Bradman effect

[the big bambino]

Really none of this should be surprising. Bedser and Tate are recalled with equal fondness. Besides DGB faced both. Larwood and Tyson both won a famous series for Eng (though Tyson has the better figures until you adjust for Bradman's effect and Tyson's gimmes against lesser teams). No surprises that Allen, Farnes, Bowes and Voce adjusted figures are comparable to Trueman, Statham and Bailey or Geary matches Brown. I dare say a similar pattern would emerge when rating the spinners too.

You only assume what you do bcos the post war group had more opportunities than the pre war group. And that fits in with the modern preference of rating players after they have played a large no. of tests. But that is the only difference btwn the group. Post war had more opportunities and therefore looks more credentialed. But in quality there is no difference btwn the 2 groups at all.

But there is something even more interesting about pre conceptions which shows up when looking at DGB's series against SA.

 

[MartinB]

To do a proper review you'd have to go through series by series and pick the best man in form. I have a few names like Rigg, Richardson, O'Brien, Nitschke in mind. But I made Arthur Chipperfield stand in all their stead as he was the one the selectors most preferred and had the best ave against England of the lot: which was about 29. If you add that score back in then the overall averages of the pre war group move up a little. I think its a less than 0.1.

I wouldn't suggest that a Lot of work for very little gain

Even that should be adjusted by your 2nd point but there isn't way I can calculate that. The speed of DGB's runs were a factor of our and his success. Remember too he saved us Border style, in Eng in 38 with 2 fine defensive innings.

A recent comparison is to look at the recent South Africa vs Australia and South Africa Vs Packistan (or New Zealand). Dale Steyn average drops from 31 to 12. While Australian Batting is hopefully stronger than Pakistan, it is still very ordinary. A lot of the difference in Steyn's/ South African averages is Down to clark's efforts.

Thanks again, it was great effort

 

[the big bambino]

That is true. Clarke's effort in that series and even the last year will most likely be the pinnacle of his career. Whereas with Bradman it was a constant factor on bowling averages. So much so that in the sample group chosen he alone represented a greater jump in class than moving from a series against say SA.

 

[the big bambino]

The funny thing abt pre conceptions is that they are selectively applied and when they are they are usually wrong. For example some people downgrade DGB's stats bcos of games he played against minnows. This is wrong as anyone who can divide 37/52 will realise he played 71% of his tests against his era's strongest opponent. Sure it was a quirk of the test program of the time but its ratio far higher than any modern player will achieve.

But when his SA opponents are rated they are called minnows. I think it is a brave man who would describe any SA side post the mid 20s as minnows. Including the one DGB played as it had just beat a strong Eng team and was fated to repeat the effort some 3 years later. To oppose that view a poster linked to the bowling averages of the men who bowled to DGB in 31/32 and summarised their stats with "Its not pretty". And thats all he said. On reading it doesn't look pretty. Vincent ave 31. Bell and Quinn 32. McMillan 34 and Morkel 45.

But did you notice what was missing? Where is all the qualifying talk abt minnows that poster was only too happy to spout in error in relation to Bradman? The SA of the era almost played exclusively against Eng and Australia. By far the best 2 teams. The bowlers listed only played 2 games against NZ who could be described as a lesser team. Although they had Dempster who averaged 60 odd. Even then Bell and Vincent did not benefit statistically as they averaged much higher in that series then they did overall. Quinn cashed in somewhat but the real beneficiary was McMillan whose leg spinners won that series against NZ and in doing so dropped his test ave from the mid 40s to mid 30s. Another 2 tests against NZ and he may have droped it to the mid 20s and all of sudden...

Of the 5 bowlers mentioned you can safely overlook McMillan and Morkel as test performers. Though to give then credit McMillan could atleast wallop NZ and the likes of him would probably be Australia's best spinner now. Morkel was really a bits and pieces cricketer. One picked to extend the batting and bowling. But against England and Australia Vincent averaged 30, Bell 31 and Quinn 38. These aren't bad figures against such strong opponents, as even the selective poster indirectly admitted. In another context he posted Grimmett's record against England. It is 32. No one doubts Grimmett's greatness but his average against a strong opponent is higher than Vincent and Bell's and only superior to Quinn's Yet the SA had an even harder job than Grimmett. They also had to bowl at the batsmen who were Grimmett's team mates. Most of all they had to bowl against Bradman. And the 2 things we've come to know is that the less you bowl at Bradman and the more you bowl at minnows the better off you'll be.

Well unfortunately for the SA bowlers of the era they got the ratio the wrong way around and that is why their figures appear unimpressive. Next up I'll post just how drastically Bradman inflated the bowling averages of Bell and Quinn.

 

[the big bambino]

I think I've deleted a post. Just to summarise it Allen, Voce, Farnes and Bowes as a group had the same ave against Aus as Trueman, Statham and Bailey. When DGB's contribution is removed the pre war group's ave is abt 3 run/wkt less. The post war group's ave against other countries is lesser by abt the same margin which shows there is little to no difference btwn the 2 groups when Bradman's contribution is deducted.

I also have SRT's record against all bowlers (up to abt 3 yrs ago). They show he averaged 34 v Donald and increased that bowler's ave against India by 1.68. If you scale up SRT's ave to 100 he increases Donald's ave against India by 4.01 runs.

 

[the big bambino]

After the distractions of the goings on in India a quick review of the SA side Bradman played in 31/32. Particularly their bowlers. If you consider Morkel and McMillan of ltd use then SA relied mainly on new ball bowlers Bell and Quinn and left arm spinner Vincent. The latter had a rough tour averaging 54 in tests. Even if you allow for the runs DGB scored off him his tour ave would still be up there. I have done a potted review of spin bowlers in Aus and concluded their best role here is to keep runs down and hopefully master atleast 1 batsman. So if Vincent averaged around 40 he would have been doing well for a spinner in Aus, historically speaking.

In tests Quinn took 13 wkts @ 39 and Bell 23 @ 27. Bell's effort on those figures alone rates as one of the better efforts for a touring pace bowler. What you don't get is an appreciation of how well both bowled. Bradman scored around 35% of Australia's runs that series. Woodfull also had a top series. Yet most of the others struggled. Kippax ave 32. McCabe 33. Ponsford was so poor against Quinn he was eventually dropped. However Bradman averaged 201. The weird impression is given that Bell and Quinn were bowling SA to almost parity at one end and getting slaughtered by DGB at the other. In fact they only managed to dismiss Bradman once btwn them. I've estimated that Quinn conceded 185 runs to DGB for a single success. Bell 206 runs for no successes.

Adjusted for Bradman's effect Quinn averaged 27 and Bell an incredbile 18. Bell's effort was almost as good as Larwood's in the following season and one of the most impressive pace bowling efforts by anyone to visit Australia. Their overall figures, when similarly adjusted wouldve fallen abt 4 points and into the 28 region. Vincent's would also have fallen into the high 20s I'm guessing as I haven't done the exercise to prove that.

Remember those averages are obtained with the negligible assistance of bowling to minnows as SA's test program was almost exclusively against Aus and Eng. I know there is a big difference in tests played but that is down to the opportunities available, but such an average compares with McDermott and Hughes.

Too many people have dismissed the SA bowling as minnow standard. Yet these people have not allowed for the fact these men rarely had the chance to bowl at minnows themselves to improve their figures. Yet they always raise the minnow issue when they believe it assists them to make the point. Even then the point they make is often in error.

The averages of Bell, Quinn and Vincent is only high bcos they bowled at Bradman and rarely at minnows. The assessment of their skill is distorted by a failure to appreciate those facts.

 

[fredfertang]

In a marketing announcement I can confirm that I have written a feature about Bill Ponsford which will, assuming I have understood the schedule correctly, appear on the front page tomorrow.

But back on point Bell and Quinn had the Indian Sign on Ponny in 31/32, so much so that he was dropped for the final Test

 

[the big bambino]

Great: Can't wait.

 

[the big bambino]

The less you bowl at Bradman and the more you bowl at weaker teams the better for it you'll be. Such a simple thing to understand and yet some people don't seem to credit it. Some try and revise DGB's achievements by denigrating his opposition, exaggerating his games against minnows or say he played too often against England. The fact these last 2 points are contradictory doesn't proclude some from stating both of them.

Most of all they try and use his success against him. Look at the bowling averages they say. Proves none of them could bowl. Well the only way to prove the quality of the attacks he faced is to remove his contribution which would reveal the kind of attacks he dominated to begin with. When that is done I have shown that there is no statistical difference btwn the prominent English pace men of the 30s and 50s. I have shown the 30s to be the 2nd lowest scoring decade by batting average. I have shown that in the 30s Australia's Bradmanless batting ave to be the 2nd lowest in ashes cricket since that time. In short all impressive recommendations for the quality of the attacks he subsequently dominated.

To my surprise that isn't enough to convince some people who nonetheless offer nothing in refutation. So I have compared the English attacks of 27-39 with certain modern attacks with and without his contribution deducted. First some parameters:

1) Comparison is the Eng attack of 27-39, Eng attack of 98-2009, SA attack 93-2004, Pak attack of 88-2000 and the Aus attack of the same time frame.

2) I have compared each team's record v Aus (except the last side of course) no. of top 10 bowlers who ave <30 v Aus, record v All, no. of top 10 bowlers with ave <30 and proportionally adjusted ave v All. I'll get to that last point in a while but it came about bcos I'm coming to the conclusion that the record v All (highlighted) is more a measure of how often you were fortunate enough to escape playing the best team moreso than overall quality.

3) I picked the 27-39 timeframe bcos it is the most relevant. It is the time dating back to DGB's fc debut up to the last test before the war. Therefore the others should be compared to a 12 year timeframe too. I picked the modern teams and their timeframes bcos no one can credibly doubt their quality. A knee jerk reaction might groan at the inclusion of the modern Eng team. However it won 2 ashes against strong Aussie teams in that era. I have concentrated on the 90s in the main (but not wholly) bcos it was an era of comparatively low batting averages.

4) I have selected each side's record v Aus bcos they were the quality opposition for all teams over both eras. This is the most important comparative stat imo bcos the no of tests each group played v Aus is approximately similar. Neither can anyone doubt Aus' quality in either era. The ave against Aus isnot distorted by proportions of games teams played against minnows as the figure v All is so corrupted. It is a direct and stark comparison showing how each side performed against quality.

5) Why deduct the Bradman Effect? Well bcos the DGB revisers invite it. It isn't fair to quote bowling stats after DGB distorted them. To make an absolute comparison btwn eras his stats have to be removed. After all no one's bowling reputation suffers at all if you take away SRT's and Lara's contribution toward them. Even then the 1st comparison will include DGB's stats.

Next post I'll explain why I have come to doubt the importance of overall averages which I highlighted in point 2.

 

[Flem274*]

Bradman was twice the batsman of his peers.

No one has ever done that before except Grace. Can't that be the deciding factor?

BTW Bambi I'm really enjoying your posts lately.

 

[the big bambino]

Basically its all about proportion. Why am I fixated on it? Bcos the DGB revisers are. It is a key to their argument that he, or Eng, benefitted greatly by bashing minnows. Therefore I have adjusted the overall averages to reflect an outcome as if they played the same no. of games against all opponents. To remove a key distorting factor on the overall averages. It shows that Eng 27-39 played the greater % of its games against stronger opponents. This contradicts what the revisers will have you think. Here is an example of what I'm getting at.

2 batsmen, Mutt and Jeff, ave 50 against team A and 40 against team B. If they split their games evenly btwn both teams their ave should be around 45. However if Mutt averages 48 and Jeff 42 it doesn't so much reveal a 6 run difference in quality as the fact Mutt benefitted by playing team A more often.

Now an example witn real players. One clown wrote that Larwood was an ordinary bowler in effect and said he was nothing on Donald for eg. After all the fmr averaged 28 and the latter 22. But they are overall figures which are distorted by proportion.

Donald played 19% of his games v Aus (his strongest opponent) but Larwood 71%. Therefore it isn't surprising that Larwood's overall figure will be skewed towards the high end. Now let us reverse the circumstances and give Donald Larwood's test program. That would be 15 tests v Aus at 31.07, 3 v Eng (as opposed to HL's SA), 2 tests v WI and a rained out test against NZ.

As a result of that configuration Donald's test ave increases dramatically to abt 28.4, or Larwood's ave. Save your bitching; there's more to come. Larwood faced Bradman. Donald never faced a batsman whose career ave remotely approached 100. Lets reverse the circumstances. Larwood's ave falls to 25.9. Donald's increases to 30.53.

Well well...

Save your bitching; there's more to come. What if we reverse the percentages for Larwood so that he faced Australia only 19% of the time? In those circumstances Larwood's ave falls to 23.72. In short the circumstances each man experienced explains their respective averages. We would probably then have some fools arguing Donald could not bowl...

Is this comparison fair? Well I can argue that it isn't. However the Bradman revisers can't. After all they superficially judge Larwood under exactly the same circumstances.

Therefore I think the overall averages are corrupted by lack of proportion. I've tried to correct for that. I'm not sure if the attempt works but it appears to have. What I have done is added up each average a team earned against each individual opponent and divided that figure by the no. of opponents. So if Team C averaged 36, 20 and 19 against 3 opponents its adjusted overall ave should be 25 if it played each opponent 33% of the time. Whereas in reality it may have played against team A most of time thereby skewering its overall ave to the higher figure, 36.

 

[the big bambino]

Bradman was twice the batsman of his peers.

No one has ever done that before except Grace. Can't that be the deciding factor?

BTW Bambi I'm really enjoying your posts lately.

Thanks Flem. I have about 2 or 3 more on this thread. By that time I'm hoping we can afford DGB the same respect every other champion deservedly has earned.

 

[harsh.ag]

Bambi has been reminding me of this guy recently:

Thug Kitchen

 

[the big bambino]

Bambi has been reminding me of this guy recently:

Thug Kitchen

I'm only doing it for their own good!

 

[fredfertang]

Basically its all about proportion. Why am I fixated on it? Bcos the DGB revisers are. It is a key to their argument that he, or Eng, benefitted greatly by bashing minnows. Therefore I have adjusted the overall averages to reflect an outcome as if they played the same no. of games against all opponents. To remove a key distorting factor on the overall averages. It shows that Eng 27-39 played the greater % of its games against stronger opponents. This contradicts what the revisers will have you think. Here is an example of what I'm getting at.

2 batsmen, Mutt and Jeff, ave 50 against team A and 40 against team B. If they split their games evenly btwn both teams their ave should be around 45. However if Mutt averages 48 and Jeff 42 it doesn't so much reveal a 6 run difference in quality as the fact Mutt benefitted by playing team A more often.

Now an example witn real players. One clown wrote that Larwood was an ordinary bowler in effect and said he was nothing on Donald for eg. After all the fmr averaged 28 and the latter 22. But they are overall figures which are distorted by proportion.

Donald played 19% of his games v Aus (his strongest opponent) but Larwood 71%. Therefore it isn't surprising that Larwood's overall figure will be skewed towards the high end. Now let us reverse the circumstances and give Donald Larwood's test program. That would be 15 tests v Aus at 31.07, 3 v Eng (as opposed to HL's SA), 2 tests v WI and a rained out test against NZ.

As a result of that configuration Donald's test ave increases dramatically to abt 28.4, or Larwood's ave. Save your bitching; there's more to come. Larwood faced Bradman. Donald never faced a batsman whose career ave remotely approached 100. Lets reverse the circumstances. Larwood's ave falls to 25.9. Donald's increases to 30.53.

Well well...

Save your bitching; there's more to come. What if we reverse the percentages for Larwood so that he faced Australia only 19% of the time? In those circumstances Larwood's ave falls to 23.72. In short the circumstances each man experienced explains their respective averages. We would probably then have some fools arguing Donald could not bowl...

Is this comparison fair? Well I can argue that it isn't. However the Bradman revisers can't. After all they superficially judge Larwood under exactly the same circumstances.

Therefore I think the overall averages are corrupted by lack of proportion. I've tried to correct for that. I'm not sure if the attempt works but it appears to have. What I have done is added up each average a team earned against each individual opponent and divided that figure by the no. of opponents. So if Team C averaged 36, 20 and 19 against 3 opponents its adjusted overall ave should be 25 if it played each opponent 33% of the time. Whereas in reality it may have played against team A most of time thereby skewering its overall ave to the higher figure, 36.

Interesting stuff - if you get a moment, and can be bothered, how do those comparisons work out if you ignore Larwood's performances in 1930, when he was never really fit enough to play

 

[the big bambino]

Interesting stuff - if you get a moment, and can be bothered, how do those comparisons work out if you ignore Larwood's performances in 1930, when he was never really fit enough to play

I'll get back to it Fred. Generally I wouldn't be sympathetic to excising performances due to poor form or injury. Those fates are common to all cricketers and they just have to take the good and the bad. After all I would then have to deduct Donald's last 3 tests v Aus as his form was pretty bad by then. However bowling to Bradman is a unique experience and does tend to distort comparsions with other eras.

Now the comparison. The columns against each team are:

a) Average v Australia (the strongest opponent for each team)
b) No of top 10 bowlers who averaged <30 v Aus
c) Average v All teams
d) Proportionally adjusted ave v All teams.
e) No of top 10 bowlers who averaged <30 v All teams.

Eng 27-39 (A) 34.84 (B) 4 (C) 30.24 (D) 28.42 (E) 8

Eng 98-09 (A) 41.05 (B) 2 (C) 34.06 (D) 31.57 (E) 3

Pak 88-00 (A) 37.66 (B) 3 (C) 29.72 (D) 30.17 (E) 4

SA 93-04 (A) 37.22 (B) 2 (C) 29.03 (D) 27.65 (E) 4

I picked 12 year periods to coincide with the length of the Bradman era up to the war. I tried to incorporate the 90s as much as possible for the modern teams. I started Pakistan in 1988 to capture their fine bowling performances v Aus and WI in that year. I started SA from the year they resumed test cricket against Australia. I selected the modern English period to take into account Flintoff's career as I think he was the driving force in 2 ashes victories when Aus really did have a team.

The 27-39 English team is superior in 3 of the 5 columns, A B and E. Including the 2 most important which involve games against the strongest opponent. From that you can extrapolate overall quality. SA win column C and D. I think column C is the least important as it is compromised by questions of proportionality.

The figure in column D attempts to smooth over distortions caused by proportionality (ratio of games against stronger and weaker opponents). If the figure in this column is smaller than the overall ave you spent more time bowling at stronger opponents. If it is higher than the overall ave you spent most of your time playing weaker opponents.

Note that the Eng team of 27-39 is only behind SA on this score. However even here they played proportionally more matches against stronger teams than did the Saffers. If I adjust for that the difference btwn the 2 teams, already negligible becomes practically non existent.

Lastly the figures for the 27-39 Eng team includes DGB's runs. They tend to show that this team is very competitive against the others even with this unique impediment.

 

[fredfertang]

I'll get back to it Fred. Generally I wouldn't be sympathetic to excising performances due to poor form or injury. Those fates are common to all cricketers and they just have to take the good and the bad. After all I would then have to deduct Donald's last 3 tests v Aus as his form was pretty bad by then.

You are absolutely right of course, and I ask only out of interest (not that that means I won't necessarily use the figures disingenuously at a later date if someone has the temerity to, again, suggest that Larwood was anything other than an all time great)

 

[the big bambino]

Fair enough Fred Maybe not tonight but I'll update it on this thread.

Note the figures in column B. They are the number of bowlers from each side's top 10 bowlers who averaged <30 against Australia. Don't let your prejudgments assume the identity of those bowlers. Larwood and Tate are not amongst them. Neither are Donald, Waqar or Shoaib. Neither are Gough, Flintoff, Hoggard and Harmison.

 

[honestbharani]

yeah.. good stuff, bambi.. Further proves my point on how stupid it is to start looking at things in isolation to back up your argument.. And it is funny how selective it gets too. When talking up the batsman they love, suddenly the bad figures are a reflection of how well said batsman played the bowlers but when The Don is in question, it is because how crap the bowlers really were..

 

[the big bambino]

Thankyou HB. I'll do this post, one more for Fred and then a last talking about the myths of the Bradman era.

First of all I agree with your post. You can isolate things til you get the outcome you want. But leaving the raw stats in place and unexplained can be misleading too. Frankly I think DGB was a victim of his own success as it allowed opportunity to revise the quality of his opponents. Therefore the only way to really judge the bowlers he faced is to take out his figures and see how much he distorted them. I think this is justifiable bcos a bat of his calibre is unique in the game's story this far. We then get an idea of how his opponents would be remembered if his distorting contribution was absent.

When I take DGB's runs from the 27-39 Eng attack they are superior to the other mentioned in every column. Their bowling ave versus Australia falls 5.48 points from 34.84 to 29.36. To put that in perspective it is greater than the difference in the Pakistani group when playing Aus as opposed to India. That ave is the 3rd lowest in ashes history since that time.

The no of bowlers who ave <30 v Aus increases from 4 to 7. The overall ave falls from 30.24 to 28.41. The no of bowlers who ave <30 v All teams goes from 8 to 9. The adjusted overall av falls 1 point to 27.4.

Compare that to the Australians of 88-2000 whose overall av was 27.92 and adjusted av is 27.32. Their no of bowlers in the top 10 with averages <30 is 9 as well.

With DGB's figures removed the 27-39 Eng is competitive with the Aus teams of 88-2000. No doubt some will huff incredulity without offering refutation. Even if we accept their complaints (and we have no reason to) we can include DGB's runs and state his opponents were competitive with Pak, SA and Eng of the periods chosen. And I don't think we can doubt their quality.