Gambling & Journalism

Just because you cover football doesn't mean you should have a picks column. Long ago I argued with my sports editor that I could do a better job than our NFL writer because I understood the difference between a favorite laying 3 and 13 and that the league never made any sense, especially against the number, and it was embarrassing to put his record at the bottom of every column when he couldn't hit 35 percent ATS. Not that I'd have hit 60-70 percent, no one does and they're lying if they say they do, but I could have provided a little bit of service to readers by digging for some real information to justify a pick, not just "the Niners are hot, they've won three straight, they'll blow out Atlanta." You'd never see so much unsubstantiated crap elsewhere in the paper as you would in the picks column. Rant over.

 

I dunno. Is anyone really "touting" a team as a 7-point favorite? It's objectively what that team is.

Sports books don't subjectively decide what they "expect" a team to do. Oddsmakers set lines using a statistical analysis that takes into account all of the available information, and they have it down as close to a science as possible today. They are actually reasonably good at it, as it turns out. To the extent they miss where the public is going take a line, they adjust very quickly based on market movements, which correct it pretty quickly. And these markets are very well populated, which usually makes for more efficiency than people realize.

Sports betting markets have a ton of participants, there is a lot of money in play and those participants are making their bets based on all of the publicly (and non publicly) available information. Efficient market hypothesis would suggest that like any other market that has those characteristics, lines in the aggregate are by and large going to be reasonaly efficient once they release a line and allow it to adjust. Every bit of empical evidence I am aware of suggests that is true, and I have looked at this, because it's something that really interests me.

Essentially, you should expect betting with a line to be a 50 /50 coin flip, except the vig -- what the book is taking to match buyers and sellers -- makes your expected value on a dollar bet less than a dollar (break even is 52.4 percent on a -110 line). Whch would suggest you can't win if you play for any length of time. The ONLY way you could possibly win is if you can find lines in which all of those people with all of that information, just have it wrong. Or if you can bet the line immediately somehow, before the public (and they release it to professional sports gamblers -- "sharps" -- before any of us see a line) has at it.

At one point, maybe 40 or 50 years ago, the betting markets weren't as efficient (way fewer people and less money involved), and Billy Walters and types like him were rooting around for info that might not have been easily available to others, they were using computer models (prior to widespread use of PCs) to integrate a lot of variables, and they were able to consistently find inefficiencies by all accounts. I think those days are long, long, long gone.

I'm not saying that market efficiency works in an up-to-the-minute way. For example, I have read empirical work that suggests that there is a small recency bias with bettors overvaluing recent success, so betting underdogs with big point spreads could be a successful strategy. ... and you'd need to even fine tune a strategy like that to get the criteria just right. But even to pull off that kind of strategy, you need to bet just about every game that fits the right criteria in a systematic way, which would require a lot of money, a lot of discipline and doesn't amount to the immediate jackpot of only picking winners that most casual bettors are looking for. And in that regard, it's not that different than a contrarian stock-picking strategy, where when something gets beaten down enough, there is often a regression back to the mean -- people overreact to the upside and downside.

In any case, I think for most (if not every) people, it's a fools' errand to think that you can spot the "rare gem on the board that the casinos missed the markt on line-wise." You are actually betting that with all of the available information, and all of the people (professional and casual) willing to gamble on it, that you know something the masses don't. I'd suggest that 1) When those bets work out, they probably weren't because you knew anything others don't, you essentially just won a 50/50 coin flip, just as when you are wrong about it, you lost the coin flip, and 2) To the extent that there are actually lines that don't efficiently capture games, in markets this well populated, anyone thinking they can parachute in and find those games consistently on a discretionary basis is probably going to lose their money if they play for any amount of time .... they may have a streak where they are up, but over time, it just isn't likely to work. To take advantage of any inefficiencies, if they actually exist (and I would guess they do), you'd need to be extemely disciplined and take EVERY bet that meets the criteria you have for whatever the inefficiency is, because any edge you might be giving yourself is going to be pretty small. You would be chipping away for slightly more wins than losses and trying to make money in a slow and steady way.