Beating Settled, Late-Week Football Lines is HARD!
Circa Media Contest “Best Bets” Went 6-19 ATS Last Week
If you’ve ever tried picking games against the spread (even in the old days of using the Latest Line in your morning newspaper), you’ve probably had something like this happen to you…
*Before the games were played, you circled a bunch of teams for picks against the line. YOU KNOW FOOTBALL! So, you had something like 20-25 picks in college and pro combined. Maybe more if you held extreme confidence in your picks (and, almost everyone doing that begins with extreme confidence in their picks!)
*On Monday of the next week, you noticed that you really NAILED some of the games…but your overall record was something like 12-13 or 21-24 if you went crazy. Maybe you had a winning mark of 14-11 or 25-20. But, those BLOWOUTS that you NAILED were the ones you liked best (we always tell ourselves). If you just limited your picks to your strongest opinions, you’d really be great!
*Next week, you cut down. But, you still finished something like 8-9 over 17 picks, or 12-13 if you dropped down from 45. Again, some were EASY WINNERS that were NEVER IN DOUBT. You knew what was going to happen! Just cut it down more and focus on the strongest opinions.
*Now you’re down to 7-10 picks, and those are also close to 50/50. But, there was still that ONE GAME that you NAILED COMPLETELY. “If I only bet my strongest opinion, I’ll never lose and make a fortune!”
*You finalize your ONE SURE THING that can’t possibly lose because “they’re gonna kill ‘em’” is so darned obvious to a trained eye like yours. Then, your favorite loses outright. (Of course, the games you “almost” settled on did GREAT, so now you decide you need to make more picks!)
I’ve been thinking about that as the VSiN/Circa media contest has evolved through the years. Matt Youmans came up with a fantastic design that would allow true expertise to show through. At first it was only FIVE picks per week. Any sides or totals, college or pro. You had the entire weekend board to choose from. If you thought the NFL was impossible to beat but college Over/Unders were easy…you could turn in your five best college totals. If you thought the colleges were too volatile, but the NFL would be a piece of cake because there are only 32 teams to follow…you could make all five picks in the pros. If you thought the market was still underestimating the impact of wind on scoring, you could pick Under in the five windiest games. The contest also asked for a single “Best Bet” that would be used in scoring. Maybe some of the contests would be GREAT at finding that one sure thing.
You long time readers know that I once asked a bunch of sharps how many picks/bets it takes before you can be sure somebody knows that they’re doing. Blind, stupid luck can lead to a winning or losing record over one season. Everybody blows hot and cold. How long before that works itself out and the record means something?
Consensus centered around 1,000 picks. If you have a winning record (better than 52.4% if you’re picking against 11/10 vigorish on point spreads), after 1K picks…then you know what you’re doing. (Imagine you’re in the rafters looking down at a bunch of roulette tables. Some will have had a bunch of red numbers showing up lately, other a bunch of black numbers…but after 1K spins, they all should be pretty close to even percentage-wise).
For the purposes of evaluating contest performance, any one person could blow hot and look like a genius, or blow cold and look like a knucklehead. It might take 9-10 years to get to 1K picks so the audience would know if that individual was any good. But, the COMPOSITE would get to 1K picks within a season. Observers could at least make a determination on whether or not “experts” as a composite were capable of beating contest lines at a 54%, 57%, or 60% rate.
Going from memory, the first year saw the composite top 50/50, but not reach 52.4%. No evidence that the selected experts as a group were good enough to beat the market…even with their FIVE FAVORITE PICKS.
Contest lines are posted late in the week after sharps have pounded them into place. There’s a long-time theory that “nobody” can beat late-week lines because the edges are gone. That composite performance provided more evidence for the theory. Even people that have created names for themselves in sports media…picking ONLY their five favorite plays…couldn’t “beat the market.”
Last year, VSiN upped the number of picks to seven. Not sure if that’s because somebody important didn’t think five was enough (“true expertise” is more likely to show over a larger sampling of games?), or what the reasons were. Same thing happened though. Good enough to be better than a coin flip, but not good enough to beat the vig.
This year’s change was to invite more people. There are now 25 invitees making seven picks a week. That will get us to 1K picks much faster in the composite. But, it’s going to be hard for the group to make a statement because so many were lined up with square sentiment that took an absolute betting bath this past weekend. You’ve probably heard by now that recreational bettors lost a fortune.
Week Two
All Picks: 77-96-2 ATS
Single Best Bets: 6-19 ATS
Season to Date
All Picks: 162-176-12
Single Best Bets: 17-32-1
I’m not saying all 25 had losing records. But, the combined mark was disappointing. And, we can’t know for sure which good extremes or bad extremes were luck influenced. Each individual is a long way from 1K picks. This year, the composite will be there after only seven weeks. We can at least draw conclusions about the composite by then.
Main lesson for now is that it’s very hard for ANYBODY to consistently pick winners against settled late-week likes. Sharps bet early in the week, sharpening any soft oddsmaker openers (that’s where the term comes from). Some also fade any overadjustments right before kickoff. Let’s say a line moves from an opener of -6.5 to -10.5 in the hour before kickoff. Sharps will have laid -6.5 and -7…but then come back on +10.5 at the very end. They don’t bet Friday’s -9…they bet the vulnerabilities furthest away from what they think the “right” line is.
Does this mean the invitees aren’t “experts.” Depends on how you define the term. Many in media “know the material” extremely well. They can tell you who the backup QB is at Indiana or what the stadium names are in the Big 12. I think most “trends” are random trivia…but many can recite trends off the top of their heads, and dig up more if you need them. Coaching trees? They know them.
But, somebody who knows everything about “Star Wars” can be an expert at “Star Wars” without being able to write the next script. Different things.
Be ready to accept the premise that NOBODY is an expert at making picks against settled late-week lines. The VSiN/Circa contest looks like it’s still proving that. Though, the laws of math point very strongly to some bounce back performances in coming weeks. The good angel on the shoulder of fate will counter-act that 6-19 “Best Bet” weekend at some point!
And, also be ready to accept that the hosts/guests you’re watching on your favorite shows may be experts on many things about football (or other sports in other seasons), but aren’t capable of hitting better than 52.4% on settled, late-week lines. No crime in that. Over a large sampling, it’s probably impossible for everyone! (One of the big issues for the future of sports betting media is that the audience wants to watch people who can consistently pick winners…but show times aren’t well-designed for when that can actually happen in real-world betting.)
Wanted to get this down on paper amidst all the current discussions about the bad weekend for bettors and pick makers. That’s it for this report. See you again Wednesday for our weekly estimate of “market” Power Ratings in the NFL.
Best of luck!
Hi Jeff. I’m in Circa/VSiN’s “Average Joe” portion of the contest. Having fun and surprising myself and wife at being second each of the first two weeks. What I’m learning from you certainly factors into my luck. Thanks.