Essay:Reflections on SasBound

From Archon Arcana - The KeyForge Wiki

This is an attempt at critical analysis of the sas-bound event format, created by Karen Brown and run recently by the folks at Swindle. It's not critical in an "I hate it" sense, because I don't. I went in very excited, and I still think it's a great format. Instead, I aim to be critical in a "how could this be better?" sense. I also doubt I'll say anything others won't also think of, but maybe I'll spark others to provide ideas I won't think of. For that reason, I may bring up problems without offering solutions, something I normally try to avoid. But please don't be fooled! I think Sas-bound is a great idea, and I'm very grateful to the event organizers who made it happen. Alright, let's go!

The biggest problem is that people with losing records are not incentivized to play. For what it's worth, any system where you are likely to schedule 4-0's with 0-4's has this problem. The player with the losing record may feel that there is no point in continuing, and, even if they intend to finish out whatever they agreed to, they feel less motivated than another 4-0 player would. That's the strength of the Swiss system. But sas-bound probably suffers even more from this: pairing a 73 cap and an 84 cap is kind of the point. In fact, you might think that this uneven pairing might give them hope to bounce back, but that doesn't seem to be the prevailing attitude I've seen. So, probably more incentive is needed.

Another potential problem I see is that a player could game the system via scheduling. The obvious thing here is that, if I wait to play my four games until my opponents have each played three of theirs, I'll face a lower average sas cap - not by much, but by some. I also could manipulate scheduling by prioritizing matchups with the higher-cap players, ensuring that, if I win, I will face a lower-cap opponent when I am also playing on a lower cap.

One obvious thing to do about absentee players is kick them and award wins to their opponents. However, this potentially punishes other, unrelated contestants in a way that is not true in Swiss format. In Swiss, a forfeit by my opponent allows me to advance, but the winner on another arm of the bracket will still have a chance to face me down. But when the event win condition is "win a game with your sas cap lower than 60", I could win without ever playing against one of the other players who is also close to victory. So, the forfeit, especially if it's when my sas cap is pretty low, and that of the forfeiting player is high, gives me a huge benefit relative to my fellow racers. Of course, the reverse is also true, if we do not award wins for no-shows: if I simply cannot play matches, I have no hope of winning the event. In the actual event, subs were introduced as a way to keep the train moving. And, while I think this was the right call in the moment, it would certainly be more fun if we find a way to keep people engaged, even when they need to catch up.

Another solution would be to pool together abandoned players, and have them play each other instead. This is much more fair, but has the downside that now we are likely pitting more low-sas players against each other, and therefore it becomes more of a sas ladder (descending) than an analog for Chainbound.

Of course, actual Chainbound solves this a different way. Each weekly event has its own stakes. In fact, in that case there are no stakes for reaching a particular number of chains, other than prestige. But even within broader stakes, having some sort of motivation to win games during a particular fortnight, even if you feel hopeless about winning the event overall, could help the event flow more smoothly. As KOR becomes more widely esteemed, it may provide this sort of meta layer on top of an event like this, to help a player value a 10th place finish over a 15th place finish, even if prizes only accrue to the top two players.

The entrance of a teammate of mine as a sub alerted me to another problem, which is that game-fixing could be tantalizing. Imagine that he gets to sas 58 and is scheduled against me, while I am at sas 82. The match certainly means much more to him at that point than it does to me (especially with a KFPL entry on the line), and it could be tempting to throw. Again, in Swiss play this situation only rarely arises, because for the most part players are paired against players with the same record. The counterexample is when a player who is guaranteed a spot in the top cut is paired down to a friend who must win in order to make the top cut, but power-of-2 player counts preclude that possibility.

Lastly, I just want to reiterate my gratitude to the Swindle folks for coming up with and running such a fun event. It takes courage to try something new, and running an event is a lot of work. I hope to participate in more sas-bound events with you all in the future!

Okay, those are my thoughts. What do you think? Did I miss any areas that could be improved? Did I miss some solutions?