Week 10 Lines, Totals, and Implied Totals
I'm having a clamato time
Week 9’s marquee matchups (Dolphins - Chiefs, Seahawks - Ravens, Cowboys - Eagles, and Bills - Bengals) delivered in a way that can only be compared to a slowly deflating balloon at a funeral… I’m reminded of the time that I ordered a coloring book for my son from Amazon, and in the box was not only a coloring book, but a copy of the new Gears Of War Video Game (that I didn’t order). I was so excited - like a kid on Christmas receiving an unexpected, totally awesome gift from their fun uncle (that their parents more than likely didn’t want them to have) - and then I realized that I don’t own an XBOX… and probably didn’t have time to play any video games… ever.
Was that a mixed metaphor that got lost somewhere along the way? Yes… Were those games mediocre, also yes. One could argue the Cowboys - Eagles contained some last minute excitement, but the Cowboys’ collective ability to embrace failure with open arms was on full display with:
A Touchdown that was overturned on fourth down because the TE fell down at the 1 inch line.
A 2 point conversion attempt (to only be down 3) where Dak went out of bounds at the 1 inch line.
3 major penalties for the Eagles, pushing the Cowboys to a First and Goal from the 9 with a minute left, down 23-28.
A sack, 2 penalties, and now a second and goal from the Eagles 25 yard line…
I’m sure I missed a few acts of futility from earlier in the game, but the point remains, the biggest matchups ended up as the biggest flops.
Week 9 Model Performance
Before diving in to the Spread/Totals performance, I’m going to victory lap my MoneyLine plays for last week.
Vikings (if Dobbs plays) - +180 (Win $18), Also live bet at +300 when Dobbs came in (Win $30)
Commanders - +150 (Win $15)
Cowboys - +135 (LOSS)
Bills - +125 (LOSS)
Each bet was $10 - so despite losing the Cowboys and Bills (who also didn’t bother covering), I netted $43 on 5-$10 bets. Live betting MLs counter to momentum swings is a great way to get better odds (see Vikings games), particularly early in games, where you could double down on your initial assessment (or hedge if you have to).
The model selected 14 lines as favorable - going 6-8 ATS. The one bad beat was the Panthers (+2.5) losing to the Colts 13-27 (11.5 point net loss). We also had three games that were right on the line - Vikings vs. Falcons (slight Vikings to slight Falcons guidance due to line movement), Saints 24 (-7.5) - Bears 17 (0.5 point net loss ATS), and the Texans 39 (-3) - Buccaneers 37 (1 point net loss ATS). In a week that was chaotic, the ML bets helped us overcome variability in spreads, even if it would have meant sacrificing a bit on returns.
The model selected 14 TOTALS as favorable - going 9-5. The UNDER carried 11-3 games, 2 of the OVERS were games with sub-40 point TOTALS on the day, so the model actually effectively identified variances in both low and high total games.
If you’re following along at home - the UNDER is now carrying 61.8% of games on the year (84-52), and we saw the second lowest PPG average (41.4) on the season. Scoring trends continue to show that the UNDER will win because Vegas is still making money with betting lines at 49% towards the OVER and 51% to the UNDER.
As we removed weighted betting, we would have netted $20 for the week on $140 wagered. Not a bad return if you were to play every game/odd, which is not something I’d ever recommend.
Five Things I Think I Think, Plus The Focus Area
International Football Sucks - I’m sure as a fan it’s great to see your team play in a foreign city, but the product feels diluted, and I’m not sure any team has figured out how the travel/logistics work.
Thursday Night Football Sucks AND Blows - Prioritizing attending a Football game knowing you have to “work” Friday + a diluted, exhausted product is just terrible all around. I’m digging in to TNF Trends for next week.
The Model HATES how well the Bengals are playing right now, I effectively had to adjust it to only take in to account their last 3 games…
The Ravens hold the distinction as the UNDER-COVER Kings - a team who will cover almost any spread in a game that fails to reach a point total.
To the point above, I spent some time this week pondering the question - what types of games do teams generally find themselves in. I.e. High Scoring Blowouts vs. Nail Biter Low Scoring Games…. So I did what comes naturally, I plotted it…
Deep Dive - Characteristics of a Game For Each Team
I stole this concept from SmolaDS on Twitter, where he mapped each teams’ average Points For + Points Against to determine how likely a team was to participate in a high scoring game. For Instance, the Dolphins score an average of 32 PPG (Points For) and their defense gives up 25 PPG (Points Against). They have the highest Total Team Points in the NFL (57 total points - Dolphins 32, Insert Opponent 25) - which means that a Dolphins game is likely to hit the OVER. On the other hand, the Jets hold one of the lowest totals (36 total points) scoring 16.5 PPG and allowing 19.5 PPG - so they’re likely to participate in a LOW scoring game (favoring the UNDER).
There is another topic here - do these teams WIN or LOSE according to their Totals. To determine this, I plotted NET Scoring, Points For minus Points Against, to determine if teams generally “Win” or “Lose”. In the example with the Jets above, their NET scoring is 16.5 PF - 19.5 PA = -3 NET PPG, so they are losing by, on average, 3 PPG.
By merging these two datasets together, we can plot if teams are likely to Win/Lose and in High or Low scoring games…
The Results?
The raw data and graph are now in the Week 10 Tracker so you can see how a given team fares - but in essence:
Y-AXIS Net Points: The “higher” a team is on the Y Axis - the more team “Wins” by, and the “lower” - the more a team loses by.
X-AXIS Average Total Points (net NFL Scoring average of 43.5 PPG) -The left-most teams participate in games with the lowest TOTAL points - the right-most teams participate in the highest scoring games.
How can you use this? Well - it’s an interesting way of aligning our “feel” about certain teams - the Dolphins tend to win and score a lot of points - and identify how they generally align against their spreads/totals.
Let’s use the RAVENS as an example (Upper-Left Quadrant) - they favor winning by a lot in low scoring games, meaning they’ll generally make their Implied Total and their Opponent will not (Ravens contribute many points, opponents do not). This is backed up by the fact that they Score, on average, 26.5 PPG, and allow a league low 13.75 PPG.
What’s the takeaway - I’d love to start factoring this easily calculated data in to the matchup models as secondary guidance. There’s definitely something to using this as another source of information in terms of the types of games teams generally participate in.
There are always exceptions - like the Dolphins 14 - Chiefs 21, the model favored the Dolphins, but as you can see here, the Chiefs (7.2) have a slightly higher net points than the Dolphins (6.7), and the TOTAL for that game (50.5) is not the mid-point between the Chiefs Average Total Points (39) and the Dolphins (57) - the average of both teams’ total points (48) leans us towards the UNDER. Using this data, we’d expect the Chiefs to cover a (-1 to -2) point spread and for the UNDER to be the bet (50.5 anticipated points against 48 average points for both teams).
To me, this is another interesting way to look at data - and yes, for all of you know it alls - the model does take Week 9 in to account which is probably why the Chiefs-Dolphins game looks the way it does.
Week 10 Picks
There are quite a few spreads that stick out to me this week in terms of Delta to the model -
Bengals (-6.5) vs. Texans - the model feels like this should be closer to Bengals -4.5, and the Bengals haven’t exactly been "Winning Big”; that said, both teams are very banged up…. Not sure I’ll be betting this one
Falcons vs. Cardinals (+1.5) - Kyler Murray is playing this week, and the line proves that Vegas assumes gamblers are all about the Cardinals. Honestly, I’d take them +1.5, I might even take the ML given how streaky the Falcons are playing right now.
Giants vs. Cowboys (-17) - The Giants have an implied total of 11, and while a 17 point spread seems difficult to cover, the implied total is the Giant’s scoring average (11.22) on the season - meaning they’ll have to “beat expectations” against a stingy Cowboys defense if they want to cover.
Washington vs. Seattle (-6) - I generally avoid East-West coast games not in prime time because travel distance might truly be a problem (just ask Seattle who got their doors blown off last week after travelling east to Baltimore). That said, their notoriously stout defense gave up a lot of big running plays, and I don’t think the Commanders have as solid of a running game… We’ll see - 6 points is a lot against a good Washington offense (and bad defense).
We’re also leaning heavily on the UNDER - with 8 matchups favoring the UNDER, 4 the OVER, and 1 NEUTRAL. We lost our socks on Thursday calling the OVER, but you can’t be mad expecting the Panthers (avg 17.5 PPG, Implied Total 17.75) and Bears (avg. 21 PPG, Implied Total 21.25 PPG) to match their season scoring average against bottom 5 defenses… Thursday Night games both suck and blow…
Good luck this week - may the odds be forever in your favor…
Share this post, leave a comment, or reach out to me on Twitter @dbfinsightsvt