When statisticians talk about a “trend”, they mean that they have a data set of trustable size which will let them predict future data.
When Malcolm Conn writes about a trend, as he does in “Runaway trend an ominous omen for Australia ahead of hosting 2015 World Cup”, he means several cricket matches which happened six years ago, four matches which happened in the last week or so, and nothing that happened in between.
Ominous omen? Sounds… significantly significant.
Or possibly it’s the sort of ominousness that can be treated like this?
The “trend” in question is Australia’s alleged tendency in one-day matches to make large scores in the first innings, only to have them beaten by their opponents later in the day.
A cynical person could suggest that, rather than being an ominous omen, Australia has played some matches (six years apart) in conditions very friendly for batting, batted first, and played second best – and if only they’d batted second in a few of those matches, Conn’s article may never have been written. And given the “pattern” of the “trend”, might not the next set of nasty results for Australia come in 2019?
The comments from both captains might have made some writers reconsider, too. George Bailey merely “played down” the high scoring from both sides – no doubt with typically Aussie understatement, given that each side hit 19 sixes. Meanwhile, MS Dhoni said: “I honestly don’t know what is a good performance now by a bowler. 10 overs, 80 (runs) is a bad performance or 10 overs 60 is a bad performance or 10 overs 100 is a bad performance? That’s what we are still trying to find out.” Or, as Malcolm Conn interpreted it: Aussie bowlers choke again – just like six years ago!
So, no graphs or numbers to thrill you with this week – just another example of cherry-picking some data to pack out a story. In last week’s post, I said we might see a bunch of articles like – sadly, Conn’s “ominous omen” piece may be an omen all of its own.Follow @newstatsman