I kept hearing something strange during the cricket yesterday. Apparently Peter Siddle has cast some intense mind-voodoo over the hapless Kevin Pietersen.
When Pietersen fell to Siddle this afternoon, it was a worthy statistic that this was the 10th such occasion in Tests. It is an achievement against an excellent player.
It far exceeds the record of other bowlers: the next best are Muttiah Muralitharan and Brett Lee, who got Pietersen six times; then totals of five from Glenn McGrath, Shane Warne, Morne Morkel, Saeed Ajmal, and Sreesanth.
Wow, people said. He’s so far ahead of all these great bowlers and Sreesanth. After they’d finished searching for hilarious ways to credit vegetarianism and bananas, they started wondering more genuinely what it is. What is this magic that allows Siddle to tame England’s big dog?
So I’ll tell you.
It’s nothing.
Peter Siddle is an excellent pace bowler. He’s not as outrageously threatening as Mitchell Johnson or Ryan Harris, but he’s courageous, skilful, and durable. He and Michael Clarke have been Australia’s only consistent performers over a fallow few years. He has learned at Test level and prospered.
But all he is delivering Pietersen is good, solid, Test-level bowling.
The people stitching these dismissals together are hunting causality in correlation. It’s an unsound premise. Pietersen’s dismissal today was that of a man lost.
He’d endured 59 balls in the blazing heat with barely a stroke played, scraped up 19 runs, then lost his rag and tried to show some dominance.
He picked a horrible ball to do it to, and a horrible shot to do it with. He might have been a man trying to get a squirrel out of a tree. But he still looked like he was going to clear mid on until Johnson proved to be half elastic, stretching so ludicrously into the sky that he deserved a wacky sound effect.
Pietersen caught at mid on pulling a rank wide short ball. The previous dismissal, in Adelaide, he was late withdrawing the bat from one just back of a length on off stump, nudged back onto his wicket. The one before that, he clobbered a half volley to a short midwicket catcher.
Before that it was Manchester earlier this year, where he played at one that was given caught behind despite no third-umpire evidence of an edge. Before that, Lord’s, where he thrashed a full wide ball to point on the square drive.
My point being? There is no pattern, no consistency, to Siddle’s recent dismissals of Pietersen. The batsman is not getting out because he’s worried about the bowler. The bowler is not preying on a particular weakness of the batsman’s. This is not McGrath to Michael Atherton.
Siddle is bowling the kind of deliveries that can get wickets. He is actually conforming to the adage of putting the ball in the right areas.
But there is nothing specifically about Siddle that is making this happen. He’s not distinctive. He essentially bowls a style of delivery Pietersen has faced tens of thousands of times.
Let’s have a think about statistical probability. If a contingency arises a lot of times, one reasonable explanation is that its requisite context is equally common.
If Siddle has dismissed Pietersen more than anyone, that might just means he’s played against him more than anyone.
Holy discovery. The pair have played 15 Tests against each other, with Big Kev walking to the crease 25 times. He only played 10 matches against Lee, Warne and Morkel; 8 against McGrath and Sreesanth; and 6 against Ajmal and Murali.
In which case, a percentage of dismissals per innings would be far more illustrative than a raw total. Siddle has got Pietersen in 40 percent of their opposing innings. Lee had 30 percent, McGrath 31, and Sreesanth 36, not big differences when dealing with such small data sets. Siddle has had a good run, but is still behind Ajmal at 50 percent, and Murali’s 54 percent.
If you want someone with a hold over KP, then, call the spinners first. But wait, Geoff, you say passionately, wiping some cream cheese from your blouse where you’ve dropped your salmon canapé in agitation. Peter Siddle has dismissed Pietersen 10 times and conceded 174 runs.
At an average of 17.4 per dismissal without even getting a calculator, doesn’t that imply complete domination?
Well, let me tell you that you’re beautiful when you’re angry. But while averaging 17 sounds great, a player isn’t just facing one bowler, while only one bowler can dismiss him. Let’s say he does equally well against all four bowlers, taking 17 from each. He’d average 68 against the group.
In matches involving Pietersen, Siddle has bowled 467 overs, just under 20 percent of the team’s total. So in a perfect statistical world, Siddle would dismiss Pietersen 20 percent of the time, not 40. But random sequences don’t work like that.
In a different life, I used to run a roulette table.
You would spin the ball and land red or black, with the occasional zero. Over a long period the results would be roughly even, but that didn’t mean you couldn’t have long streaks of one colour.
Whatever the streak, the odds for the next one being black remain exactly the same. The most manipulative invention on a roulette table is called the tree, which displays the last 14 numbers to have been spun. This makes people see patterns that aren’t actually there.
Once you rolled five blacks, then six, then seven, people would start to get interested. Roll ten and you were pulling a crowd. By the time you got to 14, filling the board, people were losing their minds. They would hammer red, sure in their bones that the streak had to break.
22 black, you would call. They would stack up on red. 17 black. Higher on red. The fizz of the ball. The tension like a long day waiting for a thunderstorm. Number 8, black. The clatter of chips as you razed small skyscrapers and they dived to build more.
My best run was 23 blacks in a row. The table was like a football match, all sides yelling and cursing disbelief like I was an umpire. A few quiet achievers sat there, riding the streak, each time sliding a few more chips on red. Both camps were wrong, in thinking that a streak even existed. It was simple probability, both results equally likely, anomalies par for the course.
Peter Siddle is as likely as anyone. He will bowl for his team until his body fails. He is good at it. It is plausible that he will get people out by doing so. It is plausible that the frequency may be anomalously high, or elsewhere may be low.
As for Pietersen, he’s not currently in very good shape.
He may retain the pouty, chest-out posture that implies he wants you to admire his nipples, but he doesn’t currently have the batting swag to back it up. On the other hand, he’s a guy who has three double centuries and 20 other hundreds. He’s no lost cause.
The real evidence came on the field today, when that shot was played, when that catch was reeled in. Peter Siddle’s face was like nothing I’ve seen on a cricket field. He clenched both fists, planted his feet and roared.
It was like all of his own skin was trying to crawl off his head. He seemed to be more tooth than man. He was a white-flanelled Langolier in bad ‘90s CGI.
That face was many things, most of them terrifying, and one of them the reason I’m still awake and writing at midnight. What it wasn’t was the face of a man enjoying mastery over a shattered opponent.
It was a bowler expressing the relief at having got rid of the opposition’s most dangerous player, and enjoying utterly his good fortune. It was the face of a man who knew the odds had gone his way.
Geoff Lemon is a writer and radio broadcaster. He joined The Roar as an expert columnist in 2010, writes the satirical blog Heathen Scripture, and tweets from @GeoffLemonSport. This article was first published by Wisden India, in a new-founded Ashes partnership.
Richard
Guest
So Siddle's got Pietersen out ten or eleven times,or whatever it is.I'd like to know Pietersen's score for each of the dismissals.There's probably a few high scores in the tally.I wouldn't exactly be calling him 'Siddle's bunny' just yet
Richard
Guest
So Siddle's got Pietersen out ten or eleven times,or whatever it is.I'd like to know Pietersen's score for each of the dismissals.There's probably a few high scores in the tally.I wouldn't exactly be calling him 'Siddle's bunny' just yet
ds
Guest
The runs count against Siddle's average regardless of whether he gets him out or not; such as when Lyon got him out yesterday.
deccas
Guest
i'd love to see that article though!
deccas
Guest
well 25% is the base percentage (4 bowlers, all with equal chance of dismissing a batsmen), so its still high but not unbelievably so. If you then factor in how well KP plays spin you could decrease the spinners chances by a bit, lets change that players chances to 15% and put the other 3 up to 28% (or right 28.33 but its a bit of a bung up job). Then you put that over the 15 tests that siddle has played against for australia he has been our best bowler in most and you adjust the percentage chance around to reflect that you are getting up toward 35, 40%.
deccas
Guest
i think the way to look at how good someone is is to firstly compare their batting average to other batsmen of their era, then look at the standard deviation from the mean batting average for top 6 batsmen over that era say 15 year periods. How much better they are than the aerate of their day is how good they are, when we assume that people are not fundamentally better now than then (or vice versa) which i think is a safe assumption.
Wasim Ranamadroota
Roar Pro
James P and Tock, those are exactly the factors I believe the ICC batting and bowling rankings take into account, except over the most recent year or two, not over a whole career.
Wasim Ranamadroota
Roar Pro
Me too. Statistics should be used with caution at the best of times, and certainly don't tell the full tale here. But Geoff has been ABUSING statistics this time. I would love to have a crack at his roulette wheel if he thinks you can magically multiply a stat by 4 and have it mean the same thing. Think that means he would pay 8:1 for Red or black - I like those odds.
Steve
Guest
Agree - big fan of Geoff usually, but Geoff, like David Lord, have a tendency to not have a clue what they are talking about when it comes to understanding and interpreting statistics.
Matt
Roar Rookie
A day late here, but the ignorance of statistics and common sense in the asssertion "He’d average 68 against the group" is absolutely staggering. Do "Experts" on The Roar ever get relegated to the right hand side when they publish something like this?
Josh
Guest
Stats are way off. Pietersen has scored 174 runs off Siddle IN ALL INNINGS. Not just the ones he was dismissed by Siddle in. Which is the reason you can't multiply 17 by 4.
moreton bait
Guest
Or perhaps this explanation offered by Michael Vaughan: "In Pietersen's case, Australia has played on his talent and ego. It stationed two men on the leg-side in the first four innings and his inability to say, ''right, I am going to play elsewhere'', was his downfall. His ego told him to play the same way and he was caught out twice. Australia has restricted his scoring by bowling as many dot balls as possible and then tested his ego with the odd short ball outside off-stump to waft at. Peter Siddle loves bowling at Pietersen. He bowls tight and regularly lands the ball on the imaginary fourth stump outside off. He tempts him with a short ball knowing he will go for the shot. Whoever came up with that plan has cracked Pietersen and the execution has been impressive." Sounds like a plan to me!
SandBox
Roar Guru
Hitchens died two years ago to this day. "Because we are pattern seeking mammals. It's part of our evolution. We look for patterns. We're designed to look for them. And, if we can't find a good explanation we'll come up with a bad one, rather than none at all. Most people would rather have a conspiracy theory than no theory. It's very observable that. There's a lot of junk science around before good science arrives. Before we have Astronomy we have Astrology. Before we have Chemistry we have Alchemy." Christopher Hitchens 1949-2011
expathack
Guest
Love to play poker against you Cam.
expathack
Guest
Hilarious, the guys are still coming down below the line, trying to argue the point. They're published writers so they must always be right. About everything! Including the stuff they apparently have no idea about. Geoff got one of the main premises of his article horribly wrong. That kind of thing is bound to happen when you don't understand what you're writing about. He's been called on it, and I'm sure he'll suck it up and move on. He likely doesn't need his fellow writers coming down alos making fools of themselves as well in a weird show of above-the-line solidarity....
James P
Guest
As JGK says. Based on the stats (using the right ones this time) Siddle does better against Pietersen than he does against the average batsman. Given the large number of matches that they have played against each other, this does suggest that Siddle has been more effective against Pietersen than we would expect and conversely, Pietersen struggles somewhat against Siddle. 17 does show some level of dominance but I'd like to see under 10 before claiming bunny status
Blaze
Guest
Hahahaha
Tony Loedi
Roar Guru
I just checked England's stats for the last year and a half and besides the tour in India they are all averaging less than 40 with the bat except for Prior. And that includes Bell and Cook which I found odd. And now that Prior is horrifically out of form I think England are a real worry with the bat. I can see them put up some performances like the aussies have been dishing out the last 2 years.
jammel
Guest
I kind of like Root at #3 - well I would if I was an Englishman. Does anyone else think perhaps KP should bat at #5 for England? If I were them, I'd have Bell at 4 and Pietersen at 5…. Not sure if that's something that Pietersen would agree to though...
bryan
Guest
As mentioned above your stats are completely off with your 4x17 = 68 thingy. I would also like to take to task your assertion that "The bowler is not preying on a particular weakness of the batsman’s. This is not McGrath to Michael Atherton." I think that is utterly wrong. I think the brains truss, and Siddle have an underrated but amazing plan against Pietersen. They are bowling back of a length, 4 inches outside off, with 2 short mid-wickets, a point and a gully. This has the effect of drying up runs, ask Glenn McGrath. It also has the effect that you cannot hit the ball often, without being overly negative. The bowl is still near the batsmen, just there. But you are still best to leave it alone. Lets look again at the wickets, and sort them out a bit more. caught at mid on pulling, clobbered a half volley to a short midwicket catcher, thrashed a full wide ball to point late withdrawing the bat from one just back of a length, given caught behind Siddle is very very clearly bowling to dry up and annoy KP, and attack his biggest weakness, that Ego. 2 of those 5 are very similiar balls outside off stump, that cannot be hit. However, he has tries anyway. Also, Siddle is a good enough bowler that the "Half Volley" or full and wide are deliberate balls, bowled directly to a KP hitting zone with multiple fielders waiting for a mistake. To further illustrate the point, I'll lay money that the majority of those dismissals have either Watto or Lyon down the other end. They are stopping KP from scoring, annoying him, and he is attacking balls that are not there to hit, or hitting to areas that are very well protected. Its the old Warne play. Bowl a long Hop legbreak that gets cut for 4, the bowl the same ball 2 balls later, but its the flipper and traps you LBW. I believe they are convincingly setting up KP, and getting their man continuously. In conclusion, I actually think Siddle is the a hugely important part of this team, and the hardest to replace. As soon as England start scoring any runs, on comes Watto, Siddle and Lyon to slow the rate down for 10-15 overs. Back comes Harris and Johnson, and wickets suddenly fall. This is mainly effecting: Carberry (3 times to 40, 3 times slowed, 3 times out KP: strike rates of 50, 42.85, 33.33, 53.33, 32.3 Cook twice to 60 and twice slowed, twice time out