It seems to me that the best path to a high GR is WTA live games, cuz there are inevitably people who either don't have the experience to stop a solo, not to mention the huge advantage you get just from having a high typing speed. Playing in a game and typing 100 WPM against someone who types only 20 or 30 is a ridiculous advantage.

My goals... If the game end in a win, be the winner or be completely eliminated. Survivng when someone else wins is the ultimate failure of a diplomat. If the game ends in a draw, be part of it. Beyond that, have fun and play the game whatever way brings you the most enjoyment. I'm trying something in one now that I can't go into detail, but the posts I'm reading in game are hillarious and fun as a result. Sort of a chainsaw diplomacy hellalt kind of thing. Will I lose? Yeah, probably. But I'll have fun dying. :-)

@uclabb - That's why I don't do 5 minute games. I top out at 45 WPM. Hey, I'm a programmer who works in a system with intellisense so I don't have to type out everything once the object has been declared in code. Oh and copy/paste is every programmer's best friend *and* worst enemy. :-)

@uclabb, thats not always true. Just because you play with beginners doesn't mean you'll win. You're just as likely, in those games, to be a 5 centre Italy holding a stalemate in the Med against a 17 centre France only to have Turkey show up and take your centres and give the game to France. So I feel like it is almost a wash.

I think the GR is just another reward system - like the points used here. It does better than the points in some areas, and worse in others. For an excellent defence of the points system, see kestas' post on the forums where TGM posted his original idea for the GRs.

Diplomacy tournaments have their own, unique reward systems too - and I'm told players turn up in them who follow personal systems different from those in the tournament. The point being, in Draug's example, if you let the little guys live, you're using a personal reward system that values this more than killing them for the 3-way draw. You made that choice because in your system, it's more rewarding than the 3-way draw, and good on you for sticking to that system - but it's useful to remember other systems may not reward it as much, or at all!

Which brings me to an interesting idea - (particularly for anyone who's ever played mansions of madness). Set up a tourney, the TD is referred to as the Keeper, and he doesn't play. He decides a reward system beforehand, per round, that is not revealed to the players. Games go on for an agreed amount of time (say 5 years), at which point he reveals his reward system, and then games continue to completion. Players progress depending on how many points they score as per the revealed reward system. The fun part is that the reward systems may be utterly ridiculous - the average sum of distances of all your units from Portugal, for example.

Im just happy to be at 722 after only 1 month. I want to keep climbing!

Yes! Jumped 90 spots!

@WhiteSammy, the page probably just loaded bad for you. Try refreshing.

@spyman, as your GR increases, you risk more and therefore playing the same level of play with the same number of wins/draws/eliminations will cause you to hit a ceiling where your loses match your gains. To keep going up you need to increase your gains by losing less or playing tougher competition and not losing more.

But, assuming you are undervalued, playing more games will jump you up to the "correct" level quicker than playing less.

And when you reach a certain point where most people are below you, you ave to win, not just draw, to keep going up. Draws actually start to hurt you when the majority of the players in your game are a lower GR than you.

@Draug: That's not true. lets say I have a GR of 300 and the other 6 have a GR of 150. That means you are a top 50 player playing guys ranked around 400. A four way draw won't hurt your rating, and a three way draw will still help you. For a player like you, you could play 6 absolute beginners and still increase your GR with a three way draw.

In fact,if you aren't in the top 20, the caliber of competition doesnt really matter and you'll keep going up as long as you draw a decent proportion of your games and win occasionally. The ghost ratings really start to spread out in the top 20 or 30, so at that point it becomes harder to move up without playing better people or having a very high solo rate.

If Ghost Rating is pretty similar to Chess Elo ratings, then the higher rated you are, the fewer games you should play.

At a very high Elo, a Chessmaster has almost nothing at all to gain from playing against low Elo players. So little to be gained that it's not even worth the time it takes to play a match. The issue is that if he gets lazy, or doesn't pay attention, or doesn't take the game seriously enough, it's possible for the low Elo player to sneak in a win and DEVASTATE his ranking. So the high Elo master has to play a game with his full attention and try very hard to win, but will gain virtually no rank from the win.

The only way an already-high Elo player can increase his Elo is by playing other similarly ranked players, or perhaps higher ranks if they exist. This way he has an actual reward for devoting his entire mental faculties to trying to win the game. He can gain at least *some* rank from winning.

No matter what, with an Elo system, each player has to try as hard as they possibly can to win (that's a good feature). But playing against players of substantially lower rank offers no reward for that effort, which is why it is a waste of time (unless you just generally enjoy stomping noobs or are trying to teach another player I suppose).

That's actually one of the policy reasons given for the point system on WebDip: it discourages the "resting on your laurels" approach that is *definitely* incentivized by the Elo system.

There's been situations in Chess or other games that use Elo where a player gets to a high or the highest rating and then stops playing altogether! To protect the rating of course.

Wizards of the Coast last month abolished their Elo-based ranking system in favor of "Planeswalker Points" (gee, that sounds pretty familiar!).

I'm not sure how many differences there are between GR and Elo but they seem pretty damn similar and I suspect the have the same problems.

However I wouldn't go as far as to say that a point system accurately describes a player's skill either, because with points it's possible to get to a high score just by beating bad players over and over again and never challenging good players.

Analogy: A quasi-famous videogame professional was able to make TONS of money playing online poker, and I believe he even set a world record for number of hands played in a month or something to that effect. But he made so much money just by playing zillions of bad players over and over and eeking out his profit just by being slightly better than average.
It makes mathematical sense: If he is just barely better than 50% of the players, but plays 100,000 hands of poker, he will make a pretty good amount of money. He could profit indefinitely.
But as I recall, that guy entered into a real world poker tournament and got destroyed (or he trolled in on purpose, I can't remember exactly). But the point is, he actually is not good at poker. He can make a ton of money simply by being better than average and playing average players a bunch of times.

The points system permits that, if you allow the analogy of points to money.

What I personally like about points is that it distinguishes between low-stakes and high-stakes matches. Anybody who has played a poker game with no money on the line, a game for $10, and a game for $1,000 would understand what I'm talking about.

I find it laughable to say that if a guy loses 10 games of poker for $10 but wins a game of poker for $10,000, he is a terrible player because he has only 9% wins, and the guy who won $100 worth of pots is a better player because he has 91% wins.

Therefore, the signs a of a good player would be a fair number of points/and a high ghost rating.

I'm not sure how Elo works but you "bet" the same amount of your GR regardless of who you are playing. So a loss to poor players doesn't hurt your rating any more than a loss to good players. Also, just from looking at the list, your ranking usually falls if you don't play. The difference could be that in chess, the emphasis is on the absolute rating while here the emphasis is on your ranking. Another difference is that this is a 7 player game so soloing against 6 beginners still adds a significant amount to your rating (around 30 , i think)

I think that both points and GR are pretty good proxies for skill although neither is perfect. Most players with a lot of points have very good GRs and vice versa, with some exceptions.

And a five way draw will hurt you. It depends on the difference in the average range of your opponents and your own GR. I've played games where my GR was so high, even a 4 way hurt me and a 3 way just broke even. I've been nice and played with new players who have no starting GR, meaning it is computed based on their GR being 100. 6 new players versus a 300 player means the 300 player needs a 3 way draw to break even. My point stands. A draw *can* hurt you when you reach a certain point if everyone in your game is below you. If you carebear like me and let players stay in the draw so it is a 5 way because they showed spunk and spirit, you screw yourself. I just 4 way drew a game where every one was below me and I could (and should) have gone for the victory. I definitely could have reduced the draw size further with my Russian ally. But I (we) were nice and drew. I will regret it in my GR, I'm almost certain. I may get a half point boost (I did the math), but that may as well have been a break even game for me. gameID=66749 if you'er interested.

meh, up 1 spot for 1 finished game.

The other thing is that both GR and points take into account a player's performance in a bunch of different games. For example, MadMarx is clearly the best WTA, non-live, full press, classic map, anon player here (i think that's more or less all he plays). But he might not be the best gunboat player, or the best non-anon player, or the best world map player or the best live player. It would cool to see seperate rankings for the different types of games. For example, Lando, Czech and Barn3tt are probably the three best live gunboaters on the site but you never see them near the top of GR. It would be neat to compile a ranking that recognized the success of these types of players. I'd be willing to try it if I could get access to Ghost's data.

I believe it is just a page that present the statistics of completed games. Maybe Ghost and Kestas would let us all see it so we could use to to run stats.

Thanks so much, G! How did your charity bike ride go?

Up to 284 . . . probably as good as I'll get, given that I've already clocked one bad loss in Oct and I'm staring at a second one. Parcheesi, where are you when I need you?

@santosh: That type of tourney sounds pretty interesting. However, I think it could be really frustrating to some players when a totally absurd reward system was set, like the distance from Portugal example. A way to reduce the absurdity while maintaining the interesting "unknown goal" feature would be to limit the number of possible reward systems by creating a "deck of objectives."

One way to proceed from there is to play like Mission Risk/Ticket to Ride. In this style, each player would draw a generic goal(s) and would have to complete the goal(s) by some end point such as a specific year or board state. Players in this way would know their victory condition and will not feel like their fate was solely due to random chance.

Another way would be to play like Civilization, where there are several known victory conditions (Conquest, Domination, Cultural, Diplomatic and Science/Technology in Civ). This would allow for varied strategies to compete. For these conditions you could have something like "Control the Seas" where you need X fleets to win. A slightly different version of this would be to have country specific victory conditions that are known to all.

Well not sure quite what happened but I jumped over 1500 places... Still only in the 600's but good jump anyway!

Thanks, Ghostmaker, I'll get that check in the mail I promised you tomorrow.

:) rank 429 now.
I must keep improving...

97. Broke the top 100. :)

@swords, Wizards changed their point system? I need to pay more attention, how do the 'Planeswalker Points' work exactly?

@Frank, a Gunboat Rating would improve this site infinitely.
@Ghost please let Frank have your data so he can he can prove that I am the best goonbatter.

So. I went up in ratings a little just because I've managed to draw out the three games I'm currently losing badly. I feel like I'm real estate speculating 5 years ago.

And November will be the bubble bursting? hehehe