MGCOLBY's Picker Thread 2024
- Thread starter mgcolby
- Start date
mgcolby
Go Eat a satchel of Richards!
I tried to get this posted before kickoff but didn’t quite make it. I want to be clear up front that there are no changes to any predictions, lines, or picks from the weekly slate I posted on Tuesday. This post is strictly informational.
So why am I reposting Week 16 picks, something regulars here know I rarely do? I finally got around to testing my confidence rating system, and I wanted to share the results.
Below is the season record heading into Week 16, broken down by confidence level.
Based on those results, I’ve added a confidence rating column to the picker and introduced two new visual alerts. You’ll notice I haven’t highlighted Shamlock’s Pick of the Week yet. I usually prefer to wait closer to kickoff so I can factor in as much late information as possible. That said, I selected it earlier than usual this week so all three alerts would be visible. Shamlock’s Pick of the Week is marked with a green shamrock. It does not distinguish between ATS or totals in this view, it's shown in the weekly view I post.
The two new alerts are represented by pots of gold. One is a full pot of gold for five-star plays, and the other is a spilled pot of gold for one-star plays.
As you can see in the chart, five-star picks are hitting at 84.6 percent while accounting for just under 10 percent of all games since Week 5, specifically 9.3 percent. The pot of gold is simply a quick, quirky visual to flag those high-confidence spots. Digging a little deeper, the model is 5-0 when both the model and Vegas are within one to two points of each other and agree on the favorite. It is 6-2-2 when the predictions still fall within that one to two point range but disagree on who the favorite is.
On the other end of the spectrum are the one-star ratings, which I’m labeling the fade zone. The spilled pot of gold signals games that are worth considering as fades. As part of my personal approach, I try to be as honest and transparent as possible without giving away the recipe. If the data suggests fading the picker in certain spots, it only makes sense to acknowledge that. If I were to reclassify these one-star picks as fades, it would add five wins to the ATS season total, pushing it to 61 percent on the year. That’s nothing to scoff at.
So, welcome to the new fade zone picks.
I will note that I have not added a totals confidence. One not sure its 48% win rate warrants one and honestly I haven't narrowed down a pattern that structurally works i.e. more stars equals higher win percentage.
So why am I reposting Week 16 picks, something regulars here know I rarely do? I finally got around to testing my confidence rating system, and I wanted to share the results.
Below is the season record heading into Week 16, broken down by confidence level.
Based on those results, I’ve added a confidence rating column to the picker and introduced two new visual alerts. You’ll notice I haven’t highlighted Shamlock’s Pick of the Week yet. I usually prefer to wait closer to kickoff so I can factor in as much late information as possible. That said, I selected it earlier than usual this week so all three alerts would be visible. Shamlock’s Pick of the Week is marked with a green shamrock. It does not distinguish between ATS or totals in this view, it's shown in the weekly view I post.
The two new alerts are represented by pots of gold. One is a full pot of gold for five-star plays, and the other is a spilled pot of gold for one-star plays.
As you can see in the chart, five-star picks are hitting at 84.6 percent while accounting for just under 10 percent of all games since Week 5, specifically 9.3 percent. The pot of gold is simply a quick, quirky visual to flag those high-confidence spots. Digging a little deeper, the model is 5-0 when both the model and Vegas are within one to two points of each other and agree on the favorite. It is 6-2-2 when the predictions still fall within that one to two point range but disagree on who the favorite is.
On the other end of the spectrum are the one-star ratings, which I’m labeling the fade zone. The spilled pot of gold signals games that are worth considering as fades. As part of my personal approach, I try to be as honest and transparent as possible without giving away the recipe. If the data suggests fading the picker in certain spots, it only makes sense to acknowledge that. If I were to reclassify these one-star picks as fades, it would add five wins to the ATS season total, pushing it to 61 percent on the year. That’s nothing to scoff at.
So, welcome to the new fade zone picks.
I will note that I have not added a totals confidence. One not sure its 48% win rate warrants one and honestly I haven't narrowed down a pattern that structurally works i.e. more stars equals higher win percentage.
mgcolby
Go Eat a satchel of Richards!
Season Record & Week 16 Results
The picker having a two week slide here ATS, ironically as the totals are starting to come around. Also, a shit week to have its first Shamlock's gold loss when it agrees with Vegas (Jets game), the same week I introduce it. Seems par for the course. Still in the green for the season and overall having a good year ATS. EPR is rock steady, I'll have a comment on that later in the thread.
The picker having a two week slide here ATS, ironically as the totals are starting to come around. Also, a shit week to have its first Shamlock's gold loss when it agrees with Vegas (Jets game), the same week I introduce it. Seems par for the course. Still in the green for the season and overall having a good year ATS. EPR is rock steady, I'll have a comment on that later in the thread.
mgcolby
Go Eat a satchel of Richards!
I'm aware and in awe of Miami as you are or will be. I looked at and two things stand out. As you know its based on the last four games, well Miami still has two good games remaining in that window. I expect them to drop significantly with an asterisk next week because the best game in that window will drop off next week. The other part of the scenario, Miami is statistically doing well in the two highest weighted categories of the formula. So yes, I fully expect them to account for an EPR loss this week and they are a bit of statistical/situational anomaly.
mgcolby
Go Eat a satchel of Richards!
Week 18
For fun I went and used ESPN's playoff time machine. I used the EPR rankings to determine each game. I did not account for teams sitting starters (Chargers being the obvious one), just straight up win/loss based on higher lower ranking.
And just for the likely hood Trey Lance doesn't come out and light up the Broncos to the tune of er one more point than they have when the clock hits zeros:
mgcolby
Go Eat a satchel of Richards!
Shamlock's Pick of the Week
Realize I have been slacking on the record:
Season record 9-4
Week 5: Detroit: -9.5 (W)
Week 6: Arizona/Indianapolis: Over 46.5 (W)
Week 7: Houston/Seattle: Over 41 (W)
Week 8: Pittsburgh: +3.5 (L)
Week 9: NE: -5.5 (L)
Week 10: Arizona/Seattle: Over 44.5 (W)
Week 11: Seattle/LA Rams: Over 48 (L)
Week 12: Dallas: +3.5 (W)
Week 13: Chicago: +7.5 (W)
Week 14: Washington +2.5 (L)
Week 15: NYJ/Jag Over 41.5 (W)
Week 16: NE/Bal Over 48.5 (W)
Week 17: NYG +1.5 (W)
mgcolby
Go Eat a satchel of Richards!
Week 18
Carolina @ Tampa Bay:The NFC South title comes down to this - the 8-8 Panthers at the 7-9 Buccaneers with both teams still alive for the division crown. Carolina clinches with a win or tie, while Tampa needs a victory combined with an Atlanta loss. Neither offense has been lighting it up lately - the Panthers average 20.3 points (25th) while the Bucs sit at 21.3 (23rd). This has the makings of a gritty, low-scoring divisional affair.
Carolina's edge is the turnover battle with a plus 0.8 margin, forcing 6 takeaways over the last four games while committing just 3 turnovers. Tampa's red zone efficiency at 72.7% (4th) gives them an edge when they get close, but their defense has allowed 5.64 yards per play (20th). The Panthers' ability to force turnovers could be the difference in a game where neither offense is explosive.
The model projects Tampa winning around 22-19, which is right at the 3-point spread. Carolina getting points with the division on the line makes this feel like a coin flip.
Seattle @ San Francisco:
The NFC West title is on the line as the 13-3 Seahawks visit the 12-4 49ers. Winner clinches the division, a first-round bye, and home-field advantage throughout the NFC playoffs. San Francisco leads the league in scoring at 38.3 points with a 64.2% third-down conversion rate - also best in the NFL. Seattle counters with the league's best defense, allowing just 8.8 points per game and 3.51 yards per play.
The 49ers' offense has been firing on all cylinders - 404 total yards per game with 7.85 yards per pass attempt (4th). Seattle's pass defense leads the league at just 5.45 yards per attempt allowed, setting up a fascinating strength-on-strength matchup. The Seahawks have forced 11 takeaways over the last four games with a plus 1.5 turnover margin (2nd), but San Francisco protects the ball with just 0.8 turnovers per game.
The model projects San Francisco winning around 40-27, which is a much larger margin than the 1.5-point spread suggests. The 49ers' league-leading offense against Seattle's elite defense is appointment viewing, but San Francisco's ability to sustain drives and finish in the red zone (72.7%, 4th) gives them the edge at home with everything on the line.
mgcolby
Go Eat a satchel of Richards!
Season Record & Week 18 Results
Not a losing week. But week 18 is always a bit rough on the picker. It did make up ground on the totals. Ended up splitting the week 8-8 against the spread a couple of tough games that got away from it but thems the breaks as they say.
