Although the Super Bowl was played just three months ago, the draft hasn’t taken place yet and training camps aren’t open, it’s still not too early to start thinking about the NFL if you play at Pinnacle Sports!
The season kicks off in less than 5 months from now when the Super Bowl Champion Pittsburgh Steelers host the Miami Dolphins in a Thursday night match-up. Pinnacle Sports has already posted lines on every game in week 1 courtesy of our in-house team of odds makers who’ve handicapped the games and created the lines. But as a player, how does one handicap the first week’s games and find an edge this far out?
As a starting point, look at the relationship between season wins and the first week point spreads. An average team would expect to win 50% of the time or 8 games in a 16-game regular season. In essence, the expected season wins for a team is just a power ranking.
How many games would you expect a team to win, if it was a 3-point favorite for each game? First you need to convert the fair no-vig moneyline (ML) into a "percentage chance of winning" for each game. For favorites, that is the (ML quote / (ML - 100)) * 100. If the fair no-vig moneyline for a 3-point favorite is -145/+145, we would expect the 3-point favorite to win (-145 /(-145-100)) * 100 = 59% of the time. If we knew a team would be a 3-point favorite for every game, we would expect it to win 16*0.59 games, or about 9.5 games.
Although not a perfect science, you can use this to convert season win lines into a game line for the first week. For every ½ game better the favorite is for season wins, it should give up an additional 1 point on the spread at a neutral site. If a 9.5-win team played an 8-win team, the 9.5-win team would be a 3-point favorite on a neutral field. After that, add 3 points for home field advantage, so the 9.5 win team would be a 6-point favorite at home, or a Pick’em on the road.
Then set a “baseline” using games from the prior year, in this case the 2005-2006 NFL season. While some people will simply start with the number of games a team won in the previous season, more sophisticated bettors use the “Pythagorean Theorem” for football. This formula reduces the affects of lucky and/or close wins, and gives a team more credit for blowouts and consistently solid performances.
For example, consider the 2005 Tampa Bay Buccaneers regular season record of 11-5, with 300 points scored for and 274 points scored against. Instead of simply using their win/loss record, using the Pythagorean Theorem for football it assumes games won = (PF^2) / (PF^2+PA^2) * 16, where PF=points for and PA=points against.
Using the Pythagorean Theorem for football, the Buccaneers’ baseline would be calculated as 300*300/(300*300+274*274) * 16 which gives an expectation of 8.7 games. This suggests that Tampa Bay was very lucky to win 11 games and if they played the same season with the same roster, 9 wins would be much more likely.
Conversely using the same formula, we can see that last year Green Bay’s record undervalued the team. The Packers finished at 4-12, with 298 points for and 344 points scored against. Their baseline would be 298*298/(298*298+344*344) * 16 = 6.9 games, nearly 3 full games better than their record from last year.
The Pythagorean Theorem is a starting point in your analysis that gives you a leg up over handicappers who don’t use it. Although originally derived by Bill James for MLB, its applications have extended across many sports by changing the exponent (2 for NFL, 1.8 for MLB, and 16.5 for the NBA).
Another adjustment you can make to the 2005 season wins baseline is the “reversion to the mean”. Basically this means that no matter what a team does in a previous season, it tends to move toward winning 50% of its games the following season. Bad teams aren’t quite as bad as people remember them and the dynasties eventually fade. A general rule of thumb is to move the baseline season wins about ½ a game toward 8 for baselines between 5.5-10.5, or a full game towards 8 for very good/bad teams outside that range.
Once the baseline is calculated, you need to consider roster changes. Is a team peaking or rebuilding? If a team has several older players retiring and being replaced with younger, inexperienced players, this suggests the team could be in a rebuilding stage. Younger players tend to contribute less in the first few years and in a majority of cases, the affect of the draft on a team can be ignored and you can instead focus on trades/free agents acquired. If many starters are inexperienced at the top level, that team can be expected to fare worse the next year, but gradually improve afterwards.
If a team’s roster is fairly stable, you generally expect the team to do as well or better the following year. On teams with a low turnover rate, the focus of the off-season is to add talented veterans to positions lacking experience and hope for an immediate impact on the team.
Adding depth (e.g. a journeyman backup QB, or a fourth cornerback) will have less of an impact, but also lowers the downside variance. A team that should have put more effort into its backup quarterbacks is the 2005 New York Jets. The Jets fared well in 2004 under Chad Pennington with a reasonable QB passer rating of 91. In 2005, the 1-1 Jets lost Pennington and backup Jay Fiedler for the season in the third game. They won only 3 of the next 14 games with Brooks Bollinger and Vinny Testaverde struggling at QB, with passer ratings of 59 and 73, respectively.
Analyzing roster changes is a very subjective matter. For each one evaluated (some handicappers ignore all changes involving third-string players or deeper), try to consider how that will affect the team’s play. If a team has a poor defense and an average offense, defensive changes will have a bigger impact – the defense simply has more room for improvement.
Once season win expectations are completed, then set the line for each game. As in the earlier example, take the difference between the two teams in season wins, multiply by 2, and add 3 for the home field advantage. If your numbers suggest a play, we are open for business at Pinnacle Sports on NFL Week 1. With our 10-cent line on NFL openers, you’ll even get up to 50% better value compared to other sportsbooks when they finally get around to posting their NFL openers...
How have the sharps bet the early week 1 NFL openers?
Miami Dolphins +5.5 at Pittsburgh Steelers
We opened the game at Miami +6.5, and took multiple limit bets from sharps on the dog. If you faded the Super Bowl Champion for the first 2 weeks from 1985 to 2005, you would be 28-13. So any opening number will draw sharp versus public betting.
Cincinnati Bengals +3.5 at Kansas City Chiefs
Our opener of Cincinnati +2.5 saw moderate lopsided betting on the Chiefs, driving the number onto and past the “3”. Using the Pythagorean Theorem, Cincinnati appears to have over-performed in 2005, where 9.5 wins would be more reflective of the team that actually went 11-5. The early betting tends to agree that the number on the Bengals was too high.