Many coaches believe when putting, we want to launch the ball with overspin. The idea being it helps to ensure less speed loss during the balls skid phase, provides more consistent skid distance, plus helps produce the most consistent roll direction.

This said, I am a believer that more overspin doesn't necessarily mean a better roll. I would rather see a golfer have consistent spin numbers numbers of let's say 0-10 RPM (revolutions per minute) than have their RPM fluctuate between 20 and 60. In fact, to me it makes sense that close to 0 RPM, resembling a baseball knuckle ball, is going to provide the golfer with the least variability in the skid phase, prior to true roll.


Many believe you can predict the amount of overspin / backspin the golf ball starts with by determining the difference between the angle of attack and dynamic loft at impact, otherwise known as spin loft. The thought process is, a negative spin loft (angle of attack greater than the dynamic loft) creates forward / overspin, whilst a positive spin loft (dynamic loft greater than the angle of attack) produces backspin. For a visual explanation, please refer to the images below. 


There is also the belief that the greater this differential, the more spin will be seen on the ball. e.g. If a ball was predicted to have 10 RPM of overspin for a spin loft of -1 degrees, if every other variable is exactly the same but the spin loft is -2 degrees, the spin prediction would be at 20RPM. 

Goal of the experiment

The above theory is used by many putting analysis software tools as a way of predicting Ball RPM through calculation. However, experience in my coaching and through using Quintic Ball Roll has led me to believe, this is a too simplified version of reality. Though spin loft has a huge influence, if the vertical strike is not centered, there can be pretty huge variances in Ball RPM. I see many golfers trying to hit 'up on the ball' at the detriment of their strike, often hitting low in the face. I wanted to investigate what kind the influence of vertical strike location has on the Ball RPM. My feeling is that vertical strike location might be a more important factor to consider for consistent Ball RPM, than many believe

So how could I prove this?

Without a robot, I was relying on my own ability to strike the ball close to my intended position on the face. I used a slow motion camera to capture the impact of every putt and assess my accuracy. Any putt that seemed away from the intended strike position was removed. The goal was to hit 6 consecutive putts at a the certain vertical positions on the face (high or low) seen in the image, plus with a certain angle of attack. e.g. The first set of putts were struck with a vertical strike location position 2 (close to centre on the face) with a small upward angle of attack (my standard stroke.) Set 2 I hit at position 1 (low on the face) with a similar angle of attack to set 1, and so on... Details for each set of 6 putts can be seen below. 

er2 2.PNG

The Results

The below chart shows the results of the experiment. For each group of six putts the attempted vertical strike location is given, represented by a number corresponding to the above image of the putter face. Additionally, calculated averages for the angle of attack, spin loft and resulting Ball RPM are given. Note that a positive Ball RPM represents forward spin and a negative number represents backspin. Also note spin loft was calculated through knowing static loft, in this case 1 degree. 

Notice the set of putts with the highest forward spin (set 3) actually had the highest spin loft, and were the only set of 6 putts where the angle of attack average was negative (hitting down on the ball.) Applying the general theory to predict the direction of spin and the amount of spin this would suggest a decent amount of backspin, however the average for these putts was 27.71 rpm of forward spin, higher than any other set of putts. The difference? All of these putts were struck high on the face at position 3.

Now look at set of putts numbers 2 and 4, the only sets to show an average of backspin. They were also the only sets to be struck below the centre of the face. Coindidence? I think not...  

data result rpm.PNG

Correlation Coefficient

Firstly, for those who do not know, a correlation coefficient is a numerical measure for a statistical relationship between two variables. Using a scatter graph, we are able to calculate how strong the correlation is between two variables, by calculating the correlation coefficient. The higher that number, the stronger that variable has on the outcome of the other. 


Below you will find 3 scatter graphs plotting the data for Ball RPM vs a few different variables. Notice the strong correlation between Ball RPM and Strike Location, a coefficient of 0.583. Vertical Strike Location had a way stronger correlation with Ball RPM than any other data parameter. By comparison, Spin Loft (the number used by many to calculate a Ball RPM prediction) showed almost no correlation with Ball RPM. It has to be said that this correlation was lower than expected. However, the extreme inconsistency in the vertical strike pattern (necessary for the experiment) might play a role in this and further investigation in this area needed

RPM vs AofA.png
RPM vs AofA (1).png


The experiment proved the strong correlation between Ball RPM and Vertical Strike Location. In fact, when strike is either extremely high or low in the face, it seemed to have a greater influence on the spin of the ball compared to Spin Loft

So what can we take from this? 

First of all consistency of vertical strike is essential if we are to produce consistent Ball RPM. There can be large difference in RPM direction (backspin or overspin) and amount, if a player has poor control of vertical strike location. Also, it is pointless to have a negative spinloft (Angle of Attack being greater than Dynamic Loft), which is advocated by many coaches, if this results in a low Vertical Strike Location. 

Plan of Action

Over the coming weeks, I am going to be teaming up with Biv Wadden (Putting Coach / Statistical Genius) and Jochem Burghouts (Putting Coach / Capto Expert) to further investigate the true correlations to Ball RPM. We believe they could even stretch even beyond strike location and spin loft, and may also include green interaction. We are currently developing a new formula to predict Ball RPM, taking Strike location into account. Furthermore we also will be dive deeper into this with different putters, balls and perhaps even surfaces.


A word of thanks for the help, support and suggestions must to Biv Wadden and Jochem Burghouts. They helped with much of the input in this article and I look forward to working more on this project together.