Is Distance Really That Important for Amateurs?

[DrvFrShow]

My 3H is a 185 carry, with a 205 rollout. It's a half degree difference in loft. The 7W I don't really know. I might readjust it to 18.5.

I can get a 210 carry with my 3W +20R -10L. I'm pushing the danger zone with it, but there's a better chance of it landing on the fairway or in the first cut. I don't hit my 3W off the deck. I think at most I'd use it once a round, and I'm giving up 10 yds off the deck by using my 3H, and I'm more consistent.

Carry distances:

3W - 210 -- the deal is that I'm playing red tees - do I really need the extra distance?

5W - 200

7W - 190 (guessing) - I haven't hit this, I just adjusted my 5W to the 7W position, but I used a 7W last year and hit it 180 carry w/ a rollout to 190. I gained SS since then.

3H - 185

4H - 175

5 - 165

6 - 155

7 - 143

8 - 133

9 - 122

PW - 111

GW - 100

56 (muscleback) - 80 (full)

60 - 70 (full)

But I rarely hit the 56 and 60 full, and I've learned to hit a partial 9 irons - GW.

Then since I'm injured right now and won't be hitting a ball for at least another month depending upon what the doctors say, and I'm starting with a different pro when I do. I found that my swing puts a lot of rotational torque on my left tibia. I need to change that. I need a swing coach to work with me on this like I did with piano lessons so I don't get reinjured. Take a lesson, work on it a week, go back take another lesson just to make sure I'm doing things the right way. I don't care if he starts me on something new every lesson, but having someone there to make sure I stay on track will be helpful. The guy I found has some very good references from his students who have been where I am now.

When I'm ready to hit the driver he can show me the best way to hit it. I think I'll have a lot of "practice" rounds in the beginning.

[natureboy]

I think you can also take it as that longer hitters are given more leeway.

If a guy hits it 250 yards at 3 degrees, he'll have a 75 rating. A longer hitter, let's say 325 yards can match his same rating but at 4 degrees. So he gets 1 degree extra leeway.

Sort of. Your example players would have ratios of 83 and 81 respectively.

The curve is correlated with handicap / average score. Generally to move up (lower average score / lower handicap) on the curve you have to be longer and straighter. The 'leeway' of greater distance is offset by fractional losses from a wider dispersion. Good drives are better, poor drives are worse (deeper in rough or junk). Furyk's 2013 ratio was at 85 right around tour average. That's why he was 16th in scoring average. He was long and straight enough to benefit from his great iron play. Generally, wider dispersion also correlates with a higher percentage of 'doubly awful' (stroke & distance) shots that really offset the distance advantage. If the longer hitter somehow had a wider dispersion and a lower than normal 'double-awful' percentage I would expect him to score lower. I think that would be a very rare case, though.

[natureboy]

I think the graph is commanding people who score 118 or higher to stop using the driver!

Pretty cool chart though- what is the source of the data? Curious where there is any data published on degrees of accuracy.

I tried to PM you as I didn't want to double post - not sure if you saw. For anyone else interested (maybe @Lihu ), the degrees offline curve is derived from limited data. Two data points in Every Shot Counts allowed a straight line fit of degrees offline according to handicap. I then tweaked the line to give it a slight sigmoid shape typical of a learning curve progression evident in other handicap / average score related data.

So it has an admittedly tenuous connection to real world data and the high end of degrees offline may not be high (wild) enough. That said, I think it is still fairly useful and not too far off. The degrees offline was used in another data chart I did that correlated stats to average score and it agreed with at least one golfer's real-world average dispersion.

[Phil McGleno]

I tried to PM you as I didn't want to double post - not sure if you saw. For anyone else interested (maybe @Lihu ), the degrees offline curve is derived from limited data. Two data points in Every Shot Counts allowed a straight line fit of degrees offline according to handicap. I then tweaked the line to give it a slight sigmoid shape typical of a learning curve progression evident in other handicap / average score related data.

So you made it up? Well gee, that's useful.

[DrvFrShow]

I tried to PM you as I didn't want to double post - not sure if you saw. For anyone else interested (maybe @Lihu), the degrees offline curve is derived from limited data. Two data points in Every Shot Counts allowed a straight line fit of degrees offline according to handicap. I then tweaked the line to give it a slight sigmoid shape typical of a learning curve progression evident in other handicap / average score related data.

So it has an admittedly tenuous connection to real world data and the high end of degrees offline may not be high (wild) enough. That said, I think it is still fairly useful and not too far off. The degrees offline was used in another data chart I did that correlated stats to average score and it agreed with at least one golfer's real-world average dispersion.

Would you please explain what that means?

[natureboy]

So you made it up?

Well gee, that's useful.

Not made up. The underlying straight line relationship (slope) is straight from Broadie's book. His data set is quite deep so I trust the accuracy of the linear relationship even if it's only two data points. I added the caveat / clarification specifically so people could add their own 'grain of salt' in evaluating the chart's usefulness for themselves.

The sigmoid shape I got by slightly tweaking the low end and high end values is very mild. It doesn't deviate much from the straight line function. The reason I tweaked the straight-line at all is the sigmoid shape is much more 'realisti'c. You expect a small improvement to affect a high handicapper's average score much less because it will marginally impact their generally poor overall game relative to say a bogey golfer. Similarly there is typically a diminishing returns effect for a particular skill for low handicappers who have optimized much of their game. That's why we revere PGA players. It's really hard to get where they are.

http://en.wikipedia.org/wiki/Learning_curve

http://en.wikipedia.org/wiki/Sigmoid_function

http://www.dumblittleman.com/2008/10/lesson-of-sigmoid-curve.html

Would you please explain what that means?

See how the line is straightish in the middle with steeper slope at the near and far end. Also see above explanation.

[natureboy]

Would you please explain what that means?

Sorry for double-post. Timed out in editor. This is a better image of the typical learning curve / sigmoid ('S') shape.

[iacas]

I'm with @Phil McGleno : you basically made it up. You added a new shape to something because you thought it made sense?

BTW, this chart has several data points and isn't particularly "sigmoidal."

[natureboy]

I'm with @Phil McGleno : you basically made it up. You added a new shape to something because you thought it made sense?

BTW, this chart has several data points and isn't particularly "sigmoidal."

I explained how I did it. If you consider that made up, fine. Doesn't worry me. I'm just sharing something I think is interesting.

I would say his chart above looks more logarithmic toward increasing skill. My chart depicts the ratio of the central average of each of those (actually I used average drive distance) and plotting them as correlated against handicap / average score. If he had wanted to account for average course ratings played, I think he could have called it a continuous baseline. I think the central average for the average score data groups keeps curving and dropping with decreasing average score to the undepicted 70-shooter & club pro ovals through to the depicted PGA pro oval.

For you and also @DrvFrShow & @RandallT Here's a chart showing the relationship of my tweaked degrees offline curve relative to a linear plot of Broadie's 2 data points. As you can see it's still essentially a linear relationship hewing close to Broadie's slope. I expect straightening this curve wouldn't change the shape of my Distance / Degrees Offline chart by much at all. I'll do it for you later.

It could be that the actual degrees offline data for higher handicappers would show a tapering plateau of maximum degrees offline rather than the gradual increase I have interpolated. The reason I think it is likelier to nose down with lower average score is that where for a high handicapper a slight improvement in an infrequent hit target rate matters less than the rest of the bad shots in the game, and for a pro with a really tight dispersion, a significant miss becomes the rare event. As Broadie says, the less something happens the less it matters. That's my reasoning anyway, but would happily defer to a more complete data set.

Just finished watching the new Big Break episode #2. Why do you think so few people went for the 5-pointer? Too much cross-slope?

[DrvFrShow]

I looked up my angle. I'm 8 degrees offline with my driver. I'm 5 degrees offline and hit my 3 wood about the same distance. According to the charts, I could be shooting 80-82 by the end of the summer simply by using my 3W off the tee instead of my driver. The charts tell me that I have no business hitting a driver. In other words, I could have a handicap of 8-9 by the end of the summer by simply not using that club.

[natureboy]

I looked up my angle. I'm 8 degrees offline with my driver. I'm 5 degrees offline and hit my 3 wood about the same distance. According to the charts, I could be shooting 80-82 by the end of the summer simply by using my 3W off the tee instead of my driver. The charts tell me that I have no business hitting a driver. In other words, I could have a handicap of 8-9 by the end of the summer by simply not using that club.

That's unusual that you hit your 3-wood about the same. How much less on average? Where do you position the ball in your stance relative to with the 3-wood? What is the loft on your driver and 3-wood?

Ditching the driver on most holes except the longest or most inviting fairways works for Henrik Stenson. It could be worth trying. Before you completely give up on driver, though, I would also experiment with gripping down (about where you would be - distance from hosel - with the 3-wood). Then you are swinging a similarly controllable length club with a larger, more forgiving face than 3-wood with a loft meant to move the ball forward. It's not uncommon that gripping down a bit with driver actually increases average drive for amateurs.

-----

For those who have expressed interest ( @Lihu , @iacas , @saevel25 , @RandallT , @Phil McGleno ) here is the Distance / Degs Offline Ratio vs Average Score Chart (for course rating of 72) I posted earlier but using only the straight line data for degrees offline v average score. It is pretty much the same with the tails slightly flatter / less curved - as expected. I added some lines (these are made up) to capture the idea of the spread around particular scoring averages (the grouping circles in Broadie's chart that Iacas posted). Basically there is more variability at the high score end reflecting the larger inconsistency in both the long game and the whole game for higher handicappers. It roughly represents the idea of an increasing variance around the mean shown in the chart. Also the fit for the average could possibly follow a power law relationship rather than a s-curve polynomial, but that curve fits the numbers I have less closely.

I really hope that Broadie goes into this area a bit more with his next book, because it would be interesting to know the 'real' shape of the distance and accuracy curves at the tail ends. Is there a diminishing returns effect for pros or is there a 'hard limit' of accuracy due to the limits of human coordination that they might be close to? If the latter, that would make distance even more important for pros, but he says that's not the case in his book. Anyway if others express a desire to see this data, it may encourage him to put it out there.

[DrvFrShow]

That's unusual that you hit your 3-wood about the same. How much less on average? Where do you position the ball in your stance relative to with the 3-wood? What is the loft on your driver and 3-wood?

Ditching the driver on most holes except the longest or most inviting fairways works for Henrik Stenson. It could be worth trying. Before you completely give up on driver, though, I would also experiment with gripping down (about where you would be - distance from hosel - with the 3-wood). Then you are swinging a similarly controllable length club with a larger, more forgiving face than 3-wood with a loft meant to move the ball forward. It's not uncommon that gripping down a bit with driver actually increases average drive for amateurs.

I play a BioCell 14.5*D 3W. The distance I'm using is off a tee played about 2" inside my left heel. I don't hit that high of a shot with it, and I carry it about 215 with a rollout to 235, and about 5 degrees off target. This tells me that I have a low spinning shot. I'm playing my driver off my left heel. The loft is 12.5*D. I carry it about 215 and I get a rollout to about 230 plus it's off target more frequently up to 8 degrees. Occasionally I'll really crush one with it, but the payoff isn't worth it. I have a 44.5" shaft on it. Standard shaft is 45.5" so it's already cut down.

I hit my friend's 13.5* Ping G15 a few weeks ago and crushed it. Although it was a women's shaft and felt really soft, I think I need the loft on the club. When my leg gets better and I start playing again, I'll start some lessons with a new coach and see what he says. I may need a refit. I'm thinking a women's club head with a shaft upgrade.

[saevel25]

I really hope that Broadie goes into this area a bit more with his next book, because it would be interesting to know the 'real' shape of the distance and accuracy curves at the tail ends. Is there a diminishing returns effect for pros or is there a 'hard limit' of accuracy due to the limits of human coordination that they might be close to? If the latter, that would make distance even more important for pros, but he says that's not the case in his book. Anyway if others express a desire to see this data, it may encourage him to put it out there.

If it really mattered what the tail ends look like because they are basically statistical outliers anyways. I would say your graph misrepresents them. I think 58 was the lowest score ever posted in a tournament round (not PGA, I think it was in Japan).

So basically your graph should go more sharply as it gets closer to, I would say 50. Just in case someone gets crazy and goes 22 under par. I would presume this value goes towards infinity as the degrees get closer to zero. You do have a ratio as your Y-Axis. Theoretically a person could get very close to zero on the accuracy. Just like someone could get close to zero on the ratio. The could have a very large degree of accuracy with a very short distance. I think the graph would taper off towards zero, not dive down to the X-AXIS.

This is just how I see it in terms of math and what is physically possible. Given I am not a fan of averaging out points. It blurs that golf is a game not just made up of distance and accuracy on tee shots, but also distance control, short game and putting as well. I think it would have been better served to leave all the data points up there and fit a line through the data points.

Besides that the graph really doesn't interest me at all.

[natureboy]

I play a BioCell 14.5*D 3W. The distance I'm using is off a tee played about 2" inside my left heel. I don't hit that high of a shot with it, and I carry it about 215 with a rollout to 235, and about 5 degrees off target. This tells me that I have a low spinning shot. I'm playing my driver off my left heel. The loft is 12.5*D. I carry it about 215 and I get a rollout to about 230 plus it's off target more frequently up to 8 degrees. Occasionally I'll really crush one with it, but the payoff isn't worth it. I have a 44.5" shaft on it. Standard shaft is 45.5" so it's already cut down.

I hit my friend's 13.5* Ping G15 a few weeks ago and crushed it. Although it was a women's shaft and felt really soft, I think I need the loft on the club. When my leg gets better and I start playing again, I'll start some lessons with a new coach and see what he says. I may need a refit. I'm thinking a women's club head with a shaft upgrade.

That's a nice distance for a 3-wood! In theory, you should be able to carry the driver in the ballpark of 240 yards.

You tee the ball slightly forward with driver and some recommend really forward - up to the left instep or toe so that you catch the ball level to ascending. You may still be hitting down on it and de-lofting the club a bit. Softer shafts tend to add dynamic loft at impact so your friend's club possibly offset a tendency to de-loft with where your ball is positioned.

If you do experiment at range with teeing driver well forward make sure to have your path slightly to the outside / right of the target line to 'zero out' the D-plane numbers (https://www.youtube.com/watch?v=rcHUF_-hWcA).

But don't change hitting down / sweeping with the 3-wood or whatever you are doing - it's clearly working.

[natureboy]

If it really mattered what the tail ends look like because they are basically statistical outliers anyways. I would say your graph misrepresents them. I think 58 was the lowest score ever posted in a tournament round (not PGA, I think it was in Japan).

So basically your graph should go more sharply as it gets closer to, I would say 50. Just in case someone gets crazy and goes 22 under par. I would presume this value goes towards infinity as the degrees get closer to zero. You do have a ratio as your Y-Axis. Theoretically a person could get very close to zero on the accuracy. Just like someone could get close to zero on the ratio. The could have a very large degree of accuracy with a very short distance. I think the graph would taper off towards zero, not dive down to the X-AXIS.

This is just how I see it in terms of math and what is physically possible. Given I am not a fan of averaging out points. It blurs that golf is a game not just made up of distance and accuracy on tee shots, but also distance control, short game and putting as well. I think it would have been better served to leave all the data points up there and fit a line through the data points.

Besides that the graph really doesn't interest me at all.

Initially, I kind of expected the same, and if I plot the underlying Degrees Offline per score and Avg Drive Distance per score functions out to the scoring edges further, the Ratio chart does look more as you describe. The asymptote seems to favor ~ 55.

However, I think the low end of score looks nice graphically but as a reflection of a possible average score across multiple rounds isn't realistic. Really accurate and long PGA pros can get up to around a ratio of 100. I doubt 200, let alone 150 is humanly achievable (without bionic limbs) as an average due the random errors inherent in the swing even for super-consistent pros. I would expect there is a practical limit or hard ceiling at the low end...somewhere above the pros.

At the high end, the extension of straight-line or log (non-sigmoid) degrees offline data results in what seems to me unrealistic accuracy values. Do you think 10.7 degrees offline is more realistic for a regular 130 shooter, or 27 degrees offline? How about 11.6 degrees offline for a 140 shooter vs. 38 degrees?

I recall my initial attempts to build a swing. It wasn't pretty. There were complete whiffs (zero distance, zero degrees offline) and lots of shots that went about as far left & right as they did forward. After a fairly short time, I got more consistent, but I think my very early "skill" curve would have started pretty close to the x-axis. I can also imagine a little kid with no experience starting out on the course with a ratio that would start pretty close to zero as well. While I do agree the dive down to the x-axis should probably be shifted out to a higher average score, I don't think the turn down is entirely unrealistic for the initial phases of play. Most people would spend a very small amount of time on that part of the curve, but it would, IMO, imply a sigmoid shape for individual players at least for the very initial learning phase.

[RandallT]

Stumbled onto a little mini-project with a friend here who wanted to analyze his drives. He didn't have Trackman, so we thought we could reverse engineer the formula to get a score from the raw data on the Trackman site.

I'm posting here because I thought it might give an insight into how Trackman (arguably an important industry standard) values distance v. accuracy.  First, here is just a bunch of raw data of drives hit, sorted by TrackMan score (some analysis below that):

Total Distance (yds)	From Centerline (ft)	Trackman score
291.0	95.4	44
293.0	92.1	46
257.0	80.6	47
264.3	83.0	47
300.0	91.6	47
286.0	82.4	50
279.0	76.4	53
285.0	75.3	55
291.0	76.7	55
294.3	75.8	56
282.0	68.9	59
291.1	68.8	60
258.0	56.8	63
268.0	57.9	64
281.2	61.0	65
283.3	61.4	65
259.2	52.8	66
257.0	50.9	67
262.0	51.9	68
291.0	59.6	68
266.0	47.9	72
290.0	53.3	73
287.0	51.9	74
289.1	51.1	75
290.0	51.2	75
262.0	39.3	78
301.7	51.0	78
272.0	37.9	82
258.0	15.9	88
283.0	35.5	88
266.0	21.9	89
276.0	27.0	91
270.0	16.2	93
270.9	17.8	93
271.0	16.2	93
275.0	22.4	93
289.0	32.0	93
269.0	8.3	94
271.0	6.6	94
279.0	23.5	94
273.0	15.0	95
275.0	15.0	95
283.0	24.1	96
280.0	2.5	98
285.0	20.2	98
300.0	29.7	98
282.1	12.2	99
282.9	14.1	99
285.9	18.5	99
286.0	19	99
288.0	0.3	100
288.0	15.6	100
289.0	5.5	100
289.0	20.7	100
290.5	2.5	100
291.9	3.8	100
292.0	21.0	100
293.0	0.1	100
293.0	22.0	100
294.0	8.4	100
294.0	17.4	100
294.0	18.8	100
296.3	14.7	100
302.3	5.6	100
308.1	10.6	100
314.9	26.1	100

From there, we reverse engineered our own little formula of distance vs. accuracy, and came up with these "sample" points to show the same TrackMan drive scores at various distances for an "average"-ish drive of 73 points. There was some error in the results, but not too much.

Total Distance (yds)	From Centerline (+/- a few ft)	ReverseEngineered TrackMan Score
220.0	0.0	73
230.0	25.0	73
240.0	32.0	73
250.0	37.0	73
260.0	42.0	73
270.0	47.0	73
280.0	52.0	73
290.0	56.0	73
300.0	58.0	73

So according to TrackMan, hitting a drive 220yds dead center is worth the same as hitting about 300yds only 20yds (~60ft) off line. Big caveat is whether or not the formula is correct, but I've double-checked it. That doesn't feel right, as I'd take the 300yd drive any day.

[saevel25]

So according to TrackMan, hitting a drive 220yds dead center is worth the same as hitting about 300yds only 20yds (~60ft) off line. Big caveat is whether or not the formula is correct, but I've double-checked it. That doesn't feel right, as I'd take the 300yd drive any day.

Trackman is wrong, your instincts are right.

How about we go to a short par 4 at 350 yards. I get to drop a ball at 50 yards in the rough. You get to drop one at 130 yards in the fairway. Which one do you think has the advantage. It isn't even close, the guy in the rough will make a better score.

[newtogolf]

Trackman is wrong, your instincts are right.

How about we go to a short par 4 at 350 yards. I get to drop a ball at 50 yards in the rough. You get to drop one at 130 yards in the fairway. Which one do you think has the advantage. It isn't even close, the guy in the rough will make a better score.

I'm not sure I agree 100% with that, there are too many variables involved.   Overall I agree that distance > accuracy but I like my chances from 130 yards out on the fairway better than I do from 50 yards in nasty rough.