M57 wrote:

A corollary to that might be that best way for a 2000 kind of player to get to 2500 or higher is to play good players.  ..and what's wrong with that?

 

Hey M,

You are correct that a higher ranked player should play those near his ranking.  However, this is nearly impossible in a practical sense.  Take my case for example, I have 1900+ rate for Wargear Warfare.  If I start a game, anyone can join and the odds are that they will be 1200 or under.  This leaves me in a situation of needing to win 75% of my games to break even.  Because 1v1 Wargear Warfare is greatly influenced by luck, it is not possible (imo) to achieve that win % to increase my point total.

So, I am left with the approach you suggested.  Play those with higher rankings.  To do that I have to hang out and wait for someone with a high ranking to start a game - which is not practical because it doesn't happen very often.  If you look at my game history you will see a significant drop off in the 1v1 games I haved played recently for this reason.

A solution to this, that I have suggested before, is to allow someone to start a game and filter opponents based on ratings (+ / - 400 for example).  Although  some will say this defeats the spirit of 'open' competition, I disagree.  It has the same effect of me waiting to play a higher rated player who starts a game.  In addition, since I am filtering players of higher rating, the system does not encourage point farming (even though as I listed above I don't think that can really happen). 

SquintGnome wrote:

Hey M,

You are correct that a higher ranked player should play those near his ranking.  However, this is nearly impossible in a practical sense.  Take my case for example, I have 1900+ rate for Wargear Warfare.  If I start a game, anyone can join and the odds are that they will be 1200 or under.  This leaves me in a situation of needing to win 75% of my games to break even.  Because 1v1 Wargear Warfare is greatly influenced by luck, it is not possible (imo) to achieve that win % to increase my point total.

But isn't everyone up against the same obstacles?  I see a fair playing field. A global rating of 2000+ should be really, really difficult to maintain. I know I can't. 

So, I am left with the approach you suggested.  Play those with higher rankings.  To do that I have to hang out and wait for someone with a high ranking to start a game - which is not practical because it doesn't happen very often.

I don't know this is true -- I'm currently rated around 2000 and I start most of my games.  Granted, I start them mostly on my own boards ..because I like them, but anyone can join them.  

If you look at my game history you will see a significant drop off in the 1v1 games I haved played recently for this reason.

As it should be.. If GR is that important to you, don't play 1v1 games, or at least not ones where dice and luck of the draw will bring you down too often.

I guess in my old age, I'm becoming less and less concerned about what my own ratings are, but more interested in how accurate they are.

A solution to this, that I have suggested before, is to allow someone to start a game and filter opponents based on ratings (+ / - 400 for example).  Although  some will say this defeats the spirit of 'open' competition, I disagree.  It has the same effect of me waiting to play a higher rated player who starts a game.  In addition, since I am filtering players of higher rating, the system does not encourage point farming (even though as I listed above I don't think that can really happen). 

Think about what that would do to your H-rating ..which I tend to value more than Global Rating.  Personally, I enjoy playing against a nice wide mix of opponents.

ratsy wrote:

So yes, the games are very diverse, but I think the really high score shows a level of engagement and probably an amount of caring put into the games. 

Just to clarify: All of these measures do successfully filter out large groups. The high scorers do have regular success against the lower scorers, regardless of category. So yes, there is something the scores are doing right. And yes, we win because of engagement, effort, skill, or whatever.

The criticisms of each type of score given so far become apparent when you go to the fine level of detail comparing the players within that top group.

Ozyman wrote:

3ET:

Lastly, Wargear seems to give a lot more respect to CPs, because that is what decides your rank under your avatar.  I agree with Toto that the rank should be computed from more variables. 

How about until we get a comprehensive achievement system, players can choose what ranking they want under their avatar.  So someone who plays a ton of team tournaments, and is really good at that, can be proud of their niche and get a "Team Tournaments: #1"

 

I would be supportive of that for sure.  Though it seems like it shouldn't be that hard to come up with a unified site ranking that factors in CPs, GR, and H rating without getting anyone too upset.  I would give the most weight to CPs, but not so much that it drowns everything else out.  I know Tom has a lot on his plate and maybe it isn't too important as of now.

Hugh wrote:

 different boards lend themselves to different winning efficiencies. Mine rises noticeably when I choose certain types of games over others.

With the strict point of being argumentative: I'd propose a counter point that maybe it's happening because you just happen to be good at that particular board.  Since the points won/lost match on every board, how does a board lend itself to giving you more points if not because you're just better than the average player at that board?

M57 wrote:

A global rating of 2000+ should be really, really difficult to maintain.

+1 x 30
(Not to pick on you Squint, because you're by no means the only one who has complained about it, you just happen to be the one who brought it up here...)

1) If a player is unhappy with the possibility of losing points in any ranking....stop playing.  This is a game.  That player might lose one round to some 900 point player, but if they're as good as their score suggests, they will earn those points back in spades if they keep playing that same player. (assuming that opponent's 900 ranking isn't inaccurately low)

2) If luck is playing too much of a role in the boards a player is playing, and that player fears the effect on their rankings, then stop playing those boards.

3) And the same is true (as #2) with 1v1 games.  To paraphrase BlackDog (who does pretty well in all rankings): Play against a bunch of players.  It removes the effects of luck in many games, and the point gain could easily outweigh the point loss to a lower seeded player.

4) Just because one player has a ranking of 2000, doesn't mean that there should be rules to benefit that player to be able to hold onto that 2000 ranking.  That player doesn't deserve that ranking, and (I feel) that in no way should this site cater to making it easier for one to hold that ranking.  Which leads back to M57's comment above....

Hey Boris,

You say above:

1) If a player is unhappy with the possibility of losing points in any ranking....stop playing.  This is a game.  That player might lose one round to some 900 point player, but if they're as good as their score suggests, they will earn those points back in spades if they keep playing that same player. (assuming that opponent's 900 ranking isn't inaccurately low)

This is not true for some boards.  In some cases, like 1v1 Wargear Warfare, luck is a significant influence, so continuing to play a low ranked player will continue to result in a loss of points, even if you're a 'legitimate' 1800 playing going against a 'legitimate' 900 player.  You need to win 80% of the games to break even, and no matter how good you are you can't do that. 

To illustrate a little more for clarification.  If you can imagine two players with skills as noted above and you set them up in an environment where luck is not a factor for a certain board - the higher rated player may win 95% of the time.  Now, have them play 1v1 Wargear Warfare, and the higher rated player will only win 65-70% of the time because of the strong influence of luck.  I am not complaining about the effect of luck, only noting that the current system allows better players to lose points in certain situations that brings them below their 'equilibrium' ranking that everyone talks about.

Which brings us to your second point where you recommend to stop playing those boards.  You are correct that that is a solution and I have done that for the most part with Wargear Warfare 1v1.  Wouldn't it be better though, if there was a solution where I didn't have to stop playing my favorite board and scenario?  That solution is allowing an opponent ranking filter.  There is no down side and only an upside (imo) - so why not do it?

BorisTheFrugal wrote:

Hugh wrote:

 different boards lend themselves to different winning efficiencies. Mine rises noticeably when I choose certain types of games over others.

With the strict point of being argumentative: I'd propose a counter point that maybe it's happening because you just happen to be good at that particular board.  Since the points won/lost match on every board, how does a board lend itself to giving you more points if not because you're just better than the average player at that board?

I think what you're asserting is that the boards that bring me down are the ones I'm not as good at while the boards that bring me up are the ones I'm good at. If that's really true, it still presents a problem for GR: A player maximizes GR by playing the games he or she is most skilled at.

But really, the boards themselves don't compare very well under this system. Copying SquintGnome, suppose a player is nearly perfectly skilled at both 1v1 A&A and 1v1 WGWF. The win percentages against a poor player, a middling player, and a good player, will be different between the two games. If the player is concerned about GR, there will be a clear choice of which game to play. This difference of win percentages given equally skilled play, in my experience, influences GR. GR is a noisy frankenstat - it never should have attained the prestige it apparently has.

Frankenstat? lol

The only real measure I see is to pick 10 boards which are as different as can be. The highest average rating on those boards is your official ranking.

Championchip rankings doesnt make any sence, since there are so many boards anyone can score high as long as time is invested. Ofcourse good players will accumulate points faster, but still it;s mainly a measure of invested time/games.

Global ranking is also not perfect, since it only measures how good you are at your best boards. Since top ranked global players do usually not play many boards they are not succesfull at.

So by choosing 10 new championchip boards with complete different styles we can compute a ranking which acts like a global ranking on 10 different field of play. It still can be gamed, but everyone can use the same tactics and new players can immediately compete as it takes a fewer games to get involved. Which I think is important. To bind new player and get them motivated they should have a tailable target.

Luieuil wrote:

The only real measure I see is to pick 10 boards which are as different as can be. The highest average rating on those boards is your official ranking.

I agree that something along these lines would be the only way to have a legitimate competition for something like "overall best player". Another previously mentioned alternative was tournaments among top players that function as something like the "Grand Slam" events of Tennis and Golf.

"It still can be gamed": If it's the same system and averaging, then yes, yes it can be! And effort would still play a large large role. But, there do exist modern, mature, ranking systems that could handle something like this. (TrueSkill comes to mind.)

Hugh wrote:

Luieuil wrote:

The only real measure I see is to pick 10 boards which are as different as can be. The highest average rating on those boards is your official ranking.

I agree that something along these lines would be the only way to have a legitimate competition for something like "overall best player". Another previously mentioned alternative was tournaments among top players that function as something like the "Grand Slam" events of Tennis and Golf.

"It still can be gamed": If it's the same system and averaging, then yes, yes it can be! And effort would still play a large large role. But, there do exist modern, mature, ranking systems that could handle something like this. (TrueSkill comes to mind.)

I immediately thought of Grand Slam events as well.  Isn't there also something like this is pretty much any ranked sport?  I think table tennis and chess also have a certain # of "sanctioned" events that players need to compete in to maintain global ranking.