Quick question....Whats the situation if one of your most 8 recent rounds (As of the update) happens to be dropped? What then is double weighted, the other 7 or does it "skip" the dropped round, still using 8 (minus the dropped one).

Also, I know you've mentioned it somewhere else, but how do you calculate your own individual SD? (So I know what rounds will get dropped.)

We first determine what rounds will be double weighted. If any of those turn out to be dropped, we don't replace it with another double weighted round.

Plug all of your round ratings into Excel and use the standard deviation function.

You're THE man Chuck! Thanks!!

Also, is the SD calculated from the new data (IE with rounds I don't know ratings of right now) or from the data as of the last update.

And its 2.5 that gets dropped, correct?

The SD is based on the rounds included in that update. So what you see is what was used. Rounds more than 2.5 SD or 100 points (which ever is smaller) below your round rating average are dropped.

If you want to do rough calcs for the next update, you'll need to include your estimated round ratings from new events that weren't in the last update and drop any that are more than 12 months before the latest event date that will be in your new set of rounds for mid-December.

Plug all of your round ratings into Excel and use the standard deviation function.



BUT BE SURE to use the correct stdev calculation!

STDEVP is the correct function. STDEVP uses the standard deviation formula for when you are using the entire population of a data set.

STDEV is the incorrect function. STDEV uses a standard deviation formula that is appropriate only when you are doing the calculation based on a SAMPLE of the entire population.

since you are using the entire population of rounds, use STDEVP.

i am almost willing to wager cash that the standard deviation calcualtion in the official ratings uses the incorrect (STDEV) formula!

i am almost willing to wager cash that the standard deviation calcualtion in the official ratings uses the incorrect (STDEV) formula!



I am almost willing to wager cash that it doesn't make a fat lot of difference.

For a typical golfer, say....you, it would be a tiny difference, multiplied by 2.5, which would make it around a single point. And if, and that's a big *if*, you happen to have a round that falls that close to your 2.5SD line, the inclusion or exlusion of that round (which is almost exactly 2.5SD from your average) will impact your overall rating by about 2 points. Or put another way, about 1 stroke every 5 rounds.

If STDEV vs STDEVP is the biggest problem the ratings have to worry about, the world is a wonderful place. But you are correct, if something is easy to get right, you might as well get it right.

But don't you have some google maps to be making or something?

true dat my friend! :D

but i cant google map til certain message board upgrades get moved into production (hopefully very soon now - testing is happening now)

also true that it would rarely make a differnce. i found this thing because i could not my spreadsheet calcs to match up with my offical rating. an 850 round my spreadsheet was dropping was showing up as included in the official numbers. believe it or not, using STDEVP made my 2.5SD = 850+ while with STDEV it was a tiny fraction below 850. so one formula inclded the 850 round while the other dropped it! the actual ratings difference was about 2 rating points if i remember correctly.

rare? for sure! worth checking into and making sure we're doing it right? absolutely!

for a typical golfer, say... me, that could make a world of difference in determining which division i'm qualified for at say,... the Mid-Nationals!!! :cool:

here's the actual real live numbers with their inclusion status:

<table border="1"><tr><td>925</td><td>Yes
</td></tr><tr><td>915</td><td>Yes
</td></tr><tr><td>971</td><td>Yes
</td></tr><tr><td>937</td><td>Yes
</td></tr><tr><td>929</td><td>Yes
</td></tr><tr><td>935</td><td>Yes
</td></tr><tr><td>948</td><td>Yes
</td></tr><tr><td>947</td><td>Yes
</td></tr><tr><td>948</td><td>Yes
</td></tr><tr><td>926</td><td>Yes
</td></tr><tr><td>939</td><td>Yes
</td></tr><tr><td>898</td><td>Yes
</td></tr><tr><td>962</td><td>Yes
</td></tr><tr><td>903</td><td>Yes
</td></tr><tr><td>900</td><td>Yes
</td></tr><tr><td>899</td><td>Yes
</td></tr><tr><td>908</td><td>Yes
</td></tr><tr><td>935</td><td>Yes
</td></tr><tr><td>904</td><td>Yes
</td></tr><tr><td>850</td><td>Yes
</td></tr><tr><td>935</td><td>Yes
</td></tr><tr><td>911</td><td>Yes
</td></tr><tr><td>924</td><td>Yes
</td></tr><tr><td>956</td><td>Yes
</td></tr><tr><td>920</td><td>Yes
</td></tr><tr><td>932</td><td>Yes
</td></tr><tr><td>912</td><td>Yes
</td></tr><tr><td>962</td><td>Yes
</td></tr><tr><td>876</td><td>Yes
</td></tr><tr><td>964</td><td>Yes
</td></tr><tr><td>895</td><td>Yes
</td></tr><tr><td>903</td><td>Yes
</td></tr><tr><td>940</td><td>Yes
</td></tr><tr><td>944</td><td>Yes
</td></tr><tr><td>902</td><td>Yes
</td></tr><tr><td>909</td><td>Yes
</td></tr><tr><td>937</td><td>Yes
</td></tr><tr><td>914</td><td>Yes
</td></tr><tr><td>826</td><td>No
</td></tr><tr><td>912</td><td>Yes
</td></tr><tr><td>913</td><td>Yes
</td></tr><tr><td>935</td><td>Yes
</td></tr><tr><td></tr></td></table>

round average: 921.3658537

stdevp: 28.47321413
2.5 SDp: 71.18303533
average - 2.5 SDp (aka cutoff): 850.1828183

stdev: 28.8183566
2.5 SDp: 72.0458915
average - 2.5 SDp (aka cutoff): 849.3199622


therefore, i conclude that the official calcs use STDEV instead of the more appropriate STDEVP.

these are from my 09/15/05 rating, pdga # 10403

The other thing to consider is that the actual calculation is done using a player's DPH (differential per hole) ratings which aren't published. The DPH is the true rating before it's converted into the round ratings you're more familiar with. These DPH numbers haven't been rounded off like the published round ratings so a variance of several points is not uncommon if you try to do the calculations using the published ratings which have been rounded.

since its hidden, i cant really consider it, can i :D

what the heck is Differential Per Hole??? i know you dont wanna reveal the calculation, but how about a general description of what it is and how it fits into the overall ratings process?

regardless, could we verify which standard deviation calc is used? just for kicks if nothing else.