"Competitive matches" and the unfair double-standard for Associates
The format of 2019’s Cricket World Cup has been public knowledge for several years now, but the thrilling qualifying tournament in Zimbabwe this year thrust the exclusionary 10-team round robin structure back into the spotlight. The insular nature of cutting teams in cricket’s premier event while other sports expand their world cups has been discussed at length, with everyone from Mike Atherton to man-of-the-tournament Sikander Raza to Scotland coach Grant Bradburn lamenting the damage to development of the 10-team world cup and pleading to the ICC for more opportunities.
Those in favour of the 10-team format tend to present 3 general arguments: that less teams means it’ll be quicker and sharper (and I actually agree that the 2015 format did drag on a little), that only the top teams participating ensures close contests, and, a little more quietly, that it may not be the best format but the ICC is bound by commercial constraints.
The time argument is the easiest to dismiss, as the 2019 CWC is scheduled to be 2 days longer than 2015, and kicking out the Associates trims just 1 match from the tournament (taking it from 49 matches to 48).
The commercial angle has been analysed in great detail by this article from Russell Degnan; his conclusion is that the format isn’t necessarily the best financial decision.
This post seeks to address the second argument - that a 10-team format will ensure close matches. As ICC CEO Dave Richardson says, "The aim is to make the major events as competitive as possible. Every match should be very competitive and having 10 teams at the 2019 World Cup will make sure that will be the case."
Having a look at the data
In order to obtain a clear picture of the state of "competitiveness" at cricket world cups, and to investigate how well the stated aim of the 10-team format actually stacks up, I decided to examine the numbers from past events, as well as craft a workable metric of "closeness" for chasing teams.
With this in mind, I decided to keep 50 runs as the cutoff for a competitive margin of victory when setting a target (as it’s vanishingly rare for a match with a higher margin of victory to be considered "close"), and compare 3 different chasing metrics - 5 wickets, 5 overs remaining, and a combination of overs and wickets where for wins with more than 5 overs to spare, a 3 wicket victory is the cutoff for close (i.e. 4 or more wickets in hand is "not close"); if reaching the target with 5 overs or less remaining, a 6 wicket margin is the cutoff for close (i.e. 7 or more wickets in hand is "not close"). This last metric is the one which I personally consider to the be most accurate indicator of competitiveness.
Using those measurements, here is the overall competitive matches breakdown across CWC history (note that only completed matches are included, and one-sided Associate victories over Full Members count as "close"):
At first glance, this would lend credence to the argument that the exclusive format will lead to less mismatches. However, there are a couple of points worth noting - firstly, that a significant majority of FM-only encounters still end with thrashings. Second, and more importantly, this data spans 40 years of international cricket with varying formats, levels of professionalism and Associate involvement. The Associate game especially (but even FM cricket) is vastly different now, with ICC development programmes bearing fruit in the form of much higher standards of competition and professionalism.
So to account for that, I’ll focus on the the 2003 World Cup onwards - 2003 was the first CWC played when the ICC had started taking development seriously. They’d made a few steps in the late 90s, but it really ramped up in the 2000s with the promotion of Bangladesh to FM status plus the development of well-structured playing pathways, targeted funding programmes, and a policy of inclusive CWCs. Here are the numbers for competitive CWC matches 2003-2015:
Furthermore, the numbers show that CWCs are especially bad for one-sided matches. Here are the 2003-present stats on all ODI matches played outside the CWC (as of 22/10/2018):
Across the larger sample size, we can observe a few important points - one-sided games are still the majority of results, and a small gap in competitiveness between FMs and AMs still exists, but teams are overall statistically more competitive outside CWCs, including AMs, whose close game percentage is very similar indeed. Certainly similar enough that I would argue it discredits the "Associates can’t compete" rhetoric.
The greater propensity for one-sided matches within tournaments is also evident in the results of last year’s Champion’s Trophy - an event that should most closely act as a predictor for next year’s CWC since (as many have observed) the 2019 event is basically just a dragged-out version of the same exclusive "best-teams-only" format. In the CT2017, a mere 4/13 (31%) completed matches were "competitive". Nor can we so easily dismiss this as an aberration due to small sample size - it’s actually a slightly higher percentage than the 7/26 (27% under all three of my metrics) competitive FM-only matches at the 2015 CWC. It’s a recurring theme of the data that FMs simply thrash each other a lot (and not much less than they thrash AMs).
A digression into the why
Why? Why are FMs so prone to thumping each other at tournaments (especially recently)? I would propose a few explanations. Firstly, that of nerves/pressure. We all know the stereotypes, where some teams simply fold under the pressure of crunch games, despite playing consistently high-quality cricket throughout the intervening years between tournaments. This would plausibly explain why both FMs and AMs underperform at CWCs, especially AMs - for many teams and players they know it could be their only opportunity to play on the big stage, and it’s no exaggeration to say that livelihoods are sometimes on the line.
Another explanation would be that teams send their very best to ICC events, while they may for various reasons select more fringe players in bilateral series outside tournaments. This is especially true of Australia, who have a history of experimenting with mediocre ODI players in bilateral series only to pull everything together with their A-team clicking in the tournament. India has a similar rotation/experimentation policy, and the West Indies’ various administrative problems would be another example of a side that generally is only full-strength at ICC events.
Related to this, I would also argue that the increasing professionalism and financial stratification of cricket’s top tier (most notably the Big 3) has accentuated the differences between FMs, with Australia/India/England able to invest more in developing talent and led to more floggings - though England’s hilarious ineptitude in 2015 and the Associates’ continuing ability to compete would seem to be a counter-argument to this point.
In any case, interesting as those hypotheses are (and I may revisit them in future), my case is that Full Members thrash each other a lot (which they do), and as such singling out Associates for criticism over one-sided results is unfair. The exact reasons why FMs fail to produce close encounters is a little beside the point.
An important note on 1992
One particular CWC is mentioned more than others in discussions about the 2019 format - the 1992 event in Australia/New Zealand. It’s often referred to as the best CWC, and used as supporting evidence for the wisdom of the 10-team league format. Certainly our old friend Dave Richardson (who played in the 1992 tournament) was talking up the comparisons in an interview after the CWCQ - "The 1992 World Cup was nine teams playing everybody and it's said that was the best World Cup, so let’s give the 10-team World Cup a chance."
So I decided to run the numbers on the mythical 1992 edition, and see whether it lives up to the hype.
Statistically, 1992 certainly produced a lot more close encounters than recent CWCs so on the surface, perhaps some of the nostalgia is justified. But is it because of the format? I’m not convinced. Here’s a look at the numbers for all 7 CWCs between 1975 and 1999:
Between 1975-1999, 5 different formats were used, and 1992’s competitiveness sits only slightly above the average. So arguing that the format was responsible for the proportion of close matches is, I think, mistaking correlation for causation. In fact, if tournament format were a predictor of competitiveness, it’s actually 1987 that we should be emulating:
But wait - 1983 had the exact same format and nearly half as many close encounters!
What’s going on? For me, it strongly implies that the closeness of individual matches at a tournament has very little to do with the format, and a lot more to do with the form of the teams involved.
Applying this idea to the CWCs of 1987 and 1992, I would argue that 2 consecutive CWCs producing strongly competitive numbers is likely an indication of an unusually strong era for competitive cricket (and that 1987 was underrated compared to 1992). Certainly the overall stats for total ODIs between 1987-1992 would bear that conclusion out:
So seen in context, the amount of close games at the 1992 CWC was actually below average for the era it was played in, and fairly average across an aggregate of 5 different formats. This indicates that 1992’s success was in spite of (or, at best for the supporters of a 10-team event, unrelated to) its format - not because of it. As this piece on the fond memories of 1992 argues, I’d say the legend of the tournament is due to a confluence of nostalgia, myth, happy coincidence and an unusually competitive era (as we just saw in the overall ODI numbers). Attempting to replicate the results by repeating the format in a completely different context 25 years later mistakes correlation for causation so badly that the ICC starts to resemble a cargo cult.
Conclusion
The fact that Associates are so competitive (whilst on a fraction of the funding as FMs) is both a testament to the good work put in by ICC’s development programmes in narrowing the gap, and a damning indictment of the decision to cut the CWC to 10 teams - AMs are producing exciting contests (or at least, avoiding dull thrashings) to such an extent that their participation makes almost no statistical difference to the amount of boring matches. The ICC’s position on "competitiveness" is absolutely untenable, no matter how you slice the numbers.
Of course, none of this is to say Associates aren’t thrashed quite a lot at CWCs. They certainly are. But, and this is the thing - so are Full Members, and at a very similar rate. One-sided matches are simply part of cricket. All teams thrash each other a lot, especially in World Cups, and especially in more recent World Cups. Associates are thrashed slightly more than Full Members thrash each other, but not enough to make a significant difference to the amount of close matches in the CWC. In fact, so far this year, FM vs. AM matches are 50% more likely to be competitive than FMs playing each other! It’s simply an unfair double-standard to complain of Associates being uncompetitive whilst ignoring the amount of one-sided Full Member games. Additionally, attempting to recapture the magic of 1992 by replicating the format is completely misguided, as tournament format has nothing to do with the closeness of matches.
Those in favour of the 10-team format tend to present 3 general arguments: that less teams means it’ll be quicker and sharper (and I actually agree that the 2015 format did drag on a little), that only the top teams participating ensures close contests, and, a little more quietly, that it may not be the best format but the ICC is bound by commercial constraints.
The time argument is the easiest to dismiss, as the 2019 CWC is scheduled to be 2 days longer than 2015, and kicking out the Associates trims just 1 match from the tournament (taking it from 49 matches to 48).
The commercial angle has been analysed in great detail by this article from Russell Degnan; his conclusion is that the format isn’t necessarily the best financial decision.
This post seeks to address the second argument - that a 10-team format will ensure close matches. As ICC CEO Dave Richardson says, "The aim is to make the major events as competitive as possible. Every match should be very competitive and having 10 teams at the 2019 World Cup will make sure that will be the case."
Having a look at the data
In order to obtain a clear picture of the state of "competitiveness" at cricket world cups, and to investigate how well the stated aim of the 10-team format actually stacks up, I decided to examine the numbers from past events, as well as craft a workable metric of "closeness" for chasing teams.
Of course, determining how competitive a match is will always be open to some interpretation, since "closeness" is more of a subjective feeling than a quantifiable statistic, so I chose to look at the data from the other direction - essentially making sure that matches falling outside my parameters would be regarded as not close. The main reason for this is that my overall argument is not that Associates deserve to be at the CWC because they’re really competitive; rather, it’s the inverse - that kicking them out due to being uncompetitive is unfair, and a complete double standard, when we consider how many one-sided games are accepted unquestioningly when Full Members play each other.
With this in mind, I decided to keep 50 runs as the cutoff for a competitive margin of victory when setting a target (as it’s vanishingly rare for a match with a higher margin of victory to be considered "close"), and compare 3 different chasing metrics - 5 wickets, 5 overs remaining, and a combination of overs and wickets where for wins with more than 5 overs to spare, a 3 wicket victory is the cutoff for close (i.e. 4 or more wickets in hand is "not close"); if reaching the target with 5 overs or less remaining, a 6 wicket margin is the cutoff for close (i.e. 7 or more wickets in hand is "not close"). This last metric is the one which I personally consider to the be most accurate indicator of competitiveness.
Using those measurements, here is the overall competitive matches breakdown across CWC history (note that only completed matches are included, and one-sided Associate victories over Full Members count as "close"):
50 runs or 5 wickets
|
50 runs or 5 overs
|
50 runs or, 3 wickets
where BR > 30, or 6 wickets where BR ≤ 30 | |
Full Member
vs. Full Member |
102/242 (42%)
|
114/242 (47%)
|
101/242 (42%)
|
Full Member
vs. Associate |
36/140 (26%)
|
37/140 (26%)
|
33/140 (24%)
|
Associate
vs. Associate |
5/10 (50%)
|
6/10 (60%)
|
6/10 (60%)
|
Total Associate
|
41/150 (27%)
|
43/150 (29%)
|
39/150 (26%)
|
At first glance, this would lend credence to the argument that the exclusive format will lead to less mismatches. However, there are a couple of points worth noting - firstly, that a significant majority of FM-only encounters still end with thrashings. Second, and more importantly, this data spans 40 years of international cricket with varying formats, levels of professionalism and Associate involvement. The Associate game especially (but even FM cricket) is vastly different now, with ICC development programmes bearing fruit in the form of much higher standards of competition and professionalism.
So to account for that, I’ll focus on the the 2003 World Cup onwards - 2003 was the first CWC played when the ICC had started taking development seriously. They’d made a few steps in the late 90s, but it really ramped up in the 2000s with the promotion of Bangladesh to FM status plus the development of well-structured playing pathways, targeted funding programmes, and a policy of inclusive CWCs. Here are the numbers for competitive CWC matches 2003-2015:
50 runs or 5 wickets
|
50 runs or 5 overs
|
50 runs or, 3 wickets
where BR > 30, or 6 wickets where BR ≤ 30 | |
Full Member
vs. Full Member |
31/106 (29%)
|
38/106 (36%)
|
34/106 (32%)
|
Full Member
vs. Associate |
20/83 (24%)
|
22/83 (27%)
|
20/83 (24%)
|
Associate
vs. Associate |
4/8 (50%)
|
5/8 (63%)
|
5/8 (63%)
|
Total Associate
|
24/91 (26%)
|
27/91 (30%)
|
25/91 (27%)
|
As we can see, with FM-FM matches hovering around 1/3 competitive, and AM involvement at roughly 1/4 competitive, the gap has narrowed significantly - though interestingly it seems to be largely due to FMs flogging each other more. Why could that be?
Well, I would reiterate my point that one-sided games are simply part of cricket. I think there’s a tendency to remember the exciting matches between our favourite teams, while boring encounters fade from memory - take for example, the semi-finals at CWC 2015. The NZ-SA thriller was rightly hailed as a classic, but it was only in digging through the stats for this post that I even remembered the dull thrashing that the Aussies handed to India in the other semi-final. With AMs, this tendency often seems to be reversed - critics point to their embarrassing defeats while ignoring the times they are competitive (and statistically, it’s not that different to FM-only matches). In my view, it’s an unfair double standard applied by people who either don’t understand the data or who are pushing an anti-Associate narrative.
Furthermore, the numbers show that CWCs are especially bad for one-sided matches. Here are the 2003-present stats on all ODI matches played outside the CWC (as of 22/10/2018):
50 runs or 5 wickets
|
50 runs or 5 overs
|
50 runs or, 3 wickets
where BR > 30, or 6 wickets where BR ≤ 30 | |
Full Member
vs. Full Member |
662/1479 (45%)
|
731/1479 (49%)
|
683/1479 (46%)
|
Full Member
vs. Associate |
58/153 (38%)
|
63/153 (41%)
|
62/153 (41%)
|
Associate
vs. Associate |
87/201 (43%)
|
90/201 (45%)
|
88/201 (44%)
|
Total Associate
|
145/355 (41%)
|
153/355 (43%)
|
150/355 (42%)
|
Across the larger sample size, we can observe a few important points - one-sided games are still the majority of results, and a small gap in competitiveness between FMs and AMs still exists, but teams are overall statistically more competitive outside CWCs, including AMs, whose close game percentage is very similar indeed. Certainly similar enough that I would argue it discredits the "Associates can’t compete" rhetoric.
The greater propensity for one-sided matches within tournaments is also evident in the results of last year’s Champion’s Trophy - an event that should most closely act as a predictor for next year’s CWC since (as many have observed) the 2019 event is basically just a dragged-out version of the same exclusive "best-teams-only" format. In the CT2017, a mere 4/13 (31%) completed matches were "competitive". Nor can we so easily dismiss this as an aberration due to small sample size - it’s actually a slightly higher percentage than the 7/26 (27% under all three of my metrics) competitive FM-only matches at the 2015 CWC. It’s a recurring theme of the data that FMs simply thrash each other a lot (and not much less than they thrash AMs).
A digression into the why
Why? Why are FMs so prone to thumping each other at tournaments (especially recently)? I would propose a few explanations. Firstly, that of nerves/pressure. We all know the stereotypes, where some teams simply fold under the pressure of crunch games, despite playing consistently high-quality cricket throughout the intervening years between tournaments. This would plausibly explain why both FMs and AMs underperform at CWCs, especially AMs - for many teams and players they know it could be their only opportunity to play on the big stage, and it’s no exaggeration to say that livelihoods are sometimes on the line.
Another explanation would be that teams send their very best to ICC events, while they may for various reasons select more fringe players in bilateral series outside tournaments. This is especially true of Australia, who have a history of experimenting with mediocre ODI players in bilateral series only to pull everything together with their A-team clicking in the tournament. India has a similar rotation/experimentation policy, and the West Indies’ various administrative problems would be another example of a side that generally is only full-strength at ICC events.
Related to this, I would also argue that the increasing professionalism and financial stratification of cricket’s top tier (most notably the Big 3) has accentuated the differences between FMs, with Australia/India/England able to invest more in developing talent and led to more floggings - though England’s hilarious ineptitude in 2015 and the Associates’ continuing ability to compete would seem to be a counter-argument to this point.
In any case, interesting as those hypotheses are (and I may revisit them in future), my case is that Full Members thrash each other a lot (which they do), and as such singling out Associates for criticism over one-sided results is unfair. The exact reasons why FMs fail to produce close encounters is a little beside the point.
An important note on 1992
One particular CWC is mentioned more than others in discussions about the 2019 format - the 1992 event in Australia/New Zealand. It’s often referred to as the best CWC, and used as supporting evidence for the wisdom of the 10-team league format. Certainly our old friend Dave Richardson (who played in the 1992 tournament) was talking up the comparisons in an interview after the CWCQ - "The 1992 World Cup was nine teams playing everybody and it's said that was the best World Cup, so let’s give the 10-team World Cup a chance."
So I decided to run the numbers on the mythical 1992 edition, and see whether it lives up to the hype.
50 runs or 5 wickets
|
50 runs or 5 overs
|
50 runs or, 3 wickets
where BR > 30, or 6 wickets where BR ≤ 30 | |
Full Member
vs. Full Member |
15/29 (52%)
|
18/29 (62%)
|
16/29 (55%)
|
Full Member
vs. Associate |
3/8 (38%)
|
4/8 (50%)
|
3/8 (38%)
|
Associate
vs. Associate |
N/A
|
N/A
|
N/A
|
Total Associate
|
3/8 (38%)
|
4/8 (50%)
|
3/8 (38%)
|
Statistically, 1992 certainly produced a lot more close encounters than recent CWCs so on the surface, perhaps some of the nostalgia is justified. But is it because of the format? I’m not convinced. Here’s a look at the numbers for all 7 CWCs between 1975 and 1999:
50 runs or 5 wickets
|
50 runs or 5 overs
|
50 runs or, 3 wickets
where BR > 30, or 6 wickets where BR ≤ 30 | |
Full Member
vs. Full Member |
71/136 (52%)
|
76/136 (56%)
|
67/136 (49%)
|
Full Member
vs. Associate |
16/57 (28%)
|
15/57 (26%)
|
13/57 (23%)
|
Associate
vs. Associate |
1/2 (50%)
|
1/2 (50%)
|
1/2 (50%)
|
Total Associate
|
17/59 (29%)
|
16/59 (27%)
|
14/59 (24%)
|
Between 1975-1999, 5 different formats were used, and 1992’s competitiveness sits only slightly above the average. So arguing that the format was responsible for the proportion of close matches is, I think, mistaking correlation for causation. In fact, if tournament format were a predictor of competitiveness, it’s actually 1987 that we should be emulating:
50 runs or 5 wickets
|
50 runs or 5 overs
|
50 runs or, 3 wickets
where BR > 30, or 6 wickets where BR ≤ 30 | |
Full Member
vs. Full Member |
14/21 (67%)
|
15/21 (71%)
|
14/21 (67%)
|
Full Member
vs. Associate |
2/6 (33%)
|
2/6 (33%)
|
2/6 (33%)
|
Associate
vs. Associate |
N/A
|
N/A
|
N/A
|
Total Associate
|
2/6 (33%)
|
2/6 (33%)
|
2/6 (33%)
|
But wait - 1983 had the exact same format and nearly half as many close encounters!
50 runs or 5 wickets
|
50 runs or 5 overs
|
50 runs or, 3 wickets
where BR > 30, or 6 wickets where BR ≤ 30 | |
Full Member
vs. Full Member |
9/21 (42%)
|
9/21 (42%)
|
8/21 (38%)
|
Full Member
vs. Associate |
4/6 (66%)
|
3/6 (50%)
|
3/6 (50%)
|
Associate
vs. Associate |
N/A
|
N/A
|
N/A
|
| Total Associate |
4/6 (66%)
|
3/6 (50%)
|
3/6 (50%)
|
What’s going on? For me, it strongly implies that the closeness of individual matches at a tournament has very little to do with the format, and a lot more to do with the form of the teams involved.
Applying this idea to the CWCs of 1987 and 1992, I would argue that 2 consecutive CWCs producing strongly competitive numbers is likely an indication of an unusually strong era for competitive cricket (and that 1987 was underrated compared to 1992). Certainly the overall stats for total ODIs between 1987-1992 would bear that conclusion out:
50 runs or 5 wickets
|
50 runs or 5 overs
|
50 runs or, 3 wickets
where BR > 30, or 6 wickets where BR ≤ 30 | |
Full Member
vs. Full Member |
185/348 (53%)
|
228/348 (66%)
|
207/348 (59%)
|
Full Member
vs. Associate |
8/24 (33%)
|
8/24 (33%)
|
7/24 (29%)
|
Associate
vs. Associate |
N/A
|
N/A
|
N/A
|
Total Associate
|
8/24 (33%)
|
8/24 (33%)
|
7/24 (29%)
|
So seen in context, the amount of close games at the 1992 CWC was actually below average for the era it was played in, and fairly average across an aggregate of 5 different formats. This indicates that 1992’s success was in spite of (or, at best for the supporters of a 10-team event, unrelated to) its format - not because of it. As this piece on the fond memories of 1992 argues, I’d say the legend of the tournament is due to a confluence of nostalgia, myth, happy coincidence and an unusually competitive era (as we just saw in the overall ODI numbers). Attempting to replicate the results by repeating the format in a completely different context 25 years later mistakes correlation for causation so badly that the ICC starts to resemble a cargo cult.
Conclusion
The fact that Associates are so competitive (whilst on a fraction of the funding as FMs) is both a testament to the good work put in by ICC’s development programmes in narrowing the gap, and a damning indictment of the decision to cut the CWC to 10 teams - AMs are producing exciting contests (or at least, avoiding dull thrashings) to such an extent that their participation makes almost no statistical difference to the amount of boring matches. The ICC’s position on "competitiveness" is absolutely untenable, no matter how you slice the numbers.
Of course, none of this is to say Associates aren’t thrashed quite a lot at CWCs. They certainly are. But, and this is the thing - so are Full Members, and at a very similar rate. One-sided matches are simply part of cricket. All teams thrash each other a lot, especially in World Cups, and especially in more recent World Cups. Associates are thrashed slightly more than Full Members thrash each other, but not enough to make a significant difference to the amount of close matches in the CWC. In fact, so far this year, FM vs. AM matches are 50% more likely to be competitive than FMs playing each other! It’s simply an unfair double-standard to complain of Associates being uncompetitive whilst ignoring the amount of one-sided Full Member games. Additionally, attempting to recapture the magic of 1992 by replicating the format is completely misguided, as tournament format has nothing to do with the closeness of matches.
Comments
Post a Comment