Monday, May 1, 2006

Here I Stand, and the hazards of big decks

An observation on the sizes of the strategy card decks in recent GMT games:


GMT Games that just happen to include 110 Strategy cards: Here I Stand, Twilight Struggle, Wellington, WWII: Barbarossa to Berlin, The Napoleonic Wars, Thirty Years War, Paths of Glory, For the People.

GMT Games that have some number other than 110 cards: Wilderness War (70), Empire of the Sun (165), Sword of Rome (156).

In addition, the two games of this type on the P500 list, Kutuzov and Unhappy King Charles, are already slated to have – you guessed it – 110 cards.

I would consider it a monumental coincidence if all 10 of these games, from two-player single-deck games to two-player individual-deck games to multi-player games, covering conflicts from the Reformation to the Cold War, with game lengths running from 3 to 20 turns, all just happened to have worked out such that they really required 110 cards to work properly. I find it far more likely that the designers were told, "you've got 110 cards to work with on the press sheet", and they used all these slots up by picking their 110 favorite events from the period and figuring out how to express them in game terms. Don't get me wrong here, these designers are all smart guys and I'm sure every reasonable effort was made to fill these 110 slots in a sensible manner with interesting cards. It just seems pretty clear to me that not every game of this type is going to function best with 110 cards, and I do wonder if issues of what deck size will produce the best pacing, tension, and level of chaos have been thoroughly considered in all cases. Hannibal, for example (which I think we can all agree was a pretty good game despite its near-total lack of exactly 110 Strategy cards), had only 64 cards, despite dealing out, over the course of the game, 142 cards – 71 per player. Successors had only 56 cards in its deck, of which each player will see 25 in 5 turns. By contrast, Here I Stand has on average about 70 cards in the deck for the first 5 turns (trying to weight for additions and removals), of which each player is likely to see 20 in total (some will see more, some less, but not by a huge amount). Not a vast difference from Successors, but still, the deck is 25% larger while players receive 20% fewer cards. For a much starker example, in Wellington players will see maybe 24-30 cards from the supply of 110 over the course of the game. So the deck size has almost doubled while the number of cards players see over the course of the game has increased only slightly.

The problem here is that more cards means more randomness, without necessarily improving historical flavor. As we have learned from Betrayl at the House on Hill, randomness isn't really flavorful, it's just random. For randomness to impart flavor, there has to be at least the outline of structure to the chaos: players need some feel for what's coming down the pike, but for it to show up in surprising and interesting ways.

In your average card-driven game, some of the events in the deck just add some light, thematic flavor to the game. Then some more are events that are important to the underlying mechanics of the game (like the Force Marches and Campaigns in Hannibal). But a number are also critical cards in terms of driving forward the "story". For these cards, a reasonable level of predictability is desirable, because they represent the game’s momentum, its direction, its “soul”, if you will. In Hannibal, it's cards like Syracuse Allies With Carthage or the various Allies Desert; in For the People, it's the Emancipation Proclamation; in Here I Stand, it's a number of the Mandatory Event and "remove after play" cards like Master of Italy, the English monarch cards, or many of the Reformation/Counter-reformation cards. In Hannibal, you realize Syracuse's loyalty is sketchy, it's going to switch sides at some point – probably an inconvenient point, if you're Rome – and this factors into your game plan. The modestly-sized deck, large numbers of cards dealt out over the course of the game, and somewhat limited reshuffling gives the randomness some structure and predictability. In Here I Stand, though, a key card like Master of Italy (which gives VPs for dominating the Italian scene) may or may not come out at all over the course of any given 4 hours of play – so how do you plan for it, or is it just a basically random 1 or 2VP bonus to a player who is otherwise pursuing a specific strategy anyway? And if you're England and through bad luck are saddled with Mary, who cripples your position by giving half your card plays to the Pope, are you going to be happy about the fact that progressing to Elizabeth I (the best monarch in the game) is a bit of a crap shoot? At that point in the proceedings, she's 1 card in 80 or so. You're only likely to be dealing out 25-30 cards per turn, and if you don't get it this turn, the odds won't improve much for next turn since you reshuffle between every turn. And that's for a mandatory event, which anyone has to play if drawn. If you, as England, are looking for an important historical event like Dissolution of the Monasteries, you're only drawing 4-ish cards to get it from an 80-ish card deck.

This was my main difficulty with Wellington. With the huge card deck, 4 players, and a comparatively small number of cards dealt out over the course of the game, there is no sense of direction at all. It's a lot closer to Munchkin than Hannibal (in fairness, I kind of liked Wellington and would even play again. It's short. But as a design, it really felt slapped-together to me).

All this raises further complications for cards that provide the game’s flavor or plot in multi-player games, because not only do you need to deal out these cards in the first place, you also need to get them in the hands of a player who will play them, plus you need more of them, tailored for more positions. Here I Stand at least partially approaches this problem of "misallocation" of cards (having a card dealt out, but to the wrong person) through diplomacy and deal-making. However, this seems only partially successful because in practice it's hard to properly reward someone else for playing one of "your" cards on your behalf. For a specific example, take the Sale of Moluccas card. This card allows the (one) player who has successfully circumnavigated to draw two cards; it's a 3 CP card. If I've drawn the card and you've circumnavigated, we can make a deal: I play the card on your behalf, you draw two cards; enabling this sort of deal will help events find the right players. But the math is tough. I give up 3 CPs to do this. You get two cards, which, on average, nets you about 5 CP-equivalents. So in order to make this deal work, you need to give me stuff worth exactly 4 CPs. We would then both come out ahead by 1CP. This is hardly worth the effort of figuring out how to do it. Maybe two cards from the deck are worth a little more than 5 CP, on the chance that it might be a cool event for you which is worth dramatically more than its CP value, but this really is not a great bet. Regardless, this is a very small profit margin for either player, and, given the risks involved in a blind draw of two cards and the difficulties involved in getting the deal to work within these narrow parameters, probably not worth the amount of time it will take to work it out. So deal-making is at best a partial solution to the problem of "misallocated" cards, unless the upsides for both players can be made more significant (in some cases in Here I Stand they are; this is just an example of a card that’s going to be tough to play except for the fairly rare case of being dealt to the right person at the right time).

So, while intuitively you might think that more players would argue for a larger deck of Strategy cards in this sort of game, because so many more cards are being dealt out in total and you can make more cards tailored for each individual position, it's not clear to me that this is true. The fact is that individual cards will have a harder time getting to the players who can use them, and each individual player is usually drawing fewer cards (if only for practical playing-time reasons). This would actually seem to argue for smaller decks as numbers of players increase – and Successors does in fact have a smaller deck than Hannibal.

All this is a long way of coming to the point that I played Here I Stand again. I still like it. I'm much more comfortable now saying that while the rulebooks is long, the complexity isn't too bad. But I have the feeling that the card deck is still a bit too chaotic with not enough focus. It's leagues better than The Napoleonic Wars or Wellington, but it still falls short of the mark achieved by Hannibal or Successors. Most of the events that actually get played are the comparatively generic ones (Mercenaries Desert, Treachery, Foul Weather); the events that push forward the game's "story" seem hard to play because the odds of the right player having them at the right time are not great and deals are hard to strike because of their narrow parameters of profitability. And so some of the flavor that is the game's great strength is not as effective as it might be.

I had two thoughts about how this might be addressed.

The first idea I had was to fiddle with how the cards come out. Perhaps add a new category of cards, maybe called "key" cards. Cards that reflect important, highly-specific events that should have a reasonable probability of happening in a timely manner, something less than a Mandatory Event but more than a generic event. At the end of the turn, any of these events that were played for CPs (i.e., not played as an event), are re-shuffled into the deck. Then instead of reshuffling all the cards each turn, you just keep drawing until a specified "reshuffle" card is played (maybe Treachery, that's a highly-distinctive event). This would then be combined with more "remove" events; a number of the events in the deck seem like reasonable one-timers (Fountain of Youth, for example), but are not. As it is now, too many of the "remove" events are hard to play and so the deck starts to bloat with the large additions of cards on turns 3 and 4. The idea would be to keep all the existing cards, but combine a greater rate of card removal with a quicker cycling of key cards to give interesting cards a better chance to be at the right place at the right time.

This, though, would require non-trivial re-tuning of the design. It might be a promising idea for the next game, but retrofitting it would be too much work.

The easier answer is just to thin the deck. So that's what I decided to experiment with.

The goal was to target two kinds of cards for elimination. Firstly, cards that are generic and weak, such that they are unlikely to ever be played as events (because the ops are generally more useful) and their loss is unlikely to be missed in any event; and cards that are so highly-specific that getting them into play is a once-in-many-games kind of thing. I realize that sort of very infrequent event has a certain appeal, but if there are too many of them, they're crowding out the important stuff and making the game more chaotic without a payoff in "normal" games. I think that with the multi-player games, for all the reasons so far discussed, a harder line has to be taken against low-frequency cards. All this has to be combined with maintaining the current CP balance in the deck.

In summary: keep the CP distribution the same, but "punch up" the deck by removing low frequency-of-play events. This will (hopefully) make the individual events more attractive because we've both a) made it more likely that specialized events will be in the right place at the right time by reducing the probability that they are sitting in the deck undrawn, and b) made it more likely that a player's hand will contain useful events. All this will also help with c) making the appearance of individual thematically-important events come out in a somewhat more reliable manner.

So, the core deck has 10 1-CP, 20 2-CP (including 4 mandatory events), 21 3-CP, 7 4-CP, and 6 5-CP cards. My idea is to remove 2 1’s, 4 2’s, 4 3’s, one 4, and one 5. This isn't quite ops-neutral, but it's close. It would remove 12 cards from the core deck, taking it from 64 to 52. There are heavy additions to the deck on turns 3 and 4, but we'll deal with those later; most of them are “remove” cards, so a smaller deck going into turn 3 will see more of those cards played as events and removed, which will hopefully prevent the deck size escalating too badly.

Before continuing, I should say that this wasn't easy. When I went through this same exercise trying to trim down the bloated card deck in The Napoleonic Wars, finding 20 cards that I felt should be toasted was not a stretch. In comparison, in the main deck for Here I Stand, there were really only a couple no-brainers. But I think thinning the deck would have enough salutary effects to make it worthwhile to remove a few of the non-obvious ones.

Anyway, here is what I came up with:
1: Arquebusiers – Not a terrible card, but a couple 1's needed to go, and this card was the least interesting, and who's going to miss it?
1: Venetian Informant – Most games have a variant on this card (except Hannibal and Successors). The chances of it being useful in Here I Stand are rather remote, and removing uncertainty is not necessarily a good thing anyway. If the Channel 4 News Team has a hard time accurately predicting Foul Weather, I’m not turning to a Venetian Informant.
2: Tercios – Tiny probability of play and, as a combat card, it is very weak compared to more general combat-related 2-ops response cards like Foul Weather, Surprise Attack or the Mercenary-related cards, cards that aren't restricted to a single faction. This is a no-brainer.
2: Scots Raid – This is a little gratuitous, I felt, and it has a frequency of play that's a bit too low. The English already have enough incentive to crush the Scots; the presence of this card just makes it almost mandatory. If the timing is bad on this it could be devastating to the tiny English hand early in the game; but otherwise, Scotland is likely to be conquered. This is the card I am least comfortable in suggesting be removed, however.
2: Fountain of Youth – One of these exploration-hosers seems enough given the deck-thinning we're doing, and Search for Cibola is the more general card. Also, the exploration process is already very random, and the Hapsburgs rely on their explorers to make up ground in points – this card is really a straight VP kick in the teeth for them. Not just that, but it potentially lowers the number of VPs which they can possibly achieve during the game, which seems much harsher than the other hoser cards. The Hapsburgs already have a lot of problems.
2: Mercenaries Grow Restless – The hazard of this whole enterprise is that I'm messing with the game balance, possibly seriously. In this case, the putative removals are changing the value of mercenaries vs. regulars. The deck is being thinned by a little under 20%. There are four mercenary-hosers in the deck. So I selected the weakest one with the lowest frequency-of-play to remove.
3: Threat to Power – This card is gratuitously violent. Remove a minor leader for the entire game? Aren't these second-in-commands really just generic, in that the ruler would just appoint someone else? Permanent removal of any of these leaders seems a bit harsh.
3: Fuggers – This effect is uninteresting and very minor. You get a card you would have gotten next turn anyway. It's a no-brainer to play, but why bother?
3: Ransom – A nice capability to have available, but it has a tiny frequency-of-play.
3: Sale of Moluccas – This was already discussed.
4: Foreign Recruits – The presence of this capability does add an interesting variable to the game, but the yo-yo effect of Sprint Deployment/Winter Return means its frequency of play seems just too low (certainly the English, Protestants, and Papacy would virtually never have any cause to play it). The capability would still be present with Charles Bourbon, a rather more interesting card.
5: Diplomatic Overture – This had the distinction of being the most narrowly-focused 5 card, so it gets axed. It's actually not a bad card which I wouldn’t mind keeping, but other game mechanics seem to cover the same ground. It's just hard to use to do something you couldn't do normally with a deal in the diplomacy phase, and has some of the same problems.

Considered but rejected:
5: Sack of Rome – The frequency of play on this sucker is going to be really low. How often is this going to be a) available in someone's hand, b) playable, and c) worth giving up 5 ops for? Hardly ever. But it's a hoser for the Papacy, of which there are few it seems. And it's a nice historical event.
3: Pirate Haven – Again, a rather low frequency of play. It would have to be drawn a) after the Hapsburgs have actually taken Algiers, not a trivial undertaking, and b) by someone who is willing to spend 3 CPs hosing the Hapsburgs, c) when the Ottomans aren't winning if the Ottomans don't draw it themselves. This would easily fall below my frequency-of-play threshold, but it seems like you need some sort of backup if Algiers actually gets taken by the Hapsburgs/Papacy. It's also an interesting self-referential problem: the existence of this card makes going after Algiers not a great option, meaning that Pirate Haven itself is unlikely to ever get played.

Cards that are added to the deck later are predominantly "remove" events with decent frequency of play, so trying to work out some to cut doesn't seem to have as much upside. Still, Lady Jane Grey seems like she has a ridiculously narrow application (even though a game on the period seems like it needs a card for her), and Halley's Comet is fairly specialized although with a decent payoff if you can set it up. But more of these cards seem like Dissolution of the Monasteries, which is a fairly nice card. It's similar to Sale of Moluccas (in that one specific player draws two cards), but has a much better shot at getting played because it benefits two players (the Protestants also get three reformation attempts, which benefits both them and the English) and it has significant enough upside for all concerned to be worth dealing for. I wish there were more cards like this and Cloth Prices Fluctuate, which affect multiple players and so are both more likely to be played and could more easily be used as the basis for diplomacy.

What is the result of this whole thought experiment? My initial impression of the Here I Stand deck was that there were just too many cards, and as a result not enough interesting and flavorful events were being played. My first couple games have seen very few of the "remove" cards played, resulting in a huge draw deck by turn 4. Having now pored over the deck, I have to say, it looks better on close inspection than it did from a distance. There are lots of good events that should see fairly regular usage, the core at-start cards seem to have found a nice spot, generic enough to be generally playable but specific enough to have flavor. But I still do worry that it's too hard to play a lot of the core thematic events (typically, the non-mandatory "remove" events), resulting in a big deck that doesn't thin that much, aggravating the problem when more cards are added. It's close. But I'm pretty convinced that the Here I Stand deck should be at least a bit smaller.

1 comment:

  1. Sadly, a bunch of good commentary on this piece was lost when HaloScan (the comment provider on my old blog) went away.

    ReplyDelete