Thursday, March 1, 2012

Ramblings: Information Cascades

There's been a lot of talk lately about Netdecking, which is The process of stealing a tournament winning list from a forum and replicating it, but Netdecking is the result of a phenomenon called information cascades.
Suppose that you are choosing a restaurant in an unfamiliar town, and based on your own research about restaurants you intend to go to restaurant A. However, when you arrive you see that no one is eating in restaurant A while restaurant B next door is nearly full. If you believe that other diners have tastes similar to yours, and that they too have some information about where to eat, it may be rational to join the crowd at B rather than to follow your own information. (...) In this case, we say that herding, or an information cascade, has occurred.
With a simple setup like this the cascade effect is obvious, but it actually happens in nearly all situations involving sequential decision making, and Warmachine/Hordes is no different.

Information Cascade

I think Pistol Wraiths are amazing but every couple of months/weeks they get run over by the cascade posters telling rookies they're expensive and will die anyway. Slaughterborn is also three points and every post about him makes him out to be a god on the battlefield, so which model do you think the rookie will end up with?

This goes for everything, and the sad fact is that there's no changing it. I've been working on my Terminus build for some time, and I see several people on the forums championing builds that look very much like mine, which is perfectly alright of course, but then it will just be my build cascading around the web. If someone comes up with something we can all agree is superior then we'll all swarm that way instead.


Netdecking is the ultimate expression of information cascades, where the rookie (or seriously unimaginative) player simply copies a winning build and fields it as his own. This works in tactically limited games but not in Warmachine/Hordes, which leads to some very frustrated players. It doesn't work because tactics, psychology, style, and skill are factors in winning, as illustrated by the good old "Knife Vs. Gun" dilemma.

If you don't want to read the article I can summarize: A trained fighter with a knife will probably win against a trained shooter with a gun, provided he begins the fight within six feet of the shooter. The knife is a weapon to be wielded, but the skills, circumstances, and ingenuity of the person doing it matters a lot. The same thing goes for models/units in a game, and you could be in a tank facing a guy with a stick, but unless you know how to work the tank he'll get you in the end.



  1. I found playing proxy games with lists I had copied from the Internet allowed me to see how those particular lists worked as a new player. That was valuable knowledge. However, the list I stettled on for Cryx was a combination of my favorite facets of other lists, that worked well with my play style. I would argue that it is very difficult to build a list that works for you, until you try other lists and see how you perform with them.

  2. What you're doing is called "Templating" and is my next article. It's the process in which you "borrow" lists and learn by them, then tweak them to your style of play :)

  3. I try to play Warmachine on a budget so I look at what others are running to avoid purchasing models that are either bad or too situational. I also have been a gamer long enough to know my own play style and I take what I read with a grain of salt. Instead of "netdecking" I will call it "research" :D

  4. One of the reasons I like Tier lists is they (so far) seem to be immune to this phenomenon. Then again Tiers force specific builds so I suppose everyone's tier looks like everyone else's. The reason I bring Tiers up is because they do force you to learn the "knife skill" you mention.

    Ever play Gaspy's Tier where you're running a bunch of MAT 5 mechanithralls that can't hit the broad side of the proverbial barn? In that meta you suddenly (re)learn tools like *slams* to improve your mechanithrall MAT. This is an example where you are forced by the constraints of the list to become a better tactician.

    Then when you wield the "netdeck" you're much more capable.

  5. The Jake, unless you copy his list to the last detail it IS research. There's nothing wrong with research, except that Information Cascades make it harder. If you find ten articles praising Slaughterborn and ten articles hating on the Pistol Wraiths, you risk ending up with the wrong solo for your list. Experience diminishes this risk, and a complete understanding of the system also helps, but even experienced players get suckered by the Cascades every now and then. :)

  6. That is a good point. I bought my nightwrentches because they were in almost every list I looked at online. When I tried them out, I found myself missing my deathrippers almost immediately.

  7. Yeah, I also tried out nightwretches, and ended up not liking them nearly as much as deathrippers. Lucky for me, I just played a number of games with them proxied, so I didn't waste money on the models. I proxy things all the time, and I feel like it's the way to go in order to learn about the models in your faction.