THE BLOG
12/23/2015 02:02 pm ET | Updated Dec 23, 2016

My New Year's Resolution: Regress to a Better Mean

Every year around this time of year, people harness something that the behavioral economist Katherine Milkman calls the "Fresh Start Effect." We use the temporal landmark of a new year to give ourselves a clean slate -- a fresh start, if you will -- and commit to a whole manner of aspirational behaviors. Unfortunately, too often these behaviors are too aspirational, and after a month or so, we revert to our old ways. So this year, instead of focusing on the extraordinary, I urge you to focus on the average. This is because over the long haul, the average is the most important thing there is.

Enter: regression to the mean.

Regression to the mean is a statistical concept that says over time, things move toward their natural average. This is why if you play enough blackjack, eventually you'll lose. The natural average of blackjack is you lose -- something like 51 percent percent of the time. This means that even if you go on a maddening hot streak (which is the best time to stop playing... forever), you should expect a slightly more maddening cold streak in the future.

Regression to the mean is also why nearly every health care intervention directed at very sick, high-utilizing patients, seems to work. If you simply took the top 5 percent of health care utilizers and did nothing but "watch" them for a year or two, on average, they'd spend a lot less money on health care. This is because once patients hit rock bottom, they tend to get better. They move toward the mean.

My last, and probably favorite, example of regression to the mean involves an investing strategy called dollar-cost averaging. Dollar-cost averaging is simple. You pick a fund, let's say the S&P 500, and each and everyday, you invest the same amount of money in that fund. This way, when the fund is down (below it's mean) you are buying more shares and when the fund is up (above it's mean) you are buying fewer shares. So long as the market goes up over time -- which it always has, and barring the end of American capitalism, always will -- you'll come out on top. If this strategy is truly an automatic winner, you might ask, why doesn't everyone use it? Because dollar cost averaging requires doing things people hate: (i) being patient and hanging tough -- i.e., continuing to invest -- when the market is down and (ii) not getting overly excited -- i.e., not investing more or trying to "pick winners" -- when the market is up.

We tend not to focus on the average because the average is boring. We crave excitement and thrill, so we delude ourselves into thinking that we can outsmart the casino, crack the code and figure out how to control health care costs in the most expensive patients, and invest better than everyone else. Regression to the mean isn't sexy, and it's certainly not click-bait. It is, however, how the world works.

One more example (that I know matters to a lot of my readers) before trying to wrap this up. Fitness development. Everyone loves to focus on the extraordinary workouts, but they actually matter the least! For every workout that you totally crush it, you'll have one that you totally bomb. Rather than get yourself all worked up in excitement, or depression, about specific workouts on polar ends of the spectrum, you'd be better off focusing on the overall progression. The big picture. Is the average workout getting a little bit better, inching on an upward trajectory, over time? Are the awful workouts, while still well below average, a little less awful?

I know I'm arriving here in a somewhat disjointed way -- sorry, it's late at night -- but my point is that when it comes to change over time, pretty much everything abides by regression to the mean. Regarding your New Year's resolutions -- whether it's exercising more, eating a certain way, being a better employee, or being a better husband -- yes, take advantage of the fresh start effect. But no, don't get overly aspirational and focus on the extremes. Instead, focus on the mean.