Optimizing a Driver's Launch Parameters

Dave Tutelman  --  January 24, 2011
This is a rework of an article written four years earlier (January 2007). That article drew criticism not so much for the conclusion as for incorrect distances in the examples. Subsequent investigation showed that the program I used to generate the numbers, Tom Wishon Trajectory Profiler 2.0, was not accurate for higher clubhead speeds. (See my study of trajectory programs which resulted from this observation.) I have more confidence in the numbers presented herein. The conclusions are basically the same as the original study, though the numbers are not quite as compelling.
We hear a lot about how the optimum launch parameters for a driver mean you should be going for higher launch angle and lower spin. This seems to be a consequence of the sudden prevalence of launch monitors in clubfitting shops. As an engineer, I have been trying to make quantitative sense of that for some time now.

I think I have finally found a way to look at it that tells me what's going on. What it tells me is that the conventional wisdom of higher launch and lower spin is an over-worrying of the situation.I'm convinced that fine-tuning the spin and launch angle is mostly a waste of time. If you get the loft right for the golfer's swing, the fine-tuning barely amounts to a hill of beans. I'll show you the numbers that lead me to this conclusion. But first let's go over what I think is the right way to fit a driver with the aid of a launch monitor.

Fitting a driver

Here's my technique, which has the backing of numbers that you'll see later.
  1. Find the shaft and head combination that the golfer hits the best. "Hits the best" means things like:
    • Highest ball speed.
    • Best consistency of impact on the sweet spot, which is centered heel-to-toe and above the middle of the clubface.
    • Consistently good swing path and clubface angle (open/closed).
    • Good subjective feel.
    Notice that I haven't said a thing yet about loft. As long as the loft is vaguely in the ball park for the golfer, you can fit the other things. Yes, loft matters -- and we'll deal with it in the next step. But the properties in this step must be correct first!
  2. Now try out the combination with a variety of lofts. If you have interchangeable heads of the model you're going to use, that's obviously a big advantage. The goal here is:
    • Longest distance, as measured by the launch monitor -- or even by eye on the range, if your range is well enough marked to do it accurately. This is most likely to give the best result.
    • If you don't have interchangeable heads, you may have to choose a loft "by table". See how the golfer's loft compares with his/her launch angle (for this you'll have to measure the launch angle), and extrapolate to the optimum launch angle in one of the common tables. There are lots of such tables around; I just whipped up the following one, which is as good as any.
      Ball speed (mph) 100 115 130 145 160 175 190
      Best launch angle 17 15 13 11 10 8.5 7
      Be sure you do this with at least the same model head and the same model and flex of shaft. For flex, you should take measurements and not just rely on the LARSX markings. BTW, this table does not deal with swing variations that change the effective loft at impact -- such as the wrist cupping common to high handicappers. That can only be accounted for by measuring the actual launch angle and spin.
  3. Optimize launch angle and spin. On second thought, don't bother! You are not going to get a lot of distance out of this step -- if you did #1 and #2 right. A little tweaking may give a little improvement. But there is much more to be gained by choosing a driver where the golfer can make consistent contact just above the middle of the clubface. Any other spin/launch tweaking is small potatoes by comparison. 
Basically, what I'm saying in #3 is: If you got the loft to give the launch angle for maximum distance for that player's ball speed, there is not much left to be optimized by further tinkering with launch angle and spin. Theoretically, perhaps there is. But equipment does not give much ability to hunt down that theoretical improvement. The high-yield ways to find it have to do with the golfer's swing: consistent contact at the proper point on the clubface and with a good angle of attack.

By the numbers

Note: The calculations in this study were done using the TrajectoWare Drive program, version 1.0. This program provides carry distance only, so that is what is being optimized here. As a practical matter, ground conditions vary so much that it is hard to optimize for total distance (carry + roll) with any great precision -- but we are working on a way to include it in the next version of the program. My rough and ready approach is to use about a degree less loft than the optimum, to reduce the spin and angle of descent.

Launch Space

When I hear the words, "You should aim for higher launch angles and lower spin," I visualize the following chart:


The numbers on the chart are yardages. Moving in the direction of the green arrow -- toward higher launch angles and lower spin -- should give the fastest improvement in distance. The yellow arrows -- one with just higher launch angle and the other with just lower spin -- show more modest improvement.

That is what the words imply. However, the real world is more complex. Here are the actual yardages, for a golfer with a 124mph ball speed. (That's a fairly typical male golfer with 85mph clubhead speed and consistently good impact with a "standard" driver. Note that, as the loft goes up, the more-glancing blow means a loss of ball speed. Thus it takes a little more clubhead speed to maintain the ball speed as the loft goes up.)

Distances for spin and launch angle
Spin
(rpm)
Launch Angle (Degrees)
8 10 12 14 16 18 20 22 24 26
5500 164 164 164 162 160 158 155 151 147 143
5000 171 173 173 173 172 170 167 164 161 157
4500 175 179 181 182 182 181 179 176 173 170
4000 177 182 186 188 189 189 188 187 184 181
3500 175 182 188 192 194 196 196 195 194 192
3000 169 179 187 192 197 199 201 201 201 200
2500 161 173 183 190 196 200 203 205 205 205
2000 151 165 176 185 192 198 202 205 207 208
1500 139 155 168 178 187 194 199 203 206 208
1000 128 144 160 170 179 187 194 199 203 206

It's a little hard to understand the shape of this "launch space" just looking at a sea of numbers. So let's enhance it graphically. Here is a 3D picture of the yardage contour.

The shape looks as if you had taken a paper rectangle and curled it along the diagonal. There is a top diagonal line going from the high-spin, low-launch corner to the low-spin, high-launch corner. The two corners bent down are high launch and spin, and low launch and spin. The whole paper is tilted along the diagonal, so the highest corner is at high launch low spin.

At first glance this does seem to say, "Go for high launch, go for low spin". But a closer look (below) is going to tell us we need to be somewhere on that high diagonal. We will see that moving off the diagonal, even if it is for higher launch or lower spin, is going to cost distance. So going for high launch and low spin helps only if we manage to get both, and in the proper proportions to stay on the diagonal.

This 3D graph shows us what is going on very intuitively. But we'll need another way of looking at the surface to get enough detail to be useful.

Here's that other way. The table below is color-coded according to the distance. The warmer the color (red is the warmest), then the longer the drive. Conversely, the purple entries are the shortest drives. This is a 2-dimensional representation of the 3-dimensional surface shown above.

Distances for spin and launch angle
Spin
(rpm)
Launch Angle (Degrees)
8 10 12 14 16 18 20 22 24 26
5500 164 164 164 162 160 158 155 151 147 143
5000 171 173 173 173 172 170 167 164 161 157
4500 175 179 181 182 182 181 179 176 173 170
4000 177 182 186 188 189 189 188 187 184 181
3500 175 182 188 192 194 196 196 195 194 192
3000 169 179 187 192 197 199 201 201 201 200
2500 161 173 183 190 196 200 203 205 205 205
2000 151 165 176 185 192 198 202 205 207 208
1500 139 155 168 178 187 194 199 203 206 208
1000 128 144 160 170 179 187 194 199 203 206

What is the "lay of the land" that we're looking at?
  • There is a "ridge" of maximum distance. Here's the same table with the ridge shown explicitly. It is the black dotted line, which is the maximum distance for a given launch angle.
  • If you move along the line, the yardage changes rather slowly. If you move away from the line, you lose yardage more quickly. That's what makes it a ridge.
  • The ridge is slanted compared to the east-west and north-south axes of the map. It goes from high-launch/low-spin to low-launch/high-spin.
  • There is no "peak" on the chart. As you move down and to the right, the distance continues to increase -- but the increase slows to a crawl at the bottom right. (A bit more exploring with TrajectoWare Drive shows a maximum just under 210 yards, at a launch angle of 30 and a spin of 1000rpm.)
Let's re-phrase that in terms of driver performance:
  • As long as your launch angle and spin are "on the ridge", you are fairly close to the maximum distance you can get with your swing.
  • If you do something to reduce (or, for that matter, increase) the spin, you must do something to also change the launch angle, or you will fall off the ridge -- that is, lose distance. Just lowering the spin by itself with not help and will probably hurt. You have change both to stay on the ridge.
  • If you just raise the launch angle things, are a little better -- but not much. You may get a slight increase for a few degrees, before the distance goes down again.


Reality Bites

Loft Distance Launch
Angle
Spin
8 149 7.5 1900
10 170 9 2400
12 183 10.5 2800
14 188 12 3300
15 189 12.8 3550
16 188 13.5 3800
18 181 15 4200
20 171 16.5 4700

So far, we're talking as if we can design a driver to place us anywhere we want in the chart. But that is not true. There is a "natural" area of the chart for drivers to live, and we have to work hard to get away from that area. Let's look at how real drivers behave. Here is the performance of a set of drivers, showing launch parameters and yardages. The table is based on a very simple set of assumptions:
  • 86mph clubhead speed (because 85mph is not enough to keep the ball speed at 124mph, at the loft we are going to need).
  • 200 gram head weight.
  • Zero angle of attack; the club is at the bottom of the swing at impact.
  • The loft listed is the "dynamic" loft, defined at the difference between the direction the clubface is pointing and the clubhead is moving at impact. That means it includes not only the loft built into the driver but also:
    • Shaft bend, tilting the head up at impact.
    • Face roll, providing more loft as the impact point is higher on the clubface.
    • Cupped or bowed left wrist at impact, to the extent that this provides loft and not angle of attack.
  • No increase nor decrease of spin due to vertical gear effect. (Here are links for an explanation of gear effect and vertical gear effect.)
  • Very normal conditions otherwise: a ball fitting the TrajectoWare Drive model, normal temperature, altitude at sea level, etc.
The only thing we vary in this table is the loft of the driver. And we record the calculated distance, launch angle, and spin. The highlighted row is the driver with the best distance. Our 86mph golfer can get 189 yards with a 15 driver. Let's see how the collection of feasible drivers maps onto the launch space, how it fits on the ridge of maximum distance.



Since the rows of the driver table include launch angle and spin, we can plot each row as a point on the launch space table. Those points are the little golf balls. The dotted red line connecting them is the path of all drivers we could build for our 86mph golfer by varying just the loft. Important points to note about this diagram:
  • Varying loft does not takes us along the ridge; it takes us almost perpendicular to it, straight across the ridge. That means that there is a fairly well-defined optimum loft. In the case of our golfer, the maximum is at 12.8 of launch angle (corresponding to 15 of loft).
  • Not surprisingly, the maximum distance occurs where the red line meets the ridge (the black line). That's not surprising because the ridge represents the best spin for a given launch angle, or the best launch angle for a given spin.
  • The maximum distance -- the intersection between the black and red lines -- is a long way in launch space from the maximum distance possible. It misses the peak (of 209.7 yards) by 17 worth of launch angle and 2500rpm worth of spin. 
  • If we could somehow traverse this distance of launch angle and spin, there is about 20 yards to gain -- about 10% of the distance of the drive. That is a lot, if we could somehow obtain it. (But, as we shall see, this is not easy. Moreover, it has less to do with club design and more to do with the golfer's swing.)
The lesson here is that the line of feasible drivers and the line of maximum distance are sloped in opposite directions. Where they intersect is the driver that gives the best carry. We may be able to do some fine tuning and add a few yards. But not many yards, and probably not with most golfers.

Before we leave this point, here is a table giving the slopes of the two lines -- the ridge and the feasible drivers -- for some representative ball speeds.

Ball
Speed
Slope of ridge Slope of feasible drivers
mph How many rpm of spin
must be removed for each
degree of increased launch?
How many rpm of spin
is added for each
degree of increased loft?
Measured at
what loft?
(near optimum)
100 mph 140 rpm 187 rpm 18
124 mph 128 rpm 232 rpm 15
150 mph 125 rpm 281 rpm 12
200 mph 120 rpm 370 rpm 8

At this point, we've worked the numbers for steps #1 and #2 of my recommended fitting procedure. That's the "lazy" use of the launch monitor. Now it is time to see just how much we can get from "fine tuning" the launch angle and spin: step #3.

Fine tuning

Now we know there is about 20 yards more distance, a full 10% of the drive, out there somewhere. Is it accessible, or is it the pot of gold at the end of the raibow? There is a fundamental fact of life we need to realize before we start fine tuning, if for no other reason than to set our expectations.

It's really hard to raise the launch angle 5 at the same time you're lowering the spin by 500 rpm. (That's what it would take for our 86mph golfer to get 10 extra yards.) The loft is your strongest tool for dealing with launch angle. But loft takes spin in the wrong direction to stay on the ridge. Increase launch angle using loft, and you'll raise spin, not lower it. You have to find other methods. Such methods are few, and they generally make more modest adjustment of spin or launch angle.

That said, let's look at the approaches club designers can use to increase launch while they decrease spin.

Loft

As we said, this story is good news, bad news:
  • First the good news: loft increases launch angle by almost a degree for every degree of loft. (Actually, it is not quite one-for-one, more like 80%-90%.)
  • Now the bad news: loft also increases the spin. For our 86mph golfer with typical driver lofts, the spin increases by 270rpm for each degree of increased launch angle.
That means that we can easily increase the launch angle, but now we not only have to reduce the spin to stay on the ridge, we have to reduce hundreds or even thousands of RPM of spin just to make up for the launch angle gain. That's a high price, considering we must accomplish both higher launch and lower spin; one without the other is distance lost, not gained.

Angle of attack

This is not a club design issue, but a golfer-training issue. It requires learning how to hit the ball on the upswing, which implies placing it further forward in the stance and teeing it higher. It works because you can increase the launch angle without increasing loft. Every degree of angle of attack gives a degree of launch angle with no effect on spin!

How much can you gain through angle of attack? I don't have a lot of data on what golfers who try to hit up on the ball accomplish. But here are a couple of ways to estimate an answer:
  • Long drive data: These guys have a lot of athletic ability, plus a strong motive to learn an upwards angle of attack; they need both in order to compete. So we can look at the actual data from the high finishers in elite long drive competition and get an idea of the AoA that they manage. I plugged the launch data from two drives from the 2006 ReMax finals into TrajectoWare (they were drives Jason Zuback and Erik Lastowka), and worked backwards to the loft and angle of attack that they must have used. In both cases, the angle of attack was under 1. I was surprised it was so small.
  • More long drive data: Tom Wishon points out that "the long drive competitors can use a head with 5-6 degs and still generate a launch angle of 11-12 degs." Even allowing for a reasonable shaft bend increasing the loft by 3, this is still 3 or 4 degrees of upwards angle of attack.
  • Geometry: If you can somehow tee up the ball 6" farther forward in your stance, the natural arc of a 45" driver will allow hitting up on the ball by an angle of 8. That is a lot. It will also involve teeing the ball up almost a half inch higher. This is probably an extreme upper limit to angle of attack, nowhere near available to most of us.
In any event, this is not a clubfitting parameter. The job of the clubfitter is to specify the best club for the swing the golfer has. That will have some amount of angle of attack (perhaps zero or even negative), so let's move on to considerations the clubfitter can do something about.

Vertical gear effect

We usually think of gear effect in terms of sidespin: hooks and slices. But the same effect can work for backspin as well. Hit the ball high on the face and the clubhead rotation will reduce the nominal spin, compared to a center hit. (TrajectoWare Drive assumes a center hit when computing launch parameters from impact information.) Conversely, a low-face hit will experience increased backspin.

How much can we expect from gear effect? Until a couple of years ago, most people (myself included) didn't think vertical gear effect was an issue at all. Even today, there does not seem to be much agreement on the specifics. For instance:
  • Tom Wishon has reported a difference of 425rpm for a 1/2" vertical difference of impact position on the clubface, with a 105mph clubhead speed. A difference of 1.5" from a very low on-face hit to a very high on-face hit allows 3/4" from a center hit to a high hit. With Wishon's numbers, that allows a little over 600rpm to be saved by vertical gear effect.
  • Dana Upshaw has reported a difference of 3300rpm between a high-face hit and a low-face hit, corresponding to a difference of 1650rpm from a center hit to a high hit. That is almost 3 times Wishon's data. Upshaw's data was at a ball speed of about 145mph, corresponding to a clubhead speed of 98mph, making the difference even more surprising.
Making use of vertical gear effect is a combination of the golfer's swing and a driver fit to that swing:
  • The golfer must have a repeatable impact high on the clubface. High enough to get maximum benefit from the gear effect, but still low enough not to lose COR.
  • The golfer can be helped a little by a low center of gravity for the clubhead. A low CG increases the range and efficacy of hitting "above the center".
  • Shaft flex. (Finally, something we clubmakers can relate to easily.) Dana Upshaw has reported an anecdote involving a long drive champion getting much better runout after landing after going to a flexible-tip shaft (generally considered high-launch and high-spin) and a lower loft in the clubhead. The explanation Dana gives is that the clubhead was more free to rotate and exercise vertical gear effect with the tip-flexible shaft. I'm not sure if I agree; my inquiry suggests that tip stiffness has much less impact on vertical gear effect than Upshaw reports. But Dana's data is what it is, and I can think of no other explanation for it.

Other factors?

There are other factors sometimes cited as ways to get high launch and low spin which, unfortunately, do not do what the proponents would have you believe:
  • Head weight distribution - A low center of gravity is reputed to give a high launch angle without adding spin. Vertical gear effect is a technical explanation for why that should be. A low CG enhances the vertical gear effect as noted above.
    Oh yeah. There's also the possibility of raising the launch angle by moving the center of gravity back away from the face. That definitely works to raise the launch angle. But it has the same effect as loft -- increasing the spin as well -- so it's not helpful this time.
  • Shaft flex - Stiff shafts, and especially tip-stiff shafts, have a reputation for being low spin. (Also low launch. That is not what we want right now, but remember that because it's important.) The biggest launch effect of shaft flex is on the dynamic loft of the clubface. So if you reduce spin, you also reduce the launch angle. The only valid use of shaft flex to reduce spin without lowering launch is as Dana Upshaw revealed; use a tip-flexible shaft in conjunction with a lower loft clubhead to take maximum advantage of gear effect. The shaft flex will increase the loft, the lower-loft clubhead will get the loft back down, and a high-face hit might get less backspin because the shaft will allow more clubhead rotation and the resulting gear effect. (I qualify it with "might", not "will". My analysis does not support the substantial effect that Upshaw noted. So I accept his data with considerable reservation.)
  • Low-spin face - I don't know of a single instance of face design or face treatment that reduces spin. (At least none that is legal. Adding some sort of of slime when you play is not legal.)

Tuning up

We are not left with many tools to do fine tuning -- to try to move along the ridge and simultaneously raise the launch angle and lower the spin:
  • Loft moves across the ridge. Wrong direction.
  • Angle of attack is either there in the golfer's swing or it isn't. If it is, we already accounted for it in step #2.
  • Vertical gear effect is all that is left. Let's look closer.
First of all, the golfer must have a sufficiently repeatable swing that we can depend on high-face contact. (Maybe not all the time, but often enough to be worth fitting to.) To do such a fitting:
  1. Use impact tape or powder to determine ball impact. If impact is scattered, don't even to try to tune any further; this golfer needs lessons and practice, or at least attention to the gross driver fitting (step #1) that would give consistent high-face impact. Continue only if the impact is "tight".
  2. Find the proper tee height for high impact that still retains full ball speed. This will take a launch monitor and impact tape or powder.
  3. Experiment with shafts to find the one that gives the greatest spin reduction for the ideal impact. We're still monitoring impact here. Don't draw conclusions based on lower or higher impact. That will "churn" your shaft selection. If you can't get stable enough impact for this step, stop trying to fit for vertical gear effect.
  4. Now that the spin is lower, we have to get back to the ridge of maximum distance by adjusting the launch angle. We use loft to do this. We may need to decrease loft; that's a good thing, because it also decreases spin. Or it may be increased loft, in which case we have to spend some of our spin reduction to get back to the ridge.


Summary

In closing, let me review my recommendations and the reasons for them:
  1. Begin by finding the length, shaft, and clubhead that the golfer hits consistently, at high ball speed, in a consistent place on the clubface. Use a loft within 2-3 of what the golfer needs, but don't worry too much about loft now -- just consistency, ball speed, and comfort. This is the step where you have to be good at the clubfitting art, and it is necessary to get this right if the rest is to follow.
  2. Now vary the loft, using these components, to give the maximum distance for the golfer.
  3. You are now within 10 yards -- 15 at the most -- of what is "on the table" for that golfer's swing. You might be able to get a small fraction of this distance using a launch monitor and different shafts. Bigger improvement can be made if you can get the golfer to hit the ball consistently high-center on the clubface.
Remember that errors in step #1 can cost 5-20 yards consistently, or even the ability to hit consistent drives. Don't get caught up in step #3, where fewer yards are harder won.

Acknowledgements

I'd like to thank Jeff Parrott of Golf Provisions (Largo, FL) for getting me serious about resolving this mystery. (Below you can see a partial transcript of the emails that got this work started.)

I'd also like to thank David Bahr (from the GolfDiscussion forum) for pointing out the problems that the Wishon program had inserted into my numerical data, and Todd Kos for providing correct real-world data so I could test the various programs I use for validity.

Last, but definitely not least, is Frank Schmidberger. Frank saw my study comparing existing trajectory programs, and started building a tool for his own use. Our correspondence grew into a joint venture that resulted in TrajectoWare Drive. It would have been very tedious to repeat all the calculations for this driver optimization project without our new program -- so much so that this update might never have happened at all.



On Jan 15, 2007, Jeff Parrott emailed me:
Hi, Dave! Have you ever looked into ideal launch angles and spin rates based on ball speed? I do a lot of Vector fittings (with good results, I might add), but wondered if you'd ever taken any interest in the subject.
I responded in part:
I have never played with shafts or weight distribution trying to change the spin at a given launch angle. From what I have seen, the strategy should be:
  • Find the shaft and head style that the golfer seems to hit well and feel good with.
  • Run lofts for that head and shaft, to find the optimum loft for the golfer. Use only center hit data, determined by impact tape and consistency of the numbers on the LM.
  • You will find a pretty broad maximum, with essentially the same distance over a 3-4* loft range. Use a number toward the lower end of that range. Don't worry about exact launch angles or spin at this point; there aren't more than 2-3 yards to be wrung out of it now anyway, and the golfer's consistency (with a head and shaft that feel good and suit his eye) are more important at this point than those last 2-3 yards.
I know that's backwards from what a lot of professional fitters do. They tend to start with a loft that gives the "proper" launch angle, then play with heads and shafts to get closer to the optimum spin for that launch angle.
I decided that I needed to a more precise justification for my method. It jibed with my experience, but I didn't have a very good grasp on why it worked that way. As I proceeded with my investigation, Jeff and I exchanged a few more notes. By Jan 17, I had written the first draft of this article and gave Jeff a private link to it. His response was:
Dave: I may have argued the importance of spin 6 months ago (which is why I emailed you in the first place), but my findings after many Vector fittings agrees with what your article states completely.  In the past, I'd get great launch/distance results, but would waste additional time with the customer because the spin rates just didn't make sense. Now I know why.

Thanks for your input (as I'm sure many others will also do). Now I can spend a lot less time trying to get loft & spin to match so closely.

I also solicited Charlie Badami's opinion, and he concurred that  was what he saw in the fitting cage. Thus encouraged, I'm making the information generally available.


Last modified - Jan 24, 2011