My Golf Spy Ball Test - General Discussion

[Jagpilotohio]

Dean Snell is officially a Demi god.

 

MYB-X is the Fastest and longest ball as well as high spin off the wedges for $30 a dozen. Nice.

[vartanick]

What I don’t get is the performance breakdown chart, what’s makes an excellent ball diferent to a good ball

[Celeras]

Phenomenal test, great job by the spy guys. Interesting that the Snell Black was basically the only ball to go offline to the right.

[bogeypro]

Was all the testing performed with a robot?

[noodle3872]

> @bogeypro said:

> Was all the testing performed with a robot?

 

Yes

 

[flushem]

I have not played avx but all my friends can't get enough of it. So I wanna give a try.

 

From the stat table, I find that B'stone Tour B RXS, Srixon QST and Wilson Duo U are identical to avx.

 

Does anyone agree?

[BB28403]

TP5X is PRETTTTYYYYYY PRETTTTTTTTYYYY

PRETTTTTTTTYYYY PRETTTTTTTTYYYYY

LONG.

[GolfTurkey]

> @BB28403 said:

> TP5X is PRETTTTYYYYYY PRETTTTTTTTYYYY

> PRETTTTTTTTYYYY PRETTTTTTTTYYYYY

> LONG.

 

And it spins like a wet Pinnacle!

 

 

[Myherobobhope]

Ended up buying the mg cv4 tour based on the tests... 2 mph slower than the mtb x but significantly lower spin, which I need... price point is good too. Bought some gloves while I was there

[Sixcat]

> @vartanick said:

> What I don’t get is the performance breakdown chart, what’s makes an excellent ball diferent to a good ball

 

Not much at all. Tony Covey even Tweeted last night that any ball in the top two categories would be an exceptional choice. I would go a step further and say the top three categories. It's more important to find the right ball that fits your game and the test results should objectively be used to that end.

[dcmidnight]

I'm surprised at some of the dispersion boxes - "shot area" - seen with some of these balls. The Snell might be the longest but - in addition to being the highest compression - it also has a dispersion area of almost 1500 sq yards. So a box roughly 39 yards by 39 yards of dispersion on a robot test? It makes me wonder if maybe one ball got rope hooked pretty good, went forever and threw off the rest of the results. Or maybe there was just one bad ball in the bunch - meaning it wasnt the robots fault it went offline. I dont know how else to explain a dispersion pattern that large on a robot test. But since the shot by shot data is not on there, you cant tell. It would be nice to see the shot by shot data and/or be able to remove the outlier shots for each ball.

 

Someone said it might be manufacturing tolerances but I'm not buying that either since some of the OEM balls have even bigger dispersion patterns.

[GooseHook]

> @dcmidnight said:

> I'm surprised at some of the dispersion boxes - "shot area" - seen with some of these balls. The Snell might be the longest but - in addition to being the highest compression - it also has a dispersion area of almost 1500 sq yards. So a box roughly 39 yards by 39 yards of dispersion on a robot test? It makes me wonder if maybe one ball got rope hooked pretty good, went forever and threw off the rest of the results. Or maybe there was just one bad ball in the bunch - meaning it wasnt the robots fault it went offline. I dont know how else to explain a dispersion pattern that large on a robot test. But since the shot by shot data is not on there, you cant tell. It would be nice to see the shot by shot data and/or be able to remove the outlier shots for each ball.

>

> Someone said it might be manufacturing tolerances but I'm not buying that either since some of the OEM balls have even bigger dispersion patterns.

 

Water on the face? Small debris on the ball or club? Wind not normalized for?

 

I can't say I've ever blamed a shot that wayward on the golf ball.

[Lodestone]

Does any of this testing address the issue of quality control? It would be difficult to test, but do we know which companies' balls are the most consistent in performance to each other?

[dcmidnight]

I agree. Then doesnt it make sense to remove that outlier from the results? If you arent going to "blame" the ball for a shot that wayward you cant "give credit" to it either if the distance number is significantly higher.

 

A lot of people have been in club fittings or launch monitors and you get one real outlier for whatever reason, maybe its even just a bad read on the LM. But its clearly an outlier - so you unclick that shot so it doesnt skew your results.

 

Or maybe I'm thinking about it all wrong. I was just surprised.

 

And this has nothing to do with just one ball, there are a lot of balls with "shot areas" over 1500 sq yds.

[TheCityGame]

To address a couple things people have brought up. . .

 

They indicated in the video that the data was checked for missing or anomalous results. Not sure what line they used to differentiate "anomalous" from "bad". I'm guessing that they threw out results where the trackman was clearly wrong, and it wasn't just a bad shot that the trackman measured, but who knows.

 

They also tested over 3 days, and switched balls often to account for environmental effects.

 

For anyone interested. . .they have a video from February where they actually tested Snell MTB-X against Pro V1, using a human and they used a 100 mph swing speed in that study.

 

 

I don't dis-believe their results at all, but I just kind of wish they didn't seem to have such a close relationship with Snell himself. They also have this old video where Snell answers questions about golf balls. It's informative.

 

 

Personally, I've been playing the OnCore Elixir for a while. It fared well in the testing. Its distance is "above" what balls of similar compression are, so might be a "better feeling" distance ball (or a "better distance" feel ball). I've just put together some good rounds with it so I've come to like it and trust it.

 

 

[Sixcat]

> @dcmidnight said:

> I'm surprised at some of the dispersion boxes - "shot area" - seen with some of these balls. The Snell might be the longest but - in addition to being the highest compression - it also has a dispersion area of almost 1500 sq yards. So a box roughly 39 yards by 39 yards of dispersion on a robot test? It makes me wonder if maybe one ball got rope hooked pretty good, went forever and threw off the rest of the results. Or maybe there was just one bad ball in the bunch - meaning it wasnt the robots fault it went offline. I dont know how else to explain a dispersion pattern that large on a robot test. But since the shot by shot data is not on there, you cant tell. It would be nice to see the shot by shot data and/or be able to remove the outlier shots for each ball.

>

> Someone said it might be manufacturing tolerances but I'm not buying that either since some of the OEM balls have even bigger dispersion patterns.

 

Somewhere in the comments I saw the MGS guys reply that one shot with the driver went massively offline which led to the results being what they were for the MTB-X. Then the result repeated itself with the 7-iron. Maybe one ball in the dozen was "out-of-round" or the cores not aligned in the center? Seam issue with the cover?

 

Then I saw on Twitter where the MGS account replied to someone saying that one result with the driver and 7-iron kept the MTB-X out of the Excellent category. I ordered a box of MTB-X. Going to compare them myself with the ProV1-X and make a decision whether a $20 per box savings is worth the possible occasional tolerance issue.

[lopey986]

> @BB28403 said:

> TP5X is PRETTTTYYYYYY PRETTTTTTTTYYYY

> PRETTTTTTTTYYYY PRETTTTTTTTYYYYY

> LONG.

 

Uh, not really. It's in the bottom half of their carry distance in testing.

[Mahamilto]

I feel like a clear winner is the Tour BX... which has tested well for me in the past too.

 

It was 2nd longest tour ball off driver with very low spin and mid-mid/high launch.

 

It had ample spin off 7iron without losing a ton of distance compared to lower spinning offerings, as some other balls did.

 

It spun very well with the wedge.

 

I wonder if the dual dimple has something to do with it. We all discuss spin ball speed, and launch, but rarely talk aerodynamics. Something is making this ball do things the others aren’t.

 

A metric I have never seen studies is “ball speed retention”. In other words, 100 yards off the driver how fast is it going? This is likened to baseball where the stadium radar guns always read higher because they pick up the ball as it’s released, where scouts use radar guns that average Initial velocity with velocity on the ball when 1/2 way to the plate.

 

When you look at scout radar guns, you quickly see that not all 95mph fast balls are the same. A similar phenomenon must exist with golf balls.

[gregkeller]

It seems like the Snell MTB-x "dispersion" rating is about left to right miss, because the standard deviation for distance is pretty much where all the other balls are falling (5.52 std. dev for carry) but the highest std. dev in the offline category (21 yds). That is interesting and I'd love to see the raw data on each of those shots. Everyone is focused on the area and assuming it's a 39x39 yd square, but based on these numbers it's not, its a gigantic rectangle that is relatively narrow in the distance axis, but really, really wide in the offline axis. If this isn't some weird data anomoly it would scare me a little bit.

[mmack067]

> @Mahamilto said:

> I feel like a clear winner is the Tour BX... which has tested well for me in the past too.

>

> It was 2nd longest tour ball off driver with very low spin and mid-mid/high launch.

>

> It had ample spin off 7iron without losing a ton of distance compared to lower spinning offerings, as some other balls did.

>

> It spun very well with the wedge.

>

> I wonder if the dual dimple has something to do with it. We all discuss spin ball speed, and launch, but rarely talk aerodynamics. Something is making this ball do things the others aren’t.

>

> A metric I have never seen studies is “ball speed retention”. In other words, 100 yards off the driver how fast is it going? This is likened to baseball where the stadium radar guns always read higher because they pick up the ball as it’s released, where scouts use radar guns that average out of the half speed and speed 1/2 way to the plate.

>

> When you look at scout radar guns, you quickly see that not all 95mph fast balls are the same. A similar phenomenon must exist with golf balls.

 

I was trying to figure out how it was so much longer than the TP5x that I currently play. The ball speed, launch, and spin at 115 are very close but the apex is different which makes me think its the aerodynamics that are making the difference. More forward speed, less up, compared to the TP5x. It's one I'm going to test in the coming weeks for sure.

[erock9174]

I’m just glad it was robot testing and not humans for this one.

[Wooderson]

If premium balls have variances in shot area then can you imagine what the $12-$25 balls are like. :#

[NRJyzr]

> @Wooderson said:

> If premium balls have variances in shot area then can you imagine what the $12-$25 balls are like. :#

 

 

It isn't always a sign of the golf balls, but a sign of just how variable hitting a golf shot is.

[arbeck]

The dispersion almost has to be a manufacturing defect. If you go watch an old Mark Crossfield video with Titleist, they have him hit balls where the dimples on one side are fractionally different than the other and the ball goes way left or right. Things like that, the core being misaligned, imperceptible damage, etc would probably cause the dispersion. No matter how good your QC is and how tight your tolerances are, everyone is going to produce a ball that is off every once in a while. Unless they do a quality check on every single ball, which would add a ton to the price. I would assume most everyone is doing random QC checks on all parts of the manufacturing process with a statistically significant sample size.

 

[dcmidnight]

> @gregkeller said:

> It seems like the Snell MTB-x "dispersion" rating is about left to right miss, because the standard deviation for distance is pretty much where all the other balls are falling (5.52 std. dev for carry) but the highest std. dev in the offline category (21 yds). That is interesting and I'd love to see the raw data on each of those shots. Everyone is focused on the area and assuming it's a 39x39 yd square, but based on these numbers it's not, _its a gigantic rectangle that is relatively narrow in the distance axis_, but really, really wide in the offline axis. If this isn't some weird data anomoly it would scare me a little bit.

 

 

I buy this idea but in practice how would that work? If their dispersion box is 1600 sq yards for a particular ball - and I'm just changing it so the math is easier - it could be 40x40 square. If its relatively narrow in the distance axis, which sounds reasonable, that means it would be say 80 yards wide by 20 yards on the distance axis.

 

To me - and I'm not a data guy - but an 80 yard horizontal dispersion on robot shots just seems improbable unless there is a real defect in the ball.

 

Although someone a few posts ago said that MGS admitted one of the shots was really really off - but it was apparently left in the final data set. Even though they also say they took "anomolous results" out of the final numbers. Dunno.

 

Good study even though I think there's a little here and there I would nitpick about. I do like that they implore people to pick a ball and play that ball. I play with so many guys - as I'm sure you all do - who just have a grab bag of anything from 2 piece balls to 5 piece balls and they will play anything at any time.

[Joker91]

So many people getting suckered in by this "test". Talking about winners and the defects.

[dcmidnight]

Also, please stop PMing me about the fact that MGS is in our filtered list. Decisions like that come straight from the owners of the site and the mods have absolutely nothing to do with it. Thanks.

[arbeck]

> @dcmidnight said:

> > @gregkeller said:

> > It seems like the Snell MTB-x "dispersion" rating is about left to right miss, because the standard deviation for distance is pretty much where all the other balls are falling (5.52 std. dev for carry) but the highest std. dev in the offline category (21 yds). That is interesting and I'd love to see the raw data on each of those shots. Everyone is focused on the area and assuming it's a 39x39 yd square, but based on these numbers it's not, _its a gigantic rectangle that is relatively narrow in the distance axis_, but really, really wide in the offline axis. If this isn't some weird data anomoly it would scare me a little bit.

>

>

> I buy this idea but in practice how would that work? If their dispersion box is 1600 sq yards for a particular ball - and I'm just changing it so the math is easier - it could be 40x40 square. If its relatively narrow in the distance axis, which sounds reasonable, that means it would be say 80 yards wide by 20 yards on the distance axis.

>

> To me - and I'm not a data guy - but an 80 yard horizontal dispersion on robot shots just seems improbable unless there is a real defect in the ball.

>

> Although someone a few posts ago said that **** admitted one of the shots was really really off - but it was apparently left in the final data set. Even though they also say they took "anomolous results" out of the final numbers. Dunno.

>

> Good study even though I think there's a little here and there I would nitpick about. I do like that they implore people to pick a ball and play that ball. I play with so many guys - as I'm sure you all do - who just have a grab bag of anything from 2 piece balls to 5 piece balls and they will play anything at any time.

 

There was almost certainly a defect in one of the balls. If their eyes told them that it really did curve that much, it makes sense to leave it in. No manufacturer is perfect and there are going to be ball defects with all of them. The sample sizes they tested in don't really tell us anything about the defect rate unless one brand had multiple defects in a single box.

[GDTBATH]

I'm really surprised at the Chrome Soft. I primarily play Snell, and have for a couple of years now. I do have a box of Chrome Softs someone gave me. Now I'm worried about hitting them. I hate not to use them as it seems wasteful, but dang if that isn't a big loss of distance. I'd guess that's part of the issue with these types of tests...now it's stuck in my subconscious, whether the difference is actually real or not.

[gregkeller]

> @dcmidnight said:

 

> I buy this idea but in practice how would that work? If their dispersion box is 1600 sq yards for a particular ball - and I'm just changing it so the math is easier - it could be 40x40 square. If its relatively narrow in the distance axis, which sounds reasonable, that means it would be say 80 yards wide by 20 yards on the distance axis.

>

> To me - and I'm not a data guy - but an 80 yard horizontal dispersion on robot shots just seems improbable unless there is a real defect in the ball.>

 

 

I agree 100%, and that's why it seems weird. The basic idea is that 68% of the trials should land within 2 standard deviations of the mean (one to each side) and 99% of results within 6 standard deviations of the mean (three to each side). The fact that the standard deviation from the horizontal dispersion is 21 yds for the mtb-x means that the balls were all over the place, and probably not just one of them. I can't find how many shots they hit with each ball, but that would be helpful in trying to work back towards how many shots had to be way off to get such a wide dispersion.