Help Me; then Win a Scorecard License

[iacas]

I like the PGA Tour's new "Putts Gained" putting statistic.

http://thesandtrap.com/forum/threads...ined-Per-Round

I'd like to incorporate this into Scorecard. We can't be quite as fancy because we can't determine how difficult the greens people putt on are, their speeds, etc., but we can figure out some basic figures.

Scorecard currently calculates your percentages from these ranges (and their closest metric versions): 0-3 feet, 4-6 feet, 7-10 feet, 11-15 feet, 16-20 feet, 21-30 feet, 31-50 feet, 51+ feet. Since "Putts Gained" is one value, though, we can show the the value and be done with it regardless of the ranges we have already. So maybe the ranges for PG (Putts Gained) are 0-2 (got to be awfully close to 1.0), 2-5, 5-8, 8-12, 12-16, 16-24, 24-30, 30-40, 40+ or something.

I know there are some articles out there, perhaps written by Dave Pelz or other people who study the game, who can find these stats so we can incorporate them. Frankly, I think basing them off the average scratch golfer should suffice - even a bogey golfer can putt as well as a scratch golfer, or that can be their "goal."

If you can "help" (the amount of help is subjective, and I'd like something of substance, but I don't intend to be SUPER stingy), I'll give you a free license for Scorecard.

[Randy4h]

Just paging through Pelz' Putting Bible. He has a graph showing the percentage of putts made in PGA tournaments and for 20 hdcps (+/-5). Nothing for scratch golfers. Interpolating the graphs:

Length/ Avg. Tour Pro/ Low Tour Pro/ 20 Hdcp
3 / 90% / 84% / 75%
6 / 50% / 44% / 26%
9 / 30% / 24% / 11%
15 / 15% / 9% / 4%
18 / 11% / 7% / 2%
21 / 9% / 5% / <1%
24 / 8% / 4% / <1%
27 / 6% / 3% / <1%
30 / 5% / 3% / <1%
42 / 5% / 2% / <1%

[StrmStlkr]

Here is a table that I used for my own analysis found at http://www.probablegolfinstruction.c...ws06-01-04.htm

Putt Length / Pro / Hcp 0-4 / Hcp 11 / Hcp 26
3 feet / 1.07 / 1.22 / 1.35 / 1.50
6 feet / 1.32 / 1.46 / 1.62 / 1.86
10 feet / 1.54 / 1.63 / 1.83 / 2.10
15 feet / 1.75 / 1.76 / 1.98 / 2.29
20 feet / 1.86 / 1.86 / 2.10 / 2.43
30 feet / 1.96 / 2.00 / 2.27 / 2.61
40 feet / 2.00 / 2.10 / 2.38 / 2.77
50 feet / 2.10 / 2.20 / 2.50 / 2.91

[aca247]

What you might be able to do is see the length of putts they made total or average, your choice, since you already have the length of putt stat. and then compare it to a scratch golfer or make a points system(exp. every 3 feet is 1 point)

[jwolf02]

The charts above look great. But to go in another direction, and this may sound too complicated, but what if you made it dynamic based off of your last 10-20 rounds?

Basically, you could keep the categories already in Scorecard the same (0-3, 4-6, 7-10), and then either a) tally avg. putts from these categories over your last x number of rounds or b) (probably stupid idea) average only the bottom 50% of "putt performances" (to create somewhat of an putting anti-handicap) from each category.

I don't think option b would make that much sense, but it would be easy to see when you are having an above average putting day. The "con" is that you'll always be putting against yourself.

I don't have Scorecard, but do players upload their rounds to a server? If so, I think the closest possible method you could get to the PGA statistic would be to either a)gather stats from each category from specific courses (would probably lead to small sample sizes) or b)gather stats from each category from courses that are grouped by course rating/slope. I realize that not all greens are equal from course to course, but I would assume that the tougher courses generally have tougher greens. I'm sure this would lead to a giant sample size that you could then sort by handicap level to put out a putts gained "goal" for a specific player with a specific handicap playing a specific rated course... with a specific length putt. I might have gotten too complicated, but hope this helped. Good luck!

[andef]

This link provides Putts Gained Per Round, Putting Avg, Putts per round, Proximity to the hole
http://s.wsj.net/public/resources/do...ttingRanks.pdf
On Pelz's website I found some drills but he uses his research to back this up.
Lag Putts (Putt from 60, 40, 80 until you hit 10 putts in total) (1 Point 3-6 feet from hole, 2 Points Inside 3 feet, 4 points Hole)
0 = 40 Hcp; then with increments of one point: 38,34,30,26,22,18,14,11,8,6,4,2,0,+2,+3,+4,+5,+6,+7,+8

You can find stats on this link: http://www.pelzgolf.com/sgh/sgh.htm
They provide In-between, Makeable, Short (6 feet), Short (3 feet) and Big Breaking.
These are all in the same format.

Hope this is somewhat helpful

[B-Con]

If your goal is to emulate the PGPR statistic, you're going to need to two things:

1) Distance gaps. (Scorecard apparently already has those.) 2) Average number of putts by a scratch player from those distances.

Just paging through Pelz' Putting Bible. He has a graph showing the percentage of putts made in PGA tournaments and for 20 hdcps (+/-5). Nothing for scratch golfers. Interpolating the graphs:

I don't have his book. What do you mean by "interpolating"? Technically, it means to fill in data between data points on a graph. Are you reading off values he listed on a graph or inserting your own data points in-between data points the author provided?

Length/ Avg. Tour Pro/ Low Tour Pro/ 20 Hdcp

Here is a table that I used for my own analysis found at

Comparing these two tables, if we're looking for a scratch-ish player, the columns "Low Tour Pro" and "Hcp 0-4" (from the respective tables) are probably the ones we want to use. Assuming a 3-ft putt is always 1 or 2-putted (reasonable assumption), the first table says that such a player makes 75% of the first putts and the second table says they make 78% of their first putts.

Which raises a natural question: What kind of statistics table can we use? A table like the one Randy4h shows (% of putts made by distance) is going to be harder to use, because we're looking for average number of putts from a given distance and that table doesn't show us how close a missed putt is recovered. Remember, in PGPR each putt is compared to the average number of putts taken from that distance, and that includes 3-putts. A player might only make 9% of their 15-footers, but what distance is their second putt and how many of them will have a 3-putt total? If we use such a table, we have to make assumptions about how many putts are required from different distances. Eg, all putts under X distance are at most 2-putted, missed first putts between X and Y distance are A% 2-putts and B% 3-putts (this could be based on the same table, assuming that a missed putt is missed by a certain distance, but that requires guessing), putts between Y distance and X distance are C% 2-putts and D% 3-putts, etc. That's a lot of guesswork, and it could be made simpler by only having two ranges, say all under 30-ft are at most 2-putted and over that half are 2-putted and half are 3-putted, but that's still guessing. I think it would be better to start with a table that uses average putts, that we we don't have to reconstruct data. We're dealing with fairly detailed numbers, we probably don't want to stretch them by guesswork.

[iacas]

I think it would be better to start with a table that uses average putts, that we we don't have to reconstruct data. We're dealing with fairly detailed numbers, we probably don't want to stretch them by guesswork.

Right. Percentages are helpful - they narrow down the search - but if we're going to guess that 78% means one-putt, what % is three-putt? 0.5%?

We can rule out four-putts (they happen, but they're statistically insignificant in a scratch golfer's game), but we need the three-putt percentages. With one- and three-putt percentages, we can figure out the numbers in their entirety.

[B-Con]

We can rule out four-putts (they happen, but they're statistically insignificant in a scratch golfer's game), but we need the three-putt percentages. With one- and three-putt percentages, we can figure out the numbers in their entirety.

Given a table like the one provided by StrmStlkr, 3-putts are factored in because it gives average putts from X distance, which includes any 3-putts. So if we start with the right table, we don't need to know any percentages for 3-putts.

[iacas]

Given a table like the one provided by StrmStlkr, 3-putts are factored in because it gives average putts from X distance, which includes any 3-putts. So if we start with the right table, we don't need to know any percentages for 3-putts.

Yes, I missed that link. That may be all we need if the stats are accurate.

[StrmStlkr]

At the end of the article on that page he has a list of references but they are just various golf books. I have not checked them as I do not own any of them and had no real reason for my purposes of using that information. If anyone has those books they may be able to help validate the accuracy of those numbers.

[aca247]

At the end of the article on that page he has a list of references but they are just various golf books. I have not checked them as I do not own any of them and had no real reason for my purposes of using that information. If anyone has those books they may be able to help validate the accuracy of those numbers.

which books are these

[B-Con]

I glanced through the actual paper, located here . (As an aside, it used to be available from the home page of one of the researchers, but it has since been redirected to this page.)

Starting on page 19, there are graphs shown for both the empirical results and the final model for what they call "putts-to-go". Putts-to-go is their measure of approximately how many putts remain until holing out from a given distance on the green. Since this graph is from the actual source paper and based on the Shotlink data, this is probably as good of a graph as you're going to be able to find.

The problem is that it's a somewhat low-res graph, as it's presented in the paper the quality will likely not be precise enough to be useful. However, it may be possible to contact one of the three researchers and ask for a more high-res graph. I would think that they'll have it, the question is if they may distribute it (I would assume so), and if they may whether they can be bothered to share it with you.

[B-Con]

Update:

I sent an e-mail to the authors asking for more detailed data and one of them just got back to me. He provided me with a very detailed set of data points that they used to create their Putts To Go model in Figure 12. The data points are in increments of 1 inch from 0 to 100 feet.

I think this is as good as anything you could possibly find, the only drawback is that it's PGA stats (from 2003-2008) and not scratch golfer stats. If you're interested in this, let me know.

[iacas]

I think this is as good as anything you could possibly find, the only drawback is that it's PGA stats (from 2003-2008) and not scratch golfer stats. If you're interested in this, let me know.

Sure am. Send them my way. I'm glad you posted too - both you and StrmStlkr should send me a PM with your name and email address so I can send out your free licenses. You've been quite helpful.