Let's start with a delicious look back at Week 3, when upsets ran rampant in the NFL. Buffalo won as a 10-to-1 underdog in Minnesota. Tennessee won as a 3.75-to-1 underdog in Jacksonville. Detroit won as a 2.5-to-1 underdog versus New England. The N.Y. Giants won as a 2.15-to-1 underdog at Houston. And if you read this column and followed last week's top money-line value picks, you would have won them all.
A $100 wager on each of those four teams would have won $1,840. And a $100 four-team money-line parlay on those four teams would have won an absurd $57,506. That almost certainly won't happen again. Nevertheless, through the NFL season's first three weeks, the weekly top five money-line value bets — all modest to major underdogs — are a sterling 8-7. A hypothetical $100 on each would have you up $2,065.
Following the value has also fared well against-the-spread, albeit without the dollar pop of heavy-underdog money-line odds. The weekly top five against-the-spread value plays are 9-5-1, or +$350.
A few fun facts for you as we head into Week 4. Since 2002 — the start of the NFL's 32-team, eight-division era — 3-0 teams make the playoffs 68% of the time, 2-1 teams (52%), 1-2 teams (24%), and 0-3 teams (0%). The 2007 Giants are the only team during that period to start 1-2 and make the Super Bowl, and they won.
Probabilities to cover and win are sourced from the 50+ models aggregated on The Prediction Tracker. Against-the-spread expected values assume bets "to win $100." Money-line expected values assume bets "that risk $100."
I. Against the Spread — Top 5 Value in Week 4
Pick, Spread, Opponent, Odds, Probability to Cover (50+ Models)
1. BUF, +9.5, at GBP, -110, 63.9%
2. IND, -1.5, vs. HOU, -110, 61.8%
3. MIA, +6.5, at NEP, -110, 61.1%
4. SFO, +10.0, at LAC, -110, 60.3%
5. CLE, +2.5, at OAK, even, 55.0%
TOP 5 VALUE - Last Week: 3-2 (+80), Season: 9-5-1 (+350)
65%+ COVERS - Last Week: 0-0, Season: 3-0 (+300)
60%+ COVERS - Last Week: 0-2 (-220), Season: 4-3-1 (+60)
II. Money Line — Top 5 Value in Week 4
Pick, Odds, Opponent, Prob. to Win (Odds), Prob. to Win (50+ Models)
1. SFO, +400, at LAC, 20%, 33%
2. BUF, +325, at GBP, 24%, 38%
3. MIA, +250, at NEP, 29%, 45%
4. NYJ, +300, at JAX, 25%, 33%
5. CIN, +175, at ATL, 36%, 43%
TOP 5 VALUE - Last Week: 4-1 (+1740), Season: 8-7 (+2065)
All selections have been underdogs
III. Searching for Value in NFL Futures Bets
I skipped futures values last week but am back with them ahead of Week 4. The biggest futures news probably rests with the San Francisco 49ers, who opened the season at 20-to-1 to win the NFC, but have since fallen to 500-to-1 after starting 1-2 and losing Jimmy Garoppolo. The odds now imply a 0.2% to win the conference. If you still think they have a better chance than that, you might find value at 500-to-1.
More serious value considerations potentially include Miami and the Jets, Baltimore and Cincinnati, and Dallas and Washington. On average, ESPN FPI, Football Outsiders, and Massey-Peabody Analytics give those teams higher probabilities to win the AFC/NFC or Super Bowl than the betting odds imply. Again, I'm not suggesting the Jets are going to win the Super Bowl — only that there might be value at 200 to 1.
Feel like a stretch? That's for you to decide. As always, remember that the expected values shown below — and all others above — are only as good as your faith in the probabilities set forth by the models.
Westgate odds as of Sep 27. Probabilities above reflect the average ESPN FPI, Football Outsiders, and Massey-Peabody probabilities. ESPN FPI does not publish probabilities to win the conference, so they are excluded for that category.
Data was compiled and analyzed by ELDORADO. All charts and graphics herein were created by ELDORADO.
The original version of this article framed expected value in terms of what a hypothetical $100 wager "would turn into." For example, if you were expected to lose $10, I showed the resulting expected value as $90, or $10 less than $100. I've since updated this to show the "expected return" on that $100 wager. So in that same example, the expected value now shows as negative $10. This is much more intuitive — positive expected values are good, and negative expected values are bad.