Australia’s head of handicapping Simon Magdulski answers the questions our readers commonly ask on why a GA Handicap is determined by averaging the best 8 of the recent 20 rounds multiplied by 0.93
Under the ‘best 8 of 20’ formula for determining a handicap, golfers don’t often play to their handicap. Wouldn’t it be better if the handicap system was changed so that a handicap was instead reflective of a golfer’s average score?
The first step in determining a player’s GA Handicap is to average the best 8 of their most recent 20 rounds; the second and final step is to multiply this average by 0.93. An outcome of determining handicaps in this way is that a golfer in a Stableford competition will only play to their handicap about 15% of the time – the percentage is a little higher for low markers and a little smaller for high markers. These percentages are a very deliberate objective of the designers of the handicap system – but it is not because we are trying to set very high benchmarks! Click here to view more detailed statistics on how often a golfer really plays to their handicap.
The starting point for GA in designing the handicap system was to set a general target for the score that should win a Stableford competition. Based on detailed club consultation, this target score was set at approximately 40 points (small-field competitions on average will exhibit a slightly lower winning Stableford score, whilst large-field competitions will exhibit a slightly higher winning Stableford score on average). And we are of course mindful that winning scores will be different from day to day and they will fluctuate with the seasons.
We could change the handicap system formulas to enable golfers to play to their handicap more often but this would have the direct effect of significantly increasing the number of Stableford scores required to win competitions. Such a shift in approach would also tilt the competitive advantage very much in favour of higher handicap players. Both of these outcomes would be in direct conflict with the clear preferences that have been expressed to us by clubs and golfers.
If the handicap system formulas were changed to make handicaps reflective of average scores, why would this tilt the competitive advantage very much in favour of higher handicap players?
There are certainly exceptions but higher handicap golfers in general return a much wider range of scores than do low markers. Put another way, low markers are more consistent than high markers. As a result, when you assess the difference between a golfer’s average score and their better scores, this difference widens as a golfer’s handicap increases.
In large field net events, it is likely that at least a small number of high handicappers will have a day at the ‘good’ end of their scoring range. And this will be a larger number of strokes from their average scores than will be the difference between the scores of the best-placed low markers from their average scores. So the higher handicappers would be at a clear advantage if it was average scores that were used to determine handicaps.
By using only scores from the ‘better end’ of a golfer’s distribution of scores we are able to increase equity in handicapping.
(Note: GA is aware that the 8 of 20 x 0.93 formula can still provide advantages for inconsistent players over consistent players. This is one of the reasons why we believe the new Soft Cap regulation to be introduced in Australia as a part of the World Handicap System will represent a strong improvement for Australian handicapping. Click here for more information on the new Soft Cap regulation.
Why do we have a 0.93 Multiplier?
The Multiplier (which could have been called “The Equaliser”) reduces every golfer’s ‘8 of 20’ average by 7%, which means that higher handicap golfers receive a larger absolute reduction than do lower handicap golfers. For example 7% of a handicap of two is essentially no reduction at all whereas 7% of a handicap of 35 represents a reduction of 2.5 strokes.
Without the Multiplier, statistical modelling demonstrates that there would be a significant competitive bias to higher handicap players. The 0.93 value works in concert with the ‘8 of 20’ average to try to achieve national results patterns for net competitions that are as equitable as possible. The 0.93 and ‘8 of 20’ settings were determined after extensive statistical analysis of millions of rounds in GOLF Link. Click here to see a statistical review of these settings.
When averaging was introduced into the GA Handicap System in 2010, we started by using the best 10 of the most recent 20 scores. Why did we then move from ’10 of 20’ to ‘8 of 20’?
• Under 10 of 20 there was a clear and very strong bias towards high markers over low markers in larger field net events. As a result, GA received widespread feedback from clubs requesting some refinement. Since the move to 8 of 20, clubs have been far more comfortable with the operation of the handicap system.
• Whilst the move to 8 of 20 decreased all handicaps, it decreased higher handicaps by a greater amount than it decreased lower handicaps.
• Under ‘8 of 20’, consistent players receive slightly higher handicaps than do comparable inconsistent players. As a result, players who are prone to returning occasional ‘one-off’ good rounds find that their net scores are slightly worse than they would have been under ’10 of 20’.
• The rate of downward adjustment of a handicap is slightly quicker under of 8 of 20 than occurred under 10 of 20. The rate of outward adjustment is also slightly slower under 8 of 20 than occurred under 10 of 20. To quantify this, handicaps on average under 8 of 20 increase 50% more slowly than they reduce. (This is more in line with the expressed expectation of Australian clubs. The 10 of 20 averaging method saw handicaps increase and reduce at the same rate.)
If 8 of 20 exhibits advantages over 10 of 20, wouldn’t it therefore be better to use the best 7 of the most recent 20, or the best 6 of the most recent 20?
As the number of scores on which the handicap calculation is based becomes lower and lower, so the handicap decreases more quickly as a result of good scores. This is generally seen as positive, particularly for high-markers. However, it also adjusts handicaps outwards more quickly when ‘counting’ scores drop off. In all of the feedback GA has received, there has been no sentiment for this sort of volatility.
High-markers tend to be more inconsistent than low markers. They are also more likely to have clusters of good scores feature in their handicap records. This means that high-markers are prone to having three or four ‘counting’ scores drop off in consecutive rounds. If a high-marker’s handicap was calculated on the basis of a low number of their best 20 scores, they would be particularly prone to volatile handicap movements. Again, in all of the feedback GA has received, there has been no sentiment for this sort of outcome.
GA believes that ‘8 of 20’ strikes the right balance in this respect.
Rather than drawing the ‘counting’ scores from the most recent 20 rounds, what would happen if we increased or decreased that number?
By increasing the number from 20 to say 40, we would reduce the volatility of the handicap of a player who goes through a poor run of form. Unfortunately, the handicap also wouldn’t react very quickly to a run of good form – and such a characteristic is contrary to expressed feedback.
If we were to decrease the number from 20 to say 15, the handicap would react more quickly to a run of good scores. However it would also increase the volatility of outward movement which would be an outcome contrary to expressed feedback.
Why not base the handicap calculation for high-markers on the best ‘6 of 20’, the calculation for mid-markers on the best ‘8 of 20’, and the calculation for low markers on the best ‘10 of 20’ (or some variation of this theme)?
This type of concept has been considered in the past and it does have its attractions. However, when we scratched beneath the surface the following factors emerged:
• As the number of scores on which the handicap calculation is based becomes lower and lower, so a handicap will decrease more quickly as a result of good scores. However, it will also adjust outwards more quickly when good scores drop off. High-markers tend to be more inconsistent than low markers. They are also more likely to have clusters of good scores feature in their handicap records. This means that high-markers are prone to having three or four ‘counting’ scores drop off in consecutive rounds, leading to volatile handicap movements. In all of the feedback GA has received, there has been no sentiment for this sort of outcome.
• Such a regulation would require the addition to the process of a series of extra calculations. The first extra layer would be to create a formula to determine what handicap category a player fitted into (ie something needs to determine if someone is a ‘low’, ‘mid’, or ‘high’ handicap golfer). The second addition is to split a step in the process whereby we have a single formula for all golfers, to having three (or more, depending on the number of handicap categories) formulas operating in parallel. Note: We do not see this as prohibitive but with many clubs pushing the simplicity theme, we believe that the added complexity would need to be offset by a demonstrably better outcome.
• When a handicap calculation process includes multiple grades, an uneven handicap distribution occurs. This occurs because the multiple calculation formulae cause some level of grouping or clustering of handicaps. Note: We did not see this as prohibitive but the resultant slight distortion this causes to competition results patterns was not positive.
• So there is a series of downsides associated with this concept.
