentary and ostensibly detailed retrospective about CI, now seems as good a time as any to provide my perspective on the story.

How it all began

There were a few seminal moments, or perhaps movements, without which Simon King would never have decided to develop the initial version of CricInfo, and I’ll focus here on those early days. Much has already been said and written about what happened thereafter – some of it remarkably off base (such as the timelines and other details published by ESPN CricInfo itself!) and some of it pretty solid (Dave Liverman’s BluWiki history).

I discovered the Internet in 1992 when I arrived at Stanford. These were exciting times. I was assigned an email account at Stanford, and could even send messages to my dad in Hong Kong. No phone calls or air mail envelopes required! Actually, for a couple of weeks I could only receive email (in emacs, using M-x rmail), because nobody could tell me how to send it, until I discovered elm (and later pine). But I digress.

In addition to an international subscription to SportStar, I had brought my trusty shortwave radio with me to Stanford, in order to pick up BBC broadcasts and stay tuned to the world of cricket and football (soccer). The latter was a particular concern, because my team, Manchester United, looked like they might actually win something for a change. Yes, that’s how long ago this was.

In my desperate search for information, I stumbled upon a program called “rn”, Usenet, and specifically rec.sport.cricket (RSC) and rec.sport.soccer (RSS). And from there, I discovered the TLA (three letter acronym) that was to dominate my life – IRC, or Internet Relay Chat.

Several kind folk in civilized nations used to frequently post score updates to RSC, and several others of us used to get on to the #cricket channel on IRC (it wasn’t twitter who invented the hashtag, folks) and, well, chat about it.

Jacques de Villiers invents ‘dougie’

In December 1992, things got a little more interesting. A student in South Africa, Jacques de Villiers, wrote a quite remarkable little program called “dougie.” Remember that this was still a time of nascency for the Internet. Jacques’ masterful software not only allowed for live scoring of a cricket match on a computer, but also enabled broad distribution of those scores. I don’t want to get into a technical dissertation here, but it was absolutely brilliant. The output of dougie was written into a user’s .plan file, and you could always view the latest scorecard by simply entering the Unix command “finger user@host.”

What’s more, dougie was open-sourced. So any of us who were so inclined could download and build on our own machines, and thus have our very own live scorecard. The hierarchical tree that Jacques created for distribution was ingenious, and he addressed the resource constraints (1992 era bandwidth and port limitations) by creating a nested master-slave model, in which the slaves of the first master became masters for subsequent slaves.

The amazing thing in all of this however, was that it was not until 1996 that CricInfo started using Dougie for live scoring – in fact Dougie, with a lot of help from Vishal Misra and Travis Basevi (now of StatsGuru fame) saved CricInfo during the 1996 World Cup. That’s right – whatever live scoring we did before 1996 was done pretty much by hand. I should know – I was the fool who first decided that we should have the scorecard updated ball-by-ball, and I duly did so by editing in emacs after every single delivery. It’s amazing to think that we actually covered more than one match in that mode.

Enabler #1 – IRC #cricket

Back to late 1992 – and around the time that Dougie appeared on the scene, Manas Mandal and myself decided that we needed to get serious about this IRC thing. We focused on establishing some sort of permanence for the #cricket channel. This seems pretty childish when I look back at it, but then I can be excused, as I was a child back then. It was a non-trivial effort – we kept terminals running all over the place and even did shifts at one point to ensure that the channel was never abandoned.

This, for me, was the first little step towards the CricInfo we know and love, as it enabled a central venue for score updates and occasionally live commentary – in many cases provided by people who were running dougie and/or fingering other people’s .plan files. It’s incredible to think back to the despair that overcame us when a network problem resulted in not just a short term interruption, but a fundamentally existential question. “Will the person providing the updates ever return, and for just how long should I sit here like an idiot pretending to work on my CS 106 assignment?”

There were tens of people, and on rare occasion a hundred or so people on IRC #cricket at any given time, almost all based in the USA, with a few notable exceptions such as George Heard in Tasmania and Anthony Waller who I believe was in Austria at the time.

Enabler #2 – VAT audio (or, the day we broke the Internet)

Jump forward to early 1993, and India were about to host England in a much anticipated Test series. The two had last met in 1990, a series dominated by Graham Gooch’s legendary 333, and the sheer brilliance of my hero of the time, Mohammad Azharuddin. Since then, India had hardly played any Test cricket – a tour of Australia in 91/92 and a tour of South Africa in late 1992 were pretty much it. And for Indian fans, those tours were best forgotten about.

So in January 1993, many of us on IRC #cricket were trying to figure out if we could do something better than the sporadic coverage we had been getting to date from the likes of David McBean in England. Several people scoured the Internet, such as it was, for any hint of a streaming audio service hosted by the BBC or the like. We had no idea that we were years away from that.

Every good story has an angel sent from the heavens above – and in our case, it was a piece of software called VAT, developed at Berkeley, that was our savior. I believe the discovery of VAT was made by Vallury Prabhakar and Manas Mandal, and since I have no information to the contrary, I’ll pin it on them for now.

What VAT did was something that seems really trivial now, but was quite remarkable back then. It allowed for the real-time transmission of audio over the Internet. For the youngsters reading this, you should think for a moment what this meant. At the time, the backbone of the entire Internet, NFSNET, had just been upgraded to a T3 line. As it were, the entire Internet had a bandwidth capacity of 45Mbps. That’s less than your personal iPhone 5 can handle over LTE. The backbone of the Internet today has a capacity of at least TWO MILLION times greater than back then. Think about scaling down by a factor of two million, and you’ll have some idea of the world we were operating in.

So at this point, there was a small group of people working feverishly together to make something of this. Each of us, being frequent IRC users, had nicknames that we went by.

• Simon King aka CoolPom, a post-doc at the University of Minnesota
• Neeran Karnik aka VKFan, a PhD student at the University of Minnesota (remarkably, he and Simon never met)
• Manas Mandal aka bakait, a faculty member Ohio State
• David McBean aka Ragga, a PhD student at Oxford University
• Vallury Prabhakar aka RustFace, a PhD student at Stanford University
• Rohan Chandran aka BritRoh, a freshman at Stanford University

For my money, those are the people you have to thank, or blame, for CricInfo. Once Vallury and Manas had discovered VAT, the next steps were simple. One of them, and I have no idea which, FedEx’ed a cable across to David at Oxford, using which he could connect his radio (it might have been his Walkman in fact) to a Sun SPARC workstation in his lab there. He would tune his radio to Test Match Special, and using VAT, the rest of us in the US would listen to the radio commentary, and then transcribe onto the #cricket channel on IRC. Vallury and myself were in the Mechanical Engineering labs at Stanford from around 8pm to 4am every night, and Simon and Neeran were nowhere near each other on the Minnesota campus. That latter factoid is not as bizarre as it might seem – think about how far you might or might not want to travel in Minnesota in the middle of the winter.

The whole thing was remarkable and surreal. I mean, here I was sitting in a lab at Stanford in the middle of the night, neglecting my homework and listening to my hero score a brilliant series-defining century at Eden Gardens, sharing that joy with first tens, then hundreds and soon thousands of people following live on #cricket. Thanks to David, I also got a musical education during some of the breaks, but that’s for another time.

One moment that really brought the enormity of it all home, was when we brought down the Internet. Yes, you read that right. Our use of VAT, apart from almost certainly being of questionable legality, was allegedly responsible for killing the transatlantic pipeline for a short while, and while it was a source of amusement later, at the time our biggest concern was interruption in service! If you didn’t really comprehend the limited bandwidth that was available back then, imagine that you’re on a Google Hangout, with voice only, with half a dozen other people, and because of that, the Internet stops working.

CricInfo is born

The audio coverage and live commentary were a phenomenal step forward, and yet led to some remarkable frustration on the part of our ‘consumers’ and subsequently ourselves. As anyone who has built a successful consumer product knows, consumers have this wonderful knack for immediately taking it for granted and letting you know what’s wrong. When the first search engines came about, the reaction wasn’t always “wow, I can search for things on the Internet,” rather it was often “this sucks! It’s not finding exactly what I’m looking for!” Similarly, we found ourselves inundated with requests for the score outside of live commentary hours, and complaints that there was no way for people who didn’t stay up all night to find out the score or any details of the ongoing match.

My memory is a little hazy on this next point, but I believe sometime during the second Test of that series, Mandar Mirashi (aka Mmmm), who was a revered IRC expert and admin, developed a bot (IRC script) that not only helped with the permanence of #cricket, but also delivered a simple capability. After the day’s play was over, you could ask it the score, and it would respond with something of the form of “India 263/4 (Azhar 114* Amre 7*)”

And that’s when Simon changed the world. I remember having many late night conversations (on IRC) with him about what he thought was possible, and frankly, I am not sure I fully understood or appreciated it at the time. Actually, I don’t know if he realized what he was starting either! Simon hunkered down for a few weeks, learned everything he could from Mandar, and, not to put too fine a point on it, built the first version of CricInfo, based on the code for Mandar’s bot.

The official launch was on March 15^th, 1993, but a few people were privileged enough to play with CricInfo for a little while before that. The launch was just after the India vs. England series had ended, and you could send CricInfo a message of the form “/msg CricInfo <filename>” and in return, the bot would dump to your terminal the file you requested. Those files included the scorecards of the Tests and ODI’s in that series, and, I think, a few articles that we gleefully copied word for word from publications such as Sportstar.

When there was a live game going on, at various intervals during the day, and at the end of the day we would compile and update the scorecard. The simple command “/msg CricInfo scorecard” would of course show you the latest scorecard that we had from the current game.

It was somewhere around this time frame that I had the not-very-bright idea to start updating that scorecard file ball by ball. Ouch. It’s quite incredible to look back at how we did this. Using VAT was not a long-term option, so we had to find other options. At Stanford, Vallury had a great relationship, which I inherited, with Bob Drewes, the guy who controlled the C-band satellite dish. Any signal that was in or out of the Caribbean in those days was an unscrambled one, and by working with Bob to scan the airwaves, we were able to catch a feed and have it broadcast on one of the campus channels. I wonder how many students knew that they could watch most International cricket for free in their dorms in the early 1990’s.

A lot of the early commentary that I did, along with scorecards that I maintained ball-by-ball, was from these illegal feeds – and to this day, the majority of CricInfo scoring and commentary is done by people watching the TV broadcasts. But sometimes, that wasn’t an option. To give you an idea of the extremes we would go to, I remember when Australia visited Pakistan in 1994-95 (the famous 1 wicket win for Pakistan was in that series), we had live commentary thanks to another Stanford student, Shehzaad Nakhoda, who got on the phone to people watching back home in Karachi, and had them relay the scores to us. Ball by ball. And nobody ever thought to claim those phone calls as an expense – because at that point, CricInfo was not an incorporated entity, just a volunteer collective.

Anyway, it was at this point that Simon invited Neeran and myself to join him in running CricInfo. This wonderful responsibility even came with a special command that we could use to take control of the bot – “/msg CricInfo MasterMe Now.” It also came with a few inherent risks, such as the time I very nearly wiped out the entire CricInfo database because I was in a different shell and didn’t realize that my “rm” command was aliased to “rm –rf”

Things pretty much mushroomed from here. Countless other tireless volunteers were recruited to work for CricInfo, and we went crazy adding older scorecards (manually typed up from Wisden and other sources) and articles, all again manually typed up by incredible individuals who sought no credit or reward for their efforts. We talk about crowdsourcing today, but what CricInfo achieved in those early days of the Internet was Wikipedia-esque when you put it in context.

The importance of the volunteer group cannot be understated. At the same time as CricInfo, with Simon’s permission I adapted the bot code to create FootInfo. Manchester United were on their way to ending a 26 year dry spell, and I thought the world needed to know about it. FootInfo survived for a year or so, but while it suffered from not being at the top of my priority list, I look back and am acutely aware that it died because I was unable to generate the broad volunteer support that CricInfo had.

Gopher and the World Wide Web

The next surge forward for CricInfo was when Neeran discovered gopher a couple of months later.

“Gopher tulip.ee.ndsu.nodak.edu” is something that the early users of CricInfo will never forget – and those of us running it will never forget the countless times we had to telnet into that machine, and old 386 (!!), because it had crashed from overload. In fact, I am quite sure that there were a lot of times when Professor KS Rao, a father figure in the world of Internet cricket who was now hosting CricInfo in North Dakota, had to be roused from his dinner table to go and physically revive the computer.

What the gopher server did was take CricInfo out of the limited scope that IRC provided, and squarely into the broader Internet. The ultimate explosion of course happened when we finally embraced the web (and later specifically with the 1996 World Cup which arrived at exactly the right time for CricInfo to blossom). Ironically it was Simon who was perhaps the most convinced that Marc Andreessen was wrong when he introduced NCSA Mosaic, the first web browser, suggesting (along with many others at the time) that this would be a better way than gopher to navigate the connected Internet.

The rest, as they say, is history, and frankly there is only so much I can write. I don’t consider myself to have been the most critical influence in the early days of CricInfo, nor do I believe that I did as much as many, many others, but all said and done I will never lose the wonderful memories of having been there from the start, and playing a small role in the evolution of something so remarkable.

Assorted Memories 

Some of my best memories from those days, apart from all the obvious ones around the events above, include, in no particular order:

• Being accused of being, at various times, anti-Indian, anti-Pakistani, anti-English and anti-Australian. Somehow, even though I did the commentary and reporting on a lot of West Indies games, nobody ever accused me of being biased against them. Travis Basevi, on the other hand, didn’t get away so lightly. Over the years I believe he collected the full set – of being accused of being both pro and anti all nine Test playing nations. Quite the achievement.

• 1994 – CricInfo’s first player interview was with Shoaib Mohammad, on IRC #cricket. Interviews didn’t become a regular or formal feature until 1996 at the Hong Kong Sixes, and you can find many of those early conversations here (http://static.espncricinfo.com/db/INTERACTIVE/INTERVIEWS/). The first official interview was slated to be with Mohammad Azharuddin in June 1996, but he backed out for reasons unknown. As far as I can recall, the first interview we officially completed on behalf of CricInfo was when Alex Balfour and myself moderated a fascinating discussion with Winston Benjamin, who was candid on the record, and brutal off it (http://static.espncricinfo.com/db/INTERACTIVE/INTERVIEWS/BENJAMIN_WKM_30JUN1996/)

• 1995 – in conjunction with a small ISP in Hong Kong, we developed a Java applet that provided the first known live video coverage of anything on the Internet. That is, if you can call a rapid sequencing of still frames “video”. Pretty awesome, whatever it was, and as a Hong Kong boy, it was a source of pride for the relatively obscure Hong Kong Sixes tournament to be at the forefront of innovation in this way.

• I made my live audio commentary debut in 1996. My co-commentator was Michael Holding. Seriously – who gets to do commentary alongside Michael Holding? Those audio files were available on CricInfo for many years but seem to have disappeared. I am very grateful for that, because as far as I can recall, in my desperate bid to avoid clichés I did the exact opposite. We did an interview with him too, just to prove to myself that it wasn’t a dream – http://static.espncricinfo.com/db/INTERACTIVE/INTERVIEWS/HOLDING_MA_21SEP1996/

• We did a deal with Mick Jagger. I mean come on, Mick Jagger. Jagged Internetworks partnered with us for the first regular live video coverage of cricket online – I think it started with a Sharjah tournament in 1997?

• Then there was Hansie Cronje – one of the nicest cricketers I’ve ever met (and I’ve been very fortunate both before and during my CricInfo days). He would constantly ask me for detailed updates from county cricket matches in particular, and often commented on how he couldn’t survive without the detailed real-time information that CricInfo provided. We used his “CricInfo is King!” quote in PR for a while (gee, I wonder why Simon liked that one!) Never really thought much of it at the time, but with the benefit of hindsight one can only wonder.

• Having thrown out the match-fixing innuendo, it would be remiss of me to not relive the moment when I lifted the lid on Salim Malik’s match-fixing. It was at the Wills International Cup (think Champions Trophy) in Dhaka in 1998. This was during a period when CricInfo worked closely with the ICC and we were able to provide coverage from the stadia for ICC events, rather than second-hand.
  • Pakistan were playing West Indies. The Pakistan team walked onto the field with only 10 players. I looked around to see who was missing, and Salim Malik was visible just below our box, talking feverishly on his mobile phone. Early in the innings, Azhar Mahmood got Stuart Williams out lbw, and for no apparent reason, Salim Malik decided to high-five Aamir Sohail. Like something out of a Billy Birmingham CD, he missed, and poked Sohail straight in the eye, resulting in a lengthy delay while treatment was administered. When it was Malik’s turn to bat later in the day, he managed to get himself out lbw attempting a reverse sweep against the deadly bowling of Keith Arthurton. Naturally, I use these three pieces of information to conclude that Malik was involved in match fixing. A complete joke at the time. Except one local newspaper didn’t realize it, and the next day identified me as the intrepid reporter who uncovered a great match-fixing plot. Had I known what I know now, I’d have feared for my life.

• Sir Gary Sobers hung out in my bedroom on the Stanford campus. If you have a cooler story than that, I’d like to hear it.