Callaway Promises Epic Ball Speed (again) with EPIC FLASH Drivers and AI-Driven Face Technology

When you’ve spent as much time as I have with brilliant R&D guys and really smart (in their own way) marketing guys, you learn most of the industry’s tricks. The line between breakthrough and bullshit isn’t particularly fine, and with the education I’ve received from both arms of the golf business, I’ve gotten pretty good at sniffing out the difference between the two.

At least I thought I had.

When it comes to Callaway’s Epic Flash line, I’ve read through the media packages, I’ve spent a good chunk of time talking with Callaway’s SVP of R&D, Dr. Alan Hocknell, and I’ve kicked the tires elsewhere as to what’s theoretically possible. I’ve thought about it from every angle, but frankly, Callaway’s latest technology story is different-enough from what I’ve heard before that I’ve gone nose blind. I’m not sure I believe it, but I can’t shoot any sizable holes in it either.

If Flash Face (Callaway’s latest world-changing innovation) is real, it might be truly groundbreaking. It’s the kind of thing that could cause a fundamental shift in how golf clubs are designed from this point forward. It could widen the gap between the industry’s haves and have-nots, while firmly cementing Callaway as both the retail and technology leader.

If it proves to be a good story without much in the way of quantifiable gains, it’s still going to help Callaway sell plenty of drivers. Even if it’s no kind of breakthrough, it won’t be bad. Either way, there’s not much downside in this for Callaway.

Flash Face Origins

Like most any driver story, Flash Face’s begins with a quest for more speed. Callaway hoped to build on the gains it got through Jailbreak and thought it could use face design in much the same way. Says Callaway’s Dr. Alan Hocknell, “We wanted to see if we could take the jailbreak effect and make it larger to get even more ball speed while staying conforming to the rules.”

For that to happen, Callaway would need to literally reshape its face, something it wasn’t having a lot of success doing. If you look back at the last several iterations of Callaway face technology, they all featured a distinctive X pattern. Though the X is unique to Callaway, the variable thickness design is common across the industry. Driver faces are generally thickest in the center and gradually thin as they move to the perimeter. You’ll find different shapes here or there, some slight variations like volcanoes or rib structures, but everybody’s formula is basically the same.

“We kept applying our own engineering school knowledge to the problem,” says Dr. Alan Hocknell, “and every time we’d come up with a similar answer.”

I suppose you could say that where face technology was concerned, Callaway was living the definition of insanity, and ironically enough, to break the cycle, Callaway decided to try something a little crazy.

AI & Machine Learning

With advances in artificial intelligence and machine learning, Callaway engineers wondered if it was possible to match a computer to learn how to design a driver face. And if that was possible, could the machine design it better than humans?

Could a computer learn faster, iterate faster, and to no small degree, stop doing the same thing over and over again? It was worth looking into.

If we give it our capabilities, FEA (finite element analysis) and knowledge of what happens at collision and add artificial intelligence software, and let the computer make the decisions a human makes, and perhaps better decisions over time, let it Iterate to gather the learning that AI needs, when it got there, would it design faces better than ours?”

“The answer is yes.” – Alan Hocknell

Before Flash Face could happen, Callaway needed to feed its computer a list of prioritized goals; the stuff the new face had to deliver:

• Maximum ball speed on center strikes
• USGA conforming CT numbers
• Reasonable stress numbers to ensure it wouldn’t implode
• As much ball speed as possible everywhere else on the face

Hocknell says his team didn’t fully understand, but since the ball speed predictions were higher than they’d gotten from conventional designs, they went ahead and made some prototypes. The numbers proved accurate, even if the new design didn’t appear to make much sense.

“It has a pretty unconventional appearance, lots of thick and thin areas. It’s actually thinnest in the center, thick and thin areas sitting right adjacent to one another, unusual humps and hollows appearing for no apparent reason in various places, when we first saw that we were a little worried that we didn’t know what role each new part of that face design played in the overall performance of the face. We didn’t know what to touch and what to do to make it do something different.” – Alan Hocknell

The team built some additional prototypes so they could poke and prod and learn what face features were contributing to what aspect of performance. The process, says Hocknell, “gave us a lot of confidence that we had actually broken new ground.”

Validating (and better understanding) the computer’s designs took the machine learning idea from research to design, and that meant Callaway needed a much faster computer. The updated hardware took what was a 4-5-week process and compressed it into something that could be completed in a weekend. To put all of this in context, it would take a typical laptop about 34 years of processing to spit out one design. Given all the time in the world, it’s still unlikely any human would have come up with this design. Says Hocknell, “We think we’ve taken a sizable step forward in design method.”

In total, the computer developed more than 15,000 iterations before arriving at the final version of Flash Face.

That’s your glorious backstory, but what will Flash Face give you in the real world? In human and robot testing, Callaway says the typical gain is 1.5-2 MPH – and we’re talking about peak, dead nuts center, ball speed. None of this off-center stuff; Callaway says, Epic Flash will make your best shots better – or at least longer – and that’s what makes the technology so compelling.

I ran some quick calculations based on 145 MPH ball speed and reasonable though not ideal launch conditions, and came up with 4-5 yards of additional carry. It’s not the mythical 10 MORE YARDS of yesteryear, but in terms of year-over-year gains, it’s substantial.

Huge, if true.

Gains among Callaway’s Tour Staff are a bit higher, with some guys seeing as much as 4 MPH of newly- minted ball speed.

Unique Designs

What’s interesting, or at least worth a mention, is that face designs are unique to each head. Alan Hocknell describes Epic Flash and Epic Flash Sub Zero as close cousins, but there’s enough of a difference in the two that the face designs needed to be different. While not intentional, it provides a degree of protection against reverse engineering. Even if you could some how pull the face of an Epic Flash and attach it to another driver, it wouldn’t work well. Compared to pathetic humans, the computer has a far deeper understanding of how the face and body work together and how to optimize those relationships.

Updated Crown Material

Callaway is using a new version of its Triaxial Carbon called TC2, which features a slightly tighter weave. It’s your boilerplate strength to weight story. The idea is that by reducing weight in the crown, you can put more mass where it’s useful, like in the sliding track or down low in the Sub Zero model.

But What About the USGA LIMIT?

Given the need to reconcile reported gains with USGA constraints, it’s perfectly reasonable to ask where the rule makers fit into all of this.

We covered this not so long ago. The shorter answer is that despite the fact that many believe the current rules are absolute; driver distance isn’t maxed out. That’s not to say with certainty that Callaway’s story is true, but just about everyone on the R&D side I’ve spoken with agrees that it’s possible to effectively decouple the relationship between CT (the newer standard) and COR (the old standard). The conventional wisdom is that designs have evolved past where they were when the CT standard was developed as a less than perfect replacement for COR. By nearly every account it’s possible to keep CT well under the limit while pushing COR past what was once the limit.

The bottom line is this: there is a tremendous difference in how a golf club reacts when a pendulum is dropped on its face (CT test) and when it collides with a golf ball at speeds often in excess of 100 MPH. Within that difference is an opportunity to increase speed while remaining within the rules.

It should be noted that the USGA can fall back to the COR test – and it may do that if ball speeds creep into excessive territory – but so far, it’s content with keeping CTs close to 239 microseconds and letting the chips fall where they may.

If the ball gets rolled back, the speed contribution from the driver won’t much matter anyway.

Two Models

For now, Callaway is offering the Epic Flash in two models; Standard and Sub Zero. There is a Star version that’s expected to be Asian-market only. No draw version will be offered this time around. The return of Adjustable Perimeter Weighting (a single 16-gram weight) has diminished the need for a separate slice-correcting model.

This time around both the Standard and Sub Zero models offer the weight track. Gone is Sub Zero’s front/back movable weight system, though a front weight port is retained for swing weighting purposes. According to Callaway, the front/back adjustability wasn’t being leveraged all that frequently. Instead, most relied on hosel-based loft adjustments to tweak launch and spin.

You can still expect about 300 RPM difference between the two models at equivalent loft.

The thinking is that average players will leverage the weight track to provide shot shape correction, while more accomplished players (more likely to be SZ players) will use the weight to tweak the starting direction and influence the last 20% of flight or so with the idea being to get the ball to fall a little bit right or left depending on preference.

Mass Properties

While Callaway is habitually non-specific about its mass properties (CG locations, and MOI in particular). It concedes that its MOI numbers are lower this year, but says they’re still in the good range (somewhere in the 8000s).

That’s the result of a more traditional shape (Flash series drivers aren’t as elongated as Rogue was) and the fact that center of gravity locations are more forward, though, according to Alan Hocknell, “not so forward that you get into weird Loft Up conditions.” Again, part of that comes from the shape, and part of it comes from the fact that between Jailbreak and Flash Face, Epic Flash is more face-heavy than the average driver.

With Epic Flash, Callaway is taking a more holistic view of forgiveness. Instead of focusing purely on MOI, it would point out that downrange dispersion is actually better with Flash than with Rogue. The improvements are attributable to more consistent ball speed (Flash Face), and the more forward CG which provides more consistent spin across the face.

What’s Real and What’s Not

Flash Face is one hell of a story, but even if the gains are real, it’s hard to say with any certainty how much is definitively because of its AI-driven face design. The speed could come from other places. Moving CG forward, for example, almost always results in more speed (and lower spin). That gets you yards.

Still, I can’t explain it all away and the fact that new drivers actually measure close to 45.5” (Epic and Rogue both measured 46” despite being listed at 45.5”), suggests Callaway is finding speed in places it hasn’t in the past.

Ultimately, if Epic Flash is faster and more forgiving, how Callaway made it happen won’t much matter.

As for whether or not the industry will view Flash as having broken new ground, time will tell. Like anything else, if it’s real, if it works, the rest of the industry – at least those that can afford it – won’t be far behind with their big computers, artificial intelligence, and machine learning. If it doesn’t, the old conventions will carry-on, and it likely won’t be long before Callaway crafts a new story.

Technology that doesn’t work usually doesn’t last.

Pricing, Specs, and Availability

Filed under Don’t Say We Didn’t Warn You, the Callaway Epic Flash and Epic Flash Sub Zero drivers retail for $529.99. Fully customized UDesign models will set you back $600.

The Standard Epic Flash is available in lofts of 9°, 10.5°, and 12°. The Sub Zero is available in 9° and 10.5°.

Stock shafts include a new green Project X EvenFlow, Project X HZRDUS Smoke, and Mitsubishi Tensei AV. I don’t want to dwell on it too much, but I’d be remiss not to point out that the first one isn’t currently listed on the PX website. You likely won’t find it sold by aftermarket custom fitters (or anyone else this side of eBay). My hunch (and that’s all it is) is that it’s likely similar to the EvenFlow Max Carry. Callaway loves to have things no one else does, but at $529, I’m hard-pressed to find any justification for having an “exclusive” shaft in the lineup. Just my 2-cents.

The stock grip is green/black version of Golf Pride’s MCC Plus4 ALIGN. It might seem a bit odd to put an ALIGN grip on an adjustable driver, so I’ll remind everyone that Callaway’s dual-cog system retains shaft and grip alignment regardless of the settings. As to whether or not a majority of golfers want an ALIGN grip (I do, I do) is an entirely different question.

Retail availability begins 2/1/2019.