Ask Chuck Kennedy

[cgkdisc]

Yeah, it looks like the R1/R2 ratings dropped more than I would have expected. Unfortunately, I can't confirm that the unofficial ratings process matches what we currently do officially even on a per round basis. No one who wrote the software free has been involved with PDGA IT for maybe 8 years now. For example, I'm not sure that propagators who shoot more than 60 points below their rating (which many did at the USDGC at least one round) get dropped in the calculation like they do in the official process. This will artificially boost the SSA for unofficial ratings if this isn't done. However, the new web interface for events which is underway will address this issue and should produce ratings more consistent with the official process since we'll know for sure how they are being calculated.

[Karl]

Chuck,

I'm thinking something's goofy in the Yetter Cup ratings (official - that just came out). I base this on the following:

a. The tentative ratings were QUITE a bit higher (like 10 rating points per round average) than the "official", and

b. The REAL reason I say this - that a score of X in, say, round 5 for a Pro equates to the same rating as a round 5 X for an Am...and they played totally different courses with totally different OBs (even if they DID play the same course - which they didn't).

What's the scoop?

Karl

[cgkdisc]

Thanks for pointing this out. I looked at the TD report and I think the course assignments were accidentally shifted one round. The TD assigned courses for the Advanced divisions in R1 thru R4. However, all of the Advanced scores were entered in columns R2 thru R5 so there's a mismatch. We'll add it to our corrections for the next update.

[Karl]

Chuck,

Hope the following helps (if any doubt, better check with the TDs)...

It's my belief that the sections (AMs and Pros) / rounds went this way:

Round 1 =
only Pros played a 27-hole round.

Round 2 =
Pros played 1 x 18-hole round on course "A" with V OBs in place.
AMs played 1 x 18-hole round on course "B" with X OBs in place.

Round 3 =
Pros played 1 x 18-hole round on course "B" with Y OBs in place.
AMs played 1 x 18-hole round on course "A" with Z OBs in place.

Round 4 =
Pros played 1 x 18-hole round on course "A-" with V OBs in place.
AMs played 1 x 18-hole round on course "B-" with X OBs in place.

Round 5 =
Pros played 1 x 18-hole round on course "B-" with Y OBs in place.
AMs played 1 x 18-hole round on course "A-" with Z OBs in place.

Because of this, I believe all 9 rounds (5 for Pros and 4 for AMs) were ALL played on different courses (from any other round)...and thus should be rated "independently".

Karl

[cgkdisc]

I think all of the info is on the report although there are only six different 18-hole layouts besides the 27 played by pros. It was just shifted by round messing it up for everyone. If you think there are eight 18-hole layouts, then contact Andrew to voice your concerns because we don't know what happened other than what we see on the report. He has Adv in R3 playing the same as pros in R2 and Adv in R5 playing the same as pros in R4.

[Dana]

Well I guess that explains why the Yetter ratings were so low. Those won't get fixed until the December update then?

Thanks, Chuck

[cgkdisc]

Yep. We're doing regular updates so often now that we haven't been doing correction runs in-between them.

[jconnell]

Originally Posted by cgkdisc

I think all of the info is on the report although there are only six different 18-hole layouts besides the 27 played by pros. It was just shifted by round messing it up for everyone. If you think there are eight 18-hole layouts, then contact Andrew to voice your concerns because we don't know what happened other than what we see on the report. He has Adv in R3 playing the same as pros in R2 and Adv in R5 playing the same as pros in R4.

You and the TD report have it correct, Chuck. Six layouts used for the collective 8 rounds.

A. 1 through 18, C-pins (Pro R2, Am R3)
B. 19 through 36, C-pins, long tee 32, OB on 36 (Pro R3)
C. 19 through 36, C-pins, short tee 32, no OB on 36 (Am R2)
D. 1 through 18, mixed pins (Pro R4, Am R5)
E. 19 through 36, mixed pins, long tee 32, OB on 36 (Pro R5)
F. 19 through 36, mixed pins, short tee 32, OB on 36 (Am R4)

So the rounds that should have been rated together were both played on the front 18 course. That course was laid out identically for the pro pool and the am pool on both Sat and Sun. The only extenuating circumstances regarding use of that course was on Sunday, when the pro pool played it in the morning in sometimes heavy rain, while the ams played it in the afternoon in drier conditions. I'd argue the difference might be enough to rate the rounds separately anyway. But Pro R2 & Am R3 can be rated together with no issues...weather was same for both.

[cgkdisc]

Perhaps surprisingly, rain alone doesn't seem to affect scores. However, many times rain also involves windier conditions. Apparently, players play a little safer in the rain and it seems to produce similar scores as when it's dry. Either way, we always calculate ratings by round in the first pass and only combine them on the same layout for the final official ratings pass if they are close enough to combine.

[kUrTp]

Hey Chuck,

When you stopped in W PA and paid us a visit you showed me a LED light that was very thin and bright...Remember??? I think you said it was a prototype, are they for sale yet?

[cgkdisc]

Apparently they are not shipping at this point. I tried to set up a distributor and they couldn't meet his order.

[bruce_brakel]

Chuck is there a simple or complicated mathmatical formula that would relate the difference in two players' ratings to the odds of the higher rated player beating the lower rated player in a single round?

[cgkdisc]

It's somewhat complicated and you have to know each player's standard deviation in their ratings. There's probably a direct calculation in the stats book. But I've built a simulator that generates 1000 random scores on a bell curve for 1 to 30 people using their SD values. Then just looks at the scores and figures who won. For example, a player with a 1033 rating and an SD of 23.7 will beat a player with a 1013 rating and SD of 29.5 72% of the time if they are playing head-to-head. Can't help beyond that.

[stevenpwest]

Originally Posted by cgkdisc

It's somewhat complicated and you have to know each player's standard deviation in their ratings. There's probably a direct calculation in the stats book. But I've built a simulator that generates 1000 random scores on a bell curve for 1 to 30 people using their SD values. Then just looks at the scores and figures who won. For example, a player with a 1033 rating and an SD of 23.7 will beat a player with a 1013 rating and SD of 29.5 72% of the time if they are playing head-to-head. Can't help beyond that.

The distribution of the difference in round ratings has a mean equal to the difference in ratings, and a variance equal to the sum of the variances (square of the standard deviations). So, just use the normal distribution function to get the probability of the difference being greater than zero. (Like 70.1433% in your example.)

For general interest purposes, it wouldn't be hard to make a table of probabilities (of one rating beating another) using "typical" standard deviations.

[LongNeck]

I have 21 rounds rated. I have two rounds dropped. a 734 and like a 780something. The rest of mine are from 860- 933. I had two god awful rounds this weekend at 808 and 811. I had to good rounds at like 950 and 912. The two rounds that are dropped are going to be over a year old now. I was wondering if you think those rounds this weekend would be dropped or not. thanks a lot. I also have like 4 other rounds not in the update from like 879- 899.

[cgkdisc]

Can't tell for sure. I don't have the individual data for players to calculate it. My ratings partner Roger has that database. You can always try to do it yourself by calculating your standard deviation using Excel and see whether those rounds are far enough below your average. However, looking at the ratings range of the rounds you just mentioned, it's apparent that rounds more than 100 points below your average will be dropped.

[qdonnelly]

Can rounds at a sanctioned PDGA tournament I played in before I became a PDGA member count towards my rating? If so, how do I get these rounds added to my "record"? Thanks.

[cgkdisc]

Send an email to: dgentry at pdga.com indicating which events you played before you got your PDGA number. Only events played in 2010 will be added, not in years before.

[thediscinmusician]

Got a quick question for you Chuck. I just played in a tournament on 11.27.2010. I know ratings get calculated in a whole year back. I'm having trouble wrapping my mind around this so I'll just give you the details and you let me know. If I don't have a tournament on that date a year ago does it go back to the last tournament I had which would have been on 11.21.09 or is everything from 11.27.09 and on figured in and everything before that date dropped? I guess what I'm asking is...is it from 11.27.09-11.27.10 or what? Thanks for clearing the confusion!

[cgkdisc]

It's 12 months prior to the date of your (and each player's) most recently rated round. However, if that's less than 8 rounds, we go back farther until we get at least 8 rounds. That 11.27.10 event may not make it in this next ratings update unless the TD files the report in the next few days.

[thediscinmusician]

If I can see the results and it says "unofficial tournament results" does that mean they were turned in for the ratings or does the td have to do something else by Tuesday? I hope I'm asking that right....

[cgkdisc]

The answer is in the Ratings FAQ: http://www.pdga.com/faq/ratings/unof...gs-and-results

[thediscinmusician]

I guess it would be easier to ask this...could you tell me if the Belch'n Turkey in Chattanooga has been turned in for this Tuesday's deadline?

[cgkdisc]

Can't tell you because I don't receive the reports. They go to PDGA HQ. You'll be able to tell if it came in and was processed if the results become Official by Thursday. If not, then it didn't come in in time for this update.

[thediscinmusician]

Ok...thanks

[chris]

Has there been any thought on "weighting" rounds based on the number of holes played during a round? If somebody shoots 2 rounds of 27 holes and averages 1050 it should really count the same as someone who shot 3 1050 rated rounds of 18 holes. (The 27 hole round should be weighted with a 1.5 factor, 24 hole round with a 1.33 factor) Right now you could have a situation where 2 people play the same 2 tournaments (1st round 27 holes, 2nd round 18 holes), player A wins both events however ends up with a lower average rating than player B who took 2nd in both events simply because he shot better on the 27 hole layout.

[cgkdisc]

The ratings have been weighted by the number of holes in each round for years now. Players' round ratings are maintained in the database on a "per hole" basis. That is multiplied by the number of holes in each of the rounds being used to produce the new rating at each update. That's why someone trying to use their published round ratings to calculate their new rating might not get exactly the same number due to rounding errors.

[throwinROCS]

So here's an interesting putting style that my brother asked me about. I think I know the ruling, but thought I should ask the expert.

Say your lie sits at 29 ft. You get up to the disc and mark it. You place your right foot so that your toe is just inside the 30 cm allowable space directly behind the lie. Your left foot is then placed behind and to the left, as in a "normal" putting stance.

Ok, so here's the question. You putt and use your left foot to push off and step through. Your left foot lands behind the "no-cross" line created by the marking of your lie. Therefore, it is further forward than your right foot. After your left foot has landed, you are in a set position showing balance and remaining behind the "no-cross" line.

Is it legal?

I think so.

[cgkdisc]

Legal. The reference distance for making the call is the distance from the pin to the back edge of your marker. As long as the following foot ends up farther from the pin than that reference distance, you're still good.

Now, let's say this player pushes off with his left foot and has it in the air at the time the disc is released. The player sort of freezes that position holding the left foot in the air for say 3 seconds and then swings it forward past the marker before touching the ground. The RC is pondering whether that visible pause will be enough "proof" of balance to meet the way the rule is currently written and not be considered a foot fault.

[throwinROCS]

If they are going to review the ruling, would it not be easier to just completely vacate the 10M rule? If we were to leave this rule behind, it would mostly only affect shots from the 6M-10M range and could definitely pick up speed of play (no measuring, asking, or walking off putts). Then, all stance violations are the same on every throw any where on the course.

Does enforcing this rule really add anything to the game? Would not enforcing it really take anything away from it?

I assume this rule was created to kind of create a 'green'. It doesn't though.