Print

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

[nopawnsintended]

I want to preface this post by saying that I love the match-play scoring system used in Gokodom.  What follows is not a critique of this system, but an alternative scoring system that might be fun to try in a small (or large) match-play tournament.

Total Victory Points Scoring.  This system has the following features:

• Each match consists of an even number of regulation games of Dominion.
• Each player has the opportunity to sit as first player the same number of times.  Alternating first seat is an option, but you could use the Gokodom seating chart, too.
• The score for the match will be determined by the player who obtains the largest number of VP across the games in the match. The player with the most VP wins the match. 
  
  • In the event of a tie in the number of VP, the player who took the fewest turns across all games wins the match.
  • If, in addition to having the same number of VP, both players took the same number of turns, the player who won the most regulation games of Dominion wins the match.
  • If the number of points, number of turns, and number of regulation Dominion wins is equal for both players, the players rejoice in their shared victory.
• In the event of an outright win, two points are awarded to the victor.  In the event of a shared victory, both players earn one point.  Non-winners get zero points.
• To make sure everyone gets to have fun, you could run some number (to be determined) of Swiss rounds where playing better means you play better opponents.  At the end of the Swiss rounds, the best players could match up against one another in a bracket-style tourney.

Basically, I got the idea by watching a re-run of a Jeopardy Teen Tournament Final.  This scoring system is to Dominion as pooling winnings across multiple days is to Jeopardy.  And, if it is good enough for Jeopardy, it is good enough for me.

Moreover, I think it would be fun to play a Tournament that is governed by these rules.  If I had the ambition and/or time, I would organize a tournament to be played through Goko's lovely interface that adheres to these rules, but I don't have time (yet... right now... we'll see). I think this tweak in the scoring would lead to some interesting changes to how people play.  To cite two examples: (1) the scoring system would tamp down risky endgame strategies that rely on praying for the perfect storm -- i.e., it will usually be optimal to end a game while behind so as not to lose by too much; and (2) the scoring system would facilitate running up the score when you could easily end the game.  Obviously, the latter scenario is an issue in fun.  A modified system could mitigate this by capping the margin of victory in any particular game, but I'd rather not modify the system that way.

Some questions for discussion:

• Would you play in a Tournament governed by a refined version of these rules (keeping the spirit of Total VP Scoring)?
• Are there natural extensions to 3P or 4P matches?  Multiplayer games can be fun, too, and I find that my only excuse to play multiplayer is if there's a multiplayer tournament (OK, not really... looking at you Polk5440 and Bella Cullen).
• Are there any fun hybrids of Total VP and Gokodom style matches?  One thought: You play 6 games, and each match has a Total VP point and a match point up for grabs.
• How often do runaway boards happen where one player wins by an insurmountable margin in one game, making the rest of the match moot?  If this probability is low enough, maybe the system is as broken as the KC-Discard-Masq pin.  If it is likely, what changes would fix it?
• Any other thoughts?

[cluckyb]

I don't really see the benefit of this. The expected margin of victory is just too varied depending on the kingdom. Things like piling on the score are just no fun for the loser, whereas taking those risks that give you the 60% chance of victory are what can make the game exciting.

Or even in simpler cases, say you're up +5 points from the last game. There are two provinces left and you're down a duchy. Duchy dancing really isn't the proper move here. If you buy a duchy, the other guy can possibly win by 9 if you get a poor hand next turn. Better to hedge your bets and just lose by three. So you wind up taking the boring safe approach.

The idea of doing the same kingdom twice (alternating first player), and adding those scores I think is reasonable. But a whole tournament like that sounds broken and not much fun.

[ftl]

I don't think adding scores works at all. The scores of different game types are just two different.

If the first game is a KC-goons-engine-colony game, a score of, say, 100 to 30 might be a fairly close game - maybe p1 got their mega-goons-turn just one turn earlier than p2 would have done the same. But good luck overcoming a 70-point difference in the upcoming five province games!

Or, a KC-KC-Monument-Militia-Masquerade pin. In that case, the winner can literally run up the score as high as they want before ending the game.

And so on and so forth. That's the extreme case, but it repeats in smaller ways in other games. The number of points you win by is often more a representation of what the game was like rather than how much better you played than your opponent.

BM-ish province games have margins of 3-9 points. Engine games have a margin of either 1 point (win on piles, buy one estate) or 25+ points (piles don't get low enough, somebody gets their engine going and wins 6 provinces to 2 or so) or 50+ points (colony games, vp chip engines, etc). Rush games have tiny margins of victory, like winning by a few estates or by a 3-5 gardens split where each gardens is worth 2. Games where you build an engine which includes a trashing attack have margins of victory which are as big as you can make them - just trash the opponent's entire deck, lock them down, buy up VP.

[Warfreak2]

Don't forget Sea Hag games where you win -2 to -7.

[nopawnsintended]

Some comments.

Or even in simpler cases, say you're up +5 points from the last game. There are two provinces left and you're down a duchy. Duchy dancing really isn't the proper move here. If you buy a duchy, the other guy can possibly win by 9 if you get a poor hand next turn. Better to hedge your bets and just lose by three. So you wind up taking the boring safe approach.

Why is that a "boring safe approach"?   Properly thought through: You have a lead of 2 with two Provinces remaining (+5 from last kingdom and -3 from this one), so buying the Province is the right play.  In a single game of Dominion, is it boring to buy a Province when you're up 2 VP? 

Consider a couple of counter points. 

(1) Suppose you're up 5 after 3 games, and in Game 4, you're behind 3 with two Provinces remaining (and no other sources of VP).  In the match, you have two games to go after this.  Do you Duchy Dance?  I haven't properly thought it through, so forgive me if that's a dumb question, but it is new decisions like this that motivated my proposal/suggestion.
(2) Go back to your example.  If you're down 5 VP going into the final game, don't you adopt a riskier strategy that won't just win if it works, but will have a chance of making up the 5 point deficit?  That's a similar principle to Duchy Dancing, but it has a longer strategy arc.  Seems exciting to me, anyway.

A meta-point.  After you've done the Duchy Dance 200 times, is it really that exciting to get into another run-of-the-mill Duchy Dance?  Don't get me wrong.  I love Dominion as is, and Duchy Dancing is part of that love, but this post is about imagining a different variant on the scoring system.

And so on and so forth. That's the extreme case, but it repeats in smaller ways in other games. The number of points you win by is often more a representation of what the game was like rather than how much better you played than your opponent.

BM-ish province games have margins of 3-9 points. Engine games have a margin of either 1 point (win on piles, buy one estate) or 25+ points (piles don't get low enough, somebody gets their engine going and wins 6 provinces to 2 or so) or 50+ points (colony games, vp chip engines, etc). Rush games have tiny margins of victory, like winning by a few estates or by a 3-5 gardens split where each gardens is worth 2. Games where you build an engine which includes a trashing attack have margins of victory which are as big as you can make them - just trash the opponent's entire deck, lock them down, buy up VP.

This is a fair point, and an admitted weakness of the total VP scoring system, but the system is fair before knowing whether you've won one of those big payoff games.  That is, before the game, either player has a chance to be the player who get the win (which on different kingdoms and different shuffling have different expected value of points).  In principle, I don't see anything wrong with some games having larger prizes (I win by 20 versus 1), but I would be worried about frequent, degenerate gameplay.  Adding VP across games just puts more weight on winning those kingdoms that have greater expected margin.

Don't forget Sea Hag games where you win -2 to -7.

That's pretty similar to a regular Province game where you win 33- 28.  Going back to ftl's point, Sea Hag games probably have a lower expected margin of victory than regular Province games.  I don't mind putting less weight on ugly low margin games, but that's my taste, and there's no accounting for taste.

Most of all, I think it would be interesting to play a tournament by a different set of rules than ordinary Dominion... just for variety's sake.  As I mentioned at the outset, I love the Gokodom setup.  Maybe something else might be fun too?

[Warfreak2]

There's also the consideration that normally when you are ahead you should end the game as soon as you have the chance; sometimes, one player just has total control over the game, and the polite thing to do is just buy a Province and empty the last pile. This would normally be one of those "win by a point" engine three-piles. In this tournament, though, if I win the Ambassador war and my opponent has 20 Coppers and rising, I should prolong the game so that I can buy all of the remaining VP on the board. That doesn't sound fun for my opponent!

[nopawnsintended]

There's also the consideration that normally when you are ahead you should end the game as soon as you have the chance; sometimes, one player just has total control over the game, and the polite thing to do is just buy a Province and empty the last pile. This would normally be one of those "win by a point" engine three-piles. In this tournament, though, if I win the Ambassador war and my opponent has 20 Coppers and rising, I should prolong the game so that I can buy all of the remaining VP on the board. That doesn't sound fun for my opponent!

Agreed that this sometimes happens and that this scenario would be no fun for your opponent. In this scenario, I suspect that non-Amb opponent would do what he could to minimize the damage, including trying to 3-pile himself (maybe that's interesting).  That situation is a bit torturous (and Amb is one of my least favorite cards).  That said, I'm all for trying to avoid a total breakdown in Dominion decorum when it comes to running up the score.  Any ideas on tweaks to the system (caps on margin of victory, etc.), or is there no way to make this fun?

[Warfreak2]

I think there's no way to make it fun. The only way to make it sane would be to play all the games with the same board, but that's no fun either.

[DG]

It's possible to create scenarios where points scoring completely fails. You could imagine a big vp chip scoring deck that just continues to accumulate points until the opponent buys out the copper, estate, and curse piles. That would be an awful method of deciding anything.

[nopawnsintended]

It's possible to create scenarios where points scoring completely fails. You could imagine a big vp chip scoring deck that just continues to accumulate points until the opponent buys out the copper, estate, and curse piles. That would be an awful method of deciding anything.

That's not the nature of the question that I asked.  Returning to regular Dominion, you can construct a board where a game takes 40 turns and it isn't possible to hold onto your VP because it keeps getting trashed (thinking... Saboteur, Knights, Rogue, Village support, and no way to counter it).  Based on that possibility, I'm not going to say that Dominion is an awful game because that Kingdom (while it sometimes comes up) doesn't happen often.

To take an example from math, it's possible to create scenarios where the simplex algorithm is terrible (worst case computational time is exponential), but in practice, it's not so bad.  And, as a result, it is taught and used by people who want to solve actual problems.

http://mathworld.wolfram.com/SimplexMethod.html

Returning to Dominion, my question was about the average case,* not the worst case.  Across random kingdoms, are the scenarios you can construct going to occur more than once out of 100 Kingdoms?  If the opponent is playing optimally (or at least trying to), are runaway victories going to be so pervasive that it would sap the fun out of it?

*Average is a bad word here.  What I am more interested in knowing is something like the 1st percentile case or the 5th percentile case.  In other words, if you play 20 matches under this point scheme with random kingdoms, how many "unfun" runaway victories will you end up with?  Anyone want to test that out?

[DG]

When considering a tournament scoring system you do have to consider the worst cases.

[Polk5440]

So, I think the analogy to Jeopardy! tournaments is not a very good parallel. The only thing that changes in Jeopardy! tournaments is the strategy for betting on daily doubles and final jeopardy. Having a Dominion tournament where VP matters changes the game so much in ways that promote stalling/prolonging the game that I can't see how on average it would be more fun. Jeopardy! doesn't have stalling problems because there are a set number of questions and a time limit. I just don't think Dominion is designed for points to be comparable across kingdoms (esp alt VP) and that's the fundamental problem.

I do like the idea of doing a Dominion variant tournament, though. I would love to do a Dominion duplicate tournament, for instance.

[ftl]

*Average is a bad word here.  What I am more interested in knowing is something like the 1st percentile case or the 5th percentile case.  In other words, if you play 20 matches under this point scheme with random kingdoms, how many "unfun" runaway victories will you end up with?  Anyone want to test that out?

The opposite case is also unfun.

It would be very depressing to pull off a clever 3-pile to win a game with a 1-0 score... and just to realize that that 1-0 win really isn't any different than a tie, because last game was a Colony game where the winner won the colony split 5-3 for a 20-point difference.

A 5-3 colony split isn't a degenerate case. It's a pretty reasonable outcome of a close colony game. And yet that's a 20-point difference - enough to completely overshadow results from other game types.

[Kirian]

This is an alternative that is suggested pretty often, actually, which means most of its flaws have been exposes and explored multiple times.  Usually it's proposed as a tiebreaker for tournaments, not as the tournament decider.  The major flaws are these:

1. Some games are going to be won by just a few points either way.  Others are going to end on a megaturn.  Winning just one megaturn game means you have a string advantage for the rest of the series.

2. Different boards create different games of Dominion.  This is the point of Dominion, but it makes this scoring system inadequate, as some games will be more important than others.

3.  It can create untenable equilibria.  Infinite loops where neither person can reasonably take steps to end the game, that sort of thing.

4. It forces people to play out a game that is an obvious win, but still will take ten minutes to finish, because it's suddenly impossible to resign.

[Mic Qsenoch]

Would it kill you people to not repeat what everyone else has already said in a thread?

[DG]

Even Donald X. has gone through the problems of aggregate points scores in the past and explained why he advised against them in official tournements.

[WanderingWinder]

I'm not going to say it's necessarily bad (man, I don't think I'd like it, but that's a matter of taste), but what I will say is that it's significantly different from Dominion. In Dominion, your goal is to win the game, which means for instance you have all kinds of endgame strategies based on maneuvering around pile-outs and such. In your game, there isn't that, but there's points-running; the objective is to no longer win the game, but to maximize your score differential.

[nopawnsintended]

I'm not going to say it's necessarily bad (man, I don't think I'd like it, but that's a matter of taste), but what I will say is that it's significantly different from Dominion. In Dominion, your goal is to win the game, which means for instance you have all kinds of endgame strategies based on maneuvering around pile-outs and such. In your game, there isn't that, but there's points-running; the objective is to no longer win the game, but to maximize your score differential.

I agree with this assessment, and maybe I should have emphasized that what I had in mind was a Dominion variant tournament rather than a Dominion tournament.  That said, it looks like I'm by myself in thinking that this would be a fun idea.

Thanks all for your input.

----

To combat the runaway points idea, I had this thought: In a 6-game match, each player can throw out the worst loss so that only 4 games would count toward the VP total (this would allow resigning once, for example).  Not sure if that would make people feel better, but I have a feeling that the answer is no.  /end{idea}

[() | (_) ^/]

Would it kill you people to not repeat what everyone else has already said in a thread?

Yeah, would it kill you people to not repeat what everyone else has already said in a thread?

 

[Warfreak2]

There is a Go tournament which uses cumulative scoring. Some of the games are broadcast on the internet, and are typically very silly. Occasionally someone suggests using point score as a tiebreaker in a real tournament, but Go would have basically the same problems as Dominion if it used this system.

Paddy, could you not repeat what everyone else has already said?

[() | (_) ^/]

...
Paddy, could you not repeat what everyone else has already said?

No.