It turns out that after reading and participating in discussions about CPs, I've come to the point where I believe a more pressing concern is the state current GR system. CPs may be in need of an update, but at least they're still functional and recognize and reward diversity of play. The GR system on the other hand, is much more gameable, too unstable and it promotes board specialization, which is something most of us on this site would agree is undesirable. Moreover, if the current GR system was re-thought and was functionally more egalitarian and encompassing, that would free up CPs to be about identifying Champions ..the cream of the crop. Depending on the breadth and strength of the GR system, CPs could be as independent of GR and exclusive as folks would like.

Put another way, currently both the GR and CP systems are rooted in the GS system, one of which (the GR) is in dire need of a re-think. If the GR could be a strong and truer indicator of strength and also reward diverse play, the CP system could potentially stand on its own - perhaps with no need for GS underpinnings.

Here is the current most discussed proposal on the table for the new GR, which I think we should call the GS. After all, it's a score and shouldn't be confused with a Ranking.

Formerly known as Option I, where GS is the current GR of every board a player has played. Starting in it's purest form and listing some variants:

1. SUM(GS-1000): The score would start at 0 and can be negative because play on ALL boards counts.
2. {SUM(GS-1000)}+1000: The score would start at 1000 (which might be more palatable for new or struggling players) and can theoretically be negative, but reaching 0 and going negative would be quite rare.
3. {SUM(GS-1000)}/C:  (It may include or omit boards with negative GRs) This is currently the most commonly discussed version of Option I because the inclusion of a constant brings the numbers down and more in-line with current CP scores.
4. Any combination of the above ideas

__________________

I'm not a fan of this idea, at least not to replace GR: I just came up with it and haven't really given too much thought about it, but [SUM(GS)]/#OfBoardsPlayed is an interesting concept. It would not be the same as straight SUM(GS-1000) in that it discourages diverse play and would still be gameable by specialists, but it would reflect the overall indicator of a player's strength and would be much a more stable score than the current GR.  So playing with it a bit..

[SUM(GS)]/(#OfBoardsPlayed+C)  where C is the constant of your choice would make it profitable to play more boards, with diminishing returns the more boards you play, but returns nonetheless. I'm not a big fan of using an arbitrary constant that has an effect on scale, but again - the concept is interesting.

{SUM(GS-1000)}/C:  (It may include or omit boards with negative GRs) This is currently the most commonly discussed version of Option I because the inclusion of a constant brings the numbers down and more in-line with current CP scores.

I'm a big fan of this. 

Mainly because it gives new people a score that they can actively influence by winning on a variety of boards and it has a "achievement" feel to it.  Currently to earn CP's on popular boards you need to win a lot of games.  The current GR system is not an achievement-like.  It's more like the outside temperature that goes up and down with the seasons.

I would divide by 50 like the current CP's and omit negatives.

Amidon37 wrote:

{SUM(GS-1000)}/C

I'm a big fan of this..

..I would divide by 50 like the current CP's and omit negatives..

If it were a CP alternative I would agree with this, but CPs are about achievement. As a replacement for GR, omitting negatives simply messes with the integrity and intent of the GR/GS system.  We should strive to make GR/GS the best indicator the site has of assessing players' abilities across ALL boards - not just the ones they are good at. Let's face reality - The CP crowd wants to crown Champions, and many want to make it even clearer with disproportionate scoring ratios for top positions, etc. CPs should be a great game inside the game the site has for finding the best and giving the site its "champions," but if its anywhere near what I described above it will never (and shouldn't) be the goto stat for gauging ability and rewarding diversity of play, etc.

So back to {SUM(GS-1000)}/C. Remember, C=50 was just an arbitrary constant we pulled out of our butts that brought things in line with the current CP numbers. C=60 would likely be even closer. Though I don't feel as strongly about the value of the constant (it doesn't change the integrity of the number one iota), I would prefer C=1, I.e. just plain old SUM(GS-1000). I'm a big fan of pure, simple and easy to interpret.

Yes the numbers with C=1 would be large compared to what we're use to, but they would be much easier to interpret. A GS of 10,000 would represent the equivalent of having a 2000 GS on 10 boards  (i.e, +1000/per board), or a GS of 1200 on 50 boards.  Let the CP people worry about replacing/fixing the CP system such that its numbers look similar to the current CP system. Not only do I see no need to do so in a GS system, but I think it would be confusing if the top GS numbers end up looking like CP numbers.

M57 wrote:

It turns out that after reading and participating in discussions about CPs, I've come to the point where I believe a more pressing concern is the state current GR system. CPs may be in need of an update, but at least they're still functional and recognize and reward diversity of play. The GR system on the other hand, is much more gameable, too unstable and it promotes board specialization, which is something most of us on this site would agree is undesirable. Moreover, if the current GR system was re-thought and was functionally more egalitarian and encompassing, that would free up CPs to be about identifying Champions ..the cream of the crop. Depending on the breadth and strength of the GR system, CPs could be as independent of GR and exclusive as folks would like.

Put another way, currently both the GR and CP systems are rooted in the GS system, one of which (the GR) is in dire need of a re-think. If the GR could be a strong and truer indicator of strength and also reward diverse play, the CP system could potentially stand on its own - perhaps with no need for GS underpinnings.

This, so much. Summarises my thoughts to perfection. :worship:

Regarding CPs, in my Championship proposal, the CPs awarded to each place in a given tournament would be related to the average GS of the players involved so that points would be awarded in accordance to the strength of the field. But it could be based on Tournament Score instead. I'd rather that tournament games would count for BS (therefore counting for GS with the new formula) and that Tournament Score would be abolished. In this case, CPs could also be awarded as a function of BS instead of GS since a given tournament is played on a given board, but whatever, let's go one step at a time.

Regarding GS, I like SUM[(BS-1000)/C]+1000 with C=50 or whatever makes it similar to board scores. I think it would be more confortable to the players since they're used to that range of scores, but any of the formulas are fine to me. The only thing I feel strong about is that, whatever the formula, it should count the boards where BS is lower than 1000.

Btw, another alternative formula is the average function.

M57 wrote:

Here is the current most discussed proposal on the table for the new GR, which I think we should call the GS. After all, it's a score and shouldn't be confused with a Ranking.

Formerly known as Option I, where GS is the current GR of every board a player has played. Starting in it's purest form and listing some variants:

1. SUM(GS-1000): The score would start at 0 and can be negative because play on ALL boards counts.
2. {SUM(GS-1000)}+1000: The score would start at 1000 (which might be more palatable for new or struggling players) and can theoretically be negative, but reaching 0 and going negative would be quite rare.
3. {SUM(GS-1000)}/C:  (It may include or omit boards with negative GRs) This is currently the most commonly discussed version of Option I because the inclusion of a constant brings the numbers down and more in-line with current CP scores.
4. Any combination of the above ideas

__________________

I'm not a fan of this idea, at least not to replace GR: I just came up with it and haven't really given too much thought about it, but [SUM(GS)]/#OfBoardsPlayed is an interesting concept. It would not be the same as straight SUM(GS-1000) in that it discourages diverse play and would still be gameable by specialists, but it would reflect the overall indicator of a player's strength and would be much a more stable score than the current GR.  So playing with it a bit..

[SUM(GS)]/(#OfBoardsPlayed+C)  where C is the constant of your choice would make it profitable to play more boards, with diminishing returns the more boards you play, but returns nonetheless. I'm not a big fan of using an arbitrary constant that has an effect on scale, but again - the concept is interesting.

Some sort of average of Board Scores to calculate GRS seems reasonable.

If we were to change Board Score to aggregate of all formats, would the following formula be the right way to do it, without having to recalculate everything from the start:

BS = (M_RNBS_RN + M_RTBS_RT + M_TPNBS_TPN + M_TPTBS_TPT)/(M_RN + M_RT + M_TPN + M_TPT),

where M is the number of games played in the associated format, R is Regular, TP is Teamplay, N is Normal and T is Tournament?

One for Public and one for Private, and GS would use the public one, ofc.