Blog Archives

Optimizing Board Games

While on vacation this past week, I had a chance to put in a few rounds of Betrayal at House on the Hill. It is an ostensibly cooperative board game that consists of exploring a haunted house by laying down tiles, rolling some dice, and then attempting to survive once the Haunt starts. Once the Haunt is triggered, usually one of the players becomes a traitor working for the monsters that show up, and thus it quickly becomes 1v3 or worse.

The game was fun for the three rounds we played it, but by the third game, I started seeing the cracks in the design.

Exploring rooms will usually cause an Item, Event, or Omen card to be pulled. Items are pretty much universally good and are a hot commodity. Events are usually bad or otherwise risky – most require you to succeed on a roll to gain stats, or you otherwise lose stats. Omen cards are usually the equivalent of good Items, but once an Omen is pulled, that person has to make a Haunt roll that surpasses the number of active Omens, else the Haunting starts. In the three games we played, the Haunt pretty much consistently occurred after the sixth Omen.

The cracks mostly show once people realize that optimization is the answer. Some of the rooms, for example, allow you to increase a stat (Might, Speed, etc) by +1 if you end your turn there. Now, the rulebook states it only works once per game, but the FAQ (PDF) makes it clear that it happens once per game per player. In other words, the moment one of these rooms open up, the optimum strategy is for everyone to stop what they are doing and go get that stat increase. Free stats are free. Considering that the Haunt can only start when an Omen card is pulled, and no Omen cards can get pulled if no new rooms are being explored, there is zero reason not to perform that strategy.

Another example is the Vault room. A player needs to roll a Knowledge check and get a result of 6+ to open the Vault and snag two Items. Rolling a 6 would be exceedingly unlikely for someone with Knowledge 3, because the dice only have 0, 1, and 2 printed on them. But, again, there is zero danger pre-Haunt as long as no one is actively exploring new rooms. It costs nobody anything let one person roll three dice until a total of six appears. Granted, there are other players with higher starting Knowledge totals who can make the roll faster, but the bottom line is that the preferred result is inevitable.

Once I realized all this, the game become significantly less fun. We didn’t do the “everyone get your +1 Sanity” trick the first two times we played, because we really didn’t know better. The third time we did. And that room might as well said “everyone gets +1 whatever” because we basically cycled through everyone’s turn 2-3 times in ten seconds to make sure people with slower Speed scores could travel there. While we didn’t quite make the Vault an auto-open situation, we could have done that too.

Another example: some rooms force you to make a Might/etc check to leave without taking damage. The FAQ points out that if you fail the roll, you can choose to not leave the room and avoid the damage. Ergo, the optimal strategy is to not leave until you win the roll, and for no one to explore any rooms until you do.

Noticing a pattern yet?

The optimal strategy makes the game less about interesting decisions, and more about whether your friends are willing to play the “right” way. This becomes especially evident once the Haunt actually starts, considering the Traitor/monsters are way more dangerous than most of the other players by default. Since the Traitor/monsters get a turn to try and kill you, suddenly turns become a precious commodity. It’s less about options and more about “we need to win this roll or be turned into a toad.” What ends up being even worse is the fact that the Haunt is pretty much over – win or lose – within like 2-3 full turns. Yeah, sometimes it takes several turns to successfully research X, or tear apart a room for Y, but you either have a strategy/house layout that gives you breathing room or you are dead.

All in all, I found Betrayal at the House on the Hill to be relatively fun for a while. It honestly reminded me of a sort of Arkham Horror-lite, in fact. But having played Arkham before, I immediately recognized how much of a difference it makes to be time-limited. There are still optimal decisions to be made in closing portals or otherwise holding back the eldritch beings, but at least the gambling in Arkham has teeth. Sometimes literally.

Arkham Horror

I have played two games of Arkham Horror in the past few days, and the experience has been interesting.


Yes, that is a door.

The first thing I learned was, yes, “Arkham” is a Lovecraftian reference that Batman utilized for Arkham Asylum. The second thing I learned is that the game is definitely in the Axis & Allies level of board game setup. Well, maybe not that long, but it still requires a dozen or so stacks of cards and such.

I do like how physical the game feels. For example, each character gets X number of dollars, and said money is represented by little rectangular pieces. Characters have their own sort of character “sheet,” but they also have cardboard character pieces that are placed on the board standing up; monsters have the same sort of thing.

The gameplay flow is… kind weird. The premise of the game is to basically close gates (to other dimensions) before the Eldar God wakes up. Alternatively, you can try and shoot the Eldar God in the face, generally with predictable results. Having played twice and looked through the various cards/abilities, I was struck with a sense that the game is remarkably balanced – nearly every system in the game has an “Eldar God wakes up” failsafe built in. On the other hand, the second game I played ended up with the Big Bad waking up on like turn 6. We actually ended up beating said god through unique circumstances – the Lurker Beyond the Threshold and a crowbar/carbine combo was da real MVP – but it was a close thing, with 2 of the 4 characters being devoured.

As I was saying though, the gameplay decisions end up being a bit weird. Outside of a few character abilities there are very few ways of regenerating health and sanity. The success system is basically rolling 5-6 on a standard six-sided die, so the odds are generally that each encounter with leave you bleeding a bit. This means that most characters can’t undergo more than two encounters before having to make pit stops in either the hospital or Asylum to recharge, and since closing gates requires you to get hit with special “other world” encounters (which can be anything, but could be nasty monsters), most of the time it feels like there is never enough time to do anything.

Which is a good thing, I suppose, when you are trying to simulate the urgency of people running around stopping an Eldar God from awakening. Still, I kinda felt like that it put an absurd dependence on A) the characters you picked to play as at the beginning (we did a shuffle,  deal 3, pick 1 deal), and B) what random items you were dealt. Everyone have a weapon? Awesome. Everyone get some bullshit tomes? Welp, maybe it’s worth starting over.

Beyond all that, I can see it being a good game to play with a group of friends, if you all have 2-4 hours to kill. One of the best aspects of the game is that it is entirely cooperative, which I think is fairly unique in terms of board games. Getting straight-up devoured isn’t even Game Over either, as there are rules for you grabbing a new Investigator to play as, assuming the game isn’t over via Eldar God thrashing yet. There are numerous expansions to the game, including literal expansions to the game board, so between that and the games themselves not lasting long (in terms of not using all the card text) the sense of the unknown is preserved pretty well.

On a final note, this is absolutely one of those games that I feel could be 100% digitized with little lost. I mean, I supposed by definition most board games could be digitized with little lost, but at least here with Arkham Horror there isn’t much interaction with the vast majority of the cards in a vast majority of the decks, such that a computer spitting out outcomes wouldn’t remove much of anything. And even if everything but direct player interaction was removed, the game itself would still take an hour or two.

So… yeah. Arkham Horror. Played it in Japan, and now I might see if there are any open spots at GenCon.