Delaware Disc Golf Challenge

[T-Rey]

Throws below SSA doesn't work except for courses within a small SSA range like the examples posted above. Rating points per throw increases as SSA decreases so you have to decide where to make SSA breaks to bundle courses together for stats comparisons. 15 under will always have a higher rating on a lower SSA course than higher SSA course. It's a natural law we discovered with the ratings process, not a factor we fabricated.

With strokes below SSA, points per stroke doesn't matter. At Iron Hill Paul shot 58 and the SSA was ~68.3, so -10.3

At Great Lakes open he shot 45 and SSA was 58, so -13

 

[Cgkdisc]

With strokes below SSA, points per stroke doesn't matter. At Iron Hill Paul shot 58 and the SSA was ~68.3, so -10.3

At Great Lakes open he shot 45 and SSA was 58, so -13

It's a start but it's not that simple when comparing courses with significantly different SSAs. The SSA establishes the course level challenge. 15 below SSA 54 will always rate better than 15 below SSA 72 but they are apples to oranges different course challenges.

Think of it in terms of percentages. Each shot saved on SSA 54 equals 1/54% (.0185). Each shot saved on SSA 72 equals 1/72% (.0139). So saving 15 shots on an SSA 54 will be a higher percentage (27.8%), i.e., rating, than saving 15 shots on an SSA 72 (20.8%). You might say that's an unfair comparison because you can only save one shot at a time, and you would be right. That's why we need to only compare relative performances among courses within a fairly narrow SSA range.

Another real world example is 3-pt shooting. 100% would be perfect and it's only been achieved in the NBA after 9 attempts. No one has gone 10 for 10 in a game. The highest season performance runs about 53% and lifetime around 45%. The bottom line is the more attempts, the lower the record level performance percentage will rate.

 

[T-Rey]

It's a start but it's not that simple when comparing courses with significantly different SSAs. The SSA establishes the course level challenge. 15 below SSA 54 will always rate better than 15 below SSA 72 but they are apples to oranges different course challenges.

Think of it in terms of percentages. Each shot saved on SSA 54 equals 1/54% (.0185). Each shot saved on SSA 72 equals 1/72% (.0139). So saving 15 shots on an SSA 54 will be a higher percentage (27.8%), i.e., rating, than saving 15 shots on an SSA 72 (20.8%). You might say that's an unfair comparison because you can only save one shot at a time, and you would be right. That's why we need to only compare relative performances among courses within a fairly narrow SSA range.

Another real world example is 3-pt shooting. 100% would be perfect and it's only been achieved in the NBA after 9 attempts. No one has gone 10 for 10 in a game. The highest season performance runs about 53% and lifetime around 45%. The bottom line is the more attempts, the lower the record level performance percentage will rate.

Is there really a difference in saving a stroke on a par 3 vs a par 4? Players get 18 chances to save a stroke no matter whether it is a par 54 or par 72.

 

[Cgkdisc]

Is there really a difference in saving a stroke on a par 3 vs a par 4? Players get 18 chances to save a stroke no matter whether it is a par 54 or par 72.

Yes there is. Saving 1 throw on a par 3.0 reduces the score 33% and on a par 4.0 only 25%. It's mathematically more impressive on the par 3 even though it's one throw in both cases. Remember that a par 3 might range from 2.1 to 3.6 SSA and a par 4 from maybe 3.1 to 4.6 due to the vagaries of holes design, length, foliage, hazards, etc. The fundamental problem is a drop-in and a 500-ft throw both count 1 on the scorecard. There's no scoring adjustment for difficulty such as a drop-in counts 2 and the 500 footer counts 0.002.

 

[teemkey]

Yes there is. Saving 1 throw on a par 3.0 reduces the score 33% and on a par 4.0 only 25%. It's mathematically more impressive on the par 3 even though it's one throw in both cases. Remember that a par 3 might range from 2.1 to 3.6 SSA and a par 4 from maybe 3.1 to 4.6 due to the vagaries of holes design, length, foliage, hazards, etc. The fundamental problem is a drop-in and a 500-ft throw both count 1 on the scorecard. There's no scoring adjustment for difficulty such as a drop-in counts 2 and the 500 footer counts 0.002.

I'm a little confused: Chuck, are you saying hole par (or course par) is a component in computing SSA?

For example, would the SSA at DeLaveaga be different if the holes weren't all par 3?

 

[T-Rey]

Yes there is. Saving 1 throw on a par 3.0 reduces the score 33% and on a par 4.0 only 25%. It's mathematically more impressive on the par 3 even though it's one throw in both cases. Remember that a par 3 might range from 2.1 to 3.6 SSA and a par 4 from maybe 3.1 to 4.6 due to the vagaries of holes design, length, foliage, hazards, etc. The fundamental problem is a drop-in and a 500-ft throw both count 1 on the scorecard. There's no scoring adjustment for difficulty such as a drop-in counts 2 and the 500 footer counts 0.002.

There's no mathematical difference when you look at score relative to par. 4-3 = 3-2

 

[biscoe]

It's a natural law we discovered with the ratings process, not a factor we fabricated.

:\

 

[biscoe]

Whichever round is the most improbable is the one which is best.

 

[JC17393]

I'm a little confused: Chuck, are you saying hole par (or course par) is a component in computing SSA?

For example, would the SSA at DeLaveaga be different if the holes weren't all par 3?

Par is an artificial construct. It's a round number because we like round numbers, but statistically, it is irrelevant. You could call every hole at DeLaveaga a par 15 and it wouldn't change the SSA one bit. SSA stands for Statistical Scoring Average. So what is relevant to SSA on a per hole basis is the average score on each hole. Those are rarely perfectly round numbers like 2 or 3 or 4. They're fractions like 2.85 or 3.12 or 3.92. So when Chuck refers to par 4s vs par 3s, in statistical terms, he's referring to that 3.92 (par 4) versus a 2.85 (par 3).

At least that's what I assume.

 

[T-Rey]

Ok, I'll post about ratings in a different thread.

Seppo 3'd the 860 foot par 5. Not even a throw in, he hit the green in 2!

 

[teemkey]

Par is an artificial construct. It's a round number because we like round numbers, but statistically, it is irrelevant. You could call every hole at DeLaveaga a par 15 and it wouldn't change the SSA one bit. SSA stands for Statistical Scoring Average. So what is relevant to SSA on a per hole basis is the average score on each hole. Those are rarely perfectly round numbers like 2 or 3 or 4. They're fractions like 2.85 or 3.12 or 3.92. So when Chuck refers to par 4s vs par 3s, in statistical terms, he's referring to that 3.92 (par 4) versus a 2.85 (par 3).

At least that's what I assume.

That's what I thought.

So I think T-Rey's question might be rephrased as asking why ratings don't reflect that more good throws per round is harder to accomplish than fewer.

(I think Mr. West might ask the same question ).

 

[Cgkdisc]

That's what I thought.

So I think T-Rey's question might be rephrased as asking why ratings don't reflect that more good throws per round is harder to accomplish than fewer.

(I think Mr. West might ask the same question ).

It's not about how many good throws were made but how many did not need to be made. If I'm supposed to make 72 good throws (par) and only make 60, I saved 12 out of 72. If I'm supposed to make 54 good throws (par) and make 42, I also saved 12 throws but that's out of 54. 12/54ths is bigger than 12/72nds.

 

[T-Rey]

It's not about how many good throws were made but how many did not need to be made. If I'm supposed to make 72 good throws (par) and only make 60, I saved 12 out of 72. If I'm supposed to make 54 good throws (par) and make 42, I also saved 12 throws but that's out of 54. 12/54ths is bigger than 12/72nds.

Ok, guess we will continue this here. In both cases 12 strokes were saved over 18 holes.

 

[Spectacledbear]

Hope there will be some footage of Jerm.

 

[JC17393]

Ok, guess we will continue this here. In both cases 12 strokes were saved over 18 holes.

And Chuck's point is those 18 holes aren't equal. This really isn't that complicated.

 

[Cgkdisc]

It's a natural law we discovered with the ratings process, not a factor we fabricated.

:\

Here's an example. We know a 1000 rated player will beat a 950 player by 5 throws on average on Course X, 5000 ft long with a 50 SSA. Right next to Course X is Course Y also 5000 ft with SSA 50. If they play Course X and then Course Y, the 1000 rated player would be expected to win by 10 throws on average for the two rounds. They just threw 36 holes totaling 10,000 feet.

Let's say we were able to remove every other basket on both courses to create an 18-hole combo course XY that stretches to 10,000 ft on this same terrain. Our two players play Course XY. What would you expect the the scoring difference to be between these two players rated 50 points apart? We know the difference averaged 5 shots on a 5000 ft course and 10 shots when that course was played twice equaling 10,000 feet. It would be logical that the throw difference would average somewhere between 5 and 10. It actually turns out to be about 7 shots based on thousands of data points, not because the ratings team made up some factor.

Assuming these players are shooting scores that average their rating playing this course that's twice as long, their round ratings should still be 50 points apart. Since we know the score difference was measured to average 7 throws, that means the rating points per throw on this much longer 10,00 ft course is 50 points / 7 throws = 7.1 pts/throw versus 10 pts/throw on the 5000 ft 50 SSA course.

 

[biscoe]

It's not about how many good throws were made but how many did not need to be made. If I'm supposed to make 72 good throws (par) and only make 60, I saved 12 out of 72. If I'm supposed to make 54 good throws (par) and make 42, I also saved 12 throws but that's out of 54. 12/54ths is bigger than 12/72nds.

but saving the 15th stroke in the course of a round is not as easy as shaving the first few. any old soul can shave the first few... the probability of shaving more and more does not remain linear once the score leaves the flat part of the bell curve.

 

[AHagglund]

The percentage system, in my opinion, is flawed. In most cases, a birdie on a hole with a SSA of 3 is not any more difficult than a birdie on a hole with a SSA of 4. Yet the current rankings system says that it is.

The logic that "shooting a 40 on a SSA 50 is more impressive than shooting 60 on a SSA 72 because 80% is less than 83%" doesn't work because such calculations use "0" as a starting point (because 40 is 80% of the way between 0 and 50). "0" is a natural starting point for most percentages, but it might as well be an imaginary number when comparing strokes in disc golf, as shooting a zero is impossible.

That is why you get such screwy results like this tournament:
McBeth shoots a 14 down, breaking a course record on a nasty course with roughly a billion trees that has been played over and over again by the best players in the world. He has circle 2 birdie putts on literally every hole. Didn't have to scramble once for an entire round! Then, he makes five or six circle 2 putts. Yet, according to round ratings, this year's Memorial saw FOURTEEN ROUNDS better than his. At last year's Memorial, his 1062 rating was met or exceeded 21 times. This is obviously ridiculous.

A formula could be built using standard deviations as its core mechanic that would fix all of the inconsistencies.

 

[brutalbrutus]

Final round vids are up...

 

[Steve West]

Whichever round is the most improbable is the one which is best.

OK.

So, how do you calculate that?