Ask Chuck Kennedy

[AviarX]

Hey Chuck -- the results in Open for the 3rd Turkey Tussle for The Caring Center are wrong. I finished 1st and Vince finished 2nd. I shot the 59 in round 2 and he shot the 60. The payout info is correct...

http://www.pdga.com/tournament_results/64432/Open

will this raise my rating one point? :-)

[cgkdisc]

Send a note to asweeton@pdga.com He has to confirm with the TD for the results to be changed.

[lonhart]

Hi Chuck,

Where can I read about the math that determines the number of rating points associated with a stroke? For example, in some rounds, a score of 54 = 950 and a 53 = 955. A later round has 54 = 950 and a 53 = 957. I am curious what drives the variation in the magnitude of the difference for a single stroke from one round to the next.

Thanks!
Steve

[kadeatkinson]

Chuck,
In the past I remember this site had a compilation of past ssa's for a course/tournament lay out. I can't seem to locate this. Do you know if this link still exits an if so where to locate it?

[cgkdisc]

@lonhart - The table is not available. You can figure it out for various courses from looking at the difference in rating between scores 1 throw apart. For courses with low SSA values under 45, it gets up to 14 rating points per throw. For high SSA courses like Winthrop it gets down to 7 points per throw.

@kadeatkinson - The Board decided to withdraw that information from the site.

[Patrick P]

Is a Final 9 rated? Would it be weighted .5x to an 18 hole round and are 27 hole rounds 1.5x weighted? Or are they all weighted equally?

[cgkdisc]

No rounds under 13 holes get official ratings. Yes, all round ratings get weighted relative to the number of holes compared with 18.

[ERicJ]

Originally Posted by cgkdisc

Yes, all round ratings get weighted relative to the number of holes compared with 18.

Wait... what?

If that were true then (ignoring double weighting most recent 25%):
Joe has two 18-hole rounds at 900 each and two 18-hole rounds at 1000 each his player rating would be 950.

Bill has two 18-hole rounds at 900 each and two 27-hole rounds at 1000 each his player rating would be 960.

Is that really the case?? I don't recall seeing anything like that in the Ratings document or FAQ.

[Patrick P]

Originally Posted by ERicJ

Wait... what?

If that were true then (ignoring double weighting most recent 25%):
Joe has two 18-hole rounds at 900 each and two 18-hole rounds at 1000 each his player rating would be 950.

Bill has two 18-hole rounds at 900 each and two 27-hole rounds at 1000 each his player rating would be 960.

Is that really the case?? I don't recall seeing anything like that in the Ratings document or FAQ.

960 is what I calculated in your scenario. I never knew that as well, I'll have to go back and recalculate my prior updated ratings, that might explain why sometimes I'm off a rating point.

[cgkdisc]

It's always been that way. Courses with a different number of holes is partly why we calculate and hold your ratings in the database on a "per hole" basis then expand them later based on the number of holes. That's why you'll be off a point or two due to rounding errors because the "per hole" values are magnified to 3 or 4 digits.

Back in the early 2000s, we didn't get good data from a Worlds report for some reason and it went in the database as a 144-hole "round rating" for every player.

[ERicJ]

Originally Posted by cgkdisc

It's always been that way.

If that's the case then may I suggest a couple improvements?

Add that "weighted" detail to:
http://www.pdga.com/faq/ratings/how-...ing-calculated
http://www.pdga.com/documents/ratings-guide

A player's "Rating Detail" page (e.g. http://www.pdga.com/player_ratings_d...8396/2011/4949) should also show the number of holes per round.

[cgkdisc]

The current position from PDGA HQ is to not provide more written details on the process than is already out there. It's not my call. Roger and I are contractors with the PDGA to provide what they ask for. Enhancing the player stats pages is on hold until after the fabled new website finally launches. There are several enhancement ideas in the pipeline such as including the number of holes in a round.

[JenniferB]

I thought PDGA members who got worlds invites would get automated emails in addition to being on the list. Not sure where I got that idea, but it seems like a good promotion strategy. I'm on the list, but didn't get an email. Is that unusual?

[cgkdisc]

Emails (or letters before email) with all of the registration details traditionally are sent in mid-March. It's only in the past 7 years or so that the list of those invited got posted in mid-February once the previous year point totals were tallied.

[Fats]

Originally Posted by cgkdisc

It's always been that way. Courses with a different number of holes is partly why we calculate and hold your ratings in the database on a "per hole" basis then expand them later based on the number of holes.

From now on, I'm only playing tournament rounds of 72 holes or more.

[pgyori]

Perhaps this has already been discussed, but with the current ratings system, aren't we going to see ratings continue to inflate?

If we had a closed pool of 10 (the actual number doesn't really matter) players that participated in tournaments together. Each time a player shoots bad enough that the round doesn't count towards his ratings (i.e. is 2 standard deviations below his rating) the overall average ratings applied for the pool of players increases. I know that the actual pool of players is not static, but with bad round outliers thrown out, it pushes the overall ratings upward.

Am i missing something?

[cgkdisc]

In theory, the average ratings of your closed group would increase. However, the effect is minimal in the bigger picture with thousands of players. Note that the average player has one round dropped out of 50. That works out to one round every three years on average since PDGA members currently average around 16 rated rounds per year.

The thing to consider is how do you know whether a round is naturally poor (normal statistical variance) or deliberately poor? We don't know for sure so we lop off the tiny bottom tail of the statistical distribution just in case to make sure the overall integrity of the system stays intact and doesn't get pulled down by a deliberately poor score that wouldn't be a "naturally" generated poor score in the system.

Back to your original premise with the closed pool of 10 players. What would happen is the SSA of the courses they played, which are dynamically generated from their ratings, would slightly decrease over time offsetting the effect of dropping a few rounds in each of their ratings which would slightly increase their ratings. Net effect over time works out to zero inflation in the average ratings in the system.

[lonhart]

Originally Posted by cgkdisc

@lonhart - The table is not available. You can figure it out for various courses from looking at the difference in rating between scores 1 throw apart. For courses with low SSA values under 45, it gets up to 14 rating points per throw. For high SSA courses like Winthrop it gets down to 7 points per throw.

Thanks Chuck. It appears to boil down to this: on an easy course (a low SSA, birdie-fest) each stroke is worth a lot of rating points (e.g., 14) while on a tough course (high SSA, hard to get par) each stroke is worth less (e.g., 7 points).

I assume this pattern is a function of a lack of spread in the scoring overall on an easy course (lots of people with similar stroke counts) while on the hard course there is a great deal of spread in actual stroke counts.

Thanks again!
Steve

[pgyori]

Thanks for the response, Chuck, although i didn't think the course SSA really had an impact on the mean ratings generated, only the shape of the distribution. I thought the average of the propagators ratings would be used to determine the base ratings, and the SSA helped shape that distribution (more narrowly/widely per stroke).

I also think you can get rid of the deliberately poor rounds if bad rounds aren't dropped. I can't think of a reason a player would shoot a deliberately poor round except in the circumstance where they are having a bad round, and then try to ensure that the bad round is bad enough not to impact their ratings (and even then, i would expect that the majority of players don't think/play that way).

[cgkdisc]

@lonhart - Yep. Think of it this way. If we are 70 points apart in ratings, on average you would beat me by 7 on a 50 SSA course, only 5 on a low SSA course and by 10 on a high SSA course. The differen in our skill level (rating) hasn't changed. Thus, the number of ratings points per throw has to change as the SSA changes so the math hangs together with actual scores.

[cgkdisc]

pgyori - I also think you can get rid of the deliberately poor rounds if bad rounds aren't dropped. I can't think of a reason a player would shoot a deliberately poor round except in the circumstance where they are having a bad round, and then try to ensure that the bad round is bad enough not to impact their ratings (and even then, i would expect that the majority of players don't think/play that way).

The fact that we do drop bad rounds prevents the potential for deliberate behavior for Ams to tank rounds to drop their ratings to get into a lower division. The fundamental reason for ratings is to prevent sandbagging and get players playing the appropriate division for their skill level. Dropping bad rounds is a key mechanism to prevent any payoff from tanking behavior.

[bruce_brakel]

Chuck, I have a question about the new world rankings formula if you know the answer. The descriptive text says, "However starting 2012, players must have completed Worlds or two other Majors during the past year to be included in the World Rankings." Are there any further unstated caveats regarding the division played at the two other majors, or whether the Major is a major that produces ratings?

[cgkdisc]

Participation must be in MPO or FPO in the Majors listed on the World Rankings page and all of those provide ratings in singles competition. No doubles, teams or match play events at this point. Do you think an event or division listing was missing in the explanation text?

[bruce_brakel]

I did not see in the explanation text that it was some divisions and not others, some majors but not all of them. I did not see exactly which divisions and majors were excluded from consideration. It might be there. I might not have read closely enough or clicked on enough links to find it whereever it might be. But I did smell it, so I thought I'd ask.

[bruce_brakel]

The three main documents seem to be

http://www.pdga.com/announcements/new-world-rankings-starting-2012
fficeffice" />>>
http://www.pdga.com/WorldRankings
>>
http://www.pdga.com/files/documents/PDGA_World_Ranking_System_2012.pdf
>>
They all have a sentence that says something like this, which is quoted from the third document:

Starting 2012, players must have completed Worlds or two other Majors within the past year (Japan Open goes two years) to be included in the rankings. A man's best four NT/ET events (three for women) counts as one Major.

I take it that what is meant by that, with clarifying additional text in colored text, is:

Starting 2012, players must have completed Worlds or two other specified Majors in the MPO or FPO division within the past year (Japan Open goes two years) to be included in the rankings. A man's best four NT/ET events (three for women) counts as one Major. Competition in a division other than MPO or FPO, or competition at a Major not on the specified list, does not satisy the requirement necessary to be included in the World Rankings.

[cgkdisc]

I'll take a look at clarifying the text. It has always indicated that "Master and older players who meet the minimum rating criteria are included. However, they only get credit when entering the Open Pro division at any event included in the rankings."

[bruce_brakel]

Right, and I'm just not understanding what that sentence means. I think I'm being confused by the word "credit." But is my clarifying additional text correct? Masters and amateurs who have the rating to get on the list, and who play Worlds or two of the other specified Majors, will be excluded from the list unless they play Worlds in MPO or FPO, or play their two other Majors in MPO or FPO. They have to play MPO or FPO at the specified major tournaments for the specified major tournaments to satisfy the criteria of having played those tournaments?

I'm fine with whatever the criteria are. I'm not arguing what the criteria should be. I just want to understand in case Kelsey cares about the list. And this is the other question she might ask, "Can a woman with the necessary rating qualify for the women's ranking list by playing MPO at Worlds, or does she have to play FPO?" Kelsey would be more inclined to play MPO than FPO, but I think her plan is to play MA1.

[cgkdisc]

World Rankings are for performance in your gender specific Open division. Women are ranked with women and men with men. While a woman could enter MPO in enough Majors to qualify for the men's ranking, no woman has yet earned a rating over 999 so it's not realistic for the time being. A woman's competition results in MPO at Majors or NTs would not qualify for the Women's ranking. Women must play in FPO in those events for potential inclusion in the women's ranking.

Yes, an amateur of either gender could potentially make their gender's World Ranking list if they played Open in Worlds or enough Majors to qualify and their rating was high enough.

[sammyshaheen]

Chuck
Do tournament directors need to take the certified officials
exam every year? Is it free if I do? My tournament is coming up
in a couple of weeks and want to be ready to go.

Also would like to know for World's registration.
Thanks

[cgkdisc]

TDs have to be certified officials and I believe your certification will be good for at least 2012-2013, maybe 2014. Yes, you need to be certified to play in NTs and Majors.