I heard that England are unchanged which is rather surprising, although I don’t know what the alternatives are. Warner seems certain to remain for Australia, so it may be Patterson in for someone- possibly Cummins? He has done a fair bit of bowling, but in the context of past series, hardly onerous.
Some totally irrelevant statistics. I fooled around with Statguru tonight regarding if batting has become that much easier- which I believed it had. So I put in scores of over 259 (should have been 250 but I am hopeless with a keyboard
and by decade the results are
2010 : 15
2000: 23
1990: 11
1980: 5
1970: 6
1960: 4
1950: 6
1930: 9
I didn’t go further back as there were to few Test playing nations to make it useless. So after those scores doubling in the nineties and doubling again after 2000 it has come a bit back. Of course there are factors such as RSA’s exclusion, Zimbabwe playing and also the player friendly Dubai. I think the answer is a pineapple.
If you’re using a count statistic like this, rather than, say, batting averages, you also have to take into account the total number of test matches being played.
1950s - 164
1960s - 186
1970s - 198
1980s - 266
1990s - 347
2000s - 464
This growth, and the increasing number of tests played between mismatched teams, probably accounts for a lot of the difference.
Covered pitches will have helped a significant amount too.
I think that is a fair assessment.
My quibble would be the descriptor “only others” … 5 of 11 is a pretty even share. 
Surely with more runs, no innings under 50 vs 3 ducks and almost twice the average of Joe Root, Marnus Labuschagne would be batting at 3.
Root’s a world-class player and I’d definitely want him in my side, but you’re right: if we’re looking at this series alone, with no knowledge of prior form, you’d probably put Labuschagne in ahead of him (nitpick: Labuschagne has had one knock under 50 - he got 11 in the second innings at Old Trafford).
To be honest, I’d think about keeping Root at 3 using Labuschagne as an opener. He’s had to come in early on multiple occasions, and has looked as comfortable as anyone except Smith and Burns against the new ball. I also wouldn’t give Root the captaincy, because he seems to bat better without it, and he’s a pretty crap captain in terms of strategy.
While we are on that theme:
From Cricinfo statsguru I took the match result of each Test played since 1930 and calculated the average runs per wicket (= average partnership or batting average)
For those of you inclined, past the figures in the four tables below into Excel and generate a graph of the batting average for winning, losing and drawn/tied Tests per decade.
Quite remarkable consistency given the changes in breadth, depth and conditions of the game
In precis, the Test batting average for the period is 32.6 (hence the batting meme, hit 350 and smile)
Batting average for teams that won is 40.1
Batting average in drawn/tied Tests is also 40.1
Batting average in lost Tests is 22.5
The gap between the average for Tests won and lost is 17.6 runs per wicket, which is 350 runs per match.
From 1950 to current that superior batting differential has been in a quite tight range of from 14.6 to 19.9.
Won
1930 41.2
1940 48.6
1950 35.6
1960 37.6
1970 37.8
1980 37.6
1990 37.8
2000 43.5
2010 41.6
Drawn/Tied
1930 39.9
1940 39.5
1950 36.2
1960 36.5
1970 39.5
1980 40.0
1990 39.5
2000 44.7
2010 42.6
Lost
1930 20.8
1940 23.4
1950 19.5
1960 23.0
1970 22.5
1980 21.5
1990 22.0
2000 23.7
2010 23.1
Total
1930 32.7
1940 35.8
1950 28.6
1960 32.3
1970 32.8
1980 32.6
1990 31.6
2000 34.2
2010 32.6
Is the relatively high average here a Bradman effect?
Australia won 14/18 Tests in the 40s with Bradman playing in all but the first one and last two but Test cricket was on hold until 1946 so maybe it’s a statistical anomaly.
To a large extent yes, but not exclusively.
AUS averaged 52.3 when winning Tests in the '40s.
Excluding Tests vs AUS, the rest of the cricketing nations averages 43.7.
There was also a fair bit of beating up on India post WWII.
ENG averaged 47.6 and WI averaged 58.2 vs India in that decade.
In that era wickets were prepared as roads on day 1 and deteriorated into cart tracks.
Win the toss and bat for 3 days was SOP.
This is a fair point, and on review, I discovered what the issue was with my (not written down at that point) composite side - I had 12 people in it, with one more batsman than I should have had.
The major issue for any composite side in this series is that one of the opener slots is essentially empty - I mentally slotted Denly in, even though I don’t really rate him and had Root at 3 because 3 is also a black hole in Australia’s line up - but then had one too many in the middle order with Smith, Labuschagne, Stokes, Wade and Paine.
Having someone open from this lot - Labu seeming like as reasonable a call as any - might be the way around it - but given Root has essentially opened all series for England, I think I settle on having him bat 2, since he’s been doing it anyway, and have Smith bat 3, since he’s been doing it anyway too.
So Burns, Root, Smith, Labuschagne, Stokes, Wade, Paine, Cummins, Broad, Hazlewood, Lyon - feels about right and gives England only 4 players in the side, which also feels right.
Can’t argue much with that selection - for me it really highlights where England missed their opportunity this series. We were never going to have a batsman as good as Smith, a spinner as good as Lyon, or (once Anderson departed) a strike bowling pair as consistent as Cummins and Hazlewood. But neither Wade nor Paine have had particularly good series, if Buttler and/or Bairstow had managed to show some of their previous form we would have been a lot closer.
Understand the sentiment and the lack of options at the top.
But I can’t pick a composite side with essentially the entire batting line playing out of position, and you can’t put in the team somebody who has been dropped. Elsewise Khawaja is the best option to open.
Lyon has filled his role adequately but despite bowling on some helpful wickets has not reached his own standards. I rarely pick a side without the best available left arm seamer but Broad bowling around the wicket is a viable replacement.
So with considerable angst and hope for some clarity to form at the Oval, my pick is:
Burns
(… to be filled by the opener with highest season aggregate after Oval Test) Denly
Root
Smith
Labuschagne
Stokes
Paine
Cummins
Siddle
Broad
Hazlewood
And just when you thought the shit was laid on a deep as feasible come the news that Warner retains his spot (because there isn’t a right hander who can open in the squad) and Head has been dropped for Mitchell Marsh.
Like what the AUS team needs is a part time bowler who can’t bat.
Siddle comes back. Tres grouse. Looks like it’s him or Starc for 12th, I’d leave Starc out.
I thought that Cummins would have been rested and Pattinson back in. Que sera, sera.
England have dropped Roy and Overton for the last test, replacing them with Sam Curran and Chris Woakes. Because of his shoulder problems, Stokes will play as a specialist batsman.
As for Australia, I’d have dropped Warner, who has been terrible this series, and opened with Labuschagne. As I said above, he’s had to face an almost completely new ball a few times already this series, and he can’t possibly do worse than Warner has been doing. And he bats right-handed.
Check out this catch.
https://twitter.com/wisdencricket/status/1171783933966594048?s=12
I mean, bloody hell.
Holy shit! I’ve seen some good two-man catches on the boundary, but that’s incredible.
The decision making is stunning.
He’s got to concentrate on performing what would have been a stunning enough catch if just done infield by himself, realising he’s going to need a relay throw to complete it, work out where his teamate is and then as he doesn’t have the time in mid dive to throw the ball back in forhand and so flicks it backhand in one motion of taking the catch.
Wowsers!
Marsh. They are doing this to make me froth at the mouth.