18 Down @ Waco: Soft?

[lyleoross]

You're correct in that rating is based on props, and is about what it rates you relative to other props - but with a deep enough dataset you develop reliability. In essence the players themselves are "raters" of the course. Each time a tournament round is played, the course is given another "rating" which we call SSA and assign to the point at which a round is rated 1000. JC17393 (hi Jeff!) mentioned that Waco had an SSA of 62, so 62 is 1000 rated. When you accumulate enough rated rounds on a course - provided you know the weather conditions and the ratings of the players competing, so that you can look at inter-rater reliability (inter-round reliability) - you can assess the true difficulty of the course through the SSA.

SSA makes sense to me, as a measure of course difficulty, but even as you've laid it out, it seems to lack depth, in terms of detailed discussions of play during a round, and is a long term measure. Par gives more information, for me, round to round and hole to hole, even if it isn't deep per event or per hole, it is deep in terms of doing in-depth analysis and for providing discussion. But all I'm doing now is repeating what Hyzer suggested.

In summary, one is an overall measure of course difficulty, the other allows in-depth discussion of play, hole to hole.

 

[ChrisWoj]

SSA makes sense to me, as a measure of course difficulty, but even as you've laid it out, it seems to lack depth, in terms of detailed discussions of play during a round, and is a long term measure. Par gives more information, for me, round to round and hole to hole, even if it isn't deep per event or per hole, it is deep in terms of doing in-depth analysis and for providing discussion. But all I'm doing now is repeating what Hyzer suggested.

In summary, one is an overall measure of course difficulty, the other allows in-depth discussion of play, hole to hole.

You're right about that. SSA has absolutely nothing to do with hole-to-hole scoring. When I think of par in this conversation, I'm mostly looking at total course par. And I'm definitely not against use of par.

 

[davetherocketguy]

I think the only thing par is good for in our sport is for players to compare each other at the same course. That's it. We have one local course where a 51 is probably a 880ish round. Another course where par is only one stroke higher at 55 and if you score a 54 that will certainly be every bit of a 890 round or higher. The rating system isn't perfect by any stretch but IMHO if you want to compare players who have never played the same course its really not too bad.

 

[lyleoross]

I think the only thing par is good for in our sport is for players to compare each other at the same course. That's it. We have one local course where a 51 is probably a 880ish round. Another course where par is only one stroke higher at 55 and if you score a 54 that will certainly be every bit of a 890 round or higher. The rating system isn't perfect by any stretch but IMHO if you want to compare players who have never played the same course its really not too bad.

I agree, especially the way it is currently used. But as a talking point, there is a lot there for me.

 

[Hyzflip10]

SSA makes sense to me, as a measure of course difficulty, but even as you've laid it out, it seems to lack depth, in terms of detailed discussions of play during a round, and is a long term measure. Par gives more information, for me, round to round and hole to hole, even if it isn't deep per event or per hole, it is deep in terms of doing in-depth analysis and for providing discussion. But all I'm doing now is repeating what Hyzer suggested.

In summary, one is an overall measure of course difficulty, the other allows in-depth discussion of play, hole to hole.

I'm digging deep into the one stats course I took many moons ago. It seems score relative to par and rating should paralel each other for a particular round. The player who shoots 20 under will have a higher rating than the guy who shoots 10 under who will have a proportional difference in player rating to the player who shoots even par.

I wonder if there is a way to use standard deviation of player rating, or score relative to par, to determine course difficulty. A higher standard deviation indicates a wider score dispersion and theoretically a more difficult course. A narrower standard deviation would indicate a lower dispersion and theoretically an easier course since the average player of the group is shooting similar scores to the elite players of the group.

 

[Steve West]

There are a lot of different measures being talked about. Let me throw a few descriptions out there and open the floor for suggested jargon. Which one is "difficulty"? Or are they all various forms of "difficulty"?

Scratch Scoring Average is the average score of a typical 1000-rated player for a particular round.

Over time, the average SSA is the pre-round estimate of how many throws the average 1000-rated player would be expected to complete the course.

SSA/foot is a measure of how many extra throws the trees, OB, uphill, and such piled onto the score. Or, when SSA/foot is low, how open and downhill the course is.

Total length is the sum of the distances of each hole.

Total par is simply a number the TD or course designer assigns. These days, it could be based on any skill level, or any of a number of rules of thumb.

Correctly set total par should be the number of good throws it takes to complete the course for an expert (such as a 1000-rated player).

The difference between SSA and correctly set par (for a hole or for a course) measures how many bad throws are likely. For poorly set par, it tells more about how par was set.

Freakishly high round ratings reflect bigger-than-normal differences between average scores and the best scores.

The sum of the lowest score anyone got on each hole (maybe excluding aces) shows the lowest possible score.

 

[_MTL_]

I just posted this on another thread and thought it was worth posting here as well:

The challenge of disc golf is to make as few mistakes as possible. The difference between the pro winner of a disc golf event and a pro who finishes 20 behind him isn't 20 shots that the lower pro isn't capable of making, just more mistakes that led to the separation of their scores.

 

[Cgkdisc]

We're at risk of our sport at the highest level converging toward bowling with a "perfect" round being like rolling a 300 game. As MTL points out, the winner is the one making the fewest mistakes in a potential set of three perfect rounds.

 

[Motorpro]

A "perfect "round would be a score of 18. I am pretty sure no ne has come close to that.

 

[Cgkdisc]

A "perfect "round would be a score of 18. I am pretty sure no ne has come close to that.

If the media and players think 18 down is "perfect" then it's perfect which is why we're having this discussion. Perfect from my standpoint would be a score 18 below the legitimate 1000-rated par on a layout OR a score of 36 on a layout with a 1000-rated par less than 54.

 

[DiscFifty]

Going back in time, I sure wish par was set at what pros would shoot. My ability stagnated because I was content shooting par or just under. I was as good as my buds, had no reason to get better. lol.. yah..it was a bit of a shock to see better players shoot way under par for the first time. lol..

 

[Motorpro]

Calling 18 under "perfect" is just silly. Both times Mc Beth shot a "perfect" 18 under he had a par on one hole. If he had birdied that hole what would it be called? Under perfect? Perfecter?

 

[Cgkdisc]

Calling 18 under "perfect" is just silly. Both times Mc Beth shot a "perfect" 18 under he had a par on one hole. If he had birdied that hole what would it be called? Under perfect? Perfecter?

McBeth did not shoot 18 under the 1000-rated par on either of those courses. So, of course, they weren't perfect rounds by what I'm suggesting should be the definition.

 

[brutalbrutus]

A "perfect "round would be a score of 18. I am pretty sure no ne has come close to that.

Someone hand Kim Jong Un a disc...

 

[Steve West]

If the media and players think 18 down is "perfect" then it's perfect...

We have a different take on the concept of Reality.

Which explains a lot.

 

[Cgkdisc]

We have a different take on the concept of Reality.

Which explains a lot.

The cognoscenti know it's "fake news" but then that's why there are pages of enlightenment.

 

[Steve West]

...Perfect from my standpoint would be a score 18 below the legitimate 1000-rated par on a layout OR a score of 36 on a layout with a 1000-rated par less than 54.

It might be a while before we see that.

I calculated the probability of a 1050-rated player getting at least one under par (or a two) for 1000-rated players, on each and every hole, for a few courses with recent events.

Fountain Hills Park = 1 in 2.9 million
Vista del Camino Park = 1 in 914 million
The Beast (Nokia) = 1 in 2.8 billion
Brazos Park East = 1 in 52 billion
Toboggan = 1 in 1.8 trillion

You can tell us what round ratings those odds equate to.

 

[Cgkdisc]

Perfect shouldn't be easy and those probabilities indicate why. Although I think if we had McBeth, Eagle, Simon and a few others play Fountain with the goal of shooting a 36, I'm thinking the odds would be better than 1 in 2.9 million.

 

[wolfmandragon]

Someone hand Kim Jong Un a disc...

Someone better call Michael Malice

 

[Steve West]

Perfect shouldn't be easy and those probabilities indicate why. Although I think if we had McBeth, Eagle, Simon and a few others play Fountain with the goal of shooting a 36, I'm thinking the odds would be better than 1 in 2.9 million.

Probabilities are tricky. It just takes a few unlikely events to crash the probability of them all happening at the same time. Doing anything 18 times in a row is unlikely.

Let's take the group: Paul McBeth, Eagle McMahon, Ricky Wysocki, Calvin Heimburg, Seppo Paju, Simon Lizotte. None of them got a 2 on holes 2, 8, and 9 in either round. So, the best estimate is the chances are zero that any of them get a 2 on every hole. Actually, zero times zero times zero.

Plug any small non-zero number in there for those holes, and the probability that one player gets 2 on all three holes is still that number times itself three times.

Then, you still have to multiply it by the 1 in 381 chance that a player will get the other 15 holes all in one round.

So, getting a lot of 2s, sure. Getting 15, well... let them play 38 rounds each and you have a 50/50 chance of having seen one of them get 15 2s. Getting 18….

My calculations in the post above include a little bit of extrapolation based on the assumption of typical scoring distributions, which can include some possibility of scores that haven't been observed. So they may actually be a little optimistic.