YES! NO! WACKY THEORIES!

Australia has suffered the most run outs in the last four years:

Run-outs since 2009:

Australia: 25
India: 20
Pakistan: 20
Sri Lanka: 19
South Africa: 18
England: 17
New Zealand: 17
West Indies: 13
Bangladesh: 11
Zimbabwe: 3

My theory? (Diligently researched and peer reviewed, naturally.) Throughout the 1990s and into the 2000s, Australia played aggressive cricket against lesser fielding sides, relentlessly challenging opponents to run them out, all the while banking big totals which rendered run outs less costly, while confident McGrath and Warne would always dismiss the opponents for less. But as Australia dominated and then dropped off the pace, other sides began to emulate Australia's aggression in the field and have become better at running out the still aggressive and now perhaps even blasé Australians. Australia does not now have the runs to waste wickets on run outs, nor do they have the bowlers to put the frighteners up the opposition. Was Gary Pratt's run out of Ricky Ponting some kind of beacon? Australian batsmen, crusing along at the crease, appear to wear run outs worse than its opponents. Do Australians suffer a sence of injustice when run out? Have opponents decided that running out Australians really get under Aussies' skin? Or maybe we've just played more Tests. Someone look that up. Fingers crossed Watson, Cowan and their chummies don't reduce us spectators to jibbering wrecks as they give their wickets away with absurd, Test-killing run outs. Katich in Adelaide, anyone?

Comments

[Russ]

The correct method for this type of analysis is, I think, run out per run run, thus excluding boundaries and adjusting for time at the crease. Also ignore the lower order, who aren't worth as many runs.

Hey! I have done that analysis already. Australia's running has always been worse than England's.

Caveat: there aren't many data points in the 50/60s.

If my database was up to date I'd identify culprits with a partnership analysis, but suffice to say when I last did it, Cook and Trott were amazing.

[Cameron]

An interesting stat would be a the ratio between difference between offense (batting) and defense (fielding).

[gillysgone]

more interesting to see who is doing the running out. I reckon there will be revenge runouts to balance the books. I will also make this prediction - whoever has less runouts will win the Ashes.

[m0nty]

Russ is the Nate Silver of the AGB.

[lou]

Russ, Cook and Trott are not fast and they are cautious players, so they'd not try and nick unsafe singles.

Punter was the worst culprit in the Aus team according to cricinfo since 2009, but I'll bet Watson was the cause of a lot of the run-outs as I swear he doesn't even call sometimes.

[Russ]

Too kind m0nty. Though I did start a sideline in predictions. makes for grim viewing.

[Tony Tea]

Athers did a man-for-man for News, which also makes for grim reading. Although, he could just as easily have matched Cook-Watson & Rogers-Root instead of Cook-Rogers & Watson-Root. Mind you, it would still be one-all, the openers.

[Oh bugger, they're all big sheilas]

Our batting is laughable if you look at stats. The ONLY thing in their favour is lack of tests to have built those stats. And that's not really a help.

[Big Ramifications]

And I'm the Long Dong Silver of the AGB.

"The correct method for this type of analysis is, I think, run out per run run, thus excluding boundaries...."

Russ, does that mean *AGH FORGEDDABOUDIT* I just realized "run run" isn't a typo and I've answered my own question.

I love the quirks of your 2nd graph. It's a bell curve.... but with lots of divots in amongst the data points.

eg. A 0-2 result is 5 times less likely than the preceding 1-2 result!! A big downward step, even though the %%% are clearly trending upwards at that point.

// Did you use the binomial distribution there, Russ?

//// Those quirks or "divots" I mention. Is that something to do with the probability distribution, or is it a result of the order in which you labelled the Y-axis? Or something else, mebbes?

[Russ]

Biggy, I couldn't work out a nice way to order it, taking into account draws. So I just left it as I calculated in the spreadsheet. You can think of it as 6 overlapping bell-curves, with 6 different draw numbers OR 6 non-over-lapping bell-curves with 6 different Australian win numbers, or etc.

Talking to myself here, probably, but if you look at the most likely outcome (the peak of the curve), iterating over the numbers for different variables you get:
Draws: 4-1,3-1,2-1,2-0,1-0,0-0
Aus Wins: 4-0,3-1,2-2,1-3,1-4,0-5
Eng Wins: 0-3,1-2,2-1,3-1,4-0,5-0
If you are a betting man 3-1 England gives a nice return. But that's the only result worth looking at.

I used a monte-carlo simulation, based on my rankings, with a set 25% draw rate (which is close to the 20 year average in England). The last two series in England have been 2-1, and the last in Australia 3-1, so it passes the eye test.

[Russ]

Athers gave a 3-3 draw in the top-6? Madness. Or a sign of the deceptiveness of such measures. I'd put Root over Watson, but if you reversed it, it is a near draw and Cook miles ahead. Same with Trott and Pietersen. Clarke can't make up the short-fall, even if Warner somehow bests Bairstow.

Don't agree on Swann/Lyon either. Lyon is very under-rated. Swann's main advantage is bowling to muppet batsmen and over his career, in better conditions. Finally, I can't for the life of me understand why Bird will be omitted. But that is for another day.

[Tony Tea]

I have said plenty of times that our attack is very soft if Siddle is our main man. That hold if he is our "Go-to man". Not saying Siddle is automatically out of our best four quicks, but injuries notwithstanding I have Pattinson, Harris, Bird as 1-2-3 with Siddle and Starc scrapping for 4. Starc wins if it can be guaranteed he's not going to distribute pies.

[Russ]

Starc, f/c average: 31
Faulkner, f/c average: 23

Faulkner is probably a slightly superior batsman too. Starc shouldn't even be in the equation but as usual it is X-factor over results. Maybe in Australia where raw pace is an asset, and in ODIs where pace is definitely an asset, but a test in England?

I'd go with Bird, Pattinson, Faulkner, Lyon with Butterworth, Siddle, Starc as respective backups. I don't trust Harris to last a whole test. Good bowler, but without a sub option he is a big risk.

Same with batting incidentally. Rogers, Cosgrove and Clarke should be the spine. Watson might justify his place if bowling. Hughes, Khawaja and Warner all have better records though. Watson is no better technically than any of them, gets stuck on the front foot too easily - that test average of 30 tells you something. Cowan's main skill seems to be to stay in longer, but he doesn't score any ore runs than the others, so I'm not that fond of him. With him and Rogers opening we'd see off the new ball more often than not though.

[Tony Tea]

To be perfectly Francis, I'm not too familiar with Faulkner's work. And naturally Harris comes with an *.

[lou]

Faulkner has got mongrel for want o a better word. Far more than Starc.

[Tony Tea]

Stark is too Mitchell.

[Big Ramifications]

"I used a Monte-Carlo simulation, based on my rankings, with a set 25% draw rate."

I'm an idiot. Binomial distribution? When there are 3 possible results?

// more if you want to split hairs with a draw / tie and factor in various "no result" endings

[m0nty]

Me and me mate Carlo have a system far superior to that Duckworth Lewis rubbish.

[Tony Tea]

Carl0.