Proposed Championship Ball Specification of the World Croquet Federation
Letter to President SA Croquet Association - 7th June 1989.
At your invitation, I would like to make the following comments:
The proposed maxima and minima of 37 and 31 inches seem somewhat unnecessarily disparate. With today's mixing and moulding techniques any competent specialist could guarantee ±1 inch, never mind 3 inches, particularly with accurately moulded milling. This degree of inequality would be too close for the average moulder, but we are presuming specialists in this case, with a slight price premium to pay for the additional expertise. Such a close specification, viz. ±1 inch would also warn quick profit seekers not to enter the competition. The present price of Eclipse balls is so high when compared with the cost price of ~£2,50 for the moulded ball that a large premium price could be offered for the suggested tighter specification and still result in a ball costing half the present typical price of Championship balls.
Again, if such a resilience specification, as the suggested ±1 inch were adopted, there would be no necessity for the additional addendum or rider to the suggested specification to the effect that within any one set the limit must be ±1 inch. This would be ipso facto fulfilled, without the necessity for some match secretary or other person to either check each ball in any set, or even check and select sets if this is not incumbent on the manufacturer. It would only be necessary to make a random check of any ball in any set.
The present Eclipse balls employ two sets of 50 orthogonal grooves while the Barlow balls employ 51 grooves per set. Both comply with the suggested specification. However, the specification has no relevance to the present design of ball with regard to frequency, shape or depth of milled groove. For example, the number of grooves is apparently irrelevant as long as their width is less than that of the lands. This can be achieved, as an extreme case, with just two orthogonal grooves each one inch wide, leaving lands almost two inches wide at the circumference. At the other end of the scale it will be noted that on both Jaques and Barlow balls those areas where the one set of grooves cut through the orthogonal set there are portions of the lands that taper away to nothing, so that the specification is impossible to fulfil if no part of the land is to be narrower than a groove. The land areas I am referring to are found between adjacent ring grooved "poles" and not on the third, polar axis at all.
Lastly, since both bounce and "pull" such as in peeling are affected to a pronounced degree by the milling, more controlled uniformity is, I believe, essential. We all know that when we line balls up for a peg out it is often better to place the balls against each other with ringed pole to ringed pole, with one set of grooves horizontal to offer less likelihood of the forward ball rolling off line. This is because of pull and milling pattern.
I would like to suggest that the number of grooves should be specified within the limits of say 45 - 55 per set. Further, taking into account that sharp edged lands will definitely spall or chip whatever plastic they are made of, consideration should be given for a radius at each land edge. This could be accommodated by rewording the specification for groove width. An example would be to say that the groove width must exceed the land width at the point where the natural projection of the side of the groove meets the circumferential projection of the land.
Groove depth is also important in that a cheap mould made in only a few sections is not capable of producing a deep groove since the ball cannot be removed from the mould unless there is a large shrinkage difference between mould and ball on cooling. A more costly mould is required to give a milled depth of say 1/50 inches. The specification should therefore stipulate a minimum of 0.5mm to give some playing effect to milling in the first place as well as to prevent low grade balls being produced which are only faintly marked. Such poorly grooved balls have already been manufactured.
Further, since the sharpness of the edge where groove and land meet affects the pull of the croqueted ball as previously mentioned, some control over the limits of the edge radius should be exercised so that such radius must not be less than say one hundredth of an inch. The upper limit is already set by the groove/land width control. Failure to exercise such control could result in different manufacturers producing balls of markedly different pull during the croquet stroke.
Lastly, if the milling specification is not tightened up, it would be possible to make the balls of otherwise inferior compositions in which unsatisfactory bounce characteristics could be corrected by the use of unusual milling dimensions and the effects such as pull would differ widely from and therefore unacceptably in comparison with known existing standards.
This specification is excellent. Since the momentum imparted by a given mallet stroke is constant within very small limits whether the ball struck is heavy or light, it follows that the heavier ball will roll more slowly than the lighter ball, but because it is heavier it will roll for a longer time. The distance traversed will be limited by grass resistance, about the same in each case. For this reason mass is not so important as bounce. The mass specification is still quite tight at 1 oz. in 16 or 6.25%, while bounce which is more important should at least be kept to the same percentage as mass. This would mean a variation of 2.125 inches on an average of 34 inches (the middle of the proposed 31 - 37 inches). The suggestion of ±1 inch is thus substantiated whether on 32, 33 or more inches as the average, being more in line with the less vital weight variation.
This concludes my somewhat lengthy but hopefully constructive criticism of the present suggestions.
Regarding the transmission of the bounce test rig sketch to the C.A., it should be emphasised that whatever device is used, it is essential that the "drop" must leave the ball free to fall without turning or twisting, for accurate testing on the pole spots and areas where the groove sets meet to give a correct ball average. Although it is mentioned in the writing on the sketches, this point should be further emphasised because repeatability and accuracy of measurement are paramount. Such a drop is accomplished by the rapid removal of the "gate" and by the small area of ball support in the "gate", such as the central hole in a small metal washer (say quarter inch or 6mm).
Thanking you for the opportunity to comment.
R. le Maitre ©1989
All rights reserved © 1989