Australia's bowling woes

The last time I gave my opinion on Australia's selection (suggesting that Krejca wouldn't take wickets) it was obviously rubbish. So here goes for a second attempt. Shane Watson is never good enough to play as a 3rd seamer....although apparently Steve Waugh thinks he could be. For me he's a good all-rounder in one-day cricket but he falls dangerously close to being not quite good enough at either discipline in Tests. Averages don't tell the full story (unless they are real averages!), but Watson averages 22 with the bat and 37 with the ball - not good enough at either, but admittedly in only a 7-match career.

He still has plenty of time to come good as he's only 27, but if he's going to play for Australia in Tests surely it has to be as a batsman that bowls and not the other way round.

Alastair Cook's career real average

Having seen that Alastair Cook was the recipient of much fortune over the last 4 England series, I checked back over his short career to see how much luck he'd had. (For Cricket Burble purposes luck is defined as the difference between a batsman's official average versus their real average - their real average taking into account dropped catches and wrong umpiring decisions). Over the last 4 series the difference for Cook had been 13 runs - by far the largest in the England side. To see the full list for England over the last 4 series, click here.

Cook has had positive luck of 11.5 runs per innings over his international career - a huge stroke of fortune. I hope for his sake that his luck doesn't turn this English summer as he's not had the greatest run - even given the luck he's had - of late.

England Test Real Averages since summer '07

Having updated the real averages for the New Zealand series (which, incidentally, placed Strauss top with a real average of 54.8 and Vaughan and Cook the lowest of England's specialist batsmen with real averages of 20.5 and 22.7 respectively), I've totted up the total for England's last 4 series. In that time England have played West Indies and India at home, and Sri Lanka and New Zealand away. The overall table is shown above and makes for interesting reading.

Only three England players can consider themselves to have been unlucky overall when batting in Tests over these four series - Pietersen, Prior and Bopara. By unlucky I mean that the consequences of wrong umpiring decisions have reduced their official average to a greater extent than dropped catches and wrong umpiring decisions have aided their official average. Of those three, the interesting thing to note is that two of them - Pietersen and Prior come out first and second in the list of top real averages with 51.7 and 42 respectively.

The luckiest player - that is to say the player who's official average has benefitted by the greatest amount from wrong umpiring decisions and dropped catches - by a decent margin, is Alastair Cook. While his official average from these four series is 43.7, if it wasn't for dropped catches and wrong decisions he would in fact average just 30.7 - a very significant difference of 13 runs per innings. The next luckiest are Vaughan and Collingwood who have both benefitted by just under 8 runs per innings.

Strauss and Bell have been very slightly lucky but with Strauss' real average being 1.6 less than his official average, and Bell's being just 0.5 less, overall luck seems to have dealt them a pretty even hand - I'm sure Strauss wasn't thinking that as he trudged off in the 2nd innings in Wellington having been wrongly given out LBW by Steve Davis to a ball from Jacob Oram that pitched outside leg stump!

My take outs from these real averages over the last 4 series are (a) that Pietersen is head and shoulders above any other England batsman (hardly a revelation but his real average above doesn't take into account the poor luck Pietersen had in the 2nd Test in New Zealand when he was run out backing up at the bowlers end), (b) that if England want a batsman/wicket keeper (in that order) Prior was very unlucky to be dropped (although I still think Read should be playing and batting at 8 or 9 ideally if/when Flintoff is fit), and (c) that Cook is leading a little bit of a charmed life. That doesn't mean that I'm calling for him to be dropped, but it should be remembered when the media are considering Strauss' - and others - positions in the side.

What are your thoughts?

"Real" averages updated

"Real" averages for the first two Test matches in New Zealand are now available here. By taking into account dropped catches and poor umpiring decisions, we're able to get a better understanding of how much luck a player has had. For England supporters, key things to note are that Collingwood is top, helped by the fact that his 66 in the 1st innings at Napier is considered a not out by real averages given that he was given LBW having hit the ball, but also hindered by the fact that his 59 in the 2nd innings at Wellington is considered a duck, given that he was dropped on 0.

Pietersen comes out as 7th in the England averages, whether looking at real averages or official averages - time for him to come good in the 3rd Test! And Strauss is 4th in the England real averages, aided by the decision to give him out LBW to Oram when the ball pitched outside leg stump, which gives him 44 not out in the 2nd innings at Wellington. According to real averages Cook, Vaughan and Pietersen have all performed worse than Strauss - interesting given that he is an opener and the two openers are both performing worse than him (not that I'd consider for a moment dropping Cook or Vaughan).

Having said that, this is only two games and you can't read too much into only two games. Real averages will be updated for the series on Saturday 29th March.

More ways of calculating batting averages

Loyal readers of Cricket Burble will remember John Wright's post about alternative ways of working out averages, many months ago: "A Mean Average". And of course you'll also know about Cricket Burble's attempts to reduce the role of luck with "real" averages. This post on a Cricinfo blog also talks about alternatives, the suggesting being to create an "EBA" or Extended Batting Average. It's equally - or perhaps even more - flawed but interesting nonetheless.

Sri Lanka v England "real" averages

It's taken a while but I've finally got round to pulling together the "real" averages from the Sri Lanka v England series. Somehow a 3 match series doesn't seem to do justice to the players - someone like Pietersen was only actually out 4 times but even his real average is only 31.5, with his official average being 25.2 after the catch that never was, meaning that officially, he was out 5 times. Likewise Bopara undoubtedly goes down as a failure in his first 3 Test matches having officially averaged 8.4 and only slightly more than that with his real average. But of the 5 times he was out, one was not out where he was given LBW, and one was run out, so in effect he was got out by the Sri Lankan bowlers only 3 times. I hope he isn't discarded based on his performances in Sri Lanka.

For Sri Lanka, Sangakarra only has a "real" average of 48.3 as opposed to an official average of 72.8, and Vandort suffered a poor decision so while his official average is a highly respectable 53.3, his "real" average is 71.

England will be hoping that the New Zealand series brings a better set of batting averages - only Bell and Prior had a "real" average of over 40, with the next best being Collingwood at 32.8.

Check out the latest "real" averages here.

Dhoni drop crucial

The "real" averages and "wrong" decisions pages have now been updated to include the England v India Test match that ended on Monday.

Michael Vaughan has come in for some stick for a slow over rate (the theory being that it was his field placings that were slow), but what I hadn't picked up on was that Dhoni was dropped on 14 by Collingwood. He is England's best fielder but in this case Collingwood's mistake cost England badly. If he had taken that catch, England would almost certainly have had the game sown up before the rain and bad light came, and Vaughan wouldn't have been criticised.

Clearly as a result of the above, Dhoni's real average is just 7, as opposed to his official average of 76. Strauss is similarly affected having been dropped on 43 in his first innings 96. And Cook and Prior benefit from real averages, as they were both given out LBW to wrong decisions. Rahul Dravid suffered the same fate, but this hasn't affected his real average much. You can view all real averages here.

(Yet) more on batting averages...

As John originally posted about, a theoretical argument for working out batting averages in a different way to try to take into account the lower risk of being out at the end of your innings the lower in the batting order you bat has been put forward.

Dave (as a lower order batsman himself), clearly felt that this way of calculating a batsman's average was unfair on lower order batsmen because they are often going well and then have to end their innings not out, only to have to play themselves in again the following innings. He felt that even if lower order batsmen are truly considered not out for the purposes of working out their average, no matter how many balls they have faced, they have the disadvantage of having to "start again" rather than being able to take advantage of the fact they have played themselves in like a top order batsman would.

This, along with the disjointed nature of the current England v India Test match, got me thinking. Every time the players come off for rain or bad light, or even at the end of a day's play, the batsmen have to play themselves back in again. Given that I assume it's a given that batsmen find it easier to bat the longer they have been in (up until a certain point when fatigue kicks in), shouldn't an innings that has involved 4 restarts gain greater reward than the same innings total scored all in one go?

That way a lower order batsman who is regularly not out is given reward, in that he gains greater value for his runs given that he has had to restart. No balls faced issues - just number of restarts, including not outs. And it is equally applicable to every batsman no matter where they bat in the order. What do you think?

Real averages

Just updated the real averages again. As predicted Chanderpaul's average is severely affected - yes, he's played well, but if England could catch he wouldn't have ended up with an incredible official average of 148.7, but a decent average of 44.2.

And who would have ended up top of the averages if they were calculated as real averages? Michael Vaughan. The top 3 would have been Vaughan, Prior, Pietersen as opposed to Chanderpaul, Pietersen, Prior.

Real averages are always available as links on the right hand side of the Cricket Burble blog. Does this not give a far better indication of how well the various batsmen have played?

Cricket Burble lists updated

I've updated the Cricket Burble lists which are always available if you click on the links to the right of the page. Interestingly, Chanderpaul's "real" average is less than 50, despite an official average of 120 after the first 3 Tests of the summer. The difference between his real and official averages is likely to increase even further after his not out century in the first innings at Chester-le-Street having been dropped on 9. So while he is batting well, he is also capitalising on large slices of luck. You can view the list of "real" averages here.

The 3rd Test featured just one decision that technology proved was wrong. There were several LBW appeals turned down where the ball was just clipping the stumps but, as explained on the "wrong" decisions page, decisions are only considered "wrong" where there is a large degree of certainty. You can view the list of "wrong" decisions since the 2007 World Cup here. The main question to be answered is "would England still have won at Old Trafford if Cook had been given out on 12 as he should have been in the 2nd innings, when he went on to make 106?" As they say, "you decide".

West Indies fighting hard

I'm amazed at the determination of the West Indies to go down fighting in this Test Match. Yes, England have dropped catches and not helped themselves, but the fight that West Indies have shown has been a delight to see given the capitulation at Headingley. No-one could have predicted at the start of their innings that they would have reached lunch on the final day requiring just 76 runs with 3 wickets in hand. An exciting game was much needed in this series, but as an England supporter I can only hope that England still come through to win.

It's going to be a time-consuming task updating the "real" averages and "wrong" decisions pages after this Test Match - rumour has it there have been a few bad decisions and there have certainly been some dropped catches - it will be interesting to see how the various individual's real averages change.

"Real" averages

Given that the cricket match I was meant to be playing in today was washed out, it has allowed me to be very sad and calculate the various "real averages" of the England and West Indies players over the first two Test Matches so far. What are "real averages"? Well I've started to think that far too much of a player's success or failure is based on luck over recent months and years. Their batting average can of course be affected by the standard of bowling they face, the quality of the wicket, the speed of the outfield, the quality of the oposition captain (i.e. does he place his fielder's in the right place), the quality of the ground fielding, the quality of catching and the quality of the umpiring. Most of these can't be accounted for - or at least they could be, but it would take a lot of analysis and work. The last two, wrong decisions by umpires and dropped catches can be accounted for pretty easily, so that's what I've tried to do below.

The person most affected from England is Collingwood - this isn't an attempt to slate him - if you have read Cricket Burble for a while you'll know that I'm a big Collingwood fan. But his official average is very different from his real average as he's been dropped in every innings he has played so far in the series. For West Indies Morton has been affected twice by real averages. His scores in the series are officially 14, 5 and 25 but given that he was dropped on 1 in his innings at Lords and he was incorrectly adjudged out by Rudi Koertzen in the 1st innings at Headingley, his scores which create his real average are 1, 5 not out and 25.

AVERAGES AFTER FIRST TWO TEST MATCHES, ENGLAND v WEST INDIES 2007.

RA = Real average
OA = Official average
* = not been out in series yet, so has no OA

England:

Pietersen, RA = 115.33, OA = 120.33
Prior, RA = 108.5, OA = 111
Vaughan, RA = 103, OA = 103
Cook, RA = 70.67, OA = 70.67
Bell, RA = 58.5, OA = 58.5
Plunkett, RA = 33, OA = 44
Strauss, RA = 24, OA = 24
Collingwood, RA = 22.67, OA = 58
Shah, RA = 5, OA = 5
Harmison, RA = 2, OA = *

West Indies:

Chanderpaul, RA = 74, OA = 74
Bravo, RA = 36, OA = 43.67
Sarwan, RA = 35, OA = 35
Ganga, RA = 31.33, OA = 31.33
Gayle, RA = 30.67, OA = 33.67
Ramdin, RA = 22.33, OA = 23.67
Powell, RA = 22, OA = 22
Taylor, RA = 22, OA = 22
Smith, RA = 21, OA = 21
Morton, RA = 15.5, OA = 14.67
Joseph, RA = 7, OA = 7
Collymore, RA = 2, OA = 2

I'm still to consider how to create real averages for bowlers...