Ghost Rating Challenge

Advertise new games you are hosting, look for new games to join, or find other players to join your invitation-only games here.
Forum rules
This area is meant for advertising your new games or for finding new games to join with other players on the forum. Please do not make multiple threads for your game or spam existing threads. Please be civil, respect other members, and follow all of the webDiplomacy Forum Rules.
Post Reply
Message
Author
Dioptre
Posts: 19
Joined: Sat Mar 28, 2020 8:10 am
Karma: 8
Contact:

Re: Ghost Rating Challenge

#121 Post by Dioptre » Sun Sep 13, 2020 5:50 am

Wusti,
If I understand the effect of Ghost Ratings correctly, playing in random public games, your expected outcome according to the Ghost Rating might be better than a 4-way draw, but your actual expected outcome based on skill & stochastic variation might be a 4 way draw or worse.

So unless you strategically pick your opponents, yes, you could find that no matter what you do your Ghost Rating tanks.

In a tournament matched on Ghost Rating, the expected outcome approaches a 7 way draw. So if you believe your skill creates a better expected outcome than that (i.e. if your actual skill is higher than your Ghost Rating), your ranking is likely to go up.

This is a tournament ideal for people who think that Ghost Rating under-estimates their skill.

Szpoti
Site Moderator
Site Moderator
Posts: 456
Joined: Sun Dec 31, 2017 1:02 pm
Location: Glasgow, Scotland
Karma: 133
Contact:

Re: Ghost Rating Challenge

#122 Post by Szpoti » Sun Sep 13, 2020 6:15 am

Are we still going to use the ranking from August?

Szpoti
Site Moderator
Site Moderator
Posts: 456
Joined: Sun Dec 31, 2017 1:02 pm
Location: Glasgow, Scotland
Karma: 133
Contact:

Re: Ghost Rating Challenge

#123 Post by Szpoti » Sun Sep 13, 2020 6:20 am

Wusti wrote:
Sun Sep 13, 2020 5:30 am
Now at 7000+ and counting.
Wusti, you're at 1072 according to the Sept ranking.

Wusti
Posts: 399
Joined: Thu Oct 19, 2017 10:12 pm
Karma: 232
Contact:

Re: Ghost Rating Challenge

#124 Post by Wusti » Sun Sep 13, 2020 6:25 am

Dioptre wrote:
Sun Sep 13, 2020 5:50 am
Wusti,
If I understand the effect of Ghost Ratings correctly, playing in random public games, your expected outcome according to the Ghost Rating might be better than a 4-way draw, but your actual expected outcome based on skill & stochastic variation might be a 4 way draw or worse.

So unless you strategically pick your opponents, yes, you could find that no matter what you do your Ghost Rating tanks.

In a tournament matched on Ghost Rating, the expected outcome approaches a 7 way draw. So if you believe your skill creates a better expected outcome than that (i.e. if your actual skill is higher than your Ghost Rating), your ranking is likely to go up.

This is a tournament ideal for people who think that Ghost Rating under-estimates their skill.
So me playing against Top 100 players should yield a ranking bump no matter what, given the formula used and gap in GR. However that is not the case.

Any ranking system that REQUIRES gaming is fucked. You made my case.

Szpoti - not on Classic FP apparently.

Dioptre
Posts: 19
Joined: Sat Mar 28, 2020 8:10 am
Karma: 8
Contact:

Re: Ghost Rating Challenge

#125 Post by Dioptre » Sun Sep 13, 2020 7:52 am

Wusti, I don't think it counts as "requires gaming" for a ranking system to encourage you to play against players at your own level, rather than weaker players. That's literally how almost every sports ranking system in the world works. If a star tennis player declined invitations to the majors, and instead played at their local club, their ranking would go down.

And if your ranking is currently incorrect, yes, playing against better players gives you an expected (note, not "no matter what", just on average) increase. Again, that's the same as something like tennis. If a lowly ranked player makes it to a major tournament, their average ranking change will be positive.

The only problem with Ghost Ranking is that there's no corresponding matchmaking system. So you have to pick the matches yourself rather than get invited to sensible matches to play in.
1

CSteinhardt
Bronze Donator
Bronze Donator
Posts: 13
Joined: Mon Jan 01, 2018 8:16 pm
Karma: 32
Contact:

Re: Ghost Rating Challenge

#126 Post by CSteinhardt » Sun Sep 13, 2020 10:01 am

What you're describing is valid, and is an artifact of two flaws in the GR system.

First, the math behind the system makes a specific, but incorrect assumption about what the skill curve looks like in Diplomacy. For example, in chess, if a 1700 plays against a 1600, the results should look the same as when an 1800 plays against a 1700 or any other pair of players 100 rating points apart. In both cases, the expected score is 63.7% for the higher-rated player.

So, suppose that we have players A, B, and C, and A beats B 63.7% of the time, and B beats C 63.7% of the time. Now we let player A play against C, because that also happens sometimes. ELO assumes that in this case, A will score 75.8% against C. But you could imagine a game where skill scales very rapidly and it's actually 90%. Or one which has a significant random component, and it's more like 70%. In either case, what this would mean is that ELO is not stable, and that you can "game" the system by choosing the right opponents. For the former, you should always play much lower-rated players, and for the latter, you should always play much higher-rated players.

GR, similarly, makes an assumption like this in expected score. And in Diplomacy, this ends up being very far off at the extremes. The most skilled Austria in the world will lose if its neighbors all want a piece of the empire in 1901 (and it's sometimes more difficult to dissuade inexperienced players). In gunboat, this problem is more extreme -- in about 1/3 of the games I play, by the end of 1902 I'll have essentially zero chance to get a good result regardless of the decisions I make. My individual skill level is only reflected in what happens in the other 2/3 of games. And this makes it unlike chess. A chess grandmaster will simply never fail to win against a novice. But the best Diplomacy player in the world will sometimes fail. It's more like duplicate bridge, which similarly has no formal random component, and yet the gameplay forces stochastic results. I will beat a world class bridge partnership over one hand regularly. I'd be very, very unlikely to beat that partnership over 20 hands, though.

Anyway, what this means for GR is indeed that playing higher-GR players will make your GR go up compared with playing lower-GR players, at least in the long-term. And this is also true for strong players, so they should avoid playing lower-rated players.

However, there's a second flaw. GR takes far, far too long to equilibrate. For almost every player on the site, if they were simply to play more games, their GR would end up further from 100 than it currently is. Most of the players posting in here probably have a GR over 100. So, that means if you just play a lot more games against identical opponents with an identical skill level, your GR will go up. So, this also factors into your choice of opponents -- since most players are underrated, but the most experienced players are less underrated, you can also game the system by playing against the high-rated players who have already played a ton of games. [This effect was once even worse for gunboat players, because gunboat games only counted for 1/4 of a regular game, so it took 4 times as many games to equilibrate.]
5

MadMarx
Gold Donator
Gold Donator
Posts: 293
Joined: Sun Dec 31, 2017 12:01 am
Karma: 148
Contact:

Re: Ghost Rating Challenge

#127 Post by MadMarx » Sun Sep 13, 2020 4:06 pm

Wusti wrote:
Sun Sep 13, 2020 5:30 am
I still don't understand how no matter what I do, my rank keeps tanking. Now at 7000+ and counting. So fuck tournaments like this.
Couldn’t agree more. If you aren’t excited to play in a tournament and able to have fun doing it, that’s not the tournament for you.

MadMarx
Gold Donator
Gold Donator
Posts: 293
Joined: Sun Dec 31, 2017 12:01 am
Karma: 148
Contact:

Re: Ghost Rating Challenge

#128 Post by MadMarx » Sun Sep 13, 2020 8:10 pm

Dioptre wrote:
Sat Sep 12, 2020 10:43 am
Is this the correct order now? Still 5 full games.

G1
1. Dejan0707 - 1.5 (GR = 439.358)
2. MadMarx - 4.5 (GR = 350.189)
3. Yonni - 5
4. ben9990 - 10 (message on Discord to start)
5. ATC - 13.5 (GR = 255.26)
6. swordsman3003 - 16.5 (GR = 248.157)
7. kgray - 17

G2
8. TheFlyingBoat - 35
9. kissemo 42
10. marshal969 - 44 (GR = 198.43)
11. davidc1 65
12. BismarckAlive 71
13. Dioptre 85
14. gimix - 89

G3
15. rdrivera2005 - 117
16. cdngooner - 138
17. Laszlo_L 199
18. Sickofpalantirs - 213
19. bo_sox - 302.5 (GR = 142.11)
20. Bonatogether - 315
21. Szpoti - 338

G4
21. gryncat - 349.5 (GR = 138.61)
22. Rayjay - 381
23. D0ublesh0ts - 686
24. JRoz 861
25. Hamilton Brian 997
26. josemurc - 1725
27. President Rich - 1783
28. HdB - 2019

G5
29. lfischl - 2041
30. e.m.c^42 - 2135
31. Dart_Attack - 3973
32. 1221 - 7076
33. Larsi205 - 7349
34. Emperor Josua - 7455
35. Tobieee911 - 7554

e.m.c^42
Posts: 6320
Joined: Thu Jan 11, 2018 7:00 pm
Location: Rated 0/5 Stars; ☆☆☆☆☆
Karma: 1726
Contact:

Re: Ghost Rating Challenge

#129 Post by e.m.c^42 » Sun Sep 13, 2020 9:36 pm

Claesar wrote:
Sat Sep 12, 2020 4:16 pm
e.m.c^42 wrote:
Sat Sep 12, 2020 3:35 pm
Claesar wrote:
Sat Sep 12, 2020 7:31 am
e.m.c^42 was ranked 2135th in January. Another cheater uncovered.

lfischl is 2041th though, so I think emc barely stays in G5.
Oh huh, I did not consider that. 2537/2563 in Jan. 2020 and 8338/8410 in Aug. 2020. The actual numbers went up from 61.57 to 63.67

Lmao. Did I actually rise in ranking comparatively.
I didn't find you in the August rating. Are you sure you were following the link in the original post i.e. using the correct ratings?
Oops, my bad. And here I was thinking, oh, spreadsheet is not loading with search, let's just go straight to link that pops up when I type in 'ghostrati-'

Then in which case, the most recent one would be from May and at 5049, so still very poor :D

Wusti
Posts: 399
Joined: Thu Oct 19, 2017 10:12 pm
Karma: 232
Contact:

Re: Ghost Rating Challenge

#130 Post by Wusti » Sun Sep 13, 2020 11:46 pm

MadMarx wrote:
Sun Sep 13, 2020 4:06 pm
Wusti wrote:
Sun Sep 13, 2020 5:30 am
I still don't understand how no matter what I do, my rank keeps tanking. Now at 7000+ and counting. So fuck tournaments like this.
Couldn’t agree more. If you aren’t excited to play in a tournament and able to have fun doing it, that’s not the tournament for you.
What an awesomely valuable contribution to make MM.

So answer me this - in a whole year of playing, I've only played 8 games of Classic FP, and 7 of those were in the Masters tournament. My GR dropped from 2000 to 7500 - is this really the way this system is meant to work?

MadMarx
Gold Donator
Gold Donator
Posts: 293
Joined: Sun Dec 31, 2017 12:01 am
Karma: 148
Contact:

Re: Ghost Rating Challenge

#131 Post by MadMarx » Mon Sep 14, 2020 5:55 am

Wusti wrote:
Sun Sep 13, 2020 11:46 pm
So answer me this - in a whole year of playing, I've only played 8 games of Classic FP, and 7 of those were in the Masters tournament. My GR dropped from 2000 to 7500 - is this really the way this system is meant to work?
There is no ‘A‘ for effort. You are not entitled to an increase in GR simply because you have played 8 games. GR is simple math, based on how you perform in your games. If seven players with a 100 rating play, a loss/survive approximately results in a loss of 6 in rating, a 5-way draw will gain you ~2, a 3-way ~8 and a solo ~34. So, losses/survives cancel out a lot of draws, if in 8 games you have more losses/survives than draws, it makes sense that your rating could decline (and clearly solos are the way to raise your GR quickly).

As far as your ranking dropping from 2000 to 7500, that’s just a matter of the massive number of people in that rating range. A rating of 95.0 has a ranking of ~3900, a rating of 94.0 a ranking of -6200, so while it’s frustrating that your ranking can drop thousands of points for a minuscule drop in rating, your ranking can just as easily jump thousands of points for a small increase in rating.

Perhaps focus on your rating initially. If you can get your rating back to 100, your rank will be ~2600. Then, you’re just a 3-way draw away from a rating of 108 and cutting your ranking in half to ~1300, or a solo away from 134 and a rating of ~500... I’ve simplified the math a lot, but you get the idea.

Or, just have fun playing and don’t worry about your GR, it is afterall a mere documentation of what has been done in the past, and the possibilities held by the present and future are often far more interesting...
9

Yonni
Silver Donator
Silver Donator
Posts: 408
Joined: Thu Oct 19, 2017 6:55 pm
Karma: 368
Contact:

Re: Ghost Rating Challenge

#132 Post by Yonni » Tue Sep 15, 2020 1:09 am

Back from a fun weekend away and glad to see we have 5 full games to get going. I'll make them and send them out tomorrow.

re: GR and these games.

The point of this is definitely not to exclude players based on GR. I always found the GR challenge games a great way to find evenly matched players. GR may not be the perfect measure of skill but, as a rule of thumb, I've found it quite useful for finding players I enjoy playing with. Hopefully all the games are competitive and enjoyable. In the future, maybe we can try some different things like dividing the games by player ID or alphabetical, etc.

That being said, I think there are many flaws with GR as people have pointed out. On top of what has been mentioned, losing a game penalizes your GR the same amount irrespective of who your opponents were. This makes it quite unlike Elo ratings. (aside, I did for a while calculate a different version of GR that took this into account but haven't post the "GR2" ratings for a while).
4

Wusti
Posts: 399
Joined: Thu Oct 19, 2017 10:12 pm
Karma: 232
Contact:

Re: Ghost Rating Challenge

#133 Post by Wusti » Tue Sep 15, 2020 3:22 am

Yonni - ELO rankings would be far more meaningful in Diplomacy and the guts of my complaint.

It is not wanting a free ride, not complaining for complaining's sake, but because GR is a horribly flawed system, and to continue to use it as the sole basis for so many tournaments on this site is cancerous.

It also engenders an "Elite Club" for high GR players, as evidenced by only high GR players defending it. Converting to ELO would be a step in the right direction to fixing it.

MadMarx
Gold Donator
Gold Donator
Posts: 293
Joined: Sun Dec 31, 2017 12:01 am
Karma: 148
Contact:

Re: Ghost Rating Challenge

#134 Post by MadMarx » Tue Sep 15, 2020 5:38 am

Wusti wrote:
Tue Sep 15, 2020 3:22 am
Yonni - ELO rankings would be far more meaningful in Diplomacy and the guts of my complaint.

It is not wanting a free ride, not complaining for complaining's sake, but because GR is a horribly flawed system, and to continue to use it as the sole basis for so many tournaments on this site is cancerous.

It also engenders an "Elite Club" for high GR players, as evidenced by only high GR players defending it. Converting to ELO would be a step in the right direction to fixing it.
Cancerous?! Dang, you take GR much more seriously than even VI! ;-)

Once you accept that any rating system has flaws, how much does it really matter? If you tweak a few variables do you really think the top 1% will become the bottom 1% and the bottom 1% will become the top 1%?

Don’t get me wrong, I love playing with numbers and the theories behind tweaking rating systems, but it’s not like the 35 people that have signed up for this tournament would be significantly redistributed based on GR2, right?

GR is a tradition around here, created by a Legend, encourages better play, and close counts (especially if you agree it’s attempting an impossible task)...
2

Wusti
Posts: 399
Joined: Thu Oct 19, 2017 10:12 pm
Karma: 232
Contact:

Re: Ghost Rating Challenge

#135 Post by Wusti » Tue Sep 15, 2020 7:29 am

Doesn't matter blah blah, nothing would change blah blah, whatever dude - but if you're going to make a comeback at least put some effort into it. All your responses boil down to "our systems is best cause we say it is and you're just sore causes your rating is bad". You can shove that where the sun doesn't shine.

Claesar
Posts: 1965
Joined: Tue Oct 03, 2017 10:34 am
Karma: 1490
Contact:

Re: Ghost Rating Challenge

#136 Post by Claesar » Tue Sep 15, 2020 7:47 am

Wusti wrote:
Tue Sep 15, 2020 7:29 am
Doesn't matter blah blah, nothing would change blah blah, whatever dude - but if you're going to make a comeback at least put some effort into it. All your responses boil down to "our systems is best cause we say it is and you're just sore causes your rating is bad". You can shove that where the sun doesn't shine.
I read both responses. I can safely say a lot more effort went into his. It has arguments rather than ad hominem aggression. The difference is actually quite staggering.

All your responses boil down to "I won't read your replies because I'm sore. I'm unwilling to admit my rating is bad because my results are bad".
2

CSteinhardt
Bronze Donator
Bronze Donator
Posts: 13
Joined: Mon Jan 01, 2018 8:16 pm
Karma: 32
Contact:

Re: Ghost Rating Challenge

#137 Post by CSteinhardt » Tue Sep 15, 2020 8:35 am

Wusti wrote:
Tue Sep 15, 2020 3:22 am
Yonni - ELO rankings would be far more meaningful in Diplomacy and the guts of my complaint.

...

Converting to ELO would be a step in the right direction to fixing it.
As I tried to explain a bit better in my earlier post (I won't repeat it here since it appears above but am happy to engage in a discussion about it), ELO wouldn't work very well in Diplomacy because the math behind it makes assumptions which aren't correct for Diplomacy.
MadMarx wrote:
Tue Sep 15, 2020 5:38 am
Once you accept that any rating system has flaws, how much does it really matter? If you tweak a few variables do you really think the top 1% will become the bottom 1% and the bottom 1% will become the top 1%?
That's certainly not true, but what's much more true is that being in the top 1% requires both high skill and a large number of games (again, my earlier post explains this in more detail so I won't repeat the lengthy argument here; when I modeled this a few years ago it took 100-300 games to equilibrate). Further, because a strong player with fewer games is more underrated than one with more games, the best way to improve your GR is to only play against people with a ton of games on the site.

Then again, GR has a similar issue to ELO with the scaling; a bunch of games between very high-GR players and below-average GR players will, over time, result in a net transfer of GR to the lower-GR players.

So, the effect of both of these is to encourage players who care about GR to seek out opponents from a small set of long-term players with high GR, and for those players to similarly only want opponents from that small set.

Note that this is not inevitable; if so desired, it would be straightforward to construct a ranking system which encourages different player behavior. As an obvious example, getting to the top of the :points: chart clearly rewards something different than GR and produces a ranking that's further from a skill rating than GR. Similarly, I could construct something which would be much closer to a skill rating than GR. Whether that's desirable, and what behaviors we want to encourage, is a separate question, and I'm going to limit myself to the math here rather than wade into that debate.
2

Wusti
Posts: 399
Joined: Thu Oct 19, 2017 10:12 pm
Karma: 232
Contact:

Re: Ghost Rating Challenge

#138 Post by Wusti » Tue Sep 15, 2020 9:16 am

CSteinhardt wrote:
Tue Sep 15, 2020 8:35 am
Wusti wrote:
Tue Sep 15, 2020 3:22 am
Yonni - ELO rankings would be far more meaningful in Diplomacy and the guts of my complaint.

...

Converting to ELO would be a step in the right direction to fixing it.
As I tried to explain a bit better in my earlier post (I won't repeat it here since it appears above but am happy to engage in a discussion about it), ELO wouldn't work very well in Diplomacy because the math behind it makes assumptions which aren't correct for Diplomacy.
MadMarx wrote:
Tue Sep 15, 2020 5:38 am
Once you accept that any rating system has flaws, how much does it really matter? If you tweak a few variables do you really think the top 1% will become the bottom 1% and the bottom 1% will become the top 1%?
That's certainly not true, but what's much more true is that being in the top 1% requires both high skill and a large number of games (again, my earlier post explains this in more detail so I won't repeat the lengthy argument here; when I modeled this a few years ago it took 100-300 games to equilibrate). Further, because a strong player with fewer games is more underrated than one with more games, the best way to improve your GR is to only play against people with a ton of games on the site.

Then again, GR has a similar issue to ELO with the scaling; a bunch of games between very high-GR players and below-average GR players will, over time, result in a net transfer of GR to the lower-GR players.

So, the effect of both of these is to encourage players who care about GR to seek out opponents from a small set of long-term players with high GR, and for those players to similarly only want opponents from that small set.

Note that this is not inevitable; if so desired, it would be straightforward to construct a ranking system which encourages different player behavior. As an obvious example, getting to the top of the :points: chart clearly rewards something different than GR and produces a ranking that's further from a skill rating than GR. Similarly, I could construct something which would be much closer to a skill rating than GR. Whether that's desirable, and what behaviors we want to encourage, is a separate question, and I'm going to limit myself to the math here rather than wade into that debate.
Thank you for your well thought out reply. The crux of my objection is your second last paragraph.

It leads to the creation and maintenance of a small cabal who only want to play this type of ranked game system because it maintains their position, and they get to wheel out "GR Tournaments" as if they are inclusive when in fact the opposite is true. This is why I used the term "cancerous" because I think that is unhealthy for the community.

As for my opinion of my own meager skills, I was never playing for GR or gaming the system in my entire time on this site. I play to win, to experiment, for entertainment value etc. What I have only recently come to understand is the damage my approach has caused, by blowing out "my GR" with a stupidly small sample set, and then restricting me from playing anyone who'll play me, due to the frequency of GR based "tournaments" and games.

To add salt to the wound, if I want to play with the "better" players (and I do concede that those in the top 200 will on average have a higher skillset than those in the next 10000, but not in 100% of cases), then I have no choice at all but to attempt to game the system to do so.

Should I choose not to do so I will be condemned to not be able to support site arranged tournaments (which I do try to do), and gradually be effectively pushed out of the community I played 430 odd games with because my GR is too low and I won't play by the rules by which GR is enforced.

I don't think it's healthy for the site or it's community. It is especially unhelpful to have the elitist cabal jump out and smackdown anyone who dare challenge the status quo, with assertions and comments like MM and Claesar's.

Anyway thanks for your input CSteinhardt, I appreciate a decent answer rather than a patronising one.

Despite my (as usual) thorny turn of phrase, it is a serious issue and the guys who run the site can listen and take it on-board or just write me off as a whinger - up to them.
1

Yonni
Silver Donator
Silver Donator
Posts: 408
Joined: Thu Oct 19, 2017 6:55 pm
Karma: 368
Contact:

Re: Ghost Rating Challenge

#139 Post by Yonni » Tue Sep 15, 2020 12:38 pm

Wusti, the reason you have a low rating has nothing to do with the fact that you haven't "gamed" the system. It has to do with the fact that you've lost the majority of the games. I just played a game with you in the Masters and I can confirm thah your tone here is 100% reflective of how your in game press. I am not surprised at all by how often you lose or how unaware you are as to why it happens. If anything, (GR or otherwise) is stopping you from joining these games then it has my full endorsement.

P.S. Elo is a name, not an acronym.

New England Fire Squad
Gold Donator
Gold Donator
Posts: 356
Joined: Thu Feb 08, 2018 3:54 am
Location: Connecticut
Karma: 263
Contact:

Re: Ghost Rating Challenge

#140 Post by New England Fire Squad » Tue Sep 15, 2020 12:55 pm

Wusti wrote:
Tue Sep 15, 2020 9:16 am
CSteinhardt wrote:
Tue Sep 15, 2020 8:35 am
Wusti wrote:
Tue Sep 15, 2020 3:22 am
Yonni - ELO rankings would be far more meaningful in Diplomacy and the guts of my complaint.

...

Converting to ELO would be a step in the right direction to fixing it.
As I tried to explain a bit better in my earlier post (I won't repeat it here since it appears above but am happy to engage in a discussion about it), ELO wouldn't work very well in Diplomacy because the math behind it makes assumptions which aren't correct for Diplomacy.
MadMarx wrote:
Tue Sep 15, 2020 5:38 am
Once you accept that any rating system has flaws, how much does it really matter? If you tweak a few variables do you really think the top 1% will become the bottom 1% and the bottom 1% will become the top 1%?
That's certainly not true, but what's much more true is that being in the top 1% requires both high skill and a large number of games (again, my earlier post explains this in more detail so I won't repeat the lengthy argument here; when I modeled this a few years ago it took 100-300 games to equilibrate). Further, because a strong player with fewer games is more underrated than one with more games, the best way to improve your GR is to only play against people with a ton of games on the site.

Then again, GR has a similar issue to ELO with the scaling; a bunch of games between very high-GR players and below-average GR players will, over time, result in a net transfer of GR to the lower-GR players.

So, the effect of both of these is to encourage players who care about GR to seek out opponents from a small set of long-term players with high GR, and for those players to similarly only want opponents from that small set.

Note that this is not inevitable; if so desired, it would be straightforward to construct a ranking system which encourages different player behavior. As an obvious example, getting to the top of the :points: chart clearly rewards something different than GR and produces a ranking that's further from a skill rating than GR. Similarly, I could construct something which would be much closer to a skill rating than GR. Whether that's desirable, and what behaviors we want to encourage, is a separate question, and I'm going to limit myself to the math here rather than wade into that debate.
Thank you for your well thought out reply. The crux of my objection is your second last paragraph.

It leads to the creation and maintenance of a small cabal who only want to play this type of ranked game system because it maintains their position, and they get to wheel out "GR Tournaments" as if they are inclusive when in fact the opposite is true. This is why I used the term "cancerous" because I think that is unhealthy for the community.

As for my opinion of my own meager skills, I was never playing for GR or gaming the system in my entire time on this site. I play to win, to experiment, for entertainment value etc. What I have only recently come to understand is the damage my approach has caused, by blowing out "my GR" with a stupidly small sample set, and then restricting me from playing anyone who'll play me, due to the frequency of GR based "tournaments" and games.

To add salt to the wound, if I want to play with the "better" players (and I do concede that those in the top 200 will on average have a higher skillset than those in the next 10000, but not in 100% of cases), then I have no choice at all but to attempt to game the system to do so.

Should I choose not to do so I will be condemned to not be able to support site arranged tournaments (which I do try to do), and gradually be effectively pushed out of the community I played 430 odd games with because my GR is too low and I won't play by the rules by which GR is enforced.

I don't think it's healthy for the site or it's community. It is especially unhelpful to have the elitist cabal jump out and smackdown anyone who dare challenge the status quo, with assertions and comments like MM and Claesar's.

Anyway thanks for your input CSteinhardt, I appreciate a decent answer rather than a patronising one.

Despite my (as usual) thorny turn of phrase, it is a serious issue and the guys who run the site can listen and take it on-board or just write me off as a whinger - up to them.
Wusti, I'm ranked #17 overall and #6 gunboat. I play with lower ranked players all the time. Play a live game sometime and chances are there will be top ranked players in them. Good luck!

Post Reply

Who is online

Users browsing this forum: No registered users and 131 guests