On my mathandpoker blog

I have a blog elsewhere where I talk about topics related to application of mathematics to poker and other gambling games. I have more than one blog because it helps me segment the things I write about. I'm not sure that's really productive, but it's what I do anyway.

I common theme that I hit on time and time again on that blog is that poker is not mathematics, mathematics is a tool that can be used to help analyze poker, but mathematical models are not the game itself.

Formulating a model, adjusting a model, validation of a model, all those topics are the main meme of the blog.

This post is about the culture of poker, so bear with me.

Anyway, I made a blog post at mathandpoker that commented on something that Ed Miller had posted on his blog.

Although I don't think much of Ed, he has a terrible track record in evaluation of the character of people, the post wasn't critical of Ed personaly in any way that I can see, although it was critical of something he'd said and the way he said it. (This statement is critical of Ed personally, not the post I originally made).

His mistake was an over-reliance on the concept of EV, something that on it's surface seems to many to be something that's impossible to do. It's not at all impossible to do. It's a topic I've talked about before.

I often hit on statistical or mathematical sacred cows on that blog. For example, the first post I made in that blog, in June 2006, was an attack on the popular idea in poker that you have to have huge sample sizes. I've made other posts pointing out the importance of looking at things other than the sample size when considering how much data you need, the most recent being yesterday.

A blind reliance on EV is also a common theme. My earliest post on that topic was in July 2006, the second month of the blog. Micheal Trick also has a nice comment on that post.

This is what I'd said in that first post on expected value.

If you really want to optimize your stratagy and maximize your win, they you just have to look at the game in a global sense. A strategic Expected Value of a collection of actions is what you need to consider, not a tactical Expected Value of one action.

Micheal Trick related my comments about poker to some airline scheduling problems he'd been working on (he's not a poker player) and he said

The problem with both the poker example and crew scheduling is that the objective is much “fuzzier” than the underlying main objective. And that makes it much harder to “optimize”.

The point is often overlooked, but the reality is that EV is often only part of the story, it's seldom the entire story in situations that have any complexity to them at all.

But EV has been elevated to religious status in the poker world. I'm not sure why. But a worship of EV has become part of the culture.

Some blog software will pick up any other blog reference to a post and put a link to it in the comments section. Ed Miller's software does that. So, my post showed up as a comment on his blog. That's fine with me, it's the way blogs should work, blogs should have software features that advance the conversation. But what I was doing wasn't really writing a comment about Ed so much as using Ed's post as a point to step off and advance an existing meme on the evils of a tunnel vision focus on EV. I don't think many of Ed's readers realized that.

The reaction I got from Ed's readers was more like something I'd have expected from Kansas school board members to somebody who pointed out that Jerry Falwell was an ignorant bigot.

It's not surprising, given the religious status of EV, and since 2+2 types think of themselves as high priests of that religion. But I think it's an interesting example of how the culture of poker works out.

Update:
One thing I don't think I made clear is that this blowback isn't about Ed Miller (or about me) it's about a religious reverence to EV = Sum(x * P(x)). Ed's wife seems to think it's about Ed and I made some comments on her blog that's probably more about Ed than it should have been.