Ask Chuck Kennedy

[Hogger]

So it is not 12 months from the update...

[cgkdisc]

No. It's rounds during the 365 days before the date of each player's most recently rated round, assuming there are at least 8 rounds in that period. For 2-day events, the date is set as the first day for all rounds in the event.

[cgkdisc]

Are you at liberty to discuss these unresolved issues?

I could but won't for the moment. Too many people involved from different angles. Highbridge is apparently open and hopefully will make some money to eventually pay off what they owe.

[Hogger]

No. It's rounds during the 365 days before the date of each player's most recently rated round, assuming there are at least 8 rounds in that period. For 2-day events, the date is set as the first day for all rounds in the event.

Good to know.
I have learned something today. Before the day starts to go downhill I should go back to bed.

[janttila]

Not likely. There are unresolved issues left over from Worlds that will keep that from happening, at least this year.

This is sad Bayfield county is beautiful and it is nice to bee able to jam out Madeline Island and then head to a great disc golf facility. It creates more incentive to make the trip when there are some tourney's. Does John know it's possible to go off the grid and hold unsanctioned events? Just a thought....

[cgkdisc]

Mt Ashwabay near Bayfield may be getting one or two courses this summer that will be right on the ski trails. Waiting to hear how they came out on budget after a good winter.

[ANHYZER]

Chuck, what is the NT web address?

[cgkdisc]

Not sure they have one this year at the moment because they've been working on the updated website.

[cgkdisc]

Posted by David Devine:
I was looking at the round ratings for courses like Paw Paw, Pawtapsco (Patapsco C pins) and DeLa which got me to wondering about the "dilution" factor for longer or more challenging courses where each shot becomes equal to roughly 7 rating points as opposed to 10, and whether non-linearities in the translation from score to rating are more important for these challenging courses.

Take I-5 at DeLa as an example. After your drive you may have only one tight route to the basket to save a three and many, many shots that will give you a four. Since a low round requires a large number of these high quality shots, it would seem that assuming a linear relation between score and round rating would fail at some point, and that point of non-linearity would come sooner for more technical, challenging courses.

In other words, the closer your score gets to a "magic round", the more points each shot should represent, especially for courses which require more technical shots.

The "toughness" of a shot is related to the skill level/rating of a player, not necessarily on an absolute scale. We might find making an accurate open field 150 ft shot easy but not necessarily for a beginner with a lower rating. Likewise, as you or I might be approaching that magic round of 1050 for us, the shots would be more routine for a 1030 player.

With regard to fewer points per shot on longer courses, it's just the lines on the graph getting farther apart the more shots that are made. Let's say two players are 50 pts apart meaning one will on average beat the other by 5 shots on an SSA 50 course. Now let's say they play two rounds on the same course. We would expect the one player to beat the other player by 10 shots for the two rounds combined, assuming both shoot their rating. If we now consider these two courses are like a "monster" course that's say 12,000 feet long, each throw is only worth 5 points instead of 10 points. Why? Because the players still have the same 50 pt spread in their ratings but their scores were 10 throws apart on this monster course. We take 50 points divided by 10 throws and it's 5 points.

Essentially, as courses get longer and tougher such that the SSA goes up, players who shoot only 5 throws apart on shorter courses will gradually shoot scores that are farther and farther apart. However, their ratings still have the same difference - that's fixed (in the short term). If their scores are farther apart, points per throw has to become smaller.

Now, to bring it to the real world, consider that the average length per throw decreases as the SSA increases. With par 4s and 5s, the second and third throws are usually shorter than the drive which brings down the overall average. How tough those shots are is taken care of by the SSA increasing if the par 4s and 5s are tougher on one course than the other.

[the_kid]

So your basically saying the system blows for longer courses? I mean Coda shoots 10 strokes better than the SSA on a par 70 course and barely gets 1050. Heck Eric shot a 65 the 1st round that was rated lower than his 2nd round but I wouldn't say the 2nd round was nearly as good it just happened to be on an easier course. Anyway you have said it yourself that the ratings tighten when playing a longer course but can you give me one reason this is a good thing? I played some pretty horrible shots with more than a handful of OBs yet it is rated 980+ and Eric shoots a very good rounds for a round just to have it be a little above his rating.

[MTL21676]

Chuck,

check this tournament out

http://www.pdga.com/tournament/tournamen..._ratings=1#Open

We played one layout rounds 1 and 3 and another layout 2 and 4.

Day 1 the weather was perfect. Day 2 the winds were in the 50 MPH mark and it clearly affected the scoring.

I know that tournaments are now comparing all rounds played on the same layout, which I agree with b/c it gives you more of a rating compared to the course, but this was a little ridiculous.

My 4th round was initially rated 997 and it dropped 20 something points?

Is there not a system that kicks in when conditions drastically change?

[cgkdisc]

Why does something blow because the rating range probabilities narrow as a course gets longer? It's a natural principle. Why aren't all of your rounds the same as your best round rating? Because you aren't likely to play at that high of a level over a longer time period. It's not that you can't do it, just that it's less likely. It's possible to shoot an 1100 round on a 70 SSA course. But it may take 100 years worth of rounds for that to happen. But it's possible mathematically.

Consider that thousands more rounds have been played on courses under SSA 54 compared to SSA 66+. We've seen what's possible with Skinner's 1117 on a 53+ SSA course. At the rate we're going with courses, it could take 100 or more years for the same number of rounds to be played on 66+ courses to see if someone can do it.

[cgkdisc]

My 4th round was initially rated 997 and it dropped 20 something points?

Is there not a system that kicks in when conditions drastically change?

TD can indicate that round scores shouldn't be blended together by setting up a separate layout for the windy round. In addition, we're looking at automating the process by Roger determining whether SSAs are less than 1.5 between rounds on the same layout and they will be combined. If any SSA is more than 1.5 higher or lower than the others, those rounds will be calculated separately. Recognize that it doesn't make too much difference because the average rating of those rounds will be the same either way they are calculated.

[the_kid]

It blows because a 1030 rated round on a huge course is better than a 1030 on a shorter course yet we have no way to tell between the two. If I took my best round ever (1060) and subbed that play into a par 72 course it would drop 20pts or more just because of the higher SSA.

[ddevine]

Howdy Chucker:

Since the ratings are primarily based on rounds played at courses with SSA under 54, would you expect non-linear effects to become more important on more challenging courses that have an SSA much greater than 54?

According to the "dilution factor" that you described, 24 holes at an SSA 54 course are equivalent to 18 holes at an SSA 72 course, however the latter tests course management skills that the former does not.

I don't have any idea how to quantify this feeling I have that the more challenging courses are not treated faithfully with the current system. Since there are so few courses with SSA > 60 it is hard to see if the numbers are skewed for these top rated courses. All I know is that a 1000 rated round at Paw Paw or DeLa or USDGC feels a whole lot better than a 1000 rated round at Morley.

[ddevine]

Howdy Chucker:

After thinking about it some more I have decided it is irrelevant. Numbers will never capture the quality of a round, just like par on the local municipal course is not equivalent to par at the US Open. But I must say that the ratings system is consistent and does a good job of capturing that part of the game that can be quantified. Triemstra, Snapper and I can usually guess to within 5 points based on how the round "felt".

Cheers, and enjoy the road trip!

[cgkdisc]

It blows because a 1030 rated round on a huge course is better than a 1030 on a shorter course yet we have no way to tell between the two.

Based on what? Yours is opinion, mine is objective with actual scores converted to ratings.

[cgkdisc]

Since the ratings are primarily based on rounds played at courses with SSA under 54, would you expect non-linear effects to become more important on more challenging courses that have an SSA much greater than 54?

According to the "dilution factor" that you described, 24 holes at an SSA 54 course are equivalent to 18 holes at an SSA 72 course, however the latter tests course management skills that the former does not.

The curves for ratings are slightly non-linear but consistently changing. Look at this graph. www.pdga.com/competition/ratings/ExpectedScoresAndCompression%20graph.pdf
I think every player should print it out, laminate it and carry it with them. This graph has been essentially unchanged since 1998. However, we now have more rounds on higher SSA courses to get a sense of the differences between par 3 and real golf courses.

It's not so much that higher rated SSA courses are treated less fairly, it's that lower SSA courses are treated more liberally. You would see the same effect in ball golf if the PGA guys played a several par 54-60 golf courses in smaller scale events on their tour. There would be lots of whining because the guys playing the lower par courses would likely get "better" points and stats than Tiger and the gang who would play more of the big events with par 72.

We see this effect now with our top guys thinking they get the shaft because they play the tougher courses on tour and the local pros get "pumped up" ratings from lower SSA courses in B & C-tiers. We checked it out and no one on average plays any better or worse on higher SSA courses than lower SSA courses. The range of ratings is somewhat wider on lower SSA courses but you can shoot your average just as easily on any SSA course.

[ddevine]

Howdy Chucker

I hope you did not misunderstand, I in no way feel like I am getting shafted by the ratings system. On the contrary, I think it works remarkably well on a wide variety of courses.

You answered my original post with the curves that you referenced. Being a scientist I can't help but wonder about "error bars" since there are so few courses with SSA > 60. How much of this is extrapolation? Also, it is especially amazing that the curves have not changed much since 1998 since there must not have been hardly any courses with SSA > 54 back then.

Cheers, and hope to cya sometime this year! DD

[cgkdisc]

Here's one of the earlier graphs we produced to determine the function of length and foliage to determine SSA.
For a given length course, the SSA is higher or lower than the red line based on the amount of foliage. Although you don't see too many data points at the higher end, subsequent graphing has indicated that the forecast for longer courses using shorter course data has panned out as expected. We tried alternative curved functions to see if they fit better but the straight line seems to fit the best and continues to do so. If you want a quick estimate of SSA for an 18-hole course, take the listed course length, divide by 285 and add 30.

[ddevine]

Thanks Chucker!

I would guess that the courses with lengths greater than 10,000 ft are the "Fly 18" types which are vastly different from courses with obstacles. I would bet these "driving ranges" retain the linear fitting that works so well for the SSA < 54 courses more faithfully than "real" DG courses. Impossible to tell with the current data.

Thanks again Chucker for your continuing work on the ratings! DD

[cgkdisc]

Much of the challenge is getting good data on course lengths. Much of the course length info we get on the TD reports is suspect, not really measured properly or not included at all. Not sure how to really get quality info unless you've done it yourself or with designers who are participating in the study.

[cgkdisc]

From MCOP: WHY in HECKS name would a pro worlds be a money drain?

Where is the money coming from to run it? Entry fees don't pay for anything. Most expenses are paid from sponsorship, merch sales, out of volunteers' pockets and the PDGA contributions of money and staff. Very few non-disc golf industry manufacturers have been significant sponsors over the years. Who knows whether Bell's would have come onboard if K-zoo was a Pro Worlds only or at least at that dollar amount.

[keithjohnson]

http://www.pdga.com/schedule/event.php?TournID=7363

There are at least 5 gators over 800 in ma3, do you know why it is not showing ratings for any of them? FW3 played the same layout also so they could get ratings. Please let me know what's up.

Thanks, Keith

[ANHYZER]

Why was the next rating update changed from 7/1 to 6/24? Will the Golden State NT be included in that update?

[the_kid]

Why was the next rating update changed from 7/1 to 6/24? Will the Golden State NT be included in that update?

Yeah I'm kind of bummed because that mean that the High Plains challenge horrible temp course ratings will stick in there a little longer.

[cgkdisc]

There are at least 5 gators over 800 in ma3, do you know why it is not showing ratings for any of them?

If you click on each of their Ratings Details, you'll see that there aren't five gators with ratings based on at least 8 rounds. If they played the same layout in any of the other rounds, then they'll get official ratings, presuming you think the weather was similar for those rounds.

[cgkdisc]

The date was changed since the end of June is when we do that update. If it was June 31st instead of July 1, we might have left it there. Actually, we also wanted to get out the new ratings a little faster for our northern new members who may just be coming out of hibernation and didn't have any rounds before the last update.

[keithjohnson]

If you click on each of their Ratings Details, you'll see that there aren't five gators with ratings based on at least 8 rounds. If they played the same layout in any of the other rounds, then they'll get official ratings, presuming you think the weather was similar for those rounds.

Crackel 846 - 18 rounds BEFORE event #3
Shimkus 851 - 8 rounds BEFORE event #3
Arnson 889 - 12 rounds BEFORE event #3
Quarles 866 - 8 rounds BEFORE event #3
Taylor 868 - 7 Rounds BEFORE event #3 (making 8 after 1 round and making him a counting gator)

Both rounds played same course with ZERO WIND and were virtually identical except for a 7-10 degree temp increase from the start of the event until the end...FW3 also played same course as MA3.

Where is ALL of the of the REC ratings?

[cgkdisc]

Click on each player's Rating Detail and it shows the "Official" rounds a player has on file at the time of the event. That's what counts to determine props for the unofficial calcs. Quarles rating for example is based on just four rounds: www.pdga.com/tournament/player_ratings_detail.php?PDGANum=34486&year=2008