What do you do with a broken spearhead? That's the question for England as they try to pick up the pieces after Brisbane.
There seems to be a widespread assumption that Steve Harmison will play in Adelaide. But it isn't his kind of pitch: it's too flat and too slow, with short square boundaries just inviting the cut and the pull, if the batsman can reach the ball. And he is in his worst form since the beginning of the last Ashes tour.
It's certainly possible that everything will suddenly click back into place, as it did at Old Trafford in July. But would you bet on it? Harmison's troubles at the Gabba didn't come out of a clear blue sky. This is the third winter in a row that he has struggled: in his last 11 Tests for England overseas, he now has 27 wickets at an average of 52 and a strike rate of 96, with no four-fors, let along five-fors. His action, his radar and his confidence have all gone, and there is no Troy Cooley in the camp to put them right.
The main argument being advanced against dropping Harmison is that it would be handing the Aussies a psychological boost. There's some truth in this. But last Thursday, the same effect was achieved by playing him.
England need to ask themselves a simple question: who should take the new ball? Which pair of bowlers offers the best hope of those vital early wickets? The first choice would surely be Andrew Flintoff himself, reluctant as he is to push himself forward. The second, Matthew Hoggard. He may have been nervous and toothless early on at the Gabba, but he bounced back later with a superb little spell of 2 for 7 in six overs, he has conquered most conditions over the past few years, and he is a natural foil to Flintoff - different length, different pace, different shape. Last time England won abroad, in Mumbai in March, these two opened the bowling.
There seems to be a widespread assumption that Steve Harmison will play in Adelaide. But it isn't his kind of pitch
In a perfect world, there would be the option of Jon Lewis, the English Stuart Clark. In a perfect world, there would also be a spare top-order batsman, so that England could move Ian Bell back down to six, the already overstretched Flintoff to seven and Geraint Jones to eight, while replacing Jimmy [James] Anderson with Monty Panesar and putting their trust in four skilful bowlers rather than five rusty ones. On this scenario, they would have to pick their four best bowlers, and that wouldn't include Harmison - it would mean Flintoff, Hoggard, Panesar and Sajid Mahmood.
Last time England played at Adelaide, they won the toss, batted, waltzed to 246 for 2, and still lost by an innings. That was with five bowlers. There is a mysterious thing about the fifth bowler: it often doesn't help. It worked a treat in England in 2005, but then Simon Jones was fit and firing as the Platonic ideal of the fourth seamer. Since he succumbed to injury, England have been unbeaten with four bowlers (Oval 2005, all four home Tests v Pakistan 2006) while losing more than they have won with five (the nine Tests in between, plus the Brisbane drubbing - won two, lost four).
Four bowlers would rule out two spinners, unless England were prepared to have Paul Collingwood as the third seamer. And it would be taking a risk with Flintoff's fitness. But that risk would not be as great as the planned punt on Harmison's form. And the negative message sent out would not be as bad as the one sent out last week, as loud and clear as all those tedious PA announcements, by the dropping of Panesar.
I'm not saying they should drop Harmison - I'm saying they should rest him. Send him ahead to Perth and take up Dennis Lillee's generous offer of remedial work. He can have a break, work with Lillee, and chat to Michael Vaughan, who handles him well. He can also acclimatise ready for the third Test, which will be played on a surface that, even if it is not what it was, should suit him better than Adelaide. And Flintoff will have one less problem on his plate.