Written By: Tony Covey

Can You Really Have the Best Golf Club Testing System Without Using Robots?

It has been our goal from day 1 at MyGolfSpy to create the best golf club review process on the planet. We thought if we use a range of handicaps, swing speeds, and swing types, provide more data, more detail, and be more analytic than anyone else, we could do just that.

I believe we have.

Using humans, imperfect as most of you are, to test golf clubs was a no-brainer for us. We’ve put hundreds of hours into developing, tweaking, and refining our review process. We spent hours on the phone with designers, engineers, performance specialists and club fitters trying to make our process even better. At every step of the way we were certain that, despite a total lack of consistency (even the best players in the world can’t touch a robot for repeatable precision), golf clubs that are played by humans need to be tested by humans.


The reality is humans, play golf, not robots. Until we reach a point where golf is a game played by machines while their human owners drink at the bar, we’ll keep relying on imperfect humans to provide the data for our club reviews.

Golf is not Real Steel.

It may sound arbitrary…even wrong, but I assure you, the decision to rely exclusively on humans was not made without doing plenty of homework.

The Voices of Dissent

Not everyone agrees. The one perpetual sticking point with our process…the one issue that gets raised more than any other is our reliance on humans to provide the data for our reviews.

I can’t count the number of times a reader has told us that without robots our review process can’t be trusted. Despite our rigorous and consistent testing protocols, and our performance formulas which take into account things the average reader might never consider, the argument persists.

We’ve heard it time and time again.

  • Humans are too inconsistent.
  • Humans are unreliable.
  • You should be using robots.
  • Your results are invalid.

Only the consistency of a robot can tell you how well a golf club really performs.

Human results, they say, are almost meaningless.

The truth of the matter is that almost from the very first moment the “robot issue” was raised we starting looking into it. If it truly made sense to use robots for our testing, we’d find a way to make it happen. We did our research, we talked to experts from across the industry, and at every step of the way there was complete agreement.

When it comes to evaluating how a golf club will perform on the golf course, robot results, we were told time and time again, are almost meaningless.

We’ve shared bits and pieces of those findings in our reviews, but we’ve deiced that in the interest of putting the robot issue to bed and once and for all we’d lay it all out for you right here.

When the goal is to determine how a golf club will perform on the golf course, data collected from human testers without question, provides the most reliable insight into actual real-world performance.

The Problem With Humans

There are countless reasons why golf companies use robots to test the performance of their golf clubs and golf balls. Humans are inconsistent. We get tired. We have bad days where we can’t hit the ball to save our lives. Sometimes we hit it on the toe. Sometimes we hit it on the heel. Occasionally we even find the sweet spot.

We are…human.

A robot can hit balls a day long without a single bad swing. Robots don’t sweat. They don’t get tired. A robot hits the ball precisely where it’s told.

There isn’t a single golf robot on the planet that’s got a bug up its metal ass about TaylorMade, or Callaway, or anybody else that some real humans really love to hate.

The robot is always as objective. That is to say the robot is always as objective as the guys programming it and interpreting the data.

Robots can do things humans can’t.

“We can hit 3/4” toe shots over and over again and see how both the ball and the club reacts…Robots allow us to test the same club with the same swing, but at different swing speeds.  We can also just vary launch angle, as an example, with robots to see where a particular club may excel or not as well”. – Mike Vrska, Global Director of Golf R&D, Wilson Golf

Try doing any of that with a pathetic human.

As design tools robots are indispensable. They are essential.

If you need to test a specific impact location, or find out exactly how a change in CG placement from one location to another changes launch parameters, you’re going to need a robot. We’re not disputing that.

“We look at launch conditions up and down and all across the face.  This data is used to tune bulge and roll radii, face thicknesses, structural and mass properties”  – Nate Radcliffe, Director of Engineering for Nike Golf

The thing is, at MyGolfSpy we’re not in the club design business.

The Problem With Robots

There are some things – some more obvious than others – that you simply can’t get from a robot. For any aspect of design and performance that is not absolutely quantifiable, robots are absolutely useless.

“Robots are not good for heads when it comes to sound/feel, adjustability, workability, left/right directional tendencies, shapes, and visual aspects of setup like face angle, crown decals etc.”. –Tom Olsavsky, Senior Director of Product Creation for Metalwoods, TaylorMade-adidas Golf

Olsavsky certainly isn’t alone in his assertions that the human element is indispensable ingredient in the club design process.

“Only players can tell us if a new S-Flex shaft is too whippy or feels like a board.  Only players can tell us that a Fairway Wood feels great, except on heel hits when there is a weird sound or vibration on that impact.  Only players can tell us if the topline looks too thick or too thin. All of the ball speed, launch angle and spin rate data from players is very, very helpful from a design and development perspective, but talking with them during and after the testing can be just as valuable.” –Mike Vrska, Global Director of Golf R&D, Wilson Golf

The fact of the matter is that we don’t concern ourselves too much with that sort of subjective stuff either. Sound and feel issues should be worked out before the club hits retail (you know…during the design phase). Other stuff, shape, and color, and decals; there are no absolutes. As a golf company you design for your audience and hope for the best.

What matters most  – the reason why almost everyone relies so much on human testing (despite the fallibility of humans) – is that even the most advanced incarnation of Iron Byron doesn’t swing like a human, and as a result, doesn’t produce the same results that humans do.

Humans vs. Robots

Guys…especially the YOU NEED A ROBOT crowd let’s take a brief timeout so you can prepare yourself for what’s going to happen next.

Lots of really smart guys, guys who know more about testing golf clubs than probably any other group of guys on the planet – guys (including the guy who basically invented the golf robot) are about to tell you, without reservation or hesitation, that when it comes to determining how a golf club will actually perform on a golf course – robots ain’t got nothing on humans.

One of the first people we spoke with about robot testing was Dick De La Cruz. For those who don’t know De La Cruz, he’s the innovative mastermind behind some of the most famous club designs in history. He’s one of the founders of Hickory Stick Golf (which ultimately became Callaway Golf). He developed many of the tools and gauges golf companies use today, and as it happens, he helped design and bring to life the modern swing robot.

De La Cruz understands that because of their precision, many golfers consider robots “the ultimate answer” when it comes to club testing, but he also tells us that in his experience, there aren’t many people who actually know much about how robots work.

The problem, as De La Cruz explains, is that robot and human swings are fundamentally very different.

“To begin with a human uses two arms 2 hands and two wrists when swinging a golf club while standing on two legs. A robot has 1 arm, 1 wrist, 1 hand and stands on four legs. The robot is fabricated from metal and is very rigid as a human is very supple and flexible in the golf swing”. – Dick De La Cruz

I know…it’s all very obvious, and your first instinct might be to say “so what, robots are still more awesomer, and your data is still invalid”, but as it happens, those basic differences that Dick De La Cruz just spelled out for you…they make all the difference in the world.

Simply put, it’s not a question of consistency and precision, and knowing what happens when you hit the ball .73mm from the sweet spot; even when all other variables are equal, humans and robots produce different results.

The biggest shortcoming [of robot testing] is it is possible to get great results on the robot and then a club doesn’t perform as well with players. –Mike Vrska, Global Director of Golf R&D, Wilson Golf

We asked club designer and custom fitting expert Tom Wishon to explain why it is that a machine engineered to mimic the human golf swing produces results so inconsistent with those produced by actual humans. This is what he told us:

“Robots do not swing like humans in one VERY key aspect of the golf swing.   With humans, once they unhinge the wrist cock release, the arms begin to slow down as the clubhead accelerates to its max speed.  This happens because when the wrist cock is unhinged, that action sends the energy of the arms to the club.  With the energy of the arms sent to the club, the arms then have no other recourse but to slow down, while the clubhead after receiving the energy of the arms begins to accelerate to its max speed.

This action of the arms slowing down while the hands are still firmly holding the grip causes the mass of the clubhead to push the shaft into a forward bend position.  This forward flexing of the shaft causes the dynamic loft (the actual loft of the head at impact) of the head to increase and thus bring about the shaft’s contribution to the launch angle, spin and trajectory of the shot.

The typical swing robot’s “arm” is mechanically driven all the way through the swing.  Therefore when the robot unhinges its “wrist cock angle” the arms do NOT slow down as they do in a human.  This means that the shaft cannot act the same manner of forward bending as it will with a human.  And from this, the dynamic loft will not be the same for a robot swing as it will be for a human swing.  That in turn means the launch angle, spin and trajectory won’t be the same for a robot hit shot as with a human hit shot.

Added to this is the fact that the robot’s “hands” are completely rigid because the “hands” are made from metal and not flesh, tissue and muscle.  In humans the forward bending action of the shaft is dampened by the fleshy and supple makeup of the hands.  Not so in robots, so this too affects how the clubhead is delivered to the ball and changes the shot result of a robot vs a human. ” – Tom Wishon, Owner Tom Wishon Golf Technology

Say that again Tom…

“That in turn means the launch angle, spin and trajectory won’t be the same for a robot hit shot as with a human hit shot.”

And that brings me to one of the most fundamental issues with robot testing:

Robots Can’t Test Shafts

This isn’t my opinion; it’s the well-educated opinion of TaylorMade’s Tom Olsavsky, Callaway’s Luke Williams, Tom Wishon, and basically anyone who actually knows anything about testing golf equipment with robots.

I think we’d all agree that the shaft is a pretty big and important piece of the golf club, and as it turns out, robots are all but useless when it comes to shaft testing. Robots show almost zero distinction between shafts of different flex, weight, torque, or any of the other engineering details that go into shaft design and ultimately play a role in how they perform in human hands.

It begs the question; if you can’t test the shaft, how do you test the finished club?

You might as well just stick a rigid steel rod in everything you test…which is exactly what Tom Wishon does when he tests his heads.

The completely rigid rod prevents the robot’s arm action from creating any atypical results. It allows Wishon to collect valuable information about the performance characteristics of his designs. What it doesn’t tell him is how the club will perform for actual humans.

For Wishon’s work as a designer, that’s fine…probably perfect even. Robots are invaluable tools in the design process, but for the average guy who wants an idea of how a fully assembled, off-the-rack driver (yes…the reality is the average guy still buys off the rack) will perform for him, the robot data, it turns out, isn’t particularly helpful.

We Don’t Buy Heads, We Buy Clubs

The average golfer also probably isn’t interested in having his next driver outfitted with a rigid steel shaft. The average golfer isn’t buying a head; he’s buying a complete club.

While a robot can tell you how a given head will perform at a certain speed, to a man, the experts we spoke with agree that it can’t tell you exactly how the assembled club will perform in human hands.

For that you need humans – living, breathing, wholly imperfect humans.

Robots Don’t Buy Clubs, and They Don’t Play Golf

In the last several years I’ve had countless casual conversations with designers, engineers, and other performance and fitting specialists. I’ve asked nearly every last one of them which – human or robot – yields more valuable information about the actual performance of their golf clubs.

To date, not a single person has answered “robot”. Not one.

It’s Actually Very Simple

Finally, we asked the experts a very simple question:

Which method of testing provides the best indicator of how a club will perform on the golf course?

Here’s what they told us:

Don’t take our word for it…take it from the experts…all of them. When it comes to determining how a golf club will perform for an actual human, robots are a poor substitute for actual humans, which is exactly why MyGolfSpy has and will continue to test golf clubs using real live human testers.