The VP counter options should be: only on, prefer on, prefer off, only off (or, prefer on/off, require preference/don't). I would default to prefer off. "Prefer off" would match you against absolutely anyone available - it's only a preference - but would pair you with an only-off or prefer-off when possible. A mix of prefers with no onlies would result in off.
I expect what would happen is most people would be using the default. If that's "prefer off", then if you select "prefer on" then you're mostly going to be playing without the point counter just because your opponents will mostly be running default settings. That means if you actually want to play with the point counter, you'll need to select "only on". So "prefer on" would be semi-useless if "prefer off" is default.
I don't think "prefer on" would be useless, but whatever; Dominion, the game Dominion, has no point counter. If people tend to have no point counter in online Dominion, that's good, that's the game as intended. The point counter is there to be friendly to people who want it. If you can force it on or off then everyone who cares is happy, and no-one is harmed by having a "prefer on" option available.
I wasn't clear in my previous post, but I was trying to compare to a 3 option solution. My argument is that the proposed 4-option solution effectively works the same as the 3-option solution.
In my opinion, the canonical 3 option solution is for the options to be: yes, no, don't care. The default is "don't care". Players whose choices for "yes" and "no" differ won't be matched with each other. If I remember correctly, this is what isotropic used, where if two "don't care"s were matched it was random whether they had counter on or off. But let's say that if two "don't care"s are matched, that the counter is off, for sake of discussion.
Pros:
- Everybody who cares always plays with the setting they want.
- Everybody who doesn't care or keeps the default can be matched into a game with anybody (making their matchmaking better/faster, hopefully).
- All players, no matter what choice they made, can be matched with players using the default option.
Cons:
- People who have a mild preference for yes must select "yes" or most of their games will be non-counter, but then they can't be matched with people who dislike the counter so strongly that they select "no".
Now, my view is that the con is so mild that it's not worth worrying about. If there are a decent number of players using the client, there will be enough people using the default option that matchmaking can still be fast-ish even if "yes" players can't be matched with "no" players.
But let's say we want to try to fix the con. Does the proposed 4 option solution fix it? I argue that it doesn't.
As proposed, the 4 options are require yes, prefer yes, prefer off, require off. The default is "prefer off", and if two "prefer" players are matched, then the counter is off.
Now, imagine you are a player that prefers counter on but are willing to sometimes play without the counter. You might try selecting "prefer yes". If you do, every time you are matched with a player who didn't change the default selection, you will play without counter. In general, most people don't change default options; let's say 80% stay on default without considering alternatives and 20% make a decision about the option (which could be to stay with "prefer off"). Then as the player who selected "prefer yes", you will play <20% of your games with the point counter.
As the player who prefers to play with counter, if you are playing <20% of your games without counter, you are not happy. I'd expect most such players to switch to "require yes". After all, you can still be matched with everybody except those who chose "require off", and
if you believe the poll then <1% of players would choose "require off".
If every player follows the same thought process, then we end up where players pick one of "require yes", "prefer off", "require off". If those are the only options in use, it's exactly equivalent to the 3-option "yes", "don't care", "no" system. Given a choice between two systems which have the same effect, I'd pick the system with fewer options for simplicity.
If you go to a 5-option system ("require yes", "prefer yes", "don't care", "prefer no", "require no"), with "don't care" the default, then it does give more meaningful choices than the 3-option system. Is it worth the complication of having more options? I don't know. My guess is it's not worth it.
In addition to having fewer checkboxes, this has the advantages of not segmenting the player base and not further complicating the matchmaking algorithms.
Not letting me force VP counter on? When I and others quit Dominion, the player base will be "segmented" into those who have quit and those who enjoy Concentration.
I doubt that the number of players who quit over this (occasionally having to play without a VP counter) would be large. And any new players wouldn't know any different.
Well,
according to the poll, 13% of f.ds members say they only play with the VP counter on. If you think it's okay to kick those people off the service, well, consider an alternative of having the point counter on for every game with no choice to do otherwise. You'd drive off the people who voted "I only play with it off". How many of those are there? Zero. 95 people voted, and of those 0 refuse to ever play with the counter. Plus, instead of having to do convoluted matchmaking systems, it's super simple, you just always play with it on.
I respect the arguments that it should be possible to play without the point counter (not least because the game's designer considers that the canonical way to play), but if I worked at MF and in charge of making UI decisions, the decision would be a no-brainer. 85% of people who bother to vote either don't care or prefer the counter on. Having it on is a dealbreaker to 0% of the people. Although this is a poll on a hardcore community, it's safe to assume that the non-hardcore would mostly fall into the "don't care" camp. Why would I spend development resources on making counter-off an option when so many other things about the service don't work? It would go to the "do this sometime later maybe" category of tasks.