It looks like a 49 at Renny Gold would be about 1117. Comparing strokes under course SSA as listed above:

Skinner's 1117 was 12.7 under the SSA
A 49 at Renny Gold would be a whopping 21.4 under the SSA!

A difference of 8.7 strokes under the SSA to get the same rating...

And 49 would have 3 eagles and 15 birdies.

And 49 would have 3 eagles and 15 birdies.



and thats on a course where par is at or lower than the actual SSA... not usually the case at a lot of courses.

I got 2 eagles at Renny my last 2 rounds. that place is EASY!!!!

I had a 3rd Eagle opp yesterday, but laid up from 35 feet, since it was a death putt. I'm gullible, but not stupid.

I got 2 eagles at Renny my last 2 rounds. that place is EASY!!!!

I had a 3rd Eagle opp yesterday, but laid up from 35 feet, since it was a death putt. I'm gullible, but not stupid.



I'm trying to think what would be crazy stupid good there without throw ins or aces....

323332322333422233 = 48 (-22). Eagles on 2, 11, 14, and 18

Peckham... i'm only seeing 1 eagle for you (hole 2)... where's the other one?

Howdy Chucker

Thanks for the data. It seems pretty clear that the linear fit breaks down as the SSA differential increases. Barry takes 19 more strokes on a course with an SSA that is 26.2 strokes greater than Feldberg's, yet Barry's rating for that round is 14 points lower.

This suggests that the further you are from the SSA the more non-linear the fit should be between shots and round ratings. In other words, one stroke should equate to a different number of rating points as the distance from SSA increases.

The problem is the uncertainty of what fit to apply and the fact that anomalies are hard to spot with the low number of SSA> 54 courses in our sport.

How do the rounds tend to cluster around the SSA? Are there similarities in the distributions of scores about the SSA for all courses, or are those similarities built in with the linear fit assumptions currently invoked? Are there differences between the distributions for the Am vs Pro when the same course is played? I would guess that the standard deviation might be broader for higher SSA courses, but it is hard to tell, especially for a course like Winthrop Gold that features only the best of the best.

Great fun, we should sit down and chat about this stuff sometime!

Cheers, DD

careful dave, you're making sense...

Schweb's round data is unofficial but I iincluded it since his round is what triggered my effort to develop a better way to look at our best rounds.



Official/Unofficial, doesn't maTTer. That was a sick round of golf and was more impressive than Skinny's round on the Darkside.

Thanks Chuck for finding a way to display "HOT" rounds on tough courses

D Silver

what was the previous best on the layout?

skinner shot 5 strokes better than the best ever on a course with hundreds of rounds on it by thousand plus players. as mtl said somewhere- schweb's round is great but will conceivably be shot again- 41 on the darkside is untouchable.

I eagled 1 last week, and had a 35 footer on 14, but decided to lay up.

Official/Unofficial, doesn't maTTer. That was a sick round of golf and was more impressive than Skinny's round on the Darkside.




I would disagree 100%

Skinners round of 41 is unheard of. I'm happy with anything under par out there.

... 41 on the darkside is untouchable.



Only a couple of 46's, along with some 48's and 49's have been shot on the Darkside during the last 2 Va Opens. Untouchable is right on the money.

Official/Unofficial, doesn't maTTer. That was a sick round of golf and was more impressive than Skinny's round on the Darkside.




I would disagree 100%

Skinners round of 41 is unheard of. I'm happy with anything under par out there.



I'm with MTL on this one -- 41 is hard to imagine on the Dark Side... yes, almost every hole on the Dark side is deuce-able (unlike Renny), but (like Renny) almost every hole has numerous places for a tiny bit of schwag to turn a great drive into a save-for-three. The number of perfect throws required to hit a 41 is hard to even fathom...

I was thinking about ways to categorize rounds, and I think there is an inherent flaw in Chuck's proposed simple calculation of SSA minus players score... the flaw is that it fails to give enough weight to the difficulty of throwing of a really hot score on a shorter course... because on a shorter course, every imperfect shot is more costly because there are less opportunities to make perfect shots.

I was thinking about trying a different -- very simple -- method to calculate "hot rounds" -- where you calculate the percentage of throws taken compared to SSA. I.E., if a course has an SSA of 50 and you shoot a 40, your percentage of throws is 80% (40/50) -- this means you could compare percentage scores to one another, and the linear scale seems to work across different SSA courses.

It makes sense that if you shoot only 4 throws for every 5 throws of a player shooting a 1000-rated round, it is a no-brainer that you are throwing some really hot golf. Likewise, if I shoot 9 throws for every 10 throws of a 1000-rated player, I'm playing some very good golf, but not awesome.

In my thinking, this system seems makes a lot of sense, scales well across all SSA courses, is easy to calculate and still helps highlight the best imaginable rounds.

I resorted Chuck's initial post, re-ranked by percentage points. I don't know how to make the chart all pretty and lined up like his, so I cut out the extra columns to make it easier to read. You can see that this produces a very different list than either Chuck's calculation of stroke difference or the PDGA round rating calculation.

You also see that based on a percentage system, Skinner's 41 is just unbelievably impressive. He was throwing approximately 3 throws for every 4 throws of a player shooting a 1000-rated round. Which is just unbelievable; and in that sense, notably more difficult to achieve than any other round on record. You will see that all the other hot scores -- even the ones on courses with similar or lower SSAs than the Dark Side -- are several percentage points higher, and all fall in a fairly tight distribution to one another.

Anyway, look at the numbers, or import Chuck's data into Excel and rearrange it yourself -- and see what you think of this system for ranking Hot Rounds.

Chuck's Rank Last, First Rating *** Percent
6 Skinner Brian 1117 *** 0.763500931
10 Feldberg David 1098 *** 0.795847751
1 Schultz Barry 1076 *** 0.798319328
8 Climo Ken 1091 *** 0.798722045
2 Climo Ken 1078 *** 0.801144492
13 Hatfield Coda 1096 *** 0.808678501
3 Schweberger Brian 1075 *** 0.809659091
4 Climo Ken 1078 *** 0.81120944
16 Arthur Phil 1094 *** 0.811881188
11 Schultz Barry 1092 *** 0.812274368
4 Climo Ken 1075 *** 0.813953488
16 Schultz Barry 1092 *** 0.815533981
23 Schultz Barry 1099 *** 0.815899582
6 Anthon Josh 1072 *** 0.817790531
25 Bubis Aleksey 1099 *** 0.820734341
9 K�llstr�m Markus 1072 *** 0.824742268
24 Stokely Scott 1092 *** 0.826446281
16 Grider Nolan 1081 *** 0.831858407
20 Schweberger Brian 1078 *** 0.833333333
21 Jernigan Justin 1081 *** 0.834845735
26 Duron Dan 1090 *** 0.836909871
27 Leiviska Cale 1089 *** 0.838709677
19 Gurthie Garrett 1079 *** 0.839041096
28 Feldberg David 1090 *** 0.840707965
12 Lundmark Jesper 1071 *** 0.843949045
14 Dorius Micah 1071 *** 0.844155844
14 Schweberger Brian 1070 *** 0.846645367
21 Lundmark Jesper 1071 *** 0.848585691

comparable to Renny Gold.... looking back @ past results from there... i've seen a 62, two 63s and a few 64s

there has never even been a practice round better than 46 to my knowledge- i was playing with john fowler a few years back and he was on track for 45 but bogeyed 18.

gotcha... thats were theres a difference... i believe there has been a 57 in practice and I def. know of a few 59s (granted this is before the changes to 18 but still a significant comparison between the 2 amazing rounds)

yep...no denying they're both fantastic rounds.

there have been 2 57s shot in practice rounds at Renny Gold, by Stan and Myself. and there is not much room for improvement, granted i think the course plays a stroke easier than it use to due to some cutting on the power line right of ways by Duke Power, nonetheless it is on of the most impressive rounds of tourney golf i have personally witnessed and should be ranked right up there with most anything else.
mj

One flaw with my alternate method is that "shots under SSA" become tougher once the SSA drops below 54 like Skinner's round with a 53+ SSA. If you imagine a player being able to birdie every hole relative to SSA, at 54 SSA, they would shoot a 36 in a "perfect" round. With a 47 SSA, they can only shoot 36 again which means they could only shoot 11 under SSA at "perfect" best.

So, I'm not sure the "shots under SSA" is necessarily "fair" for SSAs under 54 which are the courses in zone D and E on my chart.

Here's another aspect of this "hot" round ranking problem and that has to do with the individual scratch hole averages on two courses with the same total SSA. In the simplest example, consider that two courses have 16 holes with identical scoring averages. However, one course has scratch scoring averages of 3.5 and 3.5 on hole 17 & 18. The other one has scratch scoring averages of 3.0 and 4.0. Obviously it can depend on the hole designs but for purposes of illustration let's say a player can only get a birdie 3 on both holes of one course but can get a birdie 2 and birdie 3 on the other one. Both courses have the same SSA but the player has a better shot at scoring one shot better on the second course.

This is related to the frustration with getting better "hot" ratings at Castle Hayne which has been discussed before. There's not easy way around this since it's a course design issue not a ratings calc issue. It would be hard to require holes to average only between say x.6 to y.1 and exclude x.1 to x.6 holes.

Chuck,
How do you feel about my proposed percentage method for calculating hot rounds -- percentage of total throws compared to SSA? This is a method that is easy to understand, and I think should scale, no matter what the SSA of the course, and seems to give a distribution that "feels right." -- at least from my subjective perspective.

From a more objective standpoint, it charts very well, with each percentage point getting progressively less-likely as you move away from the SSA... I defer to your ratings guru-ness, as I'm not a true statistician -- just a writer who occasionally needs to stretch the left side of my brain.

I think a perfect system would take into account the fact that it is harder to improve upon good scores than bad scores. For example: for Skinner to have improved upon his 41 by one stroke is much harder than for some other golfer to improve upon a 54 by one stoke. as such, I think that as scores get lower, the amount of ratings points awarded per stroke should increase. By this model, somebody shooting a 110 on a par 54 course would get a nearly identical rating to somebody shooting a 111, or expanding ad extremum somebody shooting an infinity and somebody shooting infinity +1 get exactly the same rating (as they should, and it should be 0). Kind of like a SSA*1000/x model (but with some tweaks, obviously)

Realized I just proposed same thing as haines, just flipped the fraction, essentially.

/still think it's valid.

I'll take a closer look when I get back home next week. There's no perfect solution and that's why it's interesting to look at different ways to determine the best when it becomes more apples to oranges comparisons. In a way it's like comparing great athlete performances from an older era versus a newer era when equipment and rules have changed.

Chuck, will you be updating this again?

Just curious where Shoestring's 1083 would fall.
http://www.pdga.com/tournament/tournament_results.php?TournID=7449#Open

Thanks,

Mike.

Looks like that one might be third best ever in that 54-60 SSA range behind Feldberg's recent round and Schultz both in the 1090s, but beating two 1081s currently listed as tied for third.

I think a perfect system would take into account the fact that it is harder to improve upon good scores than bad scores. For example: for Skinner to have improved upon his 41 by one stroke is much harder than for some other golfer to improve upon a 54 by one stoke.

I dunno, I played with Skinny that round and he missed two putts on back to back holes from ~35 feet and ~25 feet...the dude almost shot a 39, which would have given him what, an 1140 rated round??? That round, despite the two misses, is still untouchable (at least on that course at any rate)
Top it off, he almost aced hole 1 (with holes 1 and 2 to finish)
Almost canned hole two from about 45. Good times.

Looks like Matt Orum just put up a 1097 at Worlds 2008 in Michigan @ Victory-Albion, I am not sure of the SSA for that course but it looks like maybe 49-51 range.

Looks like they just changed the par for the course online and now the rating dropped to a 1095!

changing par would not have an effect- they probably changed the courses by pool somehow.

How about Ken Climo's 1117 @ the Fun 'n Sun...He shot -23 on a 24 hole layout! He missed #1 and then carded 23 birdies in a row!

How about Ken Climo's 1117 @ the Fun 'n Sun...He shot -23 on a 24 hole layout! He missed #1 and then carded 23 birdies in a row!

prelim ratings are gone now that the scores have gone official but if it holds up it will become numero uno, joining brian skinner's 41 on the grange darkside as the only 2 1100+ rounds ever. Chuck may be able to offer some insight on where it may land when official.

Just waiting for the official updated Best Rounds file from Roger and I'll post. Climo will have an 1117 record separate from Skinner because the equivalent SSA for that Fun 'n Sun course falls in the under 48 SSA course category. Second place all time in that group is a 1099 by both Schultz and Bubis so you can see how good Climo's round was and he had to keep it going for 24 holes versus 18. Skinner's record is in the 48-53.99 SSA course group which also has the rounds from Schwebbie (1106) and Matty-O (1102) over 1100 in second and third.

Kris Orrick: 17-May-2009: First Tennessee Bank Lakefront Open: Rd 3: Round Rating 1084

Wow! :)

http://www.pdga.com/tournament-results?TournID=8699&year=2009&include_ratings=1#Open

That's a great round on one of my favorite courses when I travel! This course SSA category has a tough crowd though with at least a 1090 round rating required to break into the top 10 rounds all-time. A 54 would have done it.

Here's the latest update to the top 10 Best Rounds in five SSA ranges for both Men and Women. The SSA ranges are: under 48, 48-53.9, 54-59.9, 60-65.9 & 66+. The yellow highlighted rows are new entries since the last update in mid-2008. Notice that Climo is the first to top 1100 in the under 48 SSA category. In the 66+ category, it's not that Winthrop is such an "easy" course that many of the records are shot there, but it's more the case it's one of the few courses with an SSA over 66 regularly played in high level competition by many elite players - and now Red Hawk is gone.

http://www.pdga.com/files/documents/BestRounds%20May09.pdf

Here's the latest update to the top 10 Best Rounds in five SSA ranges for both Men and Women. The SSA ranges are: under 48, 48-53.9, 54-59.9, 60-65.9 & 66+. The yellow highlighted rows are new entries since the last update in mid-2008. Notice that Climo is the first to top 1100 in the under 48 SSA category. In the 66+ category, it's not that Winthrop is such an "easy" course that many of the records are shot there, but it's more the case it's one of the few courses with an SSA over 66 regularly played in high level competition by many elite players - and now Red Hawk is gone.

http://www.pdga.com/files/documents/BestRounds%20May09.pdf

Do you have any idea what 'gator has the highest differential between rated round and player rating? The best personal best, so to speak?

We haven't done that directly but we do a list like that before posting each ratings update showing the highest differential between a round rating and a player's rating as a means to check for course layout coding problems. My guess is that if we did that type of sort on historical data, we would find that the top ones on that differential were probably older "mistakes" where say a 750 player got a 925 rated round because the course layout coding was mismatched.

One great round to point out from last weekend's VA Open. 971-rated Derek Britt shot a 44 at The Grange on the Darkside, for an unofficial round-rating of 1086... Still three strokes off of Skinner's record there, but still the 2nd best round of all time on the course, and a heluva score for a player who had never had a personal rating of 970+ until the May update.... his former best personal rating was 964 back in Jan 2008, yet he's been shooting 1000+ golf the past few months.

Huge props to Derek, and best of luck to him this season.

We haven't done that directly but we do a list like that before posting each ratings update showing the highest differential between a round rating and a player's rating as a means to check for course layout coding problems. My guess is that if we did that type of sort on historical data, we would find that the top ones on that differential were probably older "mistakes" where say a 750 player got a 925 rated round because the course layout coding was mismatched.My brother Jon got a bogus 970 player rating that way back in 1999 when ratings first came out. Everyone in Michigan started out with bogus ratings, high or low depending on whether they played longs or shorts at the No Foolin that year.

My best differential was 99 points, non-bogus.

So the ratings are not congruent across the board? It depends on SSA now? Or has it always been that way?

Another question, how many times has the ratings formula been changed?

Thanks!

The potential to shoot your rating is the same at all SSA levels. However, the farther the SSA deviates from around 51, the narrower the range players are able to shoot above or below their rating. So, it only makes sense to track highest all time records by SSA range. You can see that by looking at the differences in ratings in the top 10 rounds in the SSA ranges shown.

The formula to calculate a round rating has only been tweaked twice. However, the way a player's rating is calculated from their round ratings has changed a few more times in terms of the criteria for throwing rounds out, double weighting and accounting for propagator rating average.

Bascially you can't compare ratings across different SSA's, thus the need to break them up into different classes. That just seems faulty.

But you can compare ratings across all SSAs. For example, the batting averages and other stats for baseball players are calculated right away. A player might have a .500 batting average in April, .375 by the All-Star break and end up with a .317 for the year. Those averages are all correct and calculated the same way. However, you can't really compare a .500 average for one month with a .317 season average or a .305 lifetime average. The more at-bats are involved, the harder it is to achieve a .400 average. The more throws that are made in a disc golf round, the harder it is to shoot an 1100 round. In both baseball and disc golf, a .400 and 1100 can be achieved mathematically, but it's very hard or becomes impossible to do at a certain point.

I can see why you're concerned about the ratings ranges with Rodney messing with you on the Iowa Board. We specifically looked at whether players could shoot their rating no matter what the SSA of the course is. And the answer is yes. Players rated 850 average 850 on courses ranging from SSA 41 thru 70. Players averaging 900 ratings average 900 on courses from SSA 41 thru 70. And we did the calculations for the same five SSA ranges we use for the highest round records.

Likewise for 950 players, 1000 players and 1035 players. They all average their ratings. We checked this using the whole PDGA database, not just some selected players. If a 1025 round or a 925 round rating were not equivalent on all SSA courses, players would not be able to shoot their ratings. And yet they do. Players who primarily play courses with SSAs over 60 have no more or less advantage over those who play courses around 48 SSA in terms of what rating they can achieve.

The only difference is that an established 950 rated player normally playing higher SSA courses will have a tighter range of round ratings (standard deviation) than a 950 rated player who plays mostly lower SSA courses. The best single round rating for the player who plays the lower SSA courses will likely be higher than the best round rating of the player who plays the higher SSA courses. Likewise, that player will also be more likely to have a lower rated worst round including one so bad it gets excluded compared with the player who plays the higher SSA courses. But they'll still average the same rating overall.

The latest list of top ten rounds in five SSA ranges for Men & Women has been posted here: http://www.pdga.com/player-ratings-search/top-10-best-rounds-ever-5-ssa-ranges-men-women

We're now up to seven rounds of 1100 rating or higher with five of them on courses in the 48-53.99 SSA range. Perhaps not coincidentally, more tournament rounds are played on courses in this SSA range.

I can see why you're concerned about the ratings ranges with Rodney messing with you on the Iowa Board. We specifically looked at whether players could shoot their rating no matter what the SSA of the course is. And the answer is yes. Players rated 850 average 850 on courses ranging from SSA 41 thru 70. Players averaging 900 ratings average 900 on courses from SSA 41 thru 70. And we did the calculations for the same five SSA ranges we use for the highest round records.

Likewise for 950 players, 1000 players and 1035 players. They all average their ratings. We checked this using the whole PDGA database, not just some selected players. If a 1025 round or a 925 round rating were not equivalent on all SSA courses, players would not be able to shoot their ratings. And yet they do. Players who primarily play courses with SSAs over 60 have no more or less advantage over those who play courses around 48 SSA in terms of what rating they can achieve.

But isn't this just circular reasoning? I thought the course SSAs for each round were based on the ratings of the players, and how they played that day. OF COURSE players are going to shoot their rating no matter what the SSA if the SSA is derived from the ratings of the players who play the course.

From what you are saying, this in no way confirms the validity of the SSA system.

I thought the course SSAs for each round were based on the ratings of the players, and how they played that day. OF COURSE players are going to shoot their rating no matter what the SSA if the SSA is derived from the ratings of the players who play the course.
The average rating of the players who propagate the SSA will average exactly that same average for their round ratings. That's a cornerstone of the ratings system. That's the fundamental circularity, if you will, of the system. However, that wouldn't necessarily mean individuals were treated "fairly" within that structure. We didn't know until we tested it.

Validation overall is based on consistent reproduction of the same SSA for the same layout with a completely different set of players/propagators. That has been verified within normal statistical variance. We reduce that variance even more by combining scores on the same layout in different rounds.

Validation for individuals is that players at all rating levels and courses of all SSAs from 41 to 70 are able to average their ratings. Until we did the validation studies, it's possible maybe lower rated players averaged lower ratings on high SSA courses or maybe earned higher ratings on lower SSA courses; or vice versa. But that didn't happen. The system appears to work as intended for all player ratings and all SSAs.

Is there a list based on tiers?

Like top 10 C-tiers, Bs, and so on?

No need to do it. SSA is what matters. Roger just took a look to see if there were differences in A, B & C-tier ratings. The average was the same for each tier by all players who played at least 8 rounds in each tier.

No need to do it. SSA is what matters. Roger just took a look to see if there were differences in A, B & C-tier ratings. The average was the same for each tier by all players who played at least 8 rounds in each tier.



It would still be nice to see........I am jut wondering because I had a round rated 1083 at a C-tier this weekend and no matter what you say it is harder to get that compared to if it had been an NT.

This is the data for players who had at least 8 rounds in each tier category over a 2-year period. I'm checking with Roger as to how many players this amounted to:
Tier / Rnds / Avg Round Rating
A / 15.1 / 936.1
B / 25.0 / 935.3
C / 20.5 / 935.4

This is the data for players who had at least 8 rounds in each tier category over a 2-year period. I'm checking with Roger as to how many players this amounted to:
Tier / Rnds / Avg Round Rating
A / 15.1 / 936.1
B / 25.0 / 935.3
C / 20.5 / 935.4

Ok?

The point being that players who play many rounds in all tier levels do not play any better or worse regardless of tier level meaning that tier doesn't matter in the ratings. The only thing that is different between tier levels is the number of propagators which would typically be fewer at C-tiers than A or B-tiers. That means the ratings at lower tiers might be more variable but no more or less than the other tiers on average.

The point being that players who play many rounds in all tier levels do not play any better or worse regardless of tier level meaning that tier doesn't matter in the ratings. The only thing that is different between tier levels is the number of propagators which would typically be fewer at C-tiers than A or B-tiers. That means the ratings at lower tiers might be more variable but no more or less than the other tiers on average.

So then do you have a list of more than the top 10 rounds of each SSA range?


Like the top 100 ever so I can make my own list based on tier.

Not available for publication and I don't even have the master data file. Roger generates the most recent top 10 in each SSA range from the database after each update.

Not available for publication and I don't even have the master data file. Roger generates the most recent top 10 in each SSA range from the database after each update.

You had the database back in 2003 when you posted a list of all the top rounds......

That was an export from the data that I asked Roger for. I don't have any of the live PDGA ratings databases to pull reports. That's all Roger and Gentry.

That was an export from the data that I asked Roger for. I don't have any of the live PDGA ratings databases to pull reports. That's all Roger and Gentry.

I would still like to see it whenever it is possible.

Looking at the top 10 for the separate SSA ranges it seems like maybe 2 were at C-tiers and I'm just guestimating (don't remember if the Midweek Mace events were B or C's) and the Thanksgiving classic in OR along with Cale's 1089 at the WI tour warm up.

No matter how many times you ask him he still isn't the one who has what you are asking for.

Even if you ask him 5 more times he still isn't going to have what you are asking for.

I wasn't asking him in the last few.....more like letting him know I would like to see it if there ever is a list available....

Anyway I think it would be a decent stat and almost as important than the SSA ranges.