Bidding on Nations

[MaxWilson]

There was a thread a while ago about bidding on nations in MP by bargaining away pretender points. HoneyBadger recently expressed the opinion that a lot of the nations could stand to use some rebalancing of their militaries to make some of the underused units useful. Traditionally wargames get balanced through bidding. Thus.

Balancing in single-player (which I play exclusively) is easy--just create a mod and incrementally alter every nation until they are all equally attractive. For me, since I like big maps, this would mean a lot of races would have capitol-only restrictions removed. I'll know I've succeeded when I'm happy playing on Nation: Random (knowing that any choice I get will be fun). However, this doesn't do much to balance the races to competitive play. It just customizes the game to my playstyle.

In MP, you can't offer bonuses to weak nations, say EA Ulm, as simply because EA Ulm has more than one opponent and they might disagree on how much to offer. I might need to dig through the multi-agent decision literature to find a good auction method, but at least one simple one comes to mind which is inspired by the traditional problem of cutting cakes fairly: I cut it in half and you choose which half you want, so my incentive is to cut it so fairly that it doesn't matter which half you choose. You could adapt this for the multi-player case as follows:

Suppose there are four players in the game, A, B, C, and D. A gets to declare any improvements to any nations in the game that he wants. For instance, Jotunheim PD is improved to 1/2 Jotun Javelinist per point, and Marverni Druids get move 2/9. B gets to add any modifications that he wants, but can't take away any of A's modifications[1]. C adds some more, and then D doesn't get to mod any but gets first choice of which nation to play. C chooses from the nations which remain, then B, then A. There's still some potential for abuse here--C could triple the hit points of every unit belonging to both Ulm and C'tis, ensuring that D will pick one and he'll pick the other, which basically wipes A and B out of the game--but C can't put himself at an advantage relative to D. You can deal with the potential abuse either by taking turns being A, B, etc. or perhaps by ruling that you can never play any nation that's been modified by you.

It's tough to predict whether the nations would converge over time to new balances (differing depending upon the size and type of game) or whether every game would be different than the one preceding. I predict that the nations would still remain generally thematically true, even though you could technically "improve" Sauromatia to be identical to LA Ermor.

Another possibility which is a little bit simpler: everybody submits a list of potential modifications (call them "themes") to every nation in the era. Players are randomly assigned to nations, and get to choose any theme for that nation that's been submitted, or the vanilla nation. For instance, "Gorgon pretender is available to every land nation" is a theme for any nation, but can't be combined with a theme from another player that says "White Centaurs get scale mail armor." Downside to this method is that you don't get to choose your nation. Upside is that no player is at a particular disadvantage (as A was to C before) and so you don't need extra complexity to make it fair.

I've no idea if anyone is actually interested in bidding in MP games, but this is at least one way this could be done.

-Max


[1] This is kind of hard to define in practice--is changing Spear to Shortsword on a unit that's already had Defense modified "undoing" a change?--but you could say that A always gets a choice of his version, B's version, or C's version, while B can choose from B's or C's versions and C can take his own or nothing. So C can add options for A but not remove them--the idea is all nations are supposed to be equally acceptable to you once you've had your chance to mod them and any extras your opponents add are just gravy to be exploited.

Edit: added a second bidding method.

[DrPraetorious]

This is a lot of work, and also rather drastic.

Firstly, I don't think trying to reward underused units is wise. Some units are useful only under certain niche circumstances, and that's good and shouldn't be "messed with".

Now, the game does a remarkable job of balancing the different *nations*, which is what is actually important. However, it isn't perfect - especially under non-"standard" play conditions.

So, here's the deal -
First, everyone bids on each nation they want (gold or initial gems or whatever). Secret bidding. You bid on (number of players) + 1 nations.

Nations go to the high bidder one at a time until everyone has been assigned to a nation.

Then, everyone gets initial gold (or some other bonus) for each opponent that did not bid on the nation to which they ended up assigned (or if the total of all bids on that nation is low, or something like that.)

So you are encouraged to bid on weak nations, on the expectation that opponents will not do so, and you will get an intial bonus, also you can probably get them with a lower bid.

Thoughts?

[MaxWilson]

It's a nice, simple scheme. It does assume that any putative imbalance could be addressed with an initial infusion of gold (or whatever). Since Dominions seems to work fine already without any bidding at all I think this is a reasonable assumption to make, but Dominions supports nation design well enough that unit changes are also quite feasible as bids. The technical challenge is designing the bidding process to be fair, and the complexities (and most of the "lot of work") come from trying to address that. I think I mentioned that I need to search the multi-agent literature to see if there's not a simpler bidding mechanism, and if I find one I'll report back even though perhaps no one but me will ever use Dominions to play this type of metagame. If you're not interested[1] in nation-design it's a lot of work for no payoff, and you might as well just use gold for bids.

Now, I don't claim to be any kind of an expert on the game, but there are some units that seem to be to be genuinely underpowered. I know as Helheim I've never built a spear-carrying Huskarl because they're nearly as expensive as a Hirdman and have five points less in Defense, 2 points lower Prot, and do 3 points less damage. They're clearly inferior for their gold cost, and in any case where I'd want a cheaper unit than a Hirdman I'd either build serf warriors or indies or maybe take the axe-carrying Huskarl. The spear-Huskarl might as well not exist as far as I'm concerned, and I would be much happier if I could at least conceive of a circumstance where it would be the right unit. It needs a niche. The Marverni noble in comparison to the Ambibate noble comes close to this level of uselessness in my mind, but at least it's 20% cheaper and also thematic. (If Ambibates have nobles Marverni should, too, or have a good reason why they don't exist. Maybe they're all smiths and chainmail-salesmen.) I'm still very new to this game and I may very well be overlooking uses for both these units, but there must also be underutilized units I don't know about yet (Yomi footsoldiers?). I'm all for niches, combined arms, and comparative advantage, but for that to work there has to actually exist some comparative advantage for each unit. I intend to tweak all the nations sufficiently to give every unit-type a reason for existing in my SP games. I'll probably turn off independent pop-types, too, for kind of the same reason (e.g., Marverni slingers are generally inferior to indie slingers).

-Max

[1] I don't mean you specifically here, just the generic "you." IIRC you've heavily involved in modding, which means if a good, simple bidding method can be found you could potentially use it to help balance new mods against vanilla nations.

Edit: Added comments about underused units.

[Dedas]

Well, believe it or not, against certain enemies with low defense and morale (though both can be lowered with cheap magics like "panic" and "desiccation") a spear is a very good weapon for repelling; and an axe might even be overkill. Also, does not the huskarl have 2 javelins? Javelins are good my friend, especially if you are going to tie up some axe (lower range than spear) wielding no shield guys. Just set them to fire and they will soften the enemy up before they charge them with their spears. Slingers are also good for this, making the enemy vulnerable to low damage weapons. All this reduces the need to have heavy weapon troops with low attack, defense, shield etc, thus raising the survivability of your troops.
But of course, sometimes spears and javelins suck big time, for instance against EA Ermors tower shield troops. Because against those guys you really want as big weapons (and many at the same time) as you can get. Or maybe Morningstars (+2 to hit against shields) or pikes (no shield parry); the problem is getting trough their armor with low damage weapons such as these though, but with increased strength it is a piece of cake.

My main point is: You can't really take factor A and measure it against factor B and say that it should be raised or lowered because there is also a factor C, D and E to take into account. Heck, there is so many combinations in this game that I sometimes get dissy.

However, if a spell or a unit really is inferior in 99% of the possible scenarios imaginable I would really want a "fix" for it. A recent example of this would be glamour which does not really have any good counter (poison?).

[Dedas]

Oh I forgot this one. A good way to use some units in some cases is to mix them in them in the same squad. For example 50% of the quite vulnerable amibdex forest warrior and 50% shield maidens (they both have a AP11). They complement each other wonderfully as one deals damage and the other ensures that the squad stays above 50% losses so not to rout.

[Sombre]

Although the forest warrior probably isn't worth the cost because its attack rating is quite low. You should just go with steel maidens if you want maximum killing power.

It's all well and good to say unit X is actually ok because in situation Y it's almost as good as the unit everyone builds,... but seriously, who is going to bother trying to wring the most out of units like,.. say,... the Mictlan fire warriors. Jaguar Warriors are SO much better than them that you would pretty much NEVER want to use them. That punishes the player for wanting a bit of thematic play and variety in my opinion. Which is the reason I play CBM and I'm tempted to make my own version, CBMPlus, which further addresses underpowered / 'useless' units.

[Xietor]

What are top 5 MA races, and how much is typically bid?

What are the bottom races, that no one ever bids to get?

[MaxWilson]

As far as I know, bidding simply isn't done in Dominions3. Yet.

-Max

[Gandalf Parker]

Someone needs to come up with a way of enforcing it and paying off. There is a "god reader" in the works which would allow the host to verify that a build actually didnt use their bid points. But for the other methods Im not sure how they handle it.

The only alternative I can come up with on the fly is to give each nation something and let them bid it down. Such as, an extra 500 infantry (generic indept type). If someone bids 100 for their god then you only give them 400 extra infantry before the game starts. (using set start positions and map commands)

Gandalf Parker

[MaxWilson]

Dedas,

Glamour is quite hard to address on the strategic map. The only counter I can think of is lots of scouts, attacking each province every turn and then retreating. According to Kristoffer, glamour will be altered in the upcoming patch. Details have not yet been divulged except that it's a change to battlefield glamour, not strategic glamour.

The spear-wielding Helheim Huskarls we've mentioned before--sure, a spear can be a decent weapon for repelling, but that Huskarl is three times as expensive as a serf warrior with a spear and IIRC the same defense, and almost the same price as a Hirdman with 5 points higher defense. In order for the Huskarl's repel to be better than the defense boost, you'd need to be facing a foe with a defense and morale so low that 2 out of 3 attacks were repelled. (And his weapon has to be longer than a broadsword or else the Hirdman repels him, too.) Does such a unit even exist? And if it did, is there any reason you wouldn't be fighting it with cheap serf warriors and indy light infantry, or even better with mounted Hirdmen? Mounted Hirdmen have javelins, too, cost about double what a Huskarl costs in gold, have 11 Prot instead of IIRC 9, defense 19 instead of (14?), and have a weapon the same length as a spear and so equally good at repelling. Unless I'm misremembering the Huskarl's stats I'm still not seeing a situation where you'd want to use him.

Edit: to put it another way, if we were bidding I would cheerfully offer a spear Huskarl with Prot 11, just like the Hirdman, in full confidence that it wouldn't help you at all.

-Max