Round Ratings

[sisyphus]

I think it's fantastic that you added this feature! It will certainly be much talked about, and I plan to take the ratings with a grain of salt. Hope everyone else can learn to do so as well. So many variables go into the data we get on here. Anyway, thanks again. This is my most used internet site! :thmbup:

 

[tampora]

The process is simple, just enter your score as you have always done and if you played all the holes at a course and have chosen a tee the rating will be immediately calculated as soon as you submit your card.

More information here would be helpful. Entering scores always felt too meticulous for me, but my ego loved the idea of a rating. So, I had to try it out. Unfortunately, I had no idea what to do from only these instructions without having to poke around for a while.

First, I made sure I was logged in. Then, in the "My Dashboard" drop-down context menu, I found scorebook.

(Even though logged in, the "My Dashboard" menu is sometimes oddly not visible and replaced with "Register". I was ultimately able to find it by clicking the "Home" link)

Now, at this part I'm still lost.
1. I made sure Lincoln Park has a SEE on it's details page (it was listed for the red tees)
2. I entered my favorite recent score (45 shots) on my scorebook page for this course
3. I made sure the red tees were the selected on my submission

But, I still get a 'n/a' returned in the Rating field upon completion.

 

[timg]

It looks like you put in a quick round, the SSE is only calculated for "complete" scorecards... so you have to do the hole-by-hole.

Regarding the "My Dashboard" link, if you don't click the "Remember me" box, the forums will not know you're logged in so you'll see the non-logged in menu.

 

[cc0049]

This is a cool feature! I realize it is just an estimate and for fun and all of that, but I'm curious if there is a way for the round rating to be adjusted based on the wind speed you enter when you put your round in. Obviously a 54, for example, in strong wind should be rated higher than a 54 in calm wind. I'm not sure how easy or difficult this would be to add to the formula, but it would be an improvement in my opinion.

I must say, I find it really interesting how accurate these SSAs can be to actual PDGA ratings from competitive rounds simply based on distance and foliage. Pretty cool!

 

[timg]

Adjusting for weather conditions isn't something I want to start doing. Then you start having to deal with things like what if it was raining, snowing, etc. It starts getting too complicated.

 

[tampora]

Thank you.

 

[tampora]

From the FAQ:

The SSE is calculated using course data from DGCR only. It calculates the SSE by assuming that a scratch player will throw on average 355' on lightly wooded holes, 285' on moderately wooded holes, and 255' on heavily wooded holes. It also assumes that it takes an average 1-2/3 throws to finish up the hole (putts and short range approaches).

Where is the wooded'ness of a hole found and how is it updated? When updating a course, I see where the elevation change and water is selected per hole, but nothing regarding foliage.

 

[timg]

That's not per hole, the terrain and landscape are done on a per course basis on the main course info page.

 

[Dave242]

I like the general workings of this feature, and the ratings that I'm getting on the home course seem acceptable, but we have a course nearby that is producing artificially low ratings.

While its flat, open, and moderate in length, it has a significant amount of OB on practically every hole. On windy days, sometimes even the guy who designed it (who's about a 950 rated player) has trouble breaking 54.

So I highly doubt that a round in the upper 60's there qualifies as 750ish rated round.

As has been explained before, the SSE can't account for artificial OB, etc. Hence, the "estimate" part of it. This isn't going to work 100% for all courses.

A helpful way to think of this is that heavily wooded courses impose lots obstacles to the flight of the disc and make the average throw shorter (for everyone and especially for those who are not accurate throwers).

OB is an obstacle that stops your throw dead like a tree does......and then it also adds a 0' throw to your round (the penalty). This brings down your average throw length by a ton!

So, for the SSE calculations (and the ratings which are based on SSE), a wide open treeless course with some OB on every hole would be best to be set up as "heavily wooded". Makes sense for the SSE calculations, but no sense at all for the real reason that the foliage descriptor is there.

 

[timg]

I just adjusted all the ratings for the older rounds. Hopefully they're a little more accurate now.

 

[rd.bittle]

Thanks Tim. Cool estimation tool.

 

[timg]

Player ratings are done! You'll find them in your DGCR Profiles under the throwing style. I compared a bunch to people that had a lot of scores entered along with their PDGA #s and ratings and they were pretty darn close. You'll need at least 5 rounds recorded before a rating will be calculated for you.

I'll put the rating somewhere on the scorebook page too along with an Avg. Round Rating per course.

 

[jeverett]

Round ratings for historical rounds are looking good now! Yes, given the limited inputs for the SSE formulas (course foliage and layout length), and the reality that SSA's for events *do* fluctuate a bit based on field size, weather, etc., these round ratings (and the coming player ratings) are going to be ballpark estimates, nothing more. As I mentioned earlier, for most 18-hole courses where the 'course foliage' value matches up well with how the course actually plays (technically), round ratings are almost always going to be within ~ +/- 15 of a 'real' event round. That's not phenomenal accuracy (I was really hoping to get it down to +/- 10), but it's the best we could do without more inputs into the function (which would require timg to do a lot more database work).

 

[jeverett]

Player ratings are done! You'll find them in your DGCR Profiles under the throwing style. I compared a bunch to people that had a lot of scores entered along with their PDGA #s and ratings and they were pretty darn close. You'll need at least 5 rounds recorded before a rating will be calculated for you.

I'll put the rating somewhere on the scorebook page too along with an Avg. Round Rating per course.

Woot player ratings! Thanks Tim!

 

[timg]

The issue with creating more inputs is getting the input. If we added another field that means thousands of courses would have to be updated with the new data which would take quite some time. I think the main improvements being discussed in the past were to add more options to the Landscape. I don't use the elevation adjustment in the formula, it seemed like the SSEs are a little better without it.

 

[Dave242]

I think it is fine the way it is for several reasons:

1) Even with only 3 levels of foliage, there is no common understanding in quite a few spot-checks I have done on courses I know well. Why would having more levels be better in this regard?

2) Even if we got things set up perfectly (by adding an extra field or 2), there would still be plenty of scenarios a mathematical model could not predict. And those needed fields would not be understandable or meaningful to the masses.

3) There is already plenty of controversy with the PDGA's SSAs and Player Ratings in regard to their accuracy/consistency. Why would it or should it be expected to be any different with any system?

4) Controversy is good! I leads to discussion about course design, hole design, etc. That will lead to more people who are interested in how course design is integral to the competitive sport of DG (rather than just the recreational game).

5) Inaccurate/inconsistent DGCR SSE's/Ratings here will get many people hooked on the idea of ratings and drive them towards PDGA events so they can get "real" ratings (rather than estimated).

6) SSE is already a lot more consistent than the implementation of Par.

 

[Dave242]

Will DGCR Rating be added to the "Browse Member Results" page? P.Rating is there and is sortable. It would be really cool to have this new rating on there too.

 

[simpletwist]

Actually 2 thumbs up. My rating came out almost exactly what my PDGA rating was.

Thanks Tim and everyone involved.

 

[jeverett]

I think it is fine the way it is for several reasons:

1) Even with only 3 levels of foliage, there is no common understanding in quite a few spot-checks I have done on courses I know well. Why would having more levels be better in this regard?

There's an old thread around where we were discussing ideas for how to modify the SSE formulas (shortly after they were originally implemented), but the approach that seemed most promising (vs. the amount of extra data it would require) was to try to separate how 'technical' or punishing (on average distance) a layout was from its 'foliage value'. i.e. right now, our only metric for how technical a course is is the foliage value. What if we had another field for each course where we could input how technical it was? Yes, we'd need to come up with descriptors for each of the lightly wooded, moderately wooded, and heavily wooded values, reworded instead to represent layout technicality/risk, but it would be the simplest way to deal with the big discrepancy in round ratings that comes from when the foliage value doesn't match up with the layout technicality at all.

At the next level of complexity, the light/moderate/heavy (or technicality descriptors, as mentioned above) could be input on a hole-by-hole basis (this of course would require not 1 new field per course, but 1 new field per hole per course), placed on a Likert-like scale (where lightly wooded/technical = 1, moderate = 2, and heavy = 3), and averaged to get an overall layout technicality value (an average between 1 and 3). I've already previously worked on building an overall unified formula that would accept any value from 1-3, and output the same kinds of results as the three separate length-based formulas that Dave242 previously created. It would, however, require recalibration against 'real' PDGA SSA data, as very few course layouts of course would still be 100% 'light' or 100% 'heavy' (i.e. all course layouts would be pushed a little bit closer to the middle range, as almost no course has 100% 'light' holes or 100% 'heavy' holes.

If this gets a lot of interest, we may need to start another new thread about it.

 

[jeverett]

Actually 2 thumbs up. My rating came out almost exactly what my PDGA rating was.

Thanks Tim and everyone involved.

From perusing a fair number of profiles, it does indeed appear that DGCR ratings are coming in fairly close to listed PDGA ratings. In general, I'm noticing a trend where DGCR ratings are slightly higher than PDGA ones (generally in the 8-15 point range), and to me this makes sense. For starters, most players play better in casual rounds than sanctioned event rounds (particularly due to pressure/anxiety), and second, I would suspect that players might not record "really really bad" rounds on their scorebook with greater frequency than they are stuck with bad rounds in their PDGA rating. After all, we all use the DGCR scorebook the way we want to use it.. which is good, imo.