Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Bron]

Hello everyone.

This is my first post even though I'm visiting this forum for quite a time. Please note that I'm not a native speaker so excuse my “engrish”.
I would like to suggest a variant to this great game (or more precisely rules for a session) and want to know your opinion. This is supposed to address several features of the game that can be viewed as problems by some players (like me) – let me call them *issues*.

Disclaimer: based on http://forum.dominionstrategy.com/index.php?topic=4966.msg117607#msg117607
I tried to make the setting more general so it can be more easily scalable.

Now let me start with listing of those issues.

Issue #1: First player advantage.
Maybe not everyone will agree, but I think that the evidence is strong.
Quick fix: give the other players an artificial advantage, or while playing several games cycle the game order.

Issue #2: Strong dependance on “luck”.
Quick fix: play several games.

Issue #3: Some games end very close, yet one player wins and the other feels that it was not deserved.
Quick fix: change the criterion for a game to tie, or make those games have lower value.

Issue #4: Some games are “decided” before they end and finishing them is boring.
For example one player in 2p game is pinned.
Quick fix: encourage resigning.

We can discuss issue #1 first. In my experience it seems that number of wins as the first player in 2p game is roughly equal to the number of wins as the second player + number of ties. So if we say that a second player always wins ties I think that it will be more fair, maybe it will result in slight advantage of second player, but hopefully smaller than the one which has the first player now. So the ruling here can be:

Each player starts a game with n/10 (virtual) VP tokens, where n is his game order.

To be honest I don't know if my thoughts can generalize to multiplayer and I cannot check since councilroom is still down. Nevertheless the result here is that there will be no ties at all, which strengthens issue #3 even more so let us (re)define when two players tie.

Two players tie, if the (absolute value of the) difference of their victory points is less than 3.

Actually any constant could do, but I'd stick with integers. To fix it at 3 means that in a 2p big money mirror it is not enough for a first player to win by duchy, but it is enough for the second player. And it also means that the impact on colony games is marginal, which is good since they seem not to suffer from this issue. Actually there is one problem with this rule in a multiplayer. Tie is not an equivalence relation, in particular if A ties with B and B ties with C then it doesn't mean that A ties with C, we will deal with it later.

Now let me describe the session rules for TWO players.

Each player starts the session with 60 crowns.

At the beginning of a game each player bets 6 crowns. During a cleanup phase, just before redrawing new hand, an active player without the bet token (players start the game without one) has the opportunity to raise his bet for any amount of available crowns, but doubling it at most. If the active player raises his bet then the opponent discards his bet token (if he has any) and the active player takes one. Now the opponent either “calls” his bet (evens it) or he “passes” which means that the active player claims both his and his opponent's bet. This addresses issue #4. The game from now is pointless and can be ended right now, but can continue if both players desire so. Special rule is that a player cannot rise over the other players maximum available crowns; i.e. more than to make him “all-in”.
If the game ends without any player passing, the winner claims both bets and in the case of a tie, the bets are split evenly. I think that close games result in lower bets, what addresses issue #3.

After the game ends each player discards 3 crowns, if he doesn't have enough he is eliminated from the session and has lost, the remaining player wins (and is crowned as king of the session). The same applies if a player doesn't have enough (6) crowns for an initial bet.

Between the games, the game order should alternate.

The discard rule is to cap the maximal number of games per session. Obviously 60 starting crowns and 3 discarded crowns are parameters free to experiment with. Of course there is a possibility to give a player an advantage and let him start with more crowns.

Now how to adapt this for a multiplayer.

Variant A: Multiplayer as a set of duels.

Player splits evenly his initial bet among all opponents. This is why I suggested 6 crowns, since it works good for 2-4 players, for 5 and 6 players you'd need 0.5 and 0.2 crowns, respectively. Keep track of bets for every pair of players like in 2p game. An active player can manipulate bets at his cleanup phase for each/any of his opponents and there should be different bet tokens for each pair. At the end of the game each pair of players deals with their bets comparing only their scores. So there is no problem with non-transitive ties here. The game is pointless after all bets are claimed by someone, but if  a player has no “active” bet he can leave the game if he wishes to. In a multiplayer it is possible that more players will be eliminated at the same time so in this case the number of crowns before elimination is decisive.

You can keep track of the bets like this: Every player has somewhere in front of him cards/mats representing all of his opponents and will place all bets against him on that card/mat as well as possible bet token.
Note that for 5 players there are 10 pairs so this variant is not very desirable for more than 4 players.

Variant B: Poker-like multiplayer.

Each player has only one bet (on his winning of the game). For n-player game there are (n-1) bet tokens available, players start the game with none. Also each player starts the game with an initial bet of 6 crowns. As usual in a cleanup phase an active player can raise his bet if he doesn't have a bet token.  If he does so there are two possibilities either all other players have a bet token in which case they all discard it and the active player takes one or one or more of his opponents does not have a bet token in which case he just takes one. After raising all opponents in game order either call or pass. A player who passes is done and cannot win anything, he can end if he wishes to. At the end of the game all players that have not passed compare their score. The player with the most VP checks for tie with the player with second most, if yes then the second with third and so on. All of these players are considered to tie and they split all of the bets among themselves evenly.

Note that in this variant one may want to pose more strict caps at betting than just doubling the previous bet and it is also not so clear how to deal with all-ins, maybe exactly as in poker, or maybe ruling that a player cannot raise more than to make any non-passed player all-in.

Last two notes. I think that eliminated players should have the opportunity to continue playing as dummy players with no betting. And that since in the multiplayer it is not clear how to rotate game order between games I suggest that the player with lowest score at the end of previous game will choose his game order first and so on.

[jonts26]

Hi, welcome to the forum.

You have an interesting idea here. The betting mechanism sounds sort of like the doubling cube from backgammon and it could certainly be a fun addition. As for giving the first player a disadvantage, that has been discussed before http://forum.dominionstrategy.com/index.php?topic=20.msg101#msg101.

The only thing I disagree about is calling close games a tie. I actually think it would be somewhat unsatisfying for the player with more points to not get a win. And there is often a lot of skill involved in eeking out a win, like setting up a 3 pile ending etc. And if the small margin of victory is luck driven, well as you said before, there's a lot of luck in dominion. Just play a bunch of games.

[Bron]

Thank you for your comment.

Yes doubling cube (and poker obviously) is exactly what inspired me and it was discussed before (I provided also a link). Thanks for the link for first player advantage since it is so much written about it, I missed this particular discussion. My "first-player-advantage" fix was also secretly meant as "weird-tie-rule" fix. For example in a 2p game first player can
1) end and win
2) give up = end and loose
3) try improving his position and let opponent go.
but second player has one more option 4) end and tie. So the tie rule is from strategic point of view asymmetric. My first rule symmetrizes it by disallowing ties.

If game ends with a tie it is kind of unsatisfying for both players but if a game doesn't matter - winning a whole session is what matters - then having a (quite wide) possibility of tie can add an interesting strategic element. (ok this is hypothesis, I will try to convice my friends to test it ) So that is another reason for new tie rule. But still its nothing essntial and the rest can work well without it.

But let me add one more comment. I understand that sometimes close games are well deserved victories, but I still think that mostly no. The situation that you have described, where one player surprisingly ends on piles and wins by few points, is an example of deserved victory, but I think that you can do it mostly only against an inexperienced player, what do you think?

[Powerman]

Thank you for your comment.

Yes doubling cube (and poker obviously) is exactly what inspired me and it was discussed before (I provided also a link). Thanks for the link for first player advantage since it is so much written about it, I missed this particular discussion. My "first-player-advantage" fix was also secretly meant as "weird-tie-rule" fix. For example in a 2p game first player can
1) end and win
2) give up = end and loose
3) try improving his position and let opponent go.
but second player has one more option 4) end and tie. So the tie rule is from strategic point of view asymmetric. My first rule symmetrizes it by disallowing ties.

If game ends with a tie it is kind of unsatisfying for both players but if a game doesn't matter - winning a whole session is what matters - then having a (quite wide) possibility of tie can add an interesting strategic element. (ok this is hypothesis, I will try to convice my friends to test it ) So that is another reason for new tie rule. But still its nothing essntial and the rest can work well without it.

But let me add one more comment. I understand that sometimes close games are well deserved victories, but I still think that mostly no. The situation that you have described, where one player surprisingly ends on piles and wins by few points, is an example of deserved victory, but I think that you can do it mostly only against an inexperienced player, what do you think?

I wouldn't say that.  I would say roughly half of the games I win that are well deserved are close games.  It's the nature of the game.

[eHalcyon]

Point differential is not a good measure of how much better one player is than another.  A much stronger player might win by a single point thanks to watching the piles and managing +Buys, such that you are able to pile out on (for example) Villages and Wharfs with a single Estate, while the opponent had no chance anyway because of the other player's superior engine and purchasing power.

Similarly, a very close game could end in a massive point disparity -- suppose both players are equally matched and building towards a megaturn.  A Goons engine would be a very good example of this.  Despite how close they are, one player may get lucky and trigger their megaturn first, piling out and earning hundreds of points in the process.  The point difference is huge but the game was actually close -- a cut of the deck might have let the other player get their megaturn instead.

[Bron]

I somehow convinced myself that my original tie rule is not a good idea, I just had to put what both of you said into numbers. In case that anyone is interested, here is the reasoning. Say that we have two players playing - one is stronger i.e. he has probability of winning p>0.5 so his expected number of victory points is p+(-1)(1-p)=2p-1>0. Now we introduce a tie rule that changes some games into ties independently of the skill  (I suppose that my rule is such) - say that the fraction of ties is q. The expected number of victory points for the stronger player is (1-q)p+0q+(-1)(1-q)(1-p)=(1-q)(2p-1)<2p-1 so this rule favors weaker player, which is undesirable. Now it seems to me that it is a good idea to avoid ties completely.

Or maybe a tie rule that would involve a skill to use. What would you say to this. In addition to betting a player would have another option namely to "offer a draw", what would you say to that.

[jonts26]

What would be the strategic benefit of offering a draw? Or at least, if it is optimal for me to ever offer a draw based on scoring conditions, wouldn't it then be optimal for my opponent to refuse and vice versa? I suppose there could be skill involved in recognizing optimal times to offer draw, but it seems just like the reverse of the doubling mechanism and less interesting.

 Also, it does odd things with 3 or more players.

[chesskidnate]

I think the idea of the draw would be to mitigate luck if it was a situation where it seemed like the winner would just be decided by better shuffle luck and both players seemed to have the same winning chances. That said, it is pointless because if one person just chooses to decline draws then the addition of this rule doesn't change anything

[Bron]

1) I think that it does nothing odd in my proposed variant A.
2) Of course if offering draw is optimal for one then declining is optimal for the other, but if all players would play optimally they would always play the same strategy and the game would be pointless - the thing is that it requires a skill to assess how good is the choice I have.