I Bet You're Wrong

You don't have to be a gambler to bet.

As I see it, there are two different ways to define gambling: either everything is a gamble or there is a select class of things that are gambling.

I prefer the latter definition where gambling involves no skill only luck. You can make a bet that is skillful in games of all kinds that allow for skill or edge. You can also make a bet in games that have no skill element where pure luck governs.

So this is a bit semantics. But semantics matter and especially so when one manner of speaking can be dismissive of a larger point. It also matters in understanding the realm in which we are considering a bet—one with or without an element of skill.

To this end I fully endorse David Henderson's point of view here:

I often see people make strong claims about events they expect to happen in the future. By “strong,” I don’t mean that they make good arguments; I mean that they express certainty or almost certainty.

…

So what do I do? I propose a bet. Some people are willing to put their money where their mouths are.

…

What I find, though, is that, even for small proposed bets, most people won’t bet. This is true even I offer good odds.

One of the main reasons they give for not betting is, “I don’t gamble.”

I usually let it go.

But what I want to say is, “That’s obviously false; you gamble every day and for much larger amounts.”

I like gambling to an extent [those who know me may say the word “extent” is doing a lot of work there as I have a bigger appetite for risk taking than most]. I just don't love the broad use of the term.

Betting does not necessarily imply gambling. For me it is a gamble but not gambling. Gambling is rather an action of placing at risk something that has a known likelihood of being unfair to the bettor.

The player is gambling on roulette. The house is not. The house WILL win in the long run. The player WILL lose.

But suppose the roulette wheel had some random invisible spaces on it for which there are no available bets for the player to make that might be colorless letters but might also be several duplicate numbers with black or red colors. In the former case (letters) the house would retain an edge—one larger than on the actual roulette wheel.

In the latter case (duplicate numbers) the player betting on the correct number or corresponding parity (odd or even) or color (red or black) would actually have the edge, but no one—not the players nor the house—would know this information until the ball comes to a rest in each separate spin—with the invisible spaces randomly changing each time.

Now suppose a player does know the value or parity or color of the invisible numbers or if they are letters while the casino does not. The player both has the advantage and knows it.

There is an extremely big difference among these three worlds—normal roulette wheels, ones with extra spaces unknown ex ante, and ones with extra spaces known only to the players. Too often people sloppily confuse the various versions applying them haphazardly to one or the other conditions in the world.

People commonly define a bet Henderson is describing as if it is in the standard roulette situation with all information known up front. That is gambling in my use of the term. It is fun but uninteresting.

Often other times people derisively define a Henderson bet as if it is in the invisible, fully unknown version of the roulette wheel. That too would be gambling, and it too would be fun but rather uninteresting.

Sometimes people erroneously define a Henderson bet as if it is in the third case where the player knows something the casino does not. This is interesting, but unrealistic. This is getting closer to what I believe describes the Henderson bet—it is why he chooses to bet. People think they know something they only might know or have reasons to suspect.

So there is a fourth and much more interesting way to describe the real world. That is where there are invisible letters or numbers that many players including the casino have varying guesses as to their actual value, parity quality, and color condition. Some believe with the support of theory or evidence that the invisible spaces are colorless letters while others think they are odd reds or even blacks or such similar. Some think they know very precisely while others suspect with varying confidence a range of possibilities.

In this fourth version we get betting as I like to describe it without it being gambling. Sure, it is a gamble, but the gamble is you have better information and reasoning rather than you are lucky or trying to simply have a fortunate outcome.

This is important first because one should not dismiss betting as mere gambling (remembering that, like Henderson says, you gamble [bet] every day). It is just the attempt to make predictions about the world we live in—perhaps a natural thing for the brain to do.

Second, it is important because it quantifies and gives you skin in the game in an unmistakable way. It forces you to reason. Like Alex Tabarrok says, it is a tax on bullshit.

Third, this is how the real world works. Formalizing it into a quantified bet just makes it more concrete.

And with all of that said, I will have to concede that Scott Sumner makes excellent points in this post as to why he doesn’t bet. It doesn’t have to be worth it, and the non-pecuniary cost-benefit analysis plus transactions cost might make it undesirable. As Sumner says,

Just because you cannot see any reason why someone would reject a bet other than fear of losing, don’t assume that others feel the same way.

Just don’t call my offer to bet you gambling per se. If you do so, I will assume you have conceded that I am the casino and that you are wrong.

PS: Here is a very interesting story about the gambler who beat roulette.