GAME GOLF Inaccurate Strokes Gained Putting?

[iacas]

24 minutes ago, cutchemist42 said:

Im just curious what ideas you have for it?

That's a bit farther off topic than I'd like. Good question though and perhaps a great start to a new topic?

[glinks]

16 hours ago, iacas said:

Currently, yes.

It's pretty accurate given how much of a course rating comes from the distance.

Thanks, do you know how much round were recorded in the GG database?

[iacas]

On 10/20/2016 at 6:16 AM, glinks said:

Thanks, do you know how much round were recorded in the GG database?

Not sure what you're saying/asking.

[cutchemist42]

A thought that just came into my head. Im going to make a dummy example. Lets say GG had your strokes gained  as 1 stroke each at driving, approach, and short game to a scratch. It says you lost 3 shots putting.

If we know GG's putting stats are off, doesnt this mean the other data is off too? If we're actually losing only 2 shots putting after a fix, where does the difference get made up? 

[iacas]

13 minutes ago, cutchemist42 said:

A thought that just came into my head. Im going to make a dummy example. Lets say GG had your strokes gained  as 1 stroke each at driving, approach, and short game to a scratch. It says you lost 3 shots putting.

If we know GG's putting stats are off, doesnt this mean the other data is off too? If we're actually losing only 2 shots putting after a fix, where does the difference get made up? 

Two-part answer, and @RandallT can add to it…

1. The short game will be affected most. Other areas will be affected a bit too, but not as much as you'd think. Remember, if you had 30 putts in a round a difference as small as 0.05 per putt is 1.5 strokes. That could be spread out over many categories.
2. This is the part @RandallT can help most with… but I don't believe that the GG handicaps really ever made too much sense in the comparative sense. The numbers didn't often add up properly. A guy who shot 87 might only have 10 strokes lost in that round to a scratch golfer when you added things up. So there was a little weird math going on there anyway. This is a separate issue that, by what I just wrote (I think it tended to come up short) would actually get worse since everyone's soon going to be a better putter (losing fewer strokes in their putting category). So the 87 that lost 10 strokes but had 4 of them in putting might only lose 9 strokes now with 3 in putting… maybe the scratch comparison guy loses two strokes in putting, though, so the number actually jumps to 11 strokes difference? I don't know. Just rambling now.

Basically, yeah, your short game will be affected a little bit, and the other stats a tiny bit. Not very noticeably though. Mostly everyone's putting will just drop 1-2 strokes, I'd wager. That'll be the noticeable change.

[No Mulligans]

As an illustration to what @iacas posted above.  My Game Golf handicap is currently 13.3 but when I compare my strokes gained to a 10 handicap I surprisingly have 7.6 strokes lost.  If I didn't know better regarding the putting stats, I'd try to explain this by saying perhaps my anti-handicap is much higher than average for someone of my handicap.

[RandallT]

7 hours ago, cutchemist42 said:

A thought that just came into my head. Im going to make a dummy example. Lets say GG had your strokes gained  as 1 stroke each at driving, approach, and short game to a scratch. It says you lost 3 shots putting.

If we know GG's putting stats are off, doesnt this mean the other data is off too? If we're actually losing only 2 shots putting after a fix, where does the difference get made up? 

For my stab at the answer, I'll just discuss what I know from my own strokes gained analysis against PGA baselines (and possibly it would be the same for GAME GOLF's methods). First here's the source data:

From the table above, every hole's tee shot has an expected hole-out value. For example, at just beyond 400yds on the tee, you can see that PGA players average 4.0 shots from that point. So if you bogey a 404yd hole, you will lose 1.0 strokes for the hole. Period. That 1.0 stroke lost will be distributed between the 4 categories (drives, approaches, short game, putting), of course- but no matter what they'll add up to a single stroke lost.

Could be you gained in driving, approach, but lost in putting. Or some other combination. But always adds to 1.0 lost.

The only ways to change the distribution of which categories the strokes went to are simply:

1. Play the hole differently
2. Change the underlying numbers on the reference chart above

So I think that if you simply change the putting tables in the linked chart above (as we believe GAME GOLF will do)- such that the reference player was slightly worse, then by the calculations, the other areas would suffer. Assuming the tee shots retain their original expected hole-out value, of course.

Furthermore, since you are typically hitting the green from short game and approach shots, then those areas would likely absorb the impact, depending on where you were hitting the green from that day.  

But you're right, if all you do is edit the putting tables, and if the "expected hole-out value" for each tee shot remains the same, it's simply the distribution of strokes that will change. That's just how the math of crunching the numbers works out by looking at the starting and ending values in those tables for every shot.

 

Unrelated to this specific question, but for posterity for those interested (and I may have documented elsewhere on TST), we typically saw a distribution like this for a low 90s golfer when compared to PGA data above:

As you see, we tend to lose roughly twice as much in the full swing (here, about 16 strokes), as we do in/around the greens (here, about 8 strokes). The important thing I found was that the distribution was somewhat consistent (weighted to the left side of the chart (full swing).

I'm not saying this data is perfect or anything. It is simply anecdotal from around 30 rounds analyzed for different folks here on the SandTrap- all kinda mushed together and rounded a tad for simplicity. It's probably ballpark to whatever ultimate data set exists somewhere.

From what we saw anecdotally, GAME GOLF's distribution for a similar player typically had more strokes lost to putting by 1 or 2, but often a couple strokes fewer strokes lost in other areas. The GG chart had a much more right-ward skew basically.  Of course, they are comparing to scratch, and we were comparing to PGA data.

Just thought some might find that chart interesting.

 

[iacas]

10 minutes ago, RandallT said:

That chart almost exactly matches Broadie's numbers: your percentages are 29%, 38%, 15%, and 19% and his are a few percent of all of those, and of course his are across a wide spectrum of players.

[iacas]

Everyone go check your GG Strokes Gained.

The improvement I've hinted about a few times has occurred… and this looks a bit more like what I'd expect:

[Jeremie Boop]

Yeah, definitely looking much more accurate, even given my lack of editing the flag positions in the last few rounds I played. 

[glinks]

On 20/10/2016 at 0:16 PM, glinks said:

Thanks, do you know how much round were recorded in the GG database?

sorry for my English. I mean, do you know how many round have been registered in the GG database ?

[iacas]

9 hours ago, glinks said:

sorry for my English. I mean, do you know how many round have been registered in the GG database ?

I'm sure they do, but I doubt they just want to share that number, and if I knew, I have an NDA in place for some things.

[glinks]

On 02/11/2016 at 2:55 AM, iacas said:

I'm sure they do, but I doubt they just want to share that number, and if I knew, I have an NDA in place for some things.

Ok I understand thanks