Ask Chuck Kennedy

[cgkdisc]

I don't think an automatic rule like a percentage is necessary. In some cases, there may be legit DNFs (injury, sickness, family emergency, weather) mixed in with ones that were more suspect. However, if players knew that TDs might report them to the PDGA for disciplinary action for a pattern of DNF behavior, that may become a deterent. Like I said, some players have been contacted and some disciplined for DNFs especially when connected with bad behavior.

[Trey133]

When a player DNF's does he get to keep the rest of his round ratings? For example, if a player DNF's the last round of a 2 day tournament, does he/she keep the previous 3 round ratings?

[cgkdisc]

Yes. All completed rounds get rated.

[Jah Plastifari]

Chuck,

I am thinking of playing smaller tourneys with only forehand, roller, or overhand drives. This will improve my overall game, but my rating will take a beating. My rating does not matter to me. If I promise to use my disc powers only for the forces of good and not show up at Am Worlds playing backhand Intermediate, does it matter to you? I may end up with a new record for standard deviation.

[cgkdisc]

Doesn't matter. Any players who shoot more than 60 points below their rating don't get included in the calculation that produces ratings for everyone in the event. Good luck. Just for fun I played the second round of an event with a roller off the tee for the challenge even if it wouldn't normally make sense.

[AviarX]

Originally Posted by cgkdisc

It never hurts anyone else if a player dumps a round to "protect" their rating.

Never sounds a bit over-reaching. if their are 10 propagating players on an 18 hold par 72 course and a 1000 rated Pro DNF's what would have been a 890 round -- it can artificially make the that round's rating lower for everyone else -- and for the course (than it would be if the player finished). In short, it messes with the numbers. right?

[cgkdisc]

A propagator only "propagates" in that round if their round rating is no more than 60 points below their rating. So he wouldn't have been used anyway.

[the_kid]

Originally Posted by cgkdisc

A propagator only "propagates" in that round if their round rating is no more than 60 points below their rating. So he wouldn't have been used anyway.

But if he tanks what would have been a 950 then it does......

[AviarX]

Originally Posted by cgkdisc

A propagator only "propagates" in that round if their round rating is no more than 60 points below their rating. So he wouldn't have been used anyway.

okay, so when a player DNF's in a round that they would have shot 59 points below their rating it artificially skews the data (even more so if the sample size of propagators is small). every system has its vulnerabilities -- even [y]ours -- right?

i like the rating system -- don't get me wrong. i like it so much that i wish it was revealed under our message board ID's here at pdga.com's DISCussion Board by default with every post (and those who wish to not show their rating would have to uncheck its being shown with their posts)

btw -- did you hear that Pieradise is for sale?

http://www.pieradisediscgolf.com/ind...d=64&Itemid=68

[cgkdisc]

okay, so when a player DNF's in a round that they would have shot 59 points below their rating it artificially skews the data (even more so if the sample size of propagators is small). every system has its vulnerabilities -- even [y]ours -- right?

No, because 59 points low is within the normal variation a player of any rating might shoot. It would be hard for a player to precisely calculate what they need to shoot 59 points low. And why would they try? They would be kicking themself for not playing bad enough to have the round dropped and not affect their rating. So a 59 pt low rating is more likely legit than deliberate.

[bruce_brakel]

Originally Posted by AviarX

i like the rating system -- don't get me wrong. i like it so much that i wish it was revealed under our message board ID's here at pdga.com's DISCussion Board by default with every post (and those who wish to not show their rating would have to uncheck its being shown with their posts)

"Click, click, back, back," which then became some insane combination of 12 steps due to thin-skinned low rated people in power, is now "click, close." I guess thats what happens when you get a couple of pros on the board?

[AviarX]

Originally Posted by cgkdisc

No, because 59 points low is within the normal variation a player of any rating might shoot. It would be hard for a player to precisely calculate what they need to shoot 59 points low. And why would they try? They would be kicking themself for not playing bad enough to have the round dropped and not affect their rating. So a 59 pt low rating is more likely legit than deliberate.

who said anything about the player calculating it out? They are shooting a bad round and they know it is below their rating and they want the rating to go up not down next update. so they drop out. granted -- sometimes they were not shooting as bad as they thought. point is that at those times when they are 30 to 59 points below their rating and when the number of propagators on their layout is less than 10 it skews the numbers.

to Bruce -- yeah, how long was it (a week?) that we actually had the ratings automatically beneathe our ID? that was a terrible thing Pat did i guess

[cgkdisc]

point is that at those times when they are 30 to 59 points below their rating and when the number of propagators on their layout is less than 10 it skews the numbers.

It does not skew the ratings because everyone has good and bad rounds on a regular statistical basis. If a 950 prop shoots 30 or 55 points low in a round, it does not affect the ratings calculation any more than a 1000 rated player or 850 propagator shooting the same amount low. For every prop who shoots low, there are other props shooting higher by that amount.

[AviarX]

Originally Posted by cgkdisc

It does not skew the ratings because everyone has good and bad rounds on a regular statistical basis. If a 950 prop shoots 30 or 55 points low in a round, it does not affect the ratings calculation any more than a 1000 rated player or 850 propagator shooting the same amount low. For every prop who shoots low, there are other props shooting higher by that amount.

it does not skew the ratings if the player in question finishes their round. but above example refers to a player that DNF's a round that would have been 30 or 50 points below their average. while in theory the amount of propagators is significantly high enough to mitigate the guy who DNF's due to interest in the next ratings update, in practice the amount of propagators is sometimes pretty small and that is when the data is most skewed by a ratings-concern-based DNF. granted -- if someone shooting a round 30 to 50 points above their rating also DNF'ed in the example your response holds true. but how comparatively often does that happen?

[cgkdisc]

It still doesn't skew the ratings. A DNF is no different than all of the props who didn't show up to play that day. Who's to say that the numbers aren't better or more accurate without that prop's score in the mix? We don't really know that. Occasional DNFs are all part of the statistical soup that produces the ratings.

[AviarX]

Originally Posted by cgkdisc

It still doesn't skew the ratings. A DNF is no different than all of the props who didn't show up to play that day. Who's to say that the numbers aren't better or more accurate without that prop's score in the mix? We don't really know that. Occasional DNFs are all part of the statistical soup that produces the ratings.

Yes, in theory that is exactly why ratings work. I can understand if you feel a disinclination to discuss ways the ratings can get skewed, but isn't one of the reasons rounds get reviewed before going official due to the fact of 'garbage in = garbage out?'

we can say a DNF caused by a golfer's wish to be rated higher than his buddies or high enough to become sponsored will come out in the wash, but -- what of the example of how a particular layout played for a particular round on a given course? (which impacts anyone who played that same layout for that round as well as the course rating for that particular round).

i am sure you (and others) can offer examples of atypical situations in which a round's rating on a given course could get skewed. Of course since a DNF never enters the statistical equation -- it cannot skew the ratings -- in the same sense evoked by the old saying: "dead men tell no tales."

pardon me for not knowing, but i think it's worth noting here: is there anything in the rules of conduct which explicitly states that DNF'ing to protect one's rating (positively) or that throwing a round to protect one's rating (negatively) is considered unprofessional and poor sportsmanship and is grounds for expulsion form the PDGA?

[cgkdisc]

I can understand if you feel a disinclination to discuss ways the ratings can get skewed, but isn't one of the reasons rounds get reviewed before going official due to the fact of 'garbage in = garbage out?'

No one knows whether including a round or not will produce a more accurate SSA value because there's no independent baseline SSA that's known, only what is generated from the propagator scores. The only rounds deemed inappropriate are those more than 60 points low because players can either deliberately produce those scores or they can arise due to injury.

If TDs report consistent DNF behavior, the PDGA Disciplinary Committee may get involved (and has).

[thediscinmusician]

Chuck, any way you could tell me if the Hartselle Ice Bowl from February 27th has been put in for the ratings update coming up? Thanks!

[cgkdisc]

Nope. I'm the last to know. Always contact the PDGA office for the answer on tourney reports if the event results are not yet posted online as "Official." If they are listed as Official without ratings, that means the report has been received and processed.

[bruceuk]

Hi Chuck

Thought you might like this that one of our UK stattos produced:
http://public.tableausoftware.com/vi...s&:toolbar=yes

Keep up the great work

[cgkdisc]

Cool graph and fun way to display performances.

[AviarX]

Chuck (or anybody), do you know where I can find full results to the distance competition held at Pro Worlds 07?

[cgkdisc]

Check your email.

[boobs]

Ratings; Early, On Time or Late?
Thanks!

[cgkdisc]

Probably tomorrow.

[boobs]

Today or Tomorrow?

[cgkdisc]

Tomorrow. It's done but Gentry has to upload it to the server.

[LongNeck]

My rating will go down like 2-3 points. I only played in one. I am playing in 3 in the next 3 weeks. 13 rounds in the next 2 months. So mine should go up next time.

[AviarX]

Originally Posted by cgkdisc

Check your email.

thank you Chuck!

i just noticed the "Players Stats" link now underneath the avatar's of individual posts -- how long has that been in place? i applaud the decision, and he/she/they that triggered it.

[cgkdisc]

Didn't see the link until you mentioned it. Must be pretty new. The web contractors have been working on several upgrades but mostly behind the scenes.