Wednesday, July 26, 2006

Ra vs. Beowulf

So, I had a chance to play both Ra and Beowulf in close proximity recently. Both are games that make my list of all-time classics. Both are auction games of a sort, although neither is straightforward as such (unlike, say, Modern Art or Medici). And both have had the same complaint leveled at them from time to time: they have too much luck. To what degree is this true? This is of particular interest to an auction game, because the fundamental, core issue that all auction games must wrestle with is not whether or not to have luck, but getting the impact of luck right. If we just auction off a lot of stuff clearly worth $15, that's not very interesting (I bid $14.99!). But if there is uncertainty about how much each lot is worth, and factors that impact its worth that are random or concealed from the bidder (or known only to a subset of the bidders), then you get an interesting game – but one with some amount of luck. Getting the luck right, so that players feel that they are taking risks and not just being jerked around, is the key to success.

Interestingly, Ra and Beowulf use luck in almost totally divergent ways. In Ra, you know how much you're bidding. You know how much the tiles are worth. But the flow of the game, what is going to become available for bid and the pace of those auctions, is random and rather fluid. In Beowulf, the lots available and the rate at which they come out are fixed and known to all. The variability is in the bids, and in what some of the lots (the scrolls) are actually worth. In Ra, you take a chance by holding out to see if something better becomes available. In Beowulf, you take a chance by pushing your luck with your bid.

The advantage of Beowulf's luck is that it enables more long-term planning. By knowing, generally, what the future looks like but by taking a chance in the here and now, the game enables more factors to affect your immediate judgement, resulting in an evaluation process that must take into account a large number of factors. So, you take a risk knowing that the gold you may win will be quite valuable in the very near future to buy an All-Iron Shield that you can then use later against the Dragon; or you hold off knowing that the fight card will be more valuable later. Plus, by making the risks more numerous and more immediate, but less individually risky, there is more of an emotional charge on each one. The downside on each risk (getting kicked out of the auction with a scratch if you blow it) is rather significant but not severe, and the upsides of succeeding at any individual risk is modest (typically just one card), so it's rare for an individual chance to be a game-breaker.

By contrast, in Ra the tension of each decision is more drawn-out. The decision to duck a bid, or to make a lot richer instead of auctioning, does not immediately reveal its brilliance or stupidity. If you crack now and buy a lot that does not have a Civilization tile, hoping you can get one later, the ultimate result of that risk may take the entire rest of the round to fully play out. Ra's risks tend to be more nuanced than Beowulf's "in or out" risks, and while the risks in Ra are unlikely to have immediate painful effects the way they can in Beowulf, you are also sometimes confronted with game-breaking risks (especially when the number of Ra tiles available before the end of the round grow short) that there really isn't much way to properly assess other than by raw gut feel. If you take a risk and get cut off by the end of the round, ultimately acquiring nothing, this is likely to be a far more severe blow than any risk Beowulf could have hammered you with.

So, what does all this mean? I think luck becomes frustrating and problematic when it's high-stakes, and when there isn't a lot you can do to affect it. I think this sort of thing manifests itself differently in Beowulf and Ra.

In Beowulf, it's not really a systemic problem per se, but because the odds of a risk succeeding are in the 50-60% range, you can see odd stuff in occasional late-game auctions when the downsides of a scratch can become negligible. It's very frustrating to have set yourself up for a win in the Dragon's Rampage episode, only to see a competitor take it away by succeeding in risk after risk because he's in a position where risking has no downside for him (because one additional scratch isn't going to matter at this point), and because he keeps getting lucky. It doesn't happen a lot, because the situation where the one scratch isn't going to make any difference and where the player succeeds in 5+ risks in a row are obviously fairly rare. And it's likely to happen only the Dragon's Rampage episode; timely play of the All-Iron Shield will tend to knock out gratuitous riskers in the final battle. But when it happens – wow, it's frustrating. For the person on the receiving end, it's like watching a car wreck.

In Ra, I think the negative impact of luck can be more systemic, and is related to the end-of-round, high-stakes, game-breaking type risks. If you have strong bid tiles, and if small auctions are coming up rapid-fire, and if the end-of-round is coming up quickly, you can end up facing a situation where you and one other player are dueling, trying to get a decent lot, while risking getting hosed by the end of the round. In this situation, through bad luck one of the two player can really get hammered. Obviously, how objectionable this is, is going to be decided by how often a player is caught between the bag and the end of the round.

With 3 players, I think Ra is brilliantly balanced, and this issue seems to hardly ever occur – if you're picking blind at the end with tight time pressure, or if two players are staring at each other with a 12 and a 13 sun and playing chicken, it seems to be your own fault most likely. Almost all rounds will still end with the clock running out, but players are rarely shut out. But as the numbers of players increase, your turn frequency (and thus your control over events) goes down, and the length of the rounds does not increase proportionally to the number of players. So time pressure gets tighter, your ability to have an impact on the flow of events decreases, you get smaller (and therefore significantly more random) lots, and your susceptibility to the hand of fate increases quite a bit.

As a result, I think the impact of luck in 5-player Ra is much greater than in any version of Beowulf. That is to say, players are going to feel jerked around more frequently, at the expense of feeling like they were just knowingly taking a risk and happened to blow it. Beowulf seems to be scaled much more cleanly for the full range of players, while Ra is definitely a very different game with 3 or with 5 (I consider it possibly the greatest 3-player game ever, while I think it's just a good game with 5). But Beowulf is a lucky game too, and the endgame auctions, where the one-scratch downside of risking in late-game auctions can sometimes be an insufficient deterrent to reckless risking, can sometimes play out in a way that isn't particularly satisfying.

So when it comes to auction games, Ra and Beowulf are going to have to call it a draw I think (Beowulf is way ahead of Ra on theme, but that's a discussion for a different day). I think 5-player Beowulf is much less lucky and has more player control and less frustration than 5-player Ra, but the sheer brilliance and perfect balance of 3-player Ra I think has to be judged to slightly outshine Beowulf, due to the latter's minor faltering (which can sometimes result in frustrating runs of luck) in the late-game competitions.

Let me put it this way, though: if my collection could only include 5 German-style games, there is a strong possibility it would include both Ra and Beowulf.

No comments:

Post a Comment