For as long as I recall, a small cluster of cricket tragics have congregated in roughly the same location, every day of every Shield contest their beloved Victoria is engaged.
They are a rag-tag bunch of social misfits, to say the least. Their banter, even their intermittent sledging which reverberates around the cavernous MCG, is well informed.
One buff diligently records every ball in his classic green scorebook, despite two massive digital scoreboards displaying all you need to know.
Why?
Perhaps it makes him feel more closely connected to the game, maybe it’s a compulsive obsessive habit. Possibly it prevents him nodding off.
I must admit, I’m as guilty as any sports nut of filling the nooks and crannies of my brain with useless trivia.
While millions are concerned with where their next meal is coming from and horrible recollections of war and oppression, I’ve stored the location of every Olympics, all the AFL/VFL premiers since 1897 and more Melbourne Cup winners than I care to know.
I’m no Rainman, but if only I could have used those portions of my intellect for something really useful!
Statistical data was once recorded largely for its own sake, a celebration of players and teams’ feats and failures. 99.94 is burned into our consciousness as an astonishing measure of near-perfection, and apart from anecdotal evidence, statistics are just about all we have to assess athletes of a bygone era.
Now, there is barely an aspect of sport that isn’t dissected to within an inch of its life. Everything is counted and everything counts.
The Yanks, once again, have a lot to answer for. Baseball set the benchmark over 100 years ago with its passion for all things numerical. American Football is not far behind.
Anyone who doubts that America is the one stop shop of hardcore statistical obsession need only log onto websites such as Statsology or Amstat and follow the myriad of links, blogs and articles.
Swinburne University’s Professor of Statistics, Stephen Clarke, is the pre-eminent sport stats researcher in Australia. Clarke’s career in the field was born during his Masters degree when in 1977 he assisted a lecturer with a dilemma facing squash over two different scoring systems.
Clarke has since authored over 125 papers and articles on a range of sports.
“I think that statistics are sometimes seen as a bit dry and mundane, and I wanted to show that not only can statistics be fun, but they relate to everyday life. I think that sports statistics have been very successful in achieving this.”
Clarke’s most satisfying reference paper investigated one day cricket run rates in the days when Marsh and Boony ambling along to 0/45 after 15 overs earned a pass mark.
Clarke’s findings highlighted the value of scoring runs faster, earlier, and set the framework for the mysterious Duckworth Lewis system for rain affected games.
“I certainly see a big future for the use of statistical modeling in betting applications, either for making life difficult for the bookies or in assisting them to set odds” says Professor Clarke. “With the growing developments in communication and betting in running (betting on events and outcomes during a game), they will need to use statistical models to set the odds with the necessary speed.”
Clarke has consulted for Tattersalls and has a spin-off company Sportsbet21 which provides odds for unique running forms of cricket and tennis betting. Surprisingly, the Australian Open tennis is the most popular non-racing event for gambling.
Does Professor Clarke believe that some sports are over-analysed to the point where the basic fun and enjoyment of a physical entertainment are diminished?
“Well yes, when I see run rates given to two decimal places – completely meaningless! But I think that some people enjoy watching sport, others analysing. I do think a lot of stats are collected, but never analysed. For example, if the Australian Open and other Grand Slams made available to scientists all the stats they collect at their Championships, they may learn a lot of new things about tennis.”
For coaches, selectors, journos and sports connoisseurs who pore over faded annuals, stats matter.
May the passion, skill, creativity, emotion and theatre in sports not be lost amongst all the numerical ‘evidence’.
Don’t be surprised if one day there’s a ranking system for those intangibles, too.
Oh yeah – that would be TV ratings and attendance figures.
davido
Guest
The average as Spiro points out is not actually the mean of runs per innings but rather runs per dismissal. The current average statistic really provides a positive bias the further down the order you bat. Batsmen who are batting below their 'true' position often have great averages that do not truly reflect their ability. That is why any comparison of batsmen (if you can even do that!) is only possible when both batsman bat at the same position. Any stat is going to have problems and this stat is by no means the worst you can have. Personally, I like to look at what percentage of runs a player has contributed to their team. Although arguably a better stat is to compare the percentage contribution to the total runs scored in the match. For example in the first text v Pakistan, Watson contributed 213 out of Australia's 679. A massive 31% of the Australian total. He scored 18% of the total runs by both teams - 10% more than the next next batsman.
Jeff Dowsing
Guest
On day 2 I saw about 6 ejected for making a snake out of empty plastic cups! Absolute harmless fun - if anything, doing the cleaning contractors a favour. Kind of supports the notion that you can support anything with statistics.
bever fever
Guest
No not at all Rob, dont think either are just a bit of fun, but the problems ethnic soccer clubs had were much more serious and socially divisive IMO than some cricket yobs.
Rob
Guest
Says it all really doesnt it.. a couple of teenagers shirtfronting each other at a club game is race riot if its football...but 91 people ejected from an international cricket game is just a bit of fun.
Dave01
Guest
if you played a soccer game for 6 hours with beer the result would be crowd carnage.
bever fever
Guest
From time to time the media has latched onto drunkeness/violence at the cricket, but IMO racial violence in the old NSL was far more newsworthy than guys/gals sitting in the sun all day at the cricket drinking then getting a bit over the top, its not like their were gangs of aussies going around beating up pakistanis.
Rob
Guest
Hows this for a statistic....91 spectators were ejected from the first days play in Melbourne....Imagine the sort of headlines we'd be reading now if that happened at a Socceroos match. Why do we never hear any comment from the Cricket media about this sort of thing?
JohnB
Guest
Brett, wickets per match is a pretty good indicator of whether someone is an ordinary, reasonable, good, very good or great bowler (although even that basic measure is subject to plenty of distortion). Whenever this sort of question comes up I think of a couple of things - John Benaud in his book relating to his time as an Australian selector saying that if it was just a matter of stats, why would you need selectors, and rugby league and the tackle count - you read old players saying that blokes like Terry Randall or Charlie Frith were such damaging defenders that the attackers wouldn't run at their side of the ruck if they could help it (which probably means their tackle counts never hit any great heights). But who cares about the count if you can put the wind up the other team?
Doug
Guest
It depends on what you want the batting average to be: how many runs they score on average between getting out, or how much on average they contributed to the team's score every time they batted. The number of runs scored between getting out seems to be a better indication of how good a batsmen is. Which is what the batting average tends to be used for. It also seems analogous to the bowler's balls per wicket or runs per wicket stats.
Jeff Dowsing
Roar Pro
Interesting viewpoint Spiro - I'm more inclined to say that if the innings concludes, and no one was able to get you out, practically speaking your idea seems a little harsh. I think the greater propensity for middle/lower order bastmen to be not out is counterbalanced by having fewer overs in which to accrue runs (notwithstanding often easier conditions, when the bowling team might be flagging). In most instances, batting averages are a pretty good indication of ability (within a defined era). I'd agree that bowling averages can be a little less indicitive. The greatest deficiencies though lie in the inability to factor in the strength of the opposition and the conditions. Hence eternal debate over the likes of Murali v Warne, modern day run machines v 70's & 80's batsmen up against so many fast bowling greats, lighter bats, no helmets, tougher pitches etc etc.
Brett McKay
Guest
What's the old line - 73% of all stats are made up?? Spiro, that's an interesting position re the NO's being included as a completed innings. As a batsman, I'm tempted to say that innings in which you get fired shouldn't count as an "out" either! And perhaps that's where the issue lies. If in a batsman's stats we replaced "Innings" with "Outs" alongside "Not Outs", we'd have a better refelction of what the figures are saying. So Bradman would then have averaged 99.94 every time he was dismissed, rather than every innings. The other side of this argument is bowling stats. Bowlers are only judged by wickets taken for runs conceded, and so there's no indication of how many wicktes taken per innings bowled. Clint McKay, for example, took no wickets in the first dig in Perth, and finished with 1/56 in the second. But his Test record isn't half a wicket per innings, rather it's the unflattering average of 101.0 My point is that cricket stats are a funny beast. On one hand we have batsmen measured per innings, when in fact that's not necessarily the case, as you clearly state. But then we have bowlers who aren't measured by innings, when perhaps it would be a better measure. It truly is a funny old game...
Spiro Zavos
Expert
I wonder what Professor Clarke would say about my pet objection to cricket's batting statistics, that they allow not outs not to be counted as completed innings. This aberration in my view unduly escalates the batting averages of middle order batsmen over those higher up in the innings. Compare for instance the number of not outs that Allan Border achieved with, say, the not outs achieved by Mark Taylor. And then see what a difference it makes to their averages if the not outs were recorded as completed innings. In my view, a not out is actually a completed innings and should be recorded as such.