Significance of the 'second innings denial' effect.

[Bolo]

I've checked the stats on two more bowlers.

Kapil Dev (standout bowler in a weak batting and bowling lineup)

38% 13.2 overs 1st innings
33% 11.7 2nd
21% 7.5 3rd
7% 2.6 4th

35 overs per match

Curtley Ambrose (standout bowler in strong batting and bowling team)

35.7% 13.4
26.1% 9.8
26.6% 10
11.4% 4.3

37.5 overs per match.

Of the players in weaker teams, Dev, Streak, and Hogg and all bowl 35 per match. Hadlee is a very different at 42

The bowlers in the stronger teams bowl between 36 and 39. This is on average slightly higher than the bowlers in weaker teams, but not particularly significant.

Distribution of wickets falls into two groups- strong teams (Mcgrath and Marshall) and weak teams (everyone else). Within these two groups distribution by innings is remarkably consistent.

It is starting to look like weakness of your own bowling lineup and weakness of your own batting lineup have fairly similar degrees of impact on your potential for wicket taking.

This conclusion will seem counterintuitive to everyone not named Miyagi, including myself. The evidence isn't conclusive- limited sample size and so many extraneous variables, but considering the fact that the data has not been cherry picked to support a conclusion, it is somewhat indicative unless someone can suggest a problem with the methodology or provide stats for comparable bowlers that indicate otherwise.

I checked Dale Steyn at the end for another example of strong bowling, strong batting and got 33.7 overs per. He's a bit of an outlier, but not enough to skew the conclusion (close to 35, and serves to balance the higher amount bowled on average by the other strong players). Just by eyeball, distribution by innings is similar to the rest of the strong team category.

 

[Starfighter]

The WSC vs Non WSC split of Hogg takes him out of the group that supports the significance of the effect and into the one that opposes it.

Though I haven't teased out the exact career figures the team batting averages from 1978 to 1994:

Aus 31.2
WI 31.9
India 33.4
England 28.5
Pak 32.1

rather put paid to the idea that Kapil had a weak batting lineup.

 

[Bolo]

I don't see a need to split Hoggs stats too much. If anything though the WSC spilt would make his inclusion more appropriate- the weakness of a weak team at its weakest makes it more relevant to the analysis.

As for Kapil, seems I'm just plain wrong. Given the fact that you have left out the teams with the weakest batting lineups, it might be appropriate to include him on the other side, which makes a mess of my conclusion. India averaging above WI? Real surprise. Something to do with home conditions?

 

[Borges]

It is starting to look like weakness of your own bowling lineup and weakness of your own batting lineup have fairly similar degrees of impact on your potential for wicket taking.

This conclusion will seem counterintuitive to everyone not named Miyagi, including myself.

It may not appear all that illogical if we consider the fact that strong teams tend to win more games; they tend to take twenty or close to twenty wickets per match more often than weak teams. Factor in that strike rates of bowlers are noticeably higher towards the end of the game when batting becomes more difficult, and this may not be a very surprising conclusion.

 

[Starfighter]

Another career to consider is Ray Lindwall. While he played in a fairly strong side through most of his career he went without being in a drawn or lost series for 29 tests up to the 51/52 WI series. Afterwards the team was conspicuously less successful losing to England three times and Pakistan in series totalling 31 tests, and unexpectedly drawing with SA at home in 52/53.

In the first 29 Australia had a batting average of 36.96 and Lindwall bowled 211 ball a match. In the second 32 Australia averaged 30.14 and Lindwall bowled 236 balls per match.

Another thing to consider for Kapil is that India drew 56% of their tests from 78 to 93, against a worldwide average excluding India of 38%. So Kapil being in a strong batting lineup and/or playing on flat pitches at home would have had his wicket taking opportunities reduced through lost time. Strong batting may actually reduce wicket taking. Hadlee had a much weaker batting lineup so drawn matches would have been rarer.

 

[Starfighter]

I don't see a need to split Hoggs stats too much. If anything though the WSC spilt would make his inclusion more appropriate- the weakness of a weak team at its weakest makes it more relevant to the analysis.

As for Kapil, seems I'm just plain wrong. Given the fact that you have left out the teams with the weakest batting lineups, it might be appropriate to include him on the other side, which makes a mess of my conclusion. India averaging above WI? Real surprise. Something to do with home conditions?

No, it's an appropriate split for Hogg, because it shows the importance of competition. Hogg came into the team as an opening bowler with only Hurst and Dymock for real competition. when WSC ended he was in competition with Lillee, Lawson and Alderman who were significantly superior to Hurst and Dymock, and Thomson at times. He no longer had the luxury of being able to get overs if he wanted them, and was no longer automatic choice to open. And the stats show he bowled far less.

 

[Bolo]

Another career to consider is Ray Lindwall. While he played in a fairly strong side through most of his career he went without being in a drawn or lost series for 29 tests up to the 51/52 WI series. Afterwards the team was conspicuously less successful losing to England three times and Pakistan in series totalling 31 tests, and unexpectedly drawing with SA at home in 52/53.

In the first 29 Australia had a batting average of 36.96 and Lindwall bowled 211 ball a match. In the second 32 Australia averaged 30.14 and Lindwall bowled 236 balls per match.

Another thing to consider for Kapil is that India drew 56% of their tests from 78 to 93, against a worldwide average excluding India of 38%. So Kapil being in a strong batting lineup and/or playing on flat pitches at home would have had his wicket taking opportunities reduced through lost time. Strong batting may actually reduce wicket taking. Hadlee had a much weaker batting lineup so drawn matches would have been rarer.

I feel the quality of Australia's bowling changed as well. They lost Miller plus the slow bowlers. He would have been required to bowl a much higher percentage of team overs. How this quality loss stacks up against the batting, I'm not sure.

Good point on Kapil. This analysis depends on games getting results. Probably best to just ignore him rather than using him as indicative in either direction though. 56% is too high to ignore, but not a high enough number to consider him as indicative of a reverse of the trend.

 

[Bolo]

No, it's an appropriate split for Hogg, because it shows the importance of competition. Hogg came into the team as an opening bowler with only Hurst and Dymock for real competition. when WSC ended he was in competition with Lillee, Lawson and Alderman who were significantly superior to Hurst and Dymock, and Thomson at times. He no longer had the luxury of being able to get overs if he wanted them, and was no longer automatic choice to open. And the stats show he bowled far less.

It's appropriate if you want to analyse him in the same way as you have done for Lindwall. It doesn't really matter for my type of analysis, assuming you consider him a standout bowler over the course of his career in balance (which im not sure of, but assumed you did, which is why I included him in the first place)

 

[Starfighter]

I feel the quality of Australia's bowling changed as well. They lost Miller plus the slow bowlers. He would have been required to bowl a much higher percentage of team overs. How this quality loss stacks up against the batting, I'm not sure.

Good point on Kapil. This analysis depends on games getting results. Probably best to just ignore him rather than using him as indicative in either direction though. 56% is too high to ignore, but not a high enough number to consider him as indicative of a reverse of the trend.

Australia's bowling average went from 23.1 to 27.63. Miller only took 2.7 wickets/match and wouldn't have been as important as may appear at first. He was used quite sparingly by Bradman in 1948. The decline in Johnston's effectiveness would have been more important, and Lindwall's own. A change would have been the advent of covered wickets starting in the early fifties, although by all accounts Australia's pitches weren't that different until the late fifties when they became very slow and dry.

The batting average declined 18.5% and the bowling average rose 19.6%, which are quite similar.

 

[Bolo]

It may not appear all that illogical if we consider the fact that strong teams tend to win more games; they tend to take twenty or close to twenty wickets per match more often than weak teams. Factor in that strike rates of bowlers are noticeably higher towards the end of the game when batting becomes more difficult, and this may not be a very surprising conclusion.

I think everyone understands the logic. It is a new concept to most of us though and is not something we've really factored into our assessments. We are very used to the idea that weak bowling = more opportunity for wickets, and I feel most people would assume this to be dominant.

 

[Bolo]

Australia's bowling average went from 23.1 to 27.63. Miller only took 2.7 wickets/match and wouldn't have been as important as may appear at first. He was used quite sparingly by Bradman in 1948. The decline in Johnston's effectiveness would have been more important, and Lindwall's own. A change would have been the advent of covered wickets starting in the early fifties, although by all accounts Australia's pitches weren't that different until the late fifties when they became very slow and dry.

The batting average declined 18.5% and the bowling average rose 19.6%, which are quite similar.

Weakly indicative, but too close to draw much of a conclusion from. Strike rate is a slightly better indicator than average here, but there will be a pretty big decline there too, so I don't imagine substituting it will help much.

 

[Borges]

Considering that most great bowlers on an average take only about five wickets per game, one could argue that strong bowling support from the other end = more opportunity for wickets. Since more wickets are 'given' than 'taken', sustained bowling pressure from both ends would tend to increase that number, for say someone like Hadlee, from five towards six.

 

[NotMcKenzie]

Let's approach this from another angle, by looking at a whole lot of other fast bowlers:

Bowler	WPI 1st Inns	WPI 2nd Inns	1st minus 2nd
R. R. Lindwall	2.23	1.77	0.56
K. R. Miller	2.14	1.40	0.74
J. B. Statham	2.17	1.69	0.48
F. S. Trueman	2.93	1.85	1.08
A. K. Davidson	2.55	1.95	0.6
W. W. Hall	2.63	1.5	1.13
G. D. McKenzie	2.62	1.68	0.94
J. A. Snow	2.73	1.58	1.15
D. K. Lillee	2.97	2.37	0.6
Imran Khan	2.80	2.21	0.59
R. J. Hadlee	3.36	2.22	1.14
A. M. E. Roberts	2.30	2.12	0.18
M. A. Holding	2.5	1.87	0.63
J. Garner	2.53	2.11	0.42
I. T. Botham	2.62	1.78	0.84
Kapil Dev N.	2.28	1.41	0.87
M. D. Marshall	2.47	2.51	-0.04
C. A. Walsh	2.13	2.16	-0.03
C. E. L. Ambrose	2.48	2.00	0.48
Wasim Akram	2.35	2.21	0.14
Waqar Younis	2.62	2.18	0.44
G. D. McGrath	2.65	1.97	0.68

I have not looked for team batting averages, however it is safe to assume that not all of the bowlers here played in weak batting teams. All except two show a drop between 1st and 2nd innings wickets, and all had fewer first innings than second.

If anything, when compared to the others, I'd say that Hadlee has an inflated first innings value; He is the only bowler with a WPI above three.

 

[TheJediBrah]

Considering that most great bowlers on an average take only about five wickets per game, one could argue that strong bowling support from the other end = more opportunity for wickets. Since more wickets are 'given' than 'taken', sustained bowling pressure from both ends would tend to increase that number, for say someone like Hadlee, from five towards six.

You could argue it but then why isn't the case? McGrath with almost identical average and SR to Hadlee had the stronger bowling support and took less wickets per match (because he bowled less).

It would seem more likely that it would push someone like Hadlee down from 5 to 4.5, because that's exactly what McGrath ended up with

 

[Bolo]

Let's approach this from another angle, by looking at a whole lot of other fast bowlers:

Bowler	WPI 1st Inns	WPI 2nd Inns	1st minus 2nd
R. R. Lindwall	2.23	1.77	0.56
K. R. Miller	2.14	1.40	0.74
J. B. Statham	2.17	1.69	0.48
F. S. Trueman	2.93	1.85	1.08
A. K. Davidson	2.55	1.95	0.6
W. W. Hall	2.63	1.5	1.13
G. D. McKenzie	2.62	1.68	0.94
J. A. Snow	2.73	1.58	1.15
D. K. Lillee	2.97	2.37	0.6
Imran Khan	2.80	2.21	0.59
R. J. Hadlee	3.36	2.22	1.14
A. M. E. Roberts	2.30	2.12	0.18
M. A. Holding	2.5	1.87	0.63
J. Garner	2.53	2.11	0.42
I. T. Botham	2.62	1.78	0.84
Kapil Dev N.	2.28	1.41	0.87
M. D. Marshall	2.47	2.51	-0.04
C. A. Walsh	2.13	2.16	-0.03
C. E. L. Ambrose	2.48	2.00	0.48
Wasim Akram	2.35	2.21	0.14
Waqar Younis	2.62	2.18	0.44
G. D. McGrath	2.65	1.97	0.68

I have not looked for team batting averages, however it is safe to assume that not all of the bowlers here played in weak batting teams. All except two show a drop between 1st and 2nd innings wickets, and all had fewer first innings than second.

If anything, when compared to the others, I'd say that Hadlee has an inflated first innings value; He is the only bowler with a WPI above three.

Hadlee is always going to be an outlier. He has the highest WPM and is the only ATG to have bowled with basically zero support I can think of. Let's look at a couple of the other guys with a high delta (above 1) in WPI- Trueman, Hall and Snow. How were their batting and bowling lineups?

 

[Mr Miyagi]

You could argue it but then why isn't the case? McGrath with almost identical average and SR to Hadlee had the stronger bowling support and took less wickets per match (because he bowled less).

It would seem more likely that it would push someone like Hadlee down from 5 to 4.5, because that's exactly what McGrath ended up with

Borges has possibly made a typo JediBrah, what he possibly meant to say is that "strong bowling support from the other = more opportunity for cheaper wickets" by bowling more to unset batsmen.

But he does seem to be conflating two things that are discrete.

 

[TheJediBrah]

Borges has possibly made a typo JediBrah, what he possibly meant to say is that "strong bowling support from the other = more opportunity for cheaper wickets" by bowling more to unset batsmen.

It still doesn't make sense though. If that logic held up in practice then McGrath would have a wpm of 5.5-6 (more than Hadlee), which he doesn't

 

[Mr Miyagi]

It still doesn't make sense though. If that logic held up in practice then McGrath would have a wpm of 5.5-6 (more than Hadlee), which he doesn't

Nopes, so ask him to explain what he means :P

 

[Mr Miyagi]

The pattern you all should end up seeing,

is that a lack of competition for wickets will result in a lot more first team innings balls bowled pro rata,

and a lack of runs will result in a lot less second team innings balls being bowled pro rata,

this is true for great seamers and great spinners. Warne and McGrath both average 18% of their balls bowled in the 4th innings. Hadlee is at 6%.

If you don't have runs to defend in the 4th innings, you don't get to bowl.

 

[TheJediBrah]

The pattern you all should end up seeing,

is that a lack of competition for wickets will result in a lot more first team innings balls bowled pro rata,

and a lack of runs will result in a lot less second team innings balls being bowled pro rata,

this is true for great seamers and great spinners. Warne and McGrath both average 18% of their balls bowled in the 4th innings. Hadlee is at 6%.

I don't think there's any new information in here that everyone didn't already know

Hadlee's wpm is higher because he bowled so much more 1st innings overs, and took a higher proportion of available wickets because the rest of his bowlers weren't great