Individual player props are available for betting at many sportsbooks for the Super Bowl and other playoff games.  Such props include Over/Under bets on how many touchdown passes a quarterback will throw, how many tackles a linebacker will make, how many yards a running back will gain on the ground and many others.  Props are also available on whether a certain player will score a touchdown or not, whether a certain player will intercept a pass and others.  Past individual statistics is useful to evaluate these props, but they can also mislead the user.  It is important for handicappers to put past individual statistics in the context of the market lines for the game in question.   Handicappers also should adjust for outliers and any recent personnel changes that may make statistics accumulated over the course of the season no longer relevant.  In this article, I will give an overview of some factors to consider when evaluating individual player props.    

Market Line versus Past Data

As of the writing of this article, the line on Super Bowl XLII is New England Patriots -12 against the New York Giants and a total of 53.  These lines imply New England is expected to score about 32.5 points and New York is expected to score about 20.5 points.    

In the 2007 NFL regular season, New England scored an average of 36.8 points and gave up an average of 17.1 points.  New York scored an average of 23.3 point and gave up an average of 21.9 points.  By using past individual statistics only, we could overestimate or underestimate based on the context of the game.  If we used unadjusted individual statistics from the regular season, we would overestimate the number of points scored by the Patriots by 4.3 points (36.8 minus 32.5).  Also, we would tend to overestimate the offensive statistics put up by the individual Patriot players compared to market expectations.

In order to conform to the market expectations, an adjustment needs to be made to season-long individual statistics.  If an adjustment is not made, it can lead to relative-value mistakes.  For example, is it better to bet that Brady will throw more than three touchdowns or is it better to bet over 32.5 points for the Patriots?  Brady averaged 3.13 touchdown passes in the regular season.  But when he did so, the Patriots scored an average of 36.8 points.  In the Super Bowl, they are expected to score 4.3 fewer points than they averaged in the regular season.  In order to score fewer points, it means the expected number of touchdowns scored by the Patriots is lower than their average during the regular season.  With a lower number of expected touchdowns, one would expect fewer passing touchdowns as well.  Thus taking Over 3 on the number of touchdowns thrown by Brady looks like a poor bet relative to taking Over 32.5 points for the Patriots.  It does not mean that taking Over 3 at even money passing touchdowns is necessarily a bad bet.  Instead, there is a better alternative bet that has similar characteristics at better value.

Outliers

There are only 16 games in a NFL regular season.  Due to this low number, even statistics over an entire season may be an outlier and not expected to be repeated in the next season.  Sometimes that is just due to randomness, sometimes it is due to the other team making adjustments based on the skill level displayed by some players.  For example, cornerbacks that are adept at intercepting balls thrown in their direction may see their opponents throw in their direction less often.  Due to fewer balls thrown in their direction, these high-quality cornerbacks will have fewer chances for interceptions.  Great cornerbacks may not have high interception numbers for this reason.

Here are all the players in the NFL that had 8 or more interceptions in a season since 2003.  The second number is the number of interceptions they had the following year (among those that played most of the 16-game season the following year).

2003 Tony Parrish, SF: 9 / 2004: 4
2003 Brian Russell, MIN: 9 / 2004: 1
2003 Corey Chavous, MIN: 8 / 2004: 1
2004: Tory James, CIN: 8 / 2005: 5
2005: Ty Law, NYJ: 10 / 2006: 4
2005: Nathan Vasher, CHI: 8 / 2006: 3 in 14 games
2005: Champ Bailey, DEN: 8 / 2006: 10
2006: Champ Baliey, DEN: 10 /  2007: 3 in 15 games
2006: Asante Samuel, NE: 10 / 2007: 6
2006: Rashean Mathis, JAC: 8 / 2007: 1 in 14 games
2006: Walt Harris, SF: 8 / 2007: 4 in 15 games
2006: Charles Woodson, GB: 8 / 2007: 4 in 14 games
2006: Nnamdi Asomugha, OAK: 8 / 2007: 1 in 15 games

The average player on this list decreased from an average of 8.6 interceptions to 3.6 interceptions the following year.  Only Champ Bailey increased his number.  Next to Bailey, the second most interceptions the following year were the six by Asante Samuel in 2007. 

This shows how random interceptions by defensive backs can be.  This information tells me not to rely heavily on any particular defender’s total interceptions for any year, especially if it is a high number.  A high number of interceptions by any player is likely to be an outlier.  They are not likely to be matched the following year; and they are not likely to be matched in one game like the Super Bowl. 

Unfortunately, this information on interceptions is not directly useful for Super Bowl XLII between New England and New York.  No Giant player had more than four interceptions, and other than Samuel with six, no Patriots player had more than three.  So I think it will be unlikely for a linemaker to make a mistake in this respect.  Save this information for next year’s playoffs!

Given that Tom Brady broke the record for most touchdown passes ever in a season, one has to wonder if his numbers in 2007 was an outlier.  Having an outlier season does not mean Brady is not a great quarterback, of course he is.  It does mean that Brady is unlikely to match or top his own record next season.  He averaged 3.13 touchdown passes per game in the 2007 regular season.  In my opinion, that would be too high even given the context of an expected high scoring Super Bowl. 

Recent Personnel Changes

Teams change over the course of a season.  In the first half of the season, Laurence Maroney, the running back for the Patriots, was not as productive as he was in the second half of the season.  Part of the reason may be that he was injured; he did miss three games.  Whatever the reason, his productivity has increased.  In Maroney’s first ten games that he played in, spanning from week 1 to week 14 (he did not play in three games due to injury), Maroney averaged only 12.6 carries per game and 52.9 rushing yards gained per game.  That is an average of 4.2 yards per carry.  He only scored two touchdowns and only once rushed for more than 100 yards in those ten games.

In the last five games (the last three games in the regular season and two in the playoffs), Maroney has rushed for more than 100 yards in four games and averaged 21.2 carries and 110 yards per game.  That is an average of 5.2 yards per carry.  He has also scored six touchdowns in those five games with at least one in each game.

Whatever the reason may be, clearly Maroney is more productive now than he was earlier in the year.  Will this continue in the Super Bowl?  The last five games were in the cold northeast winter.  With the Super Bowl in expected nice weather in Arizona, maybe the Patriots will run less and pass more as they did earlier in the year.  Or possibly they decided to use Maroney less often early in the year because they knew he had some injuries that they did not want to aggravate.  Estimating how much the Patriots use Maroney is just as important as estimating how well he performs when he does get the ball.  This is where the art of sports betting comes into play.

With the Patriots running the ball more in the playoffs, and with Maroney playing at a high level, maybe this means the Patriots will continue with the new balanced offensive attack.  In the early part of the season, the Patriots threw the ball much more often than they ran the ball.  If this game-plan adjustment continues in the Super Bowl, then we are likely to see Tom Brady pass the ball less than he did during the regular season.  This is now the third reason why taking Over 3 passing touchdowns by Tom Brady seems like a poor bet.  The first reason is the context of the market (Patriots expected to score 32.5 points), the second reason is the outlier factor and the third reason is the personnel or game-plan change by the Patriots in the latter part of the season.

Conclusion

There are many factors involved in estimating individual player proposition bets.  In this article, I have outlined some issues to think about, including putting player statistics in context of the market line and the total team expectations, adjusting for outliers and taking into account recent personnel changes.  These are just some of the issues to think about.  A talented evaluator can combine the art and the science of sports betting and possibly come up with a better estimate then the linemakers and profit from it.