Week 6 sides and totals
Plus recapping another frustratingly average Week 5
Last week, I described the Week 4 bets as a survive and advance week, happy that things just came out alright. I was on the wrong side of some games, and there weren’t a bunch of other spots I’d considered where I was watching the slate and annoyed at what made it onto my final card.
Week 5 was the opposite. We went 3-3 last week, and a couple of those misses like backing the Seahawks at home against the Giants were tough calls, but there were two bets I took off the official card while writing it up that were clean hits. That doesn’t help anyone — if you had that and a ham sandwich, all you’d have is a ham sandwich — but I don’t usually pull stuff from the official card while writing up, so it was frustrating on my end (typically my initial list isn’t super long, and I’m adding while working through it).
Both of those bets were things I alluded to elsewhere — I had Washington -3 at home against Cleveland after taking Jayden Daniels not just as OROY but also MVP last week, and his MVP odds have now slashed in about half after yet another impressive win, and then I also had the Denver-Vegas over that I mentioned in that writeup. The reason I’m starting with these two is I think it was particularly poor process to pull them both, and I think it gets at what my process really is, and is an interesting discussion.
I’ve talked before about how I do very much value data and numbers, but how I don’t have a model, and my selection process here is actually somewhat anti-analytics, if you will. It’s not that I ignore the data, though; I watch every game, every week, and I write up every game as well, in depth, on the fantasy side over at Stealing Signals. Those are data-heavy pieces that also mix in key contextual elements.
But in 2024, anything I could build into a model is already in the lines. We’ve known for years that lines don’t reflect some middle of public sentiment, but that Vegas is actively betting against you, taking positions that at times run counter to where as much as 80% of the money might be coming in on a game. Sure, in general, their goal is to find a line where they get money on both sides and don’t have huge liabilities, but if they think the public is wrong on something, they are not going to set a line that’s wrong and get hammered by late-week big bets from the sharpest gamblers. They need to keep their lines as correct as they can, and they aren’t moving lines just because a bunch of casual fans put $100 down on one side and that’s starting to add up, because they can quickly wind up in a worse position with just one or two strong bets the other way — “worse” in terms of having a liability they believe is more of a risk, whereas the public liabilities are often ones they feel OK betting against.
All of this relates back to the specifics of the different games, but part of where they are getting their perceptions is in the data. In the case of the Washington and Denver games last week, one of the factors that undoubtedly would have been holding the lines down a bit is long-term data on rookie QBs, who — outside perhaps a few names like C.J. Stroud — have to be a lucrative subset for books, as casual fans get excited about the next messiah every time they have a good game, and rookie QBs as a group consistently regress back or get found out.
It’s probably the case that I’ve been too exuberant on these situations as well. But it’s an example of how my picks tend to be sort of a post-analytics approach, where the edge lies in not ignoring data but in understanding it well enough to disagree with it. “I know what’s driving this line and I just don’t believe it’s accurate data.” Historically, I’ve been at my best when the data is less robust, in October. Certainly, if Jayden Daniels is a legit MVP candidate, Washington’s weekly spreads will better reflect that by November or December.
So why didn’t I take those picks? It was a hyper-focus on bad outcomes. My process always includes a “What if this is wrong?” element. In the Seahawks-Giants game I mentioned, I wrote in this piece about how it was something of a letdown spot for a Seahawks’ team that was scheduled to play that game on a Sunday following their Monday nighter in Detroit and with a big Thursday night game against divisional-rival San Francisco behind it (one of the most ludicrous scheduling stretches for any team this year, and they went 0-3, which damaged our futures). My note for the Giants’ pick was I didn’t think Seattle would come out flat, back at home, but defensively they did really struggle to get the Giants off the field, and New York dominated time of possession by about a 15-minute edge. Seattle ran only 18 plays total in the first half — in part due to the fumble return TD they got at the end of a Giants’ drive that took more than 10 minutes off the clock, then gave the ball right back to New York — and then they were down and going pass heavy in the second half, resulting in just 11 rush attempts all game in what was just a weird script. Anyway, I missed that game, and for reasons I realized were possible, and wrote about not necessarily buying.
But with the Washington and Denver bets I pulled, I was hyper-focused on these outcomes. In the writeup where I took Denver -2.5, I talked about this specifically:
“I’m tempted to also take the over on the low number because I think the Broncos should definitely be in the twenties and might flirt with 30 here, but what worries me is in those outcomes Denver’s defense might be able to completely clamp the Raiders down in like a 31-3 win.”
Not taking an over on a line sitting in the mid-thirties when you feel one side could flirt with 30 on their own certainly shows a lack of conviction in the take itself, and that’s part of it. I also mentioned some scenarios where Denver covered the 2.5 if they won “something like 16-13,” but I’ve made some bets where I’m doubling down on one call by playing it two ways and been wrong and it hits you twice obviously, so part of it on this one was just picking my favorite of the two ways to play Denver’s offense being more competent in this spot than the models were giving it credit for.
But this line that there was a possibility Denver could hit 30 and the game might still not go over is a fundamental misunderstanding of betting. Nothing is a sure thing. There are always outcomes to the other side. That’s a low-probability scenario I was discussing, and I gave it outsized impact on my decision, when ultimately we’re just trying to find spots we believe have something like a 60/40 chance of winning, even when we acknowledge the 40% side. What we saw in the actual game — Denver’s defense scoring a touchdown to help them to 34 points, and a 34-18 win with the Raiders’ offense actually putting up more total yards — reminded that I was ultimately being too precise here, and there are a lot of ways football games (and takes) can play out.
Similar went for the Washington game, where I got pretty focused on how bad Washington’s secondary is, and Cleveland’s high pass rate, and how Deshaun Watson had hit Amari Cooper for a long TD called back by a hold the week prior, in believing I was maybe overlooking the Browns’ potential to look vaguely competent offensively in that spot, in service of my extreme Daniels hype. But again, a bet is a collection of probabilities, and even if I thought that was possible, I was high enough on Daniels to be calling him an MVP candidate, and that was a 3-point home line against a team that is struggling mightily. It wasn’t a certainty that Cleveland would look horrendous on offense again, with Watson continuing to miss open throws that they just have to have, but it also wasn’t like the Commanders needed that outcome just to squeak by — when that happened, they dominated, 34-13. Even if the Browns had shown some life, backing Daniels might have been right.
So the lesson for me this week is we’re not trying to be perfect. We’re trying to find weighted coinflips. Getting overly precise with your analysis can talk you off any line — there’s almost always a way to pitch either side of either bet. You need to have respect for all eventualities, but it’s about pricing the probabilities correctly, with an understanding of why they exist and what the data says, but at times a willingness to just say, “I think this player (or this situation) is unique.” Can that lead to overconfidence? Absolutely. There’s gray area with all of it. A fluid, dynamic thought process is the first and most important rule.
Last week was a frustrating one for me, where most of my biggest stances were things I didn’t leverage enough. We’re definitely not just happy with .500 records around here; I want to be so far in the green this year there’s no question that last year was a bad runout. That also requires respect for the market and the probabilities, and I never want to get overzealous.
One major issue this week is I like several road favorites, but we know that’s been a tough segment for me. Still, we have 10 picks this week, as I’m feeling kind of frisky at this point of the season. Let’s get to them.
As always, all minus bets are bet to win 1 unit, all plus bets are risk 1 unit.
Jacksonville Jaguars at Chicago Bears
Pick: Under 44.5, -110 (widely available)
London games have had a tendency to go under, and this line does arguably reflect that. Jaguars’ games have tended to be higher-scoring, including a 71-point shootout with the Colts last week. The Bears also seemingly got their offense squared away last week in a 36-10 drubbing of the Panthers, and you have to be optimistic in Caleb Williams going forward.
The other relevant game last week was the one in London, which ended 23-17 for 40 total points and an under. Last year, the five games in Europe featured final totals of 30, 45, 40, 35, and 16 points. The 45-point game was Jacksonville’s second game in London after staying there a week; they beat Buffalo outright as 5.5-point ‘dogs, while the game did still go under a high total, meaning the Bills significantly underperformed their team total as the team traveling that week.
We did get a couple higher-scoring games in 2022 — that year we got 53, 49, 38, and 37 total points — but there is a clear trend that teams have a hard time sustaining a shootout pace in a game where they’ve traveled a long distance. Of course, it’s not like this stuff isn’t known by the models, which is what I alluded to in the first paragraph, where this number feels light given the way these offenses played last week.
The bet here is that neither of these teams has really figured it all the way out. Chicago has a strong defense, and the stuff Jacksonville has done offensively this year has been uninspired, while the Bears’ offensive planning under Shane Waldron has been similarly problematic. The longer-term trend of teams looking disorganized after the long travel is just half the point for me, with the other half being that both of these teams are prone to looking disorganized any week.