ELDORADO
  • Home
  • Football
  • Basketball
  • Politics & Culture
  • Contact | Subscribe
NEW: 2023 Super Bowl Squares Odds
Check out this year's probabilities!
Searching for Value in NFL Picks and Futures
​Weeks 1-4 - Weeks 5-8 - Weeks 9-12 - Weeks 13-17 - Summary
Tweet Follow @eldo_co

IMPORTANT NOTE

Almost every one of Searching for Value's NFL picks is an underdog. When betting underdogs on the moneyline (to win straight up), bettors are paid a multiple of the amount they've risked to compensate them for backing a team that the market believes has a lower or much lower likelihood of winning.

This NFL season, the system's average Tier 1 and Tier 2 Moneyline plays have had +306 odds, meaning winning bets would net 306 in profit for every 100 risked. The system's average Weekly Top 5 Moneyline plays have had +187 odds, meaning winning bets would net 187 in profit for every 100 risked.

As explained in the column, the system actually expects to lose more moneyline *bets* than it wins. (With average odds of +300, a bettor needs to win 25% of the time to break even.) But when it is earning over 300 per win and "only" losing 100 per loss, the record(s) below can produce highly positive returns.

FOR ONGOING REFERENCE (CLICK HEADERS TO READ MORE)
"SEARCHING FOR VALUE" EVALUATES WHICH MONEYLINE PLAYS HAVE THE HIGHEST PROJECTED EXPECTED VALUE 
This column attempts to quantify the expected value of each week's money-line plays (i.e., betting teams to win "straight up," not against the spread) throughout the NFL season. Generally speaking, a bettor (or model) would see positive expected value in a given bet if that bettor (or model) thought a certain outcome was more likely to occur than the sportsbooks' odds (or payout) implied.

For example, in Week 1, the sportsbooks had the Seattle Seahawks as +240 underdogs at home versus the Denver Broncos (i.e., risk $100 to profit $240 if Seattle wins; lose $100 if Seattle loses). By setting Seattle's money line at +240, the pre-game odds implied that the Seahawks had a 29.4% chance to win the game ("implied probability"). Positive expected value would exist if you (or some model) believed that Seattle had a higher probability to win than the 29.4% chance implied by those +240 odds. (Seattle won the game.)

(At the request of readers, I am also now tracking how these are performing when the same teams are taken against the spread.)
THE PERCENTILES REFERENCED BELOW ARE NEITHER PROBABILITIES TO WIN NOR CONFIDENCE LEVELS
The numbers associated with each play are percentiles. They indicate where the expected value of a hypothetical moneyline pick ranks when compared to all potential moneyline picks over the last several seasons (i.e., among several thousand possibilities).

Teams in Tier 1 ("Selections") are in the 99th percentile or above when compared to all those possibilities. Over the historical period I've studied, these have occurred approximately once every three weeks (i.e., 1 out of every 50 games, aka 1 in 100 team-games). 

Teams in Tier 2 ("Considerations") fall between the 94th and 99th percentile when compared to all those possibilities. On average, considerations should occur a little less than twice per week. They have roughly broken even over time (hence the name).

I also include a weekly Top 5. Betting all of those would have a negative return over time, though they have done well in 2022. 
​
the model expects to LOSE more MONEYLINE *BETS* than it WINS, but its wins are worth ~3x more
Underdogs are underdogs for a reason; the market thinks they're more likely to lose (sometimes a lot more likely) than they are to win. Consequently, moneyline underdogs pay more (sometimes a lot more) for victories (e.g., +300) than the amount at risk (e.g., 100).

If a bettor's average moneyline play is +300, they "only" need to win 25%+ of the time to break even, assuming equally sized plays. The average Tier 1 Selection in the period I've studied has carried +488 odds. The average Tier 2 Consideration has carried +343 odds.

The system is making plays that the market thinks should win only about 17-25% of the time, so they pay a lot when they win. When making 300 or 400 per win and losing 100 per loss, "losing records"
— which the system expects — can produce positive returns.
IF A TEAM IS LISTED BELOW, IT DOES NOT MEAN THE SYSTEM THINKS THEY ARE LIKELY TO WIN THEIR GAME
Positive expected value instead suggests that the model thinks the team is more likely to win than the current moneyline odds imply. And while this column focuses purely on numbers, whether you actually think that's the case is ultimately up to you and your gut!
All TEAMS that have qualified AS tier 1 or TIER 2 MONEYLINE PLAYS in 2022, WITH RESULTS
Full season-to-date: 
​9-19-1 (
+740 on 2800 risked, +26.4% return on dollars risked)
All underdog moneyline plays. In 2022, the system made an average of 306 per win and lost 100 per loss (100 per play). Hence the above returns.

1. TEN (+600) vs. DAL - Week 17 - 99.8 percentile - Lost 27-13 (Dobbs at QB)
2. SEA (+240) vs. DEN - Week 1 - 99.7 - Won 17-16 (Smith at QB)
3. DAL (+275) vs. CIN - Week 2 - 99.6 - Won 20-17 (Rush at QB)
4. TEN (+600) at KC - Week 9 - 99.3 - Lost 20-17 (OT) (Willis at QB)
5. HOU (+390) at DEN - Week 2 - 98.3 - Lost 19-6 (Mills at QB)
6. LAR (+245) vs. LVR - Week 14 - 98.3 - Won 17-16 (Mayfield at QB)
7. LAR (+265) vs. SEA - Week 13 - 98.2 - Lost 27-23 (OT) (Mayfield at QB)
8. ARZ (+300) vs. TB - Week 16 - 98.0 - Lost 19-16 (OT) (McSorley at QB)
9. HOU (+575) vs. PHI - Week 9 - 97.7 - Lost 29-17 (Mills at QB)
10. PIT (+380) vs. TB - Week 6 - 97.6 - Won 20-18 (Pickett at QB)
11. ARZ (+360) vs. SFO - Week 11 - 97.5 - Lost 38-10 (Blough at QB)
12. CAR (+530) vs. TB - Week 7 - 97.2 - Won 21-3 (Walker at QB; McCaffrey just traded)
13. WAS (+240) vs. PHI - Week 3 - 96.4 - Lost 24-8 (Wentz at QB)
14. SEA (+310) at SFO - Week 2 - 96.3 - Lost 27-7 (Smith at QB)
15. ARZ (+215) vs. PHI - Week 5 - 96.1 - Lost 20-17 (Murray at QB)
16. HOU (+283) vs. IND - Week 1 - 96.0 - Tied 20-20 (Mills at QB)
17. IND (+240) vs. PHI - Week 11 - 96.0 - Lost 17-16 (Ryan at QB)
18. HOU (+205) vs. LAC - Week 4 - 95.9 -  Lost 34-24 (Mills at QB)
19. LVR (+370) vs. SFO - Week 17 - 95.5 - Lost 37-34 (OT) (Stidham at QB)
20. CHI (+265) vs. SFO - Week 1 - 95.4 - Won 19-10 (Fields at QB)
21. LAR (+130) vs. DEN - Week 16 - 95.4 - Won 51-14 (Mayfield at QB)
22. DAL (+245) at PHI - Week 6 - 95.2 - Lost 26-17 (Rush at QB)
23. TEN (+115) vs. LVR - Week 3 - 95.2 - Won 24-22 (Tannehill at QB)
24. IND (+190) vs. LAC - Week 16 - 95.0 - Lost 20-3 (Foles at QB)
25. ATL (+140) vs. LAC - Week 9 - 94.6 - Lost 20-17 (Mariota at QB)
26. NYJ (+275) vs. BAL - Week 1 - 94.6 - Lost 24-9 (Flacco at QB)
27. PIT (+295) at MIA - Week 7 - 94.5 - Lost 16-10 (Pickett at QB)
28. NYJ (+460) vs. BUF - Week 9 - 94.4 - Won 20-17 (Wilson at QB)
29. ARZ (+150) vs. LAC - Week 12 - 94.3 - Lost 25-24 (Murray at QB)

FINAL RESULTS (2022-23)

Tier 1/2 Moneyline: 9-19-1 (+740 on 2800 risked, +26.4% ROI)

All underdogs to win straight up - Average odds +306

Weekly Top 5 Moneyline: 31-53-1 (+449 on 8400 risked, +5.4% ROI)
Almost all underdogs to win straight up - Average odds +187

Weekly Top 5 Against the Spread: 47-38-0 (+520 on 9350 risked, +5.6% ROI)
Almost all underdogs to win against the spread

If you are not familiar with what this stuff means - how to interpret these results - why these records are good (or lucky), at least for now - please read the explainers above

"Searching for Value" Season Summary

With the Super Bowl upon us, it seems like an appropriate time to summarize and reflect on this season's "Searching for Value" initiative. The literal summary can be found in the table below, which details all of the 2022-23 season's Tier 1 Selections and Tier 2 Considerations.  Over the regular season and playoffs, 29 teams qualified as Tier 1 (99+ percentile) or Tier 2 (94-99 percentile) moneyline plays. All 29 of them were underdogs, and 25 out of 29 were home dogs. Taken together, Tier 1 and Tier 2 qualifiers won 9 times, lost 19 times, and pushed once. 

That record doesn't sound good on its face, and I actually receive the occasional email mocking it. Hopefully the table below, and in particular the two right-most columns, helps clarify why that record is good (if not really good) when you're dealing with (in this case exclusively) modest to heavy underdogs on the moneyline. This season, the average odds for these plays was +306. With the system netting ~300 per win and "only" losing 100 per loss, that "ugly" record of 9-19-1 produced a +26.4% return on dollars risked in 2022-23.

Scanning the list of teams and quarterbacks (scroll all the way down) can make your stomach churn. Even at +245, +390, or +600, backing Kenny Pickett and the Steelers (vs. Tampa Bay, when people still thought they were good), Phillip Walker and the Panthers (a few days after trading Christian McCaffrey), Justin Fields and the Bears (vs. San Francisco), Zach Wilson and the Jets (vs. Buffalo), Cooper Rush and the Cowboys (vs. Cincinnati), and Baker Mayfield and the Rams (three times!) feels like it goes against your better judgment. The same can be said for Geno Smith in Week 1, when Seattle was considered "meh" and many expected some sort of Russell Wilson resurgence in Denver.

But every single one of the teams and quarterbacks pulled off big moneyline upsets in the games mentioned above, driving this year's profit. Left to my own subjective devices, I probably would have talked myself out of most of those plays. But the goal of this column was expected-value based objectivity. The system didn't think these teams were necessarily going to win those games. It thought the odds had moved too far against them, to the point that there was potential value in backing them. If the market is underestimating certain underdogs' odds -- perhaps especially home underdogs, oftentimes with their backup under center -- then it requires the ability to hold one's nose.

Past performance is not indicative of future results, and a game here or a game there could certainly have swung these results. Take out Carolina's +540 upset of Tampa Bay in Week 7, right after the Panthers traded the aforementioned McCaffrey, and the results would be close to flat on the year. On the flip side, four of these games went into overtime, and the system lost all of them, including Titans (Willis at QB) +600 at Chiefs, Rams (Mayfield at QB) +265 versus Seahawks, Arizona (McSorley at QB) +300 versus Bucs, and the Raiders (Stidham at QB) +370 versus 49ers. Imagine the swing if one or two of those hit. We can sit here and play the "what if" game til we're blue in the face. 

Other Markets and Results

​To round everything else up, the weekly top five moneyline plays did pretty well this year (see final records above). Though as I always point out, actually betting all five every week produced negative returns over a multi-year lookback period. (I did not begin the season by sharing the weekly top five, but readers wanted to see it, so I started posting it again.) Taking those top five teams against the spread each week generated an almost identical positive return on dollars risked, and put together a streak of 8 consecutive winning weeks along the way.


Despite this column's title, I did not end up taking any real looks at the expected value of the futures board, other than sharing five win total picks I made before Week 1, as generated by an ancillary system I produced last summer. Those picks were shared before the season (scroll down to Week 1), with the note that "[i]n a typical season, the model thinks that about three of these will hit; it does not imply a clean sweep." As luck would have it, three of them hit, and at good odds for win-total over/unders, making this little foray profitable as well.

- Arizona under 8.5 wins (-110) and 7.5 wins (+150) Won 120
- Atlanta under 4.5 wins (+115) Lost 100
- Dallas under 9.5 wins (+135) Lost 100
- Green Bay under 10.5 wins (+135) and 11.5 wins (-120) Won 109
- Tennessee under 8.5 wins (+120) Won 120
(+149 on 500 risked, +29.8% ROI)

Playoffs and Super Bowl

This year's NFL postseason has not produced any Tier 1 or Tier 2 qualifiers. The Super Bowl doesn't either, though the system does lean toward Kansas City as +105 underdogs. At those odds, the Chiefs fall in the 88th percentile, which is short of the 94th-percentile cutoff for a Tier 2 Consideration (and to make the table below), but would fall in a typical regular season week's top five (see notes above).

During Wild Card Weekend, the system thought Jacksonville (+120) and Minnesota (-155) were a little undervalued at those odds. Like the Chiefs in the Super Bowl, they did not qualify as Tier 1 Selections or Tier 2 Considerations but would likely have made a typical regular season week's top five. The Jaguars were in the 89th value percentile and beat the Chargers, while the Vikings (83rd) lost to the Giants


In the Divisional Round, the system thought that all four games were pretty fairly priced. The closest play was Cincinnati, which the system thought had a better chance to win that its +205 odds implies (68th percentile). The 49ers were next (-190, 63rd percentile). Both won.

At the start of Conference Championship week, 
San Francisco (+127) and Kansas City (+108) both fell in the 86th percentile. San Francisco stayed there, and would probably have cracked a regular season's week's top five. But as the health of Patrick Mahomes improved, the Chiefs swung all the way to -125 favorites, negating the modest value that had existed beforehand. San Francisco quarterback Brock Purdy got injured in the first quarter, dooming the 49ers, and Kansas City beat Cincinnati on a last-second field goal to return to the Super Bowl.

TO VIEW ON A MOBILE DEVICE, PLEASE TURN YOUR PHONE SIDEWAYS
Rank
Tier
Percentile
Pick
Odds
  Win Prob.
Opp.
Week
Starting QB
Score
Result
Net
1
1
99.8
TEN
+600
14%
vs DAL
17
Dobbs
27-13
​Lost
-100
2
1
99.7
SEA
+240
29%
vs DEN
1
Smith
17-16
Won
​+240
3
1
99.6
DAL
+275
27%
vs CIN
2
Rush
20-17
Won
​+275
4
1
99.3
TEN
+600
14%
at KC
9
Willis
20-17 (OT)
​Lost
-100
5
2
98.3
HOU
+390
20%
at DEN
2
Mills
19-6
​Lost
-100
6
2
98.3
LAR
+245
29%
vs LVR
14
Mayfield
17-16
​Won
​+245
7
2
98.2
LAR
+265
27%
vs SEA
13
Mayfield
27-23 (OT)
​Lost
-100
8
2
98.0
ARZ
+300
25%
vs TB
16
McSorley
19-16 (OT)
​​Lost
-100
9
2
97.7
HOU
+575
15%
vs ​PHI
9
Mills
29-17
​Lost
-100
10
2
97.6
PIT
+380
21%
vs ​TB
6
Pickett
20-18
​Won
​+380
11
2
97.5
ARZ
+360
22%
vs ​SFO
11
Blough
38-10
​Lost
-100
12
2
97.2
CAR
+530
16%
vs ​TB
7
Walker
21-3
Won
​+530
13
2
96.4
WAS
+240
29%
vs ​PHI
3
Wentz
24-8
​Lost
-100
14
2
96.3
SEA
+310
24%
vs ​SFO
2
Smith
27-7
​Lost
-100
15
2
96.1
ARZ
+215
32%
vs ​PHI
5
Murray
20-17
​Lost
-100
16
2
96.0
HOU
+283
26%
vs ​IND
1
Mills
20-20
​Tied
-
17
2
96.0
IND
+240
29%
vs ​PHI
11
Ryan
17-16
​​Lost
-100
18
2
95.9
HOU
+205
33%
vs ​LAC
4
Mills
34-24
Lost
-100
19
2
95.5
LVR
+370
21%
vs ​SFO
17
Stidham
37-34 (OT)
​Lost
-100
20
2
95.4
CHI
+265
27%
vs ​SFO
1
Fields
19-10
​Won
​+265
21
2
95.4
LAR
+130
​43%
vs ​DEN
16
Mayfield
51-14​
Won
​+130
22
2
95.2
DAL
+245
29%
at ​PHI
6
Rush
26-17
​Lost
-100
23
2
95.2
TEN
+115
47%
vs ​LVR
3
Tannehill
24-22
Won
​+115
24
2
95.0
IND
+190
34%
vs ​LAC
16
Foles
20-3
Lost
-100
25
2
94.6
ATL
+140
42%
vs ​LAC
9
Mariota
20-17
Lost
-100
26
2
94.6
NYJ
+275
27%
vs ​BAL
1
Flacco
24-9
​Lost
-100
27
2
94.5
PIT
+295
25%
at MIA
7
Pickett
16-10
Lost
-100
28
2
94.4
NYJ
+460
18%
vs ​BUF
9
Wilson
20-17
​Won
​+460
29
2
94.3
ARZ
+150
​40%
vs ​LAC
12
Murray
25-24
Lost
-100
Picture

Contact

ELDORADO

© ELDORADO, 2023

​All rights reserved​
  • Home
  • Football
  • Basketball
  • Politics & Culture
  • Contact | Subscribe