Forum
Advisor Blog
Reviews
Strategies
Ideas
Possible flaw with the new rating system
|
Kehm wrote
at 9:48 PM, Friday February 23, 2007 EST
Sorry if this is somewhat a long post, I really wanted to be clear on the subject. Skip it if you're not going to read it.
--- Don't get me wrong, I love the new system. It has been tested a lot and I believe it is a great improvement over the last one. I also want to state clearly (and now) that I do not blame players for playing the way they do in the new system (I do it too). --- Before I start breaking the rating system, I think it would be fair if everyone understood it and why there was a need for this new system in the first place. It might be obvious for some, but not so much for others. The main problem with the old system was that it rewarded players that used a "hiding strategy". It consisted on hoping your few (or 1, or couple) territory(ies) would overlast the stronger players fighting for the top spots. It was possible to score very good points by having a single territory for the whole game by being killed last out of generosity from the winner. The point of the new system, obviously, is to reward those trying to win: "The better game you played, the better score you should have." If you are dominating the game and an alliance turns against you and defeats you, you still get many points for the great game you played, and don't get shafted by that 1 guy who had 1 territory all game. --- The flaw. I believe it is very possible to play a great game, finish 2nd, even first, and get screwed on domination nonetheless, EVEN IF YOU MADE THE BEST POSSIBLE CHOICES over the course of the game, which ultimately led to a very good finish, but marginal rating gains. And why is that, you might ask? Simply because domination points are not calculated correctly. This is my personal opinion, so please bare with me. I don't not claim to have the ultimate truth on the matter, but I do believe I do. As far as I know, here's how the domination points are calculated: (I might be wrong, but the last time I tested in the sandbox, that's what Ryan told us) - At the begining of your turn, the number of connected territories are added to your grand total. When you are eliminated, your grand total is compared to all the other remaining players and your domination rank is calculated. It works fine as it is, but I truly believe it is flawed and could be easily fixed. That is simply because that system does not necessairly encourages good play (aiming to win), but aggressive and self destructive plays (taking useless chances to up your territory count). Example: - Player X has 4 territories, fully built (that expression means they all have 8 dice stacks). He's been playing very solid, but somewhat conservatively. He has a good chance to win, are take a top finish, since the board is generaly low stacked with a lot of fighting going on. There are big stacks around him, however. - Player Y has been playing wrecklessly. He's overly agressive, while luckily maintaining an avg of 7 unprotected territories. - 2 turns later: Player Y is eliminated, not having sufficient dice to protect his empire. Player X now has five 8-stack territories and a good shot at winning with the leader, player Z, having seven 8-stacks. Player Y gets -7 for rank 5, +15 for domination. (+8 total) --- So, what's wrong with that? I simply don't think player Y is given the right rating. I hope I'm not the only one seeing this. He played like an idiot trying to abuse the system. It's fine to abuse the system, I do it too. And there is nothing wrong with playing like this, under this system. Therefor, the system is flawed, rewarding poor play. What's the reason, then? Territories mean nothing. 3 territories with 8 dice on each > 7 territories with an avg of 3 dice. 3x8 = 24 dice, 7x3= 21 dice. Wasn't that simple enough? The solution is clearly to count the number of dice, NOT territories, at the begining of each turn. The goal is to have dice, afterall. Aren't we playing Dice wars? |
|
XicaDaSilva wrote
at 11:46 PM, Friday February 23, 2007 EST Nice post. Good analysis.
Average size of number of territories is not a good measure of domination. Fixing the system is not as trivial as you suggest. Number of dice is important, but position on the map, dice distribution (size of stacks on the borders with neighbors) etc are also important. Another issue, is that the fact that domination is measured at discreet points in time end of each round, I believe. This will only emphasize the problem with size in territories, as some players just finished their turn (close to their max size), while others are about to start their turn (close to their min size). |
|
|
XicaDaSilva wrote
at 12:30 AM, Saturday February 24, 2007 EST I reread some old post and it seems that the sizes are not recorded at the same size for all players, but for each right before starting the turn (first round excluded).
Can somebody confirm this? |
|
|
XicaDaSilva wrote
at 12:32 AM, Saturday February 24, 2007 EST == correction I reread some old post and it seems that the sizes are not recorded at the same *time* for all players, but for each right before starting the turn (first round excluded). |
|
|
Kehm wrote
at 12:34 AM, Saturday February 24, 2007 EST I explained how it works.
"At the begining of your turn, the number of connected territories are added to your grand total. When you are eliminated, your grand total is compared to all the other remaining players and your domination rank is calculated. " So you are right, it's at the begining of each turn. This is what we found worked best, and that's what was tested in the sandbox and, as far as I know, is used right now. It is working good. I think number of dice would work better. |
|
Grunvagr wrote
at 12:50 AM, Saturday February 24, 2007 EST The answer lies in bending your conception of what smart play is and was. I've had the honor of facing off against you Kehm and you are a worthwhile adversary. Here's all I can say to answer you:
A good player plays to the best of his / her ability to win a game - or when things go poorly, to maximize the amount of points they can scrounge up while going out in 2nd-5th. In the previous system, wise and (often) safe play did indeed mean slow, cautious expansion. However, the new system promotes a bit more action and expansion. The key is to remember that holding onto the lands is the key factor in the dominance. Thus, crazy expansion is not wise, though aggressive play is. What has yet to happen is a change in many players gameplay - I truly feel a lot of players are far too cautious at (and here's the kicker) checking the expansion of *others*. I think you might get annoyed at the fact that someone who expands a lot then lucks out and holds land gets to win later. Well, the truth of the matter is likely that person had a round or two with 2 and 3 and even 1 stacks interspersed among their 5 and 6 stacks but no one dared to check their expansion and go thin out their lines with a bigstack. The leader always has a round that passed when he or she was terrified of being checked back into line but was ignored and permitted to be powerful. The game is well thought out and does reward overall good gameplay. And if you happen to be low on land but playing cautiously, dont fret. Take a glance at the board and you should know who has good dominance and who doesnt - then fight those people or ally if you must. Your full stacks should be able to push them back at least a little at first. The game changed - thus good players must adapt their tactics to maximize their pointgain effectively (if they can''t get 1st). |
|
|
Grunvagr wrote
at 12:55 AM, Saturday February 24, 2007 EST I also meant to say, I think many players assume 8 dice to be a safe border, seeing as no foe will resupply after their turn to have more dice than them. The truth, however, is that (unless it is at the endgame stage) having an 8 stack is a luxury, not a defensive necessity.
A 5 stack is a massive barrier to enemies if all they have are 3s and 4s, a 6 stack is also a tremendous defense vs most stacks. Thus, though 8 stacks make us 'feel' truly safe because we know the best someone can have is almost a 50/50 shot at defeating it, the reality is that a comfortable and reliable defence does not necessitate having all 8 stacks. |
|
|
Kehm wrote
at 7:01 AM, Saturday February 24, 2007 EST I totally agree with what you are saying grun and I'm not here to dispute that. The point I was making was not geared towards the current state of playing the game, but rewarding players who play to win the game.
There is a major difference here. Whatever the system will be, I will do my best to take the best possible decisions to make my rating higher. What I'm saying, however, is that a dice count, instead of a territory count, is a better way to see if a player has an advantage over another, therefor noting which player is "dominating". The next game you play, check the dice count of every player (it's right beside the territory count, but we never notice it) and you will see exactly what I mean. Territories don't mean much, dice count means a lot more. |
|
|
Kehm wrote
at 7:07 AM, Saturday February 24, 2007 EST I started reflecting on the subject when I finished 2nd in a very close game. I've played the game perfectly, but some players died early because of wreckless plays, and they stole domination points.
I mean, they had no chance whatsoever to win that game, and they gained nearly as much rating than I did for finishing 2nd and having a great game. I stacked up and owned afterwards, losing only to the top player. After the game I thought about it and it just made no sense to me. I don't blame the players to play the way they do, I just don't think playing like that should be rewarded. If I go back to the same game, I've always had much more dice than these players, there is just nothing logical between their rating gains and mine. |
|
DealOrNoDeal wrote
at 12:01 PM, Saturday February 24, 2007 EST similar thing happend to me
fought by myself against an alliance of 2 destroyed one of them, but then was defeated by the other my dominance adjustment -1. |
|
|
Grunvagr wrote
at 12:15 PM, Saturday February 24, 2007 EST the problem with using a dice count to calculate dominance is that sometimes someone might have a lot of 8 stacks but all in the back, and their frontlines are all 2s and 3s, whereas someone else has fewer dice but a vast number of last and 6 stacks up front on their borders.
The first person might have more dice, but is in position to lose because by the time they have dice up front to do battle, the 2nd person will have passed enough to thicken their stacks and will thus be ready to win easily. dice count doesn't work it is too problematic - I'm not implying that the current dominance system is perfect, but I do believe it is at least better than a dice count version. Being protective of ones dice is not the point of the game - the point is to conquer the entire map. |
