We're betting to win.

And if I'm the one doing the betting, there's two keys I need to have to have the confidence to place the wager:

1. A model-backed approach, or at the very least a quantitative-based angle. But almost all of the bets in this column will be based on the outputs of models built by me, or occasionally my colleagues at ESPN Analytics.

2. A less efficient market. NFL sides and totals are voluminous markets, and betting against those mainstream numbers is like wagering on All-Madden mode. It's why I often look to props (especially on defense!) and alternate lines. Less attention means less efficient markets, and therefore more opportunities to find value.

Those two criteria make up the crux of this weekly column. Each week I'll post bets from our models in mostly lower-visibility categories with the simple goal of coming out ahead. We'll be looking at odds all across the spectrum, from -1000 to 100-1s -- as far as I'm concerned, value is value no matter the price.

One promise I'll make: we'll track the results -- by bet category -- no matter what. It's sort of scary, because losing is of course very much in the cards. But I also think it's really important. And it's what what I did a year ago in my similar "last-minute bets" column.

Last season it finished with a 6.2% ROI on 466 bets. I can't guarantee success just based off that -- it will be harder this year, especially if there are no sack props (my most profitable category) available, as there aren't currently in Week 1 -- but we'll post how we've performed, win or lose.

To continue the theme of transparency, here are the ground rules I'm setting:

1. For simplicity, I am flat-betting one unit per bet regardless of odds.

2. The story will be filed on Thursdays and posted shortly thereafter. I'll check the lines right before I file, and once I do I consider them locked, no matter if they move between then and kickoff, regardless which direction.

3. I can add bets later. So if a prop isn't offered on Thursday, but is added on Friday that I show value for, I can file that bet and it will be posted shortly thereafter, with a note indicating which day it was added. Same odds movement rules apply.

4. I'll treat no action bets as pushes.

5. All odds will be via ESPN BET. Of course, my advice to you is to always find the best line you can.

With all the housekeeping out of the way, let's talk football shall we? It's time for some Week 1 bets.

Alternate receiving lines

Wan'Dale Robinson 40+ receiving yards (+200)

Because he tends to catch short passes, Robinson doesn't particularly have the profile of a receiver that is going to vastly exceed his prop line, which is 29.5. And you can see that in the chart below, Robinson is more likely to reach 20, 30 or 40 yards than his teammates Darius Slayton and Jalin Hyatt, but less likely to have an explosion game. However, at 40+ yards this is just a good price for Robinson: it pays 2-1 for him to exceed his over/under by just 10.5 yards.

That Robinson commanded a very respectable 21.5% target share last year helps. And while not in the model, I think the fact that Robinson runs quick routes ought to be an asset in particular against the Vikings, who blitz a ton under defensive coordinator Brian Flores.

A new model I worked on this offseason is a full distribution of receiving yards for a single player in a single game.

Here are the Giants WRs in Week 1. The chart shows the probability (y) of each receiver reaching at least x yards. pic.twitter.com/zLipEJHQ2a

— Seth Walder (@SethWalder) September 5, 2024

K.J. Osborn 30+ receiving yards (+210)

One cue the model is taking is Osborn's receiving yards props from a year ago, when he averaged an over/under of 30.4. Now, situation certainly plays a big role in those numbers and his current lines. But he certainly can be a receiver whose median line is around 30 in the right circumstances. New England's receiver roles are a little murky, so we're using the uncertainty to our advantage by playing the alt line. Plus, the fact that the Patriots are 8.5-point underdogs to the Bengals means they will likely be behind -- and throwing.

Also consider:
Jauan Jennings 35+ receiving yards (+260)

Anytime touchdowns

OK, these aren't exactly under the radar bets -- I was originally planning on mostly focusing on no touchdown bets -- but they're still backed by a model and so that's good enough to fit into our criteria.

Nico Collins to score 1+ touchdowns (+185)

From the model's perspective, there's a ton to like about Collins here. He's an outside receiver in an offense with a high implied team total (25.75) and was simply excellent last season. A number the model loves to see: Collins ranked fifth in completed air yards per route (think: yards per route, but only on yards that travel through the air because those are more repeatable than YAC yards) at 1.91.

Also consider:
Jaylen Waddle 1+ touchdowns (+145)
Tyreek Hill 1+ touchdowns (-125)
Tank Dell 1+ touchdowns (+260)

Multiple touchdowns

Tank Dell to score 2+ touchdowns (+2000)

It's pretty clear my touchdown model is high on the Texans in Week 1. I'm slightly wary that this may be because it's not fully incorporating the added target competition that Stefon Diggs brings, but at the same time one thing the model also doesn't know -- that I put stock in -- is that Dell scored incredibly well in ESPN's receiver scores with an 81 open score that ranked eighth-best last year. In general, the model likes Dell for a lot of the same reasons it likes Collins. Dell ranked sixth in completed air yards per route last year, exactly one spot behind Collins. I'm treading lightly with my touchdown models because they are brand new, but for what it's worth I have Dell's 2+ touchdown chances priced at +1265.

Alternate total

Raiders at Chargers over 45.5 (+170)

This comes from a brand new FPI+ model created by my colleague Dylan Mervis on the sports analytics team. FPI+ is a translation of our generic Football Power Index into a model designed to be used in the betting market. So why the alternate over here? The combination of the Chargers' offensive FPI rating (13th) and Raiders (26th) offensive FPI rating results in the model believing the median total should be slightly higher than 40.5 line it currently is. And when looking across the spectrum of possible totals, over 45.5 is where we had the largest disagreement with the market. FPI+ prices that alternate over at +156.