Sports Betting Trends (Fade or Follow)

My question is when you see a high percentage of a certain side or total coming in over the course of the season will it regress to a natural mean? For example I read somewhere that in the last 16 games the Broncos have played in the Over has cashed at a record of 13-3. One would reason that in this sample size the Broncos are close to a statistical outlier and will eventually regress to closer of a coin flips chance on totals.

I understand roulette and how 20 blacks have no impact on the next spin but in sports betting (The Broncos Over example) I would imagine more people follow a trend like this then bet against it (my source is every Patriots point spread post-Spygate in 2007) . Unlike roulette, NFL totals are inflated or deflated based on public/sharp perception which is why I think there is opportunity to seize value.

Should a bettor follow, fade, or ignore the recent history in weeks to come? In your opinions, where does the value lie?

What keeps throwing me off when thinking about this is how the books have to adjust their numbers week to week based on recent history, this is a variable you won't find in coin tossing or roulette ball spinning.

Thanks!

I tend to think that following is the right course. Denver's offense has been significantly better than expected. Their defense has been somewhat worse than expected. That information is discovered iteratively over the course of the season. Because the true qualities of the team are exposed gradually, the line will lag their true output potential. But I'm sure not an expert, and I lose an awful lot of lunch bets with my coworkers...

I appreciate the speedy answer.

I think as long as the Broncos keep hitting on the over the value on it will decrease every weekend going forward until they start trending "unders". Is this illogical being that they are independent events though? I guess its all conjecture at this point, we have no idea how long they will trend this way.

The key thing to factor into this equation is has the book adjusted their lines to reflect the "over" or "under" trend. Remember the book is simply trying to get an equal distribution of bets on each side. If they recognize that money continues to poor in on the over resulting in 60% of the money on that side. The casino does not want the majority of money coming in on one side and then that side winning as they lose money in those situations. Thus resulting in a situation that when the line comes out next week and the linemakers believe the money will come in on the over again you may see a total drastically rise, reflecting the amount of money coming in on the over. Lets say the Broncos total for week 1 was 42, now that they have hit the over 13 out of 16 times the total is likely to be at 46 or higher.

I generally follow the trend in these situations, but I am on the look out for when that time comes that the total is just too high and the under gains some value, as the public generally will blindly follow these trends.

Hope that helps.

Other factors such as key injuries, weather, day/night, short/long week must also be considered. Totals require two teams to score, so trending on just one team is not wise.

I always assume the books move the line close to real time, to keep it in the middle of the action. In a sense, this should normalize the total, but only in as much as where the betting action dictates, not the actual total.