Compared to the kinds of speech we see on this site, you are hardly being impolite. In regard to the normalization of the yearly TAZD, I will wait to see how my final numbers compare to your final numbers to assess the impact of nonnormalization over 12 months. In regard to my Wall, the original document I pasted in it is so ancient that it would have to be updated to a new version of Word before I pasted it in and I don't know how many #^^&%$%**%'s will show up in it and the walls are noneditable.

As for actual, as opposed to theoretical, outliers, the only serious example has been noamlang1 and he attested that he was not trying to game the TAZD and he faded in the end.

As for the childproofing, I want to see if I anticipated the kinds of attacks that might be made. It would mess up the experiment if I tip my hand now. If the childproofing kicks in, I will post the scorers but mark them to indicate that they have been childproofed.

you go girl <3

This thread is extremely funny. Skrumgaer is the best troll ever, even if he does not do it intentionally.

"In regard to the normalization of the yearly TAZD, I will wait to see how my final numbers compare to your final numbers to assess the impact of nonnormalization over 12 months."

I doubt that they'll have any MAJOR impact, but it could be a deciding factor when two people's scores are very very close pre-normalisation (less than a couple hundred points). It's a simple correction that makes sense to do- it's just one extra line in my spreadsheet for each player.

"In regard to my Wall, the original document I pasted in it is so ancient that it would have to be updated to a new version of Word before I pasted it in and I don't know how many #^^&%$%**%'s will show up in it and the walls are noneditable."

Paste it in notepad then paste it in the wallpost editor to get rid of the &7236*s that would otherwise show up (I think- someone please correct me if I'm mistaken).
The fact remains that you've failed to update it for almost three years and still refer to that as the resource players should turn to when they have questions. You must admit that the post is hardly comprehensible for the average joe. It isn't so hard to put it in terms the non-mathematically minded can understand- I've done it several times.

"As for actual, as opposed to theoretical, outliers, the only serious example has been noamlang1 and he attested that he was not trying to game the TAZD and he faded in the end."

You should not score well in a measure of positive skill when you have 35% 7th places, plain and simple. It doesn't matter if you're trying to game the system or not. You can argue all you want that having a large number of 7th place finishes is a good strategy, but if you're truly so obtuse to see that it isn't then the point is moot.
Also, there are outliers in the number of games played every single month and in the yearly.

"As for the childproofing, I want to see if I anticipated the kinds of attacks that might be made. It would mess up the experiment if I tip my hand now. If the childproofing kicks in, I will post the scorers but mark them to indicate that they have been childproofed. "

As I said, a good measure of skill should not require childproofing. The TAZD* certainly doesn't. Nor do the ASR multiplier or ASR. No one cares enough to purposefully try to game your system. The best systems can't be taken advantage of. That you have to do this check speaks to the weakness of the TAZD.

And as for "I will wait to see how my final numbers compare to your final numbers"
It's shameful that you're unwilling to do the work yourself. It is literally one line per player on your spreadsheet (which you don't let people see for some reason).

... not to mention that my numbers don't match yours for other reasons besides normalisation of percentage profiles.

Ok I'm done now. Triple post fail.

more like triple post win imo. well said.

Why should you be the dictatrix who sets arbitrary boundaries as to what percentages represent skill and what percentages don't? You dictate that certain percentages are out of bounds without offering a justification based on statistics. You limit the maximum number of games that count in your TAZD* without offering a justification for it on basis of statistics. You make other tweaks in your system based on your beliefs about how much points should count without offering justification. You naiively assume that there aren't people out there that will try to wreck the system. People have tried to wreck the trophy system and Ryan has had to react. The TAZD is mine, not Ryan's and I am limited only to my powers. It's not shameful not to do work on spreadsheets if the work doesn't have to be done, and I won't know if it has to be done until I see the numbers. And the best time to do work on changes is at the start of a new period before the templates fill up with new data, just as Ryan reserves 12 hours at the start of each month. Can you hold out for three more weeks?

The TAZD is simple. It is a Pearson's chi-square. The zero datum is derived from a pool of real players reacting to real incentives of the game. No one has been seriously hurt by the TAZD. There are other systems I think might be better, for example, a 60-game moving Elo average, but in the contect of the current system with the data currently available, I stand by the TAZD.

"Why should you be the dictatrix who sets arbitrary boundaries as to what percentages represent skill and what percentages don't?"
I'm not. Everyone in this thread except you agrees with what I've said because it's very simple- lots of sevenths is bad, lots of firsts is good. You're trying to measure positive skill. Why not use POSITIVE deviation?

"You limit the maximum number of games that count in your TAZD* without offering a justification for it on basis of statistics."
Your justification for having a games multiplier at all isn't statistics based, so why should my limit on games be statistics based? Placing no limit ignores outliers completely, and I don't believe that people who spend 30 hours a week playing this game should be rewarded for it.

"You make other tweaks in your system based on your beliefs about how much points should count without offering justification."
Such as? Normalising the data? All that does is place everyone on even footing to begin with since Ryan truncates instead of rounds. To what other tweaks are you referring?

"You naiively assume that there aren't people out there that will try to wreck the system. People have tried to wreck the trophy system and Ryan has had to react."
People care about trophies. You naively assume that people care as much about the TAZD and they do about trophies. Fact is, most people do not give two shits about the TAZD. Those who do don't care enough to put in the effort to game the system.
And AS I SAID, you shouldn't NEED childproofing. Find me a way to game the TAZD*. Go ahead. Do it. Wreck the system. Try.

"It's not shameful not to do work on spreadsheets if the work doesn't have to be done, and I won't know if it has to be done until I see the numbers."
You won't see the numbers if you don't do the work. My numbers certainly won't help you- the TAZD* is a different calculation than the TAZD. It's one line per player. 35 or so lines in your spreadsheet. If you're too lazy to do that yourself, you should be ashamed. Then again, you're too lazy to put out a full update at the end of each month or to bother explaining the TAZD to people, so I guess 35 lines a lot to ask of you, isn't it?

"And the best time to do work on changes is at the start of a new period before the templates fill up with new data, just as Ryan reserves 12 hours at the start of each month. Can you hold out for three more weeks?"
We had all these arguments a year ago, skrum. No changes were made to the TAZD with the exception of your mysterious child-proofing (which, to me, just sounds like you hand remove the outliers to cover your ass since people complained about it). You didn't listen then and you certainly aren't listening now.

P1. 35% sevenths sounds arbitrary to me.
P2. My games number is the square root of the number of games because the variance goes as the number of games, the standard deviation is the square root of the variance, and the standard deviation is used to measure differences. Looks like statistics to me.
P3. The other tweaks are the ASR and the ASRm.
P4. Your system is already wrecked because of all the arbitrary restrictions you have put in it because of your personal preferences, not based on laws of statistics.
P5. I will do the work on the numbers (my numbers!) when the results of the experiment are in. Not before.
P6. The TAZD is a simple, robust system. Changes don't have to be made to it often, and these changes mostly have to do with changing the datum when the game changes. There was an error in my spreadsheet that I fixed immediately when it was discovered. The TAZD works the same way now as it did when I posted my three-year old wall document. The only changes that need to be made to the wall document is to change the datum numbers and change the number of games rule for the baseball-style standings to the square root rule. But especially for you, I will indicate on my monthly TAZD posts that "The most recent zero datum can be found in my yearly TAZD post on the Advisor Blog."