Just re-read Andernut's post - If it's basically a straight conversion of my/ATH's proposal to percentiles, then the only difference (I think) is his forces normalization across categories, although there are ways to de-normalize - for instance if the highest team score is 2385 then the highest percentile score for that category would be 2385/3424=0.70

It's a little convoluted - and there's something to be said for having as simple a system as possible so that members don't have to be math geeks to understand what they mean, but then again the resulting percentile scores are very accessible and a quick read.

Ozyman wrote:

+1 to M57s idea.  I like the idea of normalizing CP with GR to average.  I'm not sure about Team & Tourney.

O, Just to clarify..

You DO like the idea of "Pick 2" for Rank purposes, and your preference would be to NOT normalize TM and TN? (I see little harm in normalizing all 4 here. After all, it's only a rank and I think it's good for the site when as many members as possible see themselves getting 'promoted').

And on a related note - As far as an Aggregate is concerned would you be OK with counting all four, Normalizing CP to GR (whatever is highest) but again NOT normalizing TM and TN? (This is my preference - let market forces decide what's important, and it pretty much ensures that CPs are always 'coveted'.)

or just ignore all of it completely :^)

I'm following a bit loosely here, but when referencing "Pick 2" here, does that actually mean Players choose a setting of which 2 to pick, or the system automatically picks your 2 best categories?

Yertle wrote:

I'm following a bit loosely here, but when referencing "Pick 2" here, does that actually mean Players choose a setting of which 2 to pick, or the system automatically picks your 2 best categories?

Automagical - Players are promoted to the highest rank possible based on their highest two indicators.

I see Ranks as the "Fun" non-stat. It rewards participation of most every kind.  Players can chose to to be well-rounded on all boards (CP/GR)- or specialists on a small number of boards, but with the caveat that they would probably have to play them in tournaments or with teammates (GR/TM and GR/TN).  The only type of player who is not going to see a Rank bump is the single board specialist who only plays regular games, but they could simply adjust to playing in tournaments to gain rank.

Aggregate I see as possibly being more 'seriously' viewed.  If it includes ALL types of play, the top aggregate scorer on this site will have to most likely be a Premium member to participate actively in every type of play.  I can see this being the most respected of all metrics/scores the site has to offer.

M57 wrote:

Ozyman wrote:

+1 to M57s idea.  I like the idea of normalizing CP with GR to average.  I'm not sure about Team & Tourney.

O, Just to clarify..

You DO like the idea of "Pick 2" for Rank purposes, and your preference would be to NOT normalize TM and TN? (I see little harm in normalizing all 4 here. After all, it's only a rank and I think it's good for the site when as many members as possible see themselves getting 'promoted').

And on a related note - As far as an Aggregate is concerned would you be OK with counting all four, Normalizing CP to GR (whatever is highest) but again NOT normalizing TM and TN? (This is my preference - let market forces decide what's important, and it pretty much ensures that CPs are always 'coveted'.)

Yeah I agree with you on all of this, although I think it adds a bit of confusion to have an aggregate, but not have the rank based on the aggregate.   Not saying don't do it, but it feels odd to me.

Ozyman wrote:

 ..although I think it adds a bit of confusion to have an aggregate, but not have the rank based on the aggregate.   Not saying don't do it, but it feels odd to me.

I hear you and acknowledge your point. The key here is if people can understand and agree about the role of the Aggregate vs the Rank.

I can envision the stats being presented kind of like this.

                 TM         TR           CP          GR          P2         Rank        AG 

berickf     2385      2504      1188      2644      2574          ?         2180  

 ATM        1319      1510      2005      2147     2076          ?         1745

Note that in the current system, berickf's CP score puts a serious dent in his Rank, but he still does very well with an aggregate, even with TM and TR not being normalized.

I know I bring this up a lot, but if we do something like the above in M57's post, then I think we need to include Team Tournament, else we are leaving out games.  I realize that Team Tournament stats aren't readily available without clicking on each player individually, but that could be changed.

Cona Chris wrote:

I know I bring this up a lot, but if we do something like the above in M57's post, then I think we need to include Team Tournament, else we are leaving out games.  I realize that Team Tournament stats aren't readily available without clicking on each player individually, but that could be changed.

For some reason, I was thinking that somehow Team Tournament stats would be lumped in with regular Tournament Stats.  Isn't that the way it currently works?

Nope

Well then, There ya go.  We need TM TN TMTN CP (and GR?).  So in line with Hugh and  Babbalouie's suggestions (in one of the other threads), Categories are converted to a percentile, and ranks are displayed for each branch of play on the player's page. The player also receives a site rank that is determined by the highest of their ranks in all categories.

Adding to the Hypotheticals here: If we went with an arrangement like this, should we throw out GR? ..or perhaps combine each category with it's own dedicated GR?

M57 wrote:

Well then, There ya go.  We need TM TN TMTN CP (and GR?).  So in line with Hugh and  Babbalouie's suggestions (in one of the other threads), Categories are converted to a percentile, and ranks are displayed for each branch of play on the player's page. The player also receives a site rank that is determined by the highest of their ranks in all categories.

Adding to the Hypotheticals here: If we went with an arrangement like this, should we throw out GR? ..or perhaps combine each category with it's own dedicated GR?

As suggested earlier, a GR and a CP could be obtained for each area.  But now I'm confusing myself because GR and CP are both ranking the same set of games whereas TM TN and TMTN all are ranking their own subsets via a GR type rank for each only...  So, for each set, do we do the GR scores for TM TN & TMTN and then reverse to do the CP for the standard set... Or, do we do a hybrid CP/GR for each?  It's late so I can't think right now!

berickf wrote:

M57 wrote:

Well then, There ya go.  We need TM TN TMTN CP (and GR?).  So in line with Hugh and  Babbalouie's suggestions (in one of the other threads), Categories are converted to a percentile, and ranks are displayed for each branch of play on the player's page. The player also receives a site rank that is determined by the highest of their ranks in all categories.

Adding to the Hypotheticals here: If we went with an arrangement like this, should we throw out GR? ..or perhaps combine each category with it's own dedicated GR?

As suggested earlier, a GR and a CP could be obtained for each area.  But now I'm confusing myself because GR and CP are both ranking the same set of games whereas TM TN and TMTN all are ranking their own subsets via a GR type rank for each only...  So, for each set, do we do the GR scores for TM TN & TMTN and then reverse to do the CP for the standard set... Or, do we do a hybrid CP/GR for each?  It's late so I can't think right now!

Yeah, you're definitely in a different time zone from those of us in the Americas.  I just posted in the other thread where I think we are.  Hopefully people will weigh in on that..