Round Ratings

[Dave242]

I think the anomaly is probably a pure result of wonky par settings. Harry Myers is par 65, which is why I can score close to par even though it is the hardest course of the four.

UT Dallas is setup as a par 28 but probably should only be 27. I still think SSE of 19.5 on that course is ridiculously low, hence my round rating being so much lower than what I get on significantly harder courses.

Also, just as a note, the SSE formulas don't care about the 'par' value for the particular course layout/tee at all. They use (only) the total number of holes, the total layout length, and the foliage value for the course.

The reason I asked about perceived difficulty is because a good way to measure it is what players are expected to score (Par) and what the actually shoot (SSA.....and SSE estimates that).

Par should be a the score you should expect if "the player" throws the intended number of (error free) drives and then has 2 throws to clean up from close range. If par is set across all courses for "the player" being someone who can only throw 250' max, then you have non-wonky settings. But if one course's par is set for this player and another course is set for someone who can throw 450', then you have wonky par settings.....it is like comparing one golf course from the ladies tees and another course from the champion tees (makes no sense).

 

[jeverett]

I mentioned this proposed method of altering the SSE formulas previously (both in this thread, and also right after the SSE functionality was added to DGCR), but here was the proposed steps for altering the existing 3-value 'foliage' scale into a 5-point 'technicality'/difficulty/riskiness/whatever scale:

Step 1 - Rate each hole on the course from 1-5, with 1 being very open and 5 being super technical/tight.

Step 2 - Add all of these values up, and divide by the number of holes on the course.

Step 3 - Determine the three constants using the following formulas (x = your parameter from step 2):

constant_1 = 5x^2 - 55x + 405
constant_2 = -150x + 1950
constant_3 = 5400 / constant_2

Step 4 - Calculate the new SSE using the following:

SSE = length * (1 / constant_1 + 1 / constant_2) + #_holes * (1.67 - constant_3 / 18)

Basically, the difference between this method and the existing method would be:

1). Course layout difficulty would be independent from course foliage. Timg could actually initially populate these fields straight across from the existing course foliage field.

2). Instead of an effectively 3-point scale (where lightly wooded = 1, moderately wooded = 2, and heavily wooded = 3), the 'technicality' value would be 1-5, and the overall value would be the average technicality of each hole (i.e. it could be anything between 1 and 5, rather than stepwise 1,2, or 3).

The functions *may* still require a bit of calibrating against 'real' SSA data again, though, as very few courses would still be rated purely 'light' (easy) or purely 'heavy' (super technical). Also, some method of norming the chosen values would be necessary (e.g. writing a description of what a 1-technicality hole/course 'looks like' vs. a 5-technicality hole/course).

 

[Steve West]

The reason I asked about perceived difficulty is because a good way to measure it is what players are expected to score (Par) and what the actually shoot (SSA.....and SSE estimates that).

Par should be a the score you should expect if "the player" throws the intended number of (error free) drives and then has 2 throws to clean up from close range. If par is set across all courses for "the player" being someone who can only throw 250' max, then you have non-wonky settings. But if one course's par is set for this player and another course is set for someone who can throw 450', then you have wonky par settings.....it is like comparing one golf course from the ladies tees and another course from the champion tees (makes no sense).

That "2 throws to clean up from close range" thing also contributes to wonkiness. But that's another thread.

 

[DocHaole]

The DGCR rating is pretty stinkin close. My PDGA is 963 and DGCR is a 966.

However some of the individual round ratings are laughably off:

Campton Hills in St Charles IL a -5 on 9 holes is a dismal 866. which I thought was a fairly good round.

Then at Sky High in Wrightwood CA my 75 (-6) was given a very generous 998 DGCR rating. Where my tournament 74(-7) only got me a 969 PDGA rating.

However over a large sample size it all seems to work out.

Another fun stat on the site, good work.

 

[3j0hn]

The round rating formula definitely breaks down a little on Pyramids Pink where a "perfect round" of all birdies would give you an 875 rating.

 

[timg]

It breaks down on just about any very short course. I could limit it to courses with only X holes or more but I just did it globally.

 

[gcr_russell]

The round rating formula definitely breaks down a little on Pyramids Pink where a "perfect round" of all birdies would give you an 875 rating.

There is a reason the PDGA won't allow rounds of less than 13 holes (and even those are a huge anomaly) as rated rounds in a sanctioned event. Final 9 ratings are all for show and never actually count towards player ratings.

 

[ejvogie]

The -8 I tossed at the local pitch 'n' putt last week had the exact same rating as a -8 in the last tournament there.

 

[3j0hn]

It breaks down on just about any very short course. I could limit it to courses with only X holes or more but I just did it globally.

I totally understand the short comings of the existing formula/system with respect to overly short courses. I posted here after being amused at an extremely low round rating for what I though was an okay round there and not meaning to criticize the system.

Of course, my inner mathematician really wants to come up with a formula that does work for both 27 hole 10,000' courses as well as 6 hole 650' courses.

 

[jeverett]

I totally understand the short comings of the existing formula/system with respect to overly short courses. I posted here after being amused at an extremely low round rating for what I though was an okay round there and not meaning to criticize the system.

Of course, my inner mathematician really wants to come up with a formula that does work for both 27 hole 10,000' courses as well as 6 hole 650' courses.

I'm game to help with that, if you want. We'd need something that can adequately handle a lower limit, probably in a diminishing returns kind of way (i.e. a function that at the lower limit approaches 2 throws per hole, but doesn't ever reach it). Steve West's throw simulator may actually be the best place to look for data on this topic.. hmm..

 

[PuttyMcplaidPants]

sorry, I'm lost. where is this tool located?

 

[jeverett]

sorry, I'm lost. where is this tool located?

It looks like you only have three recorded rounds (those have round ratings listed for them in your scorebook), so you'll need a few more (you need at least 5 total rounds) to get an overall player rating. It will appear near the top of your scorebook statistics, though. Does that help?

 

[PuttyMcplaidPants]

It looks like you only have three recorded rounds (those have round ratings listed for them in your scorebook), so you'll need a few more (you need at least 5 total rounds) to get an overall player rating. It will appear near the top of your scorebook statistics, though. Does that help?

yes, thanks

 

[Dave242]

SSE (Scratch Scoring Estimate) is listed on each course's main page (if course length is set up)
DGCR Player Rating is listed on each players' page (see jeverett's)

DGCR Player Rating is calculated once 5 rounds are correctly entered into the scorebook (each round is automatically given a rating - if entered correctly)

I think I'll write a FAQ to submit to timg right now since things are a little slow today at work.

 

[jeverett]

I think I'll write a FAQ to submit to timg right now since things are a little slow today at work.

I sent timg a brief FAQ for how player ratings are calculated already, but it doesn't really merge in with the full process (of going from the scorebook/course page to round ratings to player ratings). Here's that FAQ page:

http://www.dgcoursereview.com/faq.php?mode=show&id=44

 

[PuttyMcplaidPants]

SSE (Scratch Scoring Estimate) is listed on each course's main page (if course length is set up)
DGCR Player Rating is listed on each players' page (see jeverett's)

DGCR Player Rating is calculated once 5 rounds are correctly entered into the scorebook (each round is automatically given a rating - if entered correctly)

I think I'll write a FAQ to submit to timg right now since things are a little slow today at work.

so certain courses won't have a rating if there is no SSE?

 

[jeverett]

so certain courses won't have a rating if there is no SSE?

Correct, a course (and tee) must have an SSE in order to generate round ratings from. What will cause a course to not have an SSE is when the hole info tab for the course page doesn't list the lengths for each hole.

 

[PuttyMcplaidPants]

so certain courses won't have a rating if there is no SSE?

nevermind, forgot to select the tee

 

[PuttyMcplaidPants]

Correct, a course (and tee) must have an SSE in order to generate round ratings from. What will cause a course to not have an SSE is when the hole info tab for the course page doesn't list the lengths for each hole.

who updates the course page with that info?

 

[Dave242]

I sent timg a brief FAQ for how player ratings are calculated already, but it doesn't really merge in with the full process (of going from the scorebook/course page to round ratings to player ratings). Here's that FAQ page:

http://www.dgcoursereview.com/faq.php?mode=show&id=44

Excellent! I had not seen that. I copied that directly into my document. I had the first 9 words written (word for word with yours) when I saw this.