Babbalouie wrote:

The 4 rank system is the best of all worlds. Players that want to continue to pursue CP's may do so. Players that want to pursue GR may do so.  Players that want to excel in team play may do so. And players that want to be top tournament players may do so. Whatever strikes your fancy. I personally do not care about CP's and it seems obvious that berickf (whom I believe is the best all around player on this site, who now seems to be semi-retired) does not either.

Perhaps some players want to be well rounded and pursue 2, 3, or all 4 areas. Cona Chris even suggested a 5th rank for tournament team play. This multi-rank system will renew competition and stimulate interest.

Tom wanted "Viper" to design the medals, but Viper had been a no show. This has delayed the implementation of the 4 Rank System. Is there someone that can take over for Viper so that this can finally be implemented? It has been a long time coming. Once up and running, modifications may be necessary. 

How many tournaments do you play per year or per month on average, taking into account when they start and when they end?

I think having an yearly championship based on tournament results would bring much more competitive gaming into the table. Players would only be able to count points from their best N tournaments played that year so that the Championship does not become a matter of who has more free time in their hands. Since you play a lot of tournaments, your average could give us a clue what N should be (we would know it has to be less than what you play :) ). This number could be revised at the start of the 2nd season as I guess there would be more tournaments firing each month.

I think other restrictions could be applied, like requiring an M number of different boards to be in the countables of the player, that is, all the N slots couldn't be occupied by tournaments from the same board or from only very few boards.

Tom tried something similar to this a while back. He offered 50 bucks to the player with the highest Monthly rankings.

I think berickf won all of them. 

ratsy wrote:

Tom tried something similar to this a while back. He offered 50 bucks to the player with the highest Monthly rankings.

I think berickf won all of them. 

Hahaha, hardly.  I just went on a short run to see how many I could put up consecutively, but many were won by others before I took notice.

berickf wrote:

ratsy wrote:

Tom tried something similar to this a while back. He offered 50 bucks to the player with the highest Monthly rankings.

I think berickf won all of them. 

Hahaha, hardly.  I just went on a short run to see how many I could put up consecutively, but many were won by others before I took notice.

You did really well there. :)  But the pattern will re-emerge with any 'reward the 1st place' scheme. Someone will come out on top, and will likely be able to stay there. 

You need something that can encourage the middle of the pack, or the bottom end.

ratsy wrote:

Tom tried something similar to this a while back. He offered 50 bucks to the player with the highest Monthly rankings.

I think berickf won all of them. 

How is it similar? Isn't that just based on the performance on the regular games?

How is monthly score calculated? All players are reset to 1000 at the start of the month for this score?

I was thinking we could even run the Championships from the outside if Tom does not implement it. I guess Ozy or someone else could put up a website that crawls the Board Score data from WarGear and calculates the GS with Amidon's formula. The website would serve to keep the Championship leaderboard up-to-date.

Players wanting to participate in the Championship would pay an entry fee of, let's say 10$, which would go to the prize money. Prize money distribution could be something like 50% to 1st place, 30% to 2nd and 20% to 3rd at the end of the Season.

Regarding points awarded from the tournament results it could be something like:

- points to 1st place in a tournament = AGS/100 where AGS is the average GS of the players involved in the tournament. Then points to lower places would scale down from there.

- we could have special tournaments where only players registered in the Championship could take part and award extra points in them.

Abishai wrote:

I would say that the current GR score is pretty straightforward. It just shows how well you do overall on public games. What meaning would you like the GR score to have? Not trying to be a jerk, I'm really trying to understand.

With the current system, a player can be top ranked in GRS by playing only on one board. How does that show his overall skills?

redshift wrote:

Abishai wrote:

I would say that the current GR score is pretty straightforward. It just shows how well you do overall on public games. What meaning would you like the GR score to have? Not trying to be a jerk, I'm really trying to understand.

With the current system, a player can be top ranked in GRS by playing only on one board. How does that show his overall skills?

This is the significant flaw the the current GR score.  It is definitely NOT an indicator of how well you do overall.  That is what CP scores were 'meant' to cover, but CPs has their own set of flaws. The problem with current CP system is that it only counts boards on which you are in the top ten.  So for the casual player the CP system can actually end up penalizing diverse play because those players are unlikely to raise up into the top ranks on any given board.

If I had to pick which system is the least useful, I would say it's the current GR system.  While it can give a somewhat of a snapshot of how well a player has been playing by virtue of how it fluctuates, it is otherwise worthless as an indicator of diversity of play and can be gamed by players who chose to specialize on boards.  This may all seem off topic - but if the current GR system could be replaced with something like (SUM)GR-1000, which recognizes diversity of play, then that frees up the CP system to be something that can truly be about "Champions."  Otherwise much of the discussion that precedes this post is truly off topic because right now, CPs are the only stat the site has that rewards any kind of diverse play.

redshift wrote:

Abishai wrote:

I would say that the current GR score is pretty straightforward. It just shows how well you do overall on public games. What meaning would you like the GR score to have? Not trying to be a jerk, I'm really trying to understand.

With the current system, a player can be top ranked in GRS by playing only on one board. How does that show his overall skills?

It's not intended to represent one's overall skills.  If that is what you are all looking for then it's not about revising CP or GR, or revising any of the other current ranks for that matter.  The only way you could create an overall rank would be to create a new aggregate rank that actually takes all ranks into consideration.  But to talk about that, at least in the past, was like flogging a dead donkey.  It just doesn't go anywhere!

As it is, the other ranks could be 'tweaked', but, they will still largely represent what they represent unless they are completely revamped to no longer be what they previously were, at which point, players who have already excelled at them will raise stink and Tom won't change them.

Right now

GR = unwillingness to lose games

CP = board whore

h+ rating = winning rate

team GR = unwillingness to lose team games

team CP = doesn't exist

individual board GR = unwillingness to lose on that board

Individual board CP = Excellence at a heavily played board or curiosity/luck/figuring out a board and then beating on others still learning it/perhaps excellence at a lesser played board

Then tournament for some of the same.

The reason why I classify GR as an unwillingness to lose is due to the ceiling.  Once a GR is high, a loss will take you back 70-80-90-100 points and you only gain 3-4-5-6 for a win, so wins become nearly meaningless and refusing to lose becomes the standard of your ceiling.  In other words, when first starting out in an GR climb, it's ok to lose and hardly effects your GR so long as you win at a decent rate as well, but, to attain greatness in GR at some point the losses become far too costly and are unacceptable.  The problem with that is when the general GR score becomes insanely high it ends up limiting a player to only playing boards that they 'know' they hardly lose at given the probability of odds, which sucks. 

An aggregate could allow such a player to balance losses to their highest rank with gains to their lowest ranks offsetting each other whereby the overall gains to an aggregate are still positive despite the losses to their top rank(s).  This would apply to everyone that excels in any rank.  They might have to risk one to increase a weaker rank to excel in an aggregate.

Since an aggregate is not necessarily changing any of the component parts, it might be easier to gain acceptance and I have never understood why it hasn't.

berickf wrote:

The reason why I classify GR as an unwillingness to lose is due to the ceiling.  Once a GR is high, a loss will take you back 70-80-90-100 points and you only gain 3-4-5-6 for a win, so wins become nearly meaningless and refusing to lose becomes the standard of your ceiling.  In other words, when first starting out in an GR climb, it's ok to lose and hardly effects your GR so long as you win at a decent rate as well, but, to attain greatness in GR at some point the losses become far too costly and are unacceptable.  The problem with that is when the general GR score becomes insanely high it ends up limiting a player to only playing boards that they 'know' they hardly lose at given the probability of odds, which sucks. 

An aggregate could allow such a player to balance losses to their highest rank with gains to their lowest ranks offsetting each other whereby the overall gains to an aggregate are still positive despite the losses to their top rank(s).  This would apply to everyone that excels in any rank.  They might have to risk one to increase a weaker rank to excel in an aggregate.

Since an aggregate is not necessarily changing any of the component parts, it might be easier to gain acceptance and I have never understood why it hasn't.

@bericf, I like your descriptors.  Your criticism of GR was new to me.  My inclination is to doubt that there's really a GR ceiling if only because I've seen the top score on boards get higher over time, but on the other hand - you of all people speak from first hand experience so you should know.

As for the aggregate, while I agree that one is a good idea, right now it would be an aggregate of the flawed component parts you describe and it would be a shame to develop an aggregate only to later decide to re-design some of it's parts and find that the aggregate formula is no longer valid.

While at first your argument that balancing play might let a player raise their aggregate may be true, in the end, and after you've achieve equilibrium, you end up where started - capped.

I'm OK with a Championship Point system that disproportionally awards the top players on each board PROVIDED that the GR/GS can handle the roll that the old CP system has also been burdened with.  I.e. rewarding diversity of play equitably across all players.  Something like SUM(GR-1000) would do the trick.

Here's a CP model that scales with the high board score. It employs the Fibonnaci/Golden Ratio, computed from the top down and and rounding down. 

Here it is with standard rounding.  I like the way these REALLY reward the #1 player of a board. You wanna see a dog fight for the top spot? I'll bet this'd do it!

With a new and improved GR/GS in place, personally I would prefer to scrap the whole notion that CPs are about public board play, and instead make them about Tournament play.  If this were to come to fruition, there's a certain sense in which I wouldn't be uncomfortable with the current state of the site where Tournament play doesn't count toward GR/(GS).

berickf wrote:

redshift wrote:

Abishai wrote:

I would say that the current GR score is pretty straightforward. It just shows how well you do overall on public games. What meaning would you like the GR score to have? Not trying to be a jerk, I'm really trying to understand.

With the current system, a player can be top ranked in GRS by playing only on one board. How does that show his overall skills?

It's not intended to represent one's overall skills.  If that is what you are all looking for then it's not about revising CP or GR, or revising any of the other current ranks for that matter.  The only way you could create an overall rank would be to create a new aggregate rank that actually takes all ranks into consideration.  But to talk about that, at least in the past, was like flogging a dead donkey.  It just doesn't go anywhere!

Then what does Global mean? I'm in favour of aggregating all public games into GRS as well.

As it is, the other ranks could be 'tweaked', but, they will still largely represent what they represent unless they are completely revamped to no longer be what they previously were, at which point, players who have already excelled at them will raise stink and Tom won't change them.

Tom shouldn't listen to egotists but I understand the fear of losing money. We can allways run alternate scoring sytems outside of WarGear's jurisdiction since we're not proposing changes to how BS is calculated.

The reason why I classify GR as an unwillingness to lose is due to the ceiling.  Once a GR is high, a loss will take you back 70-80-90-100 points and you only gain 3-4-5-6 for a win, so wins become nearly meaningless and refusing to lose becomes the standard of your ceiling.  In other words, when first starting out in an GR climb, it's ok to lose and hardly effects your GR so long as you win at a decent rate as well, but, to attain greatness in GR at some point the losses become far too costly and are unacceptable.  The problem with that is when the general GR score becomes insanely high it ends up limiting a player to only playing boards that they 'know' they hardly lose at given the probability of odds, which sucks. 

That's inherent to the ELO system but I still think it's a great system for the Board Scores. There's allways the option of playing against other high ranked players but that will in mosts cases lead to more losses. In World of Warcraft arena, many players used to stop playing for the rest of a given season once they attained #1 with enough lead on the #2. It was also the case with players that attained Gladiator with a safe margin to the cutoff but knew they weren't good enough to climb more. Rate decay was proposed (that is, if a team goes without playing for a given amount of time their rating starts to go down slowly) but can't remember if it ended up being implemented or not. Anyway, this is not a problem here.

An aggregate could allow such a player to balance losses to their highest rank with gains to their lowest ranks offsetting each other whereby the overall gains to an aggregate are still positive despite the losses to their top rank(s).  This would apply to everyone that excels in any rank.  They might have to risk one to increase a weaker rank to excel in an aggregate.

Since an aggregate is not necessarily changing any of the component parts, it might be easier to gain acceptance and I have never understood why it hasn't.

Exactly, that's what the proposed formula to calculate GRS accomplishes.

berickf wrote:

redshift wrote:

Abishai wrote:

I would say that the current GR score is pretty straightforward. It just shows how well you do overall on public games. What meaning would you like the GR score to have? Not trying to be a jerk, I'm really trying to understand.

With the current system, a player can be top ranked in GRS by playing only on one board. How does that show his overall skills?

It's not intended to represent one's overall skills.  If that is what you are all looking for then it's not about revising CP or GR, or revising any of the other current ranks for that matter.  The only way you could create an overall rank would be to create a new aggregate rank that actually takes all ranks into consideration.  But to talk about that, at least in the past, was like flogging a dead donkey.  It just doesn't go anywhere!

+1.  

I also like berickf's definition of GR as overall unwillingness to lose. I would maybe add my definition of CP would be overall tenacity to conquer maps. 

The fact being that I can go to anyone's profile page and know everything I need to about their play style by just looking at their CP, GR and h rating.

If I look at berickf's stats I see that he maybe sticks to a few favorite maps, but on these maps I do not want to mess with him. He is the shark in those waters and he is very, very good.

If I look at Mad Bomber's stats I see he has put in several years of hard, but likely very entertaining work. I can see that I am likely to have many more run ins with him and on any given game, no matter how big, he is probably my #1 threat.

The difference in these two players made for good examples. My point is that these stats already do a good job at conveying the necessary information. If you want to add an aggregate ranking or change which stat governs overall site ranking, then that is another conversation. I just don't see any need to change GR or CPs into a whole other stat .

So if everything is flawed- you end up with a super flawed aggregate.

Currently GR is a joke. You can't tell a damn thing about my play by looking at my GR, CPs, and H-Rating. My GR fluctuates between 1200 and 2200.  My CPs could be from playing a bunch of boards or specializing on a few.  There's no way to tell.  My GR is a roller coaster and my H-Rating hasn't strayed 2 points in the last 5 years. What does that mean?

A SUM:GS system has been proposed with a built-in aggregate of scores for all boards in a given category of play.  From there you could have a site-wide Aggregate of SUM:GS's for all categories of play if you wanted - tournament, team, public, etc..   Couple that with a CP system that equitably rewards the best players for each board, and then you could look at the GR's and CPs and get a much better picture of a player.

One thing we should all agree on - We will continue to flog that donkey carcass if we can't agree on a way to fix GR.  That is what is going to be the foundation of any aggregate system we end up with.

M57 wrote:

So if everything is flawed- you end up with a super flawed aggregate.

Currently GR is a joke. You can't tell a damn thing about my play by looking at my GR, CPs, and H-Rating. My GR fluctuates between 1200 and 2200.  My CPs could be from playing a bunch of boards or specializing on a few.  There's no way to tell.  My GR is a roller coaster and my H-Rating hasn't strayed 2 points in the last 5 years. What does that mean?

A SUM:GS system has been proposed with a built-in aggregate of scores for all boards in a given category of play.  From there you could have a site-wide Aggregate of SUM:GS's for all categories of play if you wanted - tournament, team, public, etc..   Couple that with a CP system that equitably rewards the best players for each board, and then you could look at the GR's and CPs and get a much better picture of a player.

One thing we should all agree on - We will continue to flog that donkey carcass if we can't agree on a way to fix GR.  That is what is going to be the foundation of any aggregate system we end up with.

Your ranking history begs to differ: http://www.wargear.net/players/info/M57/Ranking%20History

I get that you are just trying to make a point, but maybe stick to the story as it is?  And, your GR tells a great story!  You came out of the blocks with a head of steam and steadily grew to approx 2000gr then went on a spill and declined to hover in a range of 1400-1700 with occasional spikes above and a few dips below, but that general range was held for quite a time.  There was one particularly bad spell after your honeymoon range where you dipped deeply down to 1000 that coincided with a great diversity in board play where previously your range of board play was much less diverse.  So, you got curious about other boards and payed the price.  Playing all over the place hurts GR and specializing helps GR.  After you restricted your board variety again you quickly jumped back up to the 1300-1600 range again.  Then you got really serious about Go-Geared with some Anarchy mixed in and on the back of that specialization pushed your rank up to 2300!  Then, you started to mix in more board diversity again and with that your range decreased again, first holding a range of 1900-2100 and then further declining your range to 1600-1900.  When you started playing a lot of WotR it did you more harm then good at first, but, you steadily bounced back from that and even brought your rank back up while still playing a lot more WotR and got back up to a healthy 1600-1800 range, for the most.  Quantum and Ren came along, but didn't hurt your rank like the initial WotR.  Maybe your lessons from WotR allowed a smoother transition to Ren, and Quantum is a bit more 'standard' in its dynamics and your range increased to 1700-2000 where you have hovered to date.

It is pretty clear moreover that your fluctuations are not 'random' and that you can tell that your changing play style has had a concrete effect on your ups and downs.  You can call it flawed, or, just that it is what it is and coincides with one's style of play.  The same cannot be said for CP because you can get 20 points in many different ways, of varying difficulty, and it is hard to know how a 100 point CP really translates.  One interesting exercise might be to look at the mean, median and mode h ranks of the top page of GR and CP.  I haven't actually done it, but, my gut tells me that you'll find that the CP is far more variable according to the corresponding h rank whereas the gr will be much more narrow in its scope and higher.

So, does a bunch of flawed/semi flawed systems make a super flawed aggregate?  Or, maybe since the flaws lean in different directions they act to dampen the overall effect and make the aggregate more stable and with less flaws?  It's all theoretical I guess, but, I'm not sure if to paint it in one way without looking at the other side of the coin is the right approach either?

M57 wrote:

@bericf, I like your descriptors.  Your criticism of GR was new to me.  My inclination is to doubt that there's really a GR ceiling if only because I've seen the top score on boards get higher over time, but on the other hand - you of all people speak from first hand experience so you should know.

There is a GR ceiling, but, it is tied to the depth of the GR pool.  Lets say you had only two players and one was a wizard and the other atrocious.  The great player won every single game they ever played.  In this extreme example the GR ceiling is 2000, always.  Now lets say a few thousand players joined them, some good, some average, some poor.  Some players came out and lost a few hundred GR into the pool and got discouraged and left never to try and reclaim those lost points.  Two things start happening over time.  The 'average' GR of active players gradually increases and the pool of GR to play for deepens.  With the underlying dynamics changing, the top GR scores (ceilings) slowly increase as well.

Take for instance BoB, one of my favourite boards.  I took a break from WarGear and while I did Babba exploited the pool to extend his ceiling.  The only active players dropped to such low scores that when I tried to chase him down I was now only getting 2 points a game from some players because Babba had already sucked them dry and the chase was futile.  I'm now keeping a keen eye on sirdakka because while he is a decent player and can take an occasional game from me, I think when he gets to 1800ish points I can gain from him more then I'll lose and resume my chase on the title.  The pool is always in flux and you have to know how to read it and when is the right time to dive into it.  If 2-3 more players would also jump into the mix there and grow their GR's up to 1500-1600-1700 depending on their relative skill levels then the pool could be even deeper and the ceiling a little higher.  The other end of the spectrum, WarGear Warfare has 1000's of player winning and losing GR in a huge pool and because of that the GR ceiling constantly gets higher over time, but, it is also a harder board to specialize on as well making the title a little more difficult to hold once gotten.  Still, the ceiling gradually rises over time given the underlying dynamics of the pool.

Hope that makes some sense?

redshift wrote:

Then what does Global mean? I'm in favour of aggregating all public games into GRS as well.

 

Applies to all players irregardless of how/what they play?

Tom shouldn't listen to egotists but I understand the fear of losing money. We can allways run alternate scoring sytems outside of WarGear's jurisdiction since we're not proposing changes to how BS is calculated.

 

He's just conservative like that.  I don't think it is about money or even to listening to 'egoists'.  More about initial development of the site and sticking to it + not wanting to put as much work into the site anymore?  I think...

That's inherent to the ELO system but I still think it's a great system for the Board Scores. There's allways the option of playing against other high ranked players but that will in most cases lead to more losses. In World of Warcraft arena, many players used to stop playing for the rest of a given season once they attained #1 with enough lead on the #2. It was also the case with players that attained Gladiator with a safe margin to the cutoff but knew they weren't good enough to climb more. Rate decay was proposed (that is, if a team goes without playing for a given amount of time their rating starts to go down slowly) but can't remember if it ended up being implemented or not. Anyway, this is not a problem here.

 

See my other comment about BoB.  It is about observing the conditions of the 'pool' and reacting accordingly.  On some boards I relish the idea of playing the best players, but, they need to actually build their rank up to make it equitable for both of us.  Playing a 'good' player with 1000 points and nothing to lose is just a suckers game.

Exactly, that's what the proposed formula to calculate GRS accomplishes.

 

But it has less of a chance of actually being done except maybe as you propose, by players who post it outside of WarGear, but then if it is not reporting within WarGear thus it is merely an example to convince Tom to potentially incorporate and very little beyond that. Most players will not follow it or give much credence to it if it is not affecting them within the WarGear mechanics.  I think that if a simple and clean aggregate could be created it would be very little code to incorporate and could gain much of what this talk is designed to do - to get players to reach out in different directions.

 

Abishai wrote:

berickf wrote:

redshift wrote:

Abishai wrote:

I would say that the current GR score is pretty straightforward. It just shows how well you do overall on public games. What meaning would you like the GR score to have? Not trying to be a jerk, I'm really trying to understand.

With the current system, a player can be top ranked in GRS by playing only on one board. How does that show his overall skills?

It's not intended to represent one's overall skills.  If that is what you are all looking for then it's not about revising CP or GR, or revising any of the other current ranks for that matter.  The only way you could create an overall rank would be to create a new aggregate rank that actually takes all ranks into consideration.  But to talk about that, at least in the past, was like flogging a dead donkey.  It just doesn't go anywhere!

+1.  

I also like berickf's definition of GR as overall unwillingness to lose. I would maybe add my definition of CP would be overall tenacity to conquer maps. 

The fact being that I can go to anyone's profile page and know everything I need to about their play style by just looking at their CP, GR and h rating.

If I look at berickf's stats I see that he maybe sticks to a few favorite maps, but on these maps I do not want to mess with him. He is the shark in those waters and he is very, very good.

If I look at Mad Bomber's stats I see he has put in several years of hard, but likely very entertaining work. I can see that I am likely to have many more run ins with him and on any given game, no matter how big, he is probably my #1 threat.

The difference in these two players made for good examples. My point is that these stats already do a good job at conveying the necessary information. If you want to add an aggregate ranking or change which stat governs overall site ranking, then that is another conversation. I just don't see any need to change GR or CPs into a whole other stat .

You pretty much made the case of why the current system is so bad. Top players avoid each other, that's just wrong.

If the holy grail of Wargear was a Championship based on tournaments, top players would face off more often.

berickf wrote:

M57 wrote:

So if everything is flawed- you end up with a super flawed aggregate.

Currently GR is a joke. You can't tell a damn thing about my play by looking at my GR, CPs, and H-Rating. My GR fluctuates between 1200 and 2200.  My CPs could be from playing a bunch of boards or specializing on a few.  There's no way to tell.  My GR is a roller coaster and my H-Rating hasn't strayed 2 points in the last 5 years. What does that mean?

A SUM:GS system has been proposed with a built-in aggregate of scores for all boards in a given category of play.  From there you could have a site-wide Aggregate of SUM:GS's for all categories of play if you wanted - tournament, team, public, etc..   Couple that with a CP system that equitably rewards the best players for each board, and then you could look at the GR's and CPs and get a much better picture of a player.

One thing we should all agree on - We will continue to flog that donkey carcass if we can't agree on a way to fix GR.  That is what is going to be the foundation of any aggregate system we end up with.

Your ranking history begs to differ: http://www.wargear.net/players/info/M57/Ranking%20History

I get that you are just trying to make a point, but maybe stick to the story as it is?  And, your GR tells a great story!  You came out of the blocks with a head of steam and steadily grew to approx 2000gr then went on a spill and declined to hover in a range of 1400-1700 with occasional spikes above and a few dips below, but that general range was held for quite a time.  There was one particularly bad spell after your honeymoon range where you dipped deeply down to 1000 that coincided with a great diversity in board play where previously your range of board play was much less diverse.  So, you got curious about other boards and payed the price.  Playing all over the place hurts GR and specializing helps GR.  After you restricted your board variety again you quickly jumped back up to the 1300-1600 range again.  Then you got really serious about Go-Geared with some Anarchy mixed in and on the back of that specialization pushed your rank up to 2300!  Then, you started to mix in more board diversity again and with that your range decreased again, first holding a range of 1900-2100 and then further declining your range to 1600-1900.  When you started playing a lot of WotR it did you more harm then good at first, but, you steadily bounced back from that and even brought your rank back up while still playing a lot more WotR and got back up to a healthy 1600-1800 range, for the most.  Quantum and Ren came along, but didn't hurt your rank like the initial WotR.  Maybe your lessons from WotR allowed a smoother transition to Ren, and Quantum is a bit more 'standard' in its dynamics and your range increased to 1700-2000 where you have hovered to date.

It is pretty clear moreover that your fluctuations are not 'random' and that you can tell that your changing play style has had a concrete effect on your ups and downs.  You can call it flawed, or, just that it is what it is and coincides with one's style of play.  The same cannot be said for CP because you can get 20 points in many different ways, of varying difficulty, and it is hard to know how a 100 point CP really translates.  One interesting exercise might be to look at the mean, median and mode h ranks of the top page of GR and CP.  I haven't actually done it, but, my gut tells me that you'll find that the CP is far more variable according to the corresponding h rank whereas the gr will be much more narrow in its scope and higher.

So, does a bunch of flawed/semi flawed systems make a super flawed aggregate?  Or, maybe since the flaws lean in different directions they act to dampen the overall effect and make the aggregate more stable and with less flaws?  It's all theoretical I guess, but, I'm not sure if to paint it in one way without looking at the other side of the coin is the right approach either?

And the players who played against him, oftentimes gained or lost GRS disproportionaly to the caliber of player that M57 is. GRS should be less volatile to the process of learning more maps.

redshift wrote:

Abishai wrote:

berickf wrote:

redshift wrote:

Abishai wrote:

I would say that the current GR score is pretty straightforward. It just shows how well you do overall on public games. What meaning would you like the GR score to have? Not trying to be a jerk, I'm really trying to understand.

With the current system, a player can be top ranked in GRS by playing only on one board. How does that show his overall skills?

It's not intended to represent one's overall skills.  If that is what you are all looking for then it's not about revising CP or GR, or revising any of the other current ranks for that matter.  The only way you could create an overall rank would be to create a new aggregate rank that actually takes all ranks into consideration.  But to talk about that, at least in the past, was like flogging a dead donkey.  It just doesn't go anywhere!

+1.  

I also like berickf's definition of GR as overall unwillingness to lose. I would maybe add my definition of CP would be overall tenacity to conquer maps. 

The fact being that I can go to anyone's profile page and know everything I need to about their play style by just looking at their CP, GR and h rating.

If I look at berickf's stats I see that he maybe sticks to a few favorite maps, but on these maps I do not want to mess with him. He is the shark in those waters and he is very, very good.

If I look at Mad Bomber's stats I see he has put in several years of hard, but likely very entertaining work. I can see that I am likely to have many more run ins with him and on any given game, no matter how big, he is probably my #1 threat.

The difference in these two players made for good examples. My point is that these stats already do a good job at conveying the necessary information. If you want to add an aggregate ranking or change which stat governs overall site ranking, then that is another conversation. I just don't see any need to change GR or CPs into a whole other stat .

You pretty much made the case of why the current system is so bad. Top players avoid each other, that's just wrong.

If the holy grail of Wargear was a Championship based on tournaments, top players would face off more often.

Huh? Top players avoid each other?  Where do you get that from?  I think you are mixing up the circumstance of player preference for something else.  Simply put, it is not about the rank that chases players apart, but about game preference that dictates who plays who.  As my GR got higher I realized it made me too much of a target in multiplayer games, so, since I am pretty confident in my abilities in 1v1 situations, I focused on that.  Mad bomber likes collecting CP and the best way to do that is to play large player games on new/lesser played boards after 'figuring out' a board as best you can and beating others that might still perhaps be figuring it out.  You will lose some due to all the luck factors built into that equation, which is not great for one's GR, but he does pretty well for himself.  His preference vs my preference mean we don't play each other much (except for DEV games, I encounter him there all the time).  But it is not avoidance, but preference.  Don't mistake the two.  But, for other top players who like 1v1's, I have played with them severally due to the matching preference we share.  I have never avoided CC, Yuriz, Toto, Andernut, Luieuil, Babbalouie, Bex Valeur, Chele Nica, vult, Vyro, fiverocketcars, amongst other players who enjoy 1v1s whenever the chance arises.  I have even coached many of these same players, among others, to be better and up the competition level.  My record against these players speaks for itself.

Even this whole tournament thing.  I'll excel at 1v1 tournaments and MB will excel on multi player tournaments.  Preference will still dictate on which battlefields players meet and how often.

berickf wrote:

redshift wrote:

Abishai wrote:

berickf wrote:

redshift wrote:

Abishai wrote:

I would say that the current GR score is pretty straightforward. It just shows how well you do overall on public games. What meaning would you like the GR score to have? Not trying to be a jerk, I'm really trying to understand.

With the current system, a player can be top ranked in GRS by playing only on one board. How does that show his overall skills?

It's not intended to represent one's overall skills.  If that is what you are all looking for then it's not about revising CP or GR, or revising any of the other current ranks for that matter.  The only way you could create an overall rank would be to create a new aggregate rank that actually takes all ranks into consideration.  But to talk about that, at least in the past, was like flogging a dead donkey.  It just doesn't go anywhere!

+1.  

I also like berickf's definition of GR as overall unwillingness to lose. I would maybe add my definition of CP would be overall tenacity to conquer maps. 

The fact being that I can go to anyone's profile page and know everything I need to about their play style by just looking at their CP, GR and h rating.

If I look at berickf's stats I see that he maybe sticks to a few favorite maps, but on these maps I do not want to mess with him. He is the shark in those waters and he is very, very good.

If I look at Mad Bomber's stats I see he has put in several years of hard, but likely very entertaining work. I can see that I am likely to have many more run ins with him and on any given game, no matter how big, he is probably my #1 threat.

The difference in these two players made for good examples. My point is that these stats already do a good job at conveying the necessary information. If you want to add an aggregate ranking or change which stat governs overall site ranking, then that is another conversation. I just don't see any need to change GR or CPs into a whole other stat .

You pretty much made the case of why the current system is so bad. Top players avoid each other, that's just wrong.

If the holy grail of Wargear was a Championship based on tournaments, top players would face off more often.

Huh? Top players avoid each other?  Where do you get that from?  I think you are mixing up the circumstance of player preference for something else.  Simply put, it is not about the rank that chases players apart, but about game preference that dictates who plays who.  As my GR got higher I realized it made me too much of a target in multiplayer games, so, since I am pretty confident in my abilities in 1v1 situations, I focused on that.  Mad bomber likes collecting CP and the best way to do that is to play large player games on new/lesser played boards after 'figuring out' a board as best you can and beating others that might still perhaps be figuring it out.  You will lose some due to all the luck factors built into that equation, which is not great for one's GR, but he does pretty well for himself.  His preference vs my preference mean we don't play each other much (except for DEV games, I encounter him there all the time).  But it is not avoidance, but preference.  Don't mistake the two.  But, for other top players who like 1v1's, I have played with them severally due to the matching preference we share.  I have never avoided CC, Yuriz, Toto, Andernut, Luieuil, Babbalouie, Bex Valeur, Chele Nica, vult, Vyro, fiverocketcars, amongst other players who enjoy 1v1s whenever the chance arises.  I have even coached many of these same players, among others, to be better and up the competition level.  My record against these players speaks for itself.

Even this whole tournament thing.  I'll excel at 1v1 tournaments and MB will excel on multi player tournaments.  Preference will still dictate on which battlefields players meet and how often.

There was a suggestion once about obscuring player identities in games and a few scenarios of how that could be implemented.  If it were possible to 'remove' the big cross hairs that come along with a high GR (or a high CP), I could see myself testing out those waters and seeing if it worked or if players just figured it out anyways (chat) and re-attached that target.  Anonymity could go a long way to branching out to many different styles of play.

I think the best way would be to make it Anonymous right from the game lobby so one doesn't even know who they are in the game with to start with.