Critique my Method (NFL)

Nov 27:

SEA +12
TN -11
PHI -3

For those of you who follow the O/U, I show a high probability of:

SEA u 46.5
TN u 44

AJ, I only watch the games I have an interest in, e.g., Titans, Colts, Cardinals.  I've got a few sons with favorites, viz., Steelers, Phillies.  I watch the Braves and Diamondbacks.  I played football and baseball growing up.  I don't watch basketball at all, and I only watch NHL if I can find the Predators on.  You're more likely to find me reading a book on economics or theology than SI. 

So, I'll let you decide whether I qualify as a sports fan. 

Best of luck sdiinc. I agree with you on playing many games if you have the time to do the full analysis to do so. The more games you find +ev expectations with the better. Obviously your system depends on essentially creating a line with your stats, comparing to the betting line, and waging on nearlr everything that your stats show as a winning situation.

So far this season you are claiming that your system is showing what the score difference should be, but that previous performance hasn't been a good indicator of future performance. I think some of this is due to you not fully following injuries and personnel issues. With this many games you should try to include those things into more of your decisions and thus probably reduce your losing bets by 1-2.

And I think that the NFL is the toughest sport to have a system on due to the lack of games. I'll be looking at your NHL & NBA system, and MLB system when the next season starts. Cheers.

Casey, there's no doubt I could cut losses by 1 or 2 games a week if I had insight into true NFL handicapping.  It's not only injuries, but field conditions, weather, and other factors.  Even if we could just 'pass' on doubtful games, that would be desirable.  I remember a few weeks back I posted Dallas minus a big spread and didn't realize until Sunday that there was a major QB issue.  Duh!

The problem is two-fold, first, I don't have the insight to measure what truly matters and what the impact is, and secondly, I'm about maxed out right now running analysis on 3 sports.  I would welcome any input on the matter.

I have adjusted my formula to place an increasing emphasis on more recent games.  This is a result of my believing that the relevance of a game 10 weeks ago is less than that of a game 2 or 3 weeks ago.  The problem, of course, is how to weight this.  I've looked at some of the big-time firms and they're not transparent enough to discern their methodology.

You can't simply weight the most recent game higher than the immediately preceding game for reasons that should be obvious.  So I've decided to implement a rolling 3 week pool, where the most recent 3 games are weighted the highest, and then the 3 games preceding those by a factor of 90%, and so forth, to improve the relevance of recent performance.  The objective being an increase in accuracy.  I'll run both formulas and see if there is a difference and if that translates into an improvement in accuracy.

I should have the numbers later today.

Is there a data source on the net that publishes a total value of the accumulated injuries of each team in the NFL;  their effect on the outcome calculated by their individual contributions ( value ) added together weighted by the importance of their position? Perhaps incorporating weather, venues, and field conditons?

Such a factor could easily be added to your formula without you having to handicap each team and go through the  time-consumming data gathering prcess. If not, would someone consider volunteering  to perform this task?

Your efforts plus someone else's effort could lead to the 1 or 2 games positive outcome per week or month that Casey sees is possible.

If not, then is there a reliable source that gathers the needed information so that an analysis could be made? Would you consider such a factor sdiinc?

 

I always read one or two 'previews' for each game I'm going to bet on (unless I've gotten the information I need through radio or whatnot). That at least lets you know if there are any major issues. I haven't yet incorporated weather into my picks based on spread. Sure, if its wet you'd expect the team that can run the ball better to do slightly better than under normal conditions, but I just haven't incorporated it for a few reasons.

Anyhow, any person with a system or any handicapper (they can be one and the same really, depends on who you ask), should always check when their numbers vastly differ from what is coming out of Vegas. See what the reasons might be. Why is the public leaning a certain way? Doesn't mean you should skip the pick, but you should try to figure out why there's a difference in thought.

Lastly, I haven't heard of any professional sports investor that has made a profit year after year purely on statistical analysis.

As for how to weigh the games, I weigh them based on who they played and how soon. I don't use a weighing number like you so I can't help you there. But what I do do, is I make sure to see if any of the games can be considered outliers. For example, I did not bet on Dallas when Romo came back. But if I had, I would have had to mostly ignore their previous several offensive performances, focused mostly on certain defense statistics, and look at early season offensive statistics. Other outliers might be major personnel changes, key player coming back from injuries, some bad coaching or official calls, Donovan McNabb being pulled early in a still winnable game. Also there's future outliers, such as when a team's season is essentially over.

Week 13:

CIN +7
CHI +3.5
DEN +9
NE -1
OAK -3
SD -4.5
JAC +3

Dabar, I like your thinking.  I'm embarrassed to admit I don't have any injury consideration in my formula, but I just don't have a easy way to manage it.  Of course, if a player has been absent for one or more games, their absence would be reflected in the team's performance during those events, and so in that way it is considered.  My new weighting method may also help if a key player played early in the season but is now missing.

I'll put some feelers out, sometimes my 'friends' in the industry are a little more helpful when I ask a very precise, narrow question such as this.

Casey, reading those profiles might be good for you, but for me I've found information overload and conflicting insight by analysts.  If there was a real game-changer, that would be of value, but I always assume that is reflected in the line.

One thing I have noticed is that my picks do better the larger the deviation from Vegas.  Put another way, if I show SD as a 9 point favorite this weekend, and the line is only 4.5, those picks tend to outperform say, NE, who I only show as winning by 1.6.

All that being said, I think I need someone who is a real sports nut to look at the picks and say, "I'm really concerned about X due to _______."

Cincy +7 -- I don't see it happening. Baltimore is putting up a lot of points and keeping teams to low scoring. I don't think that Cincy can put up too many against Baltimore, but the Ravens will put up 30+. I love the Bengal's head coach but the team is in such turmoil (mostly due to factors not in his control) that I think they'll get smacked.

Chicago +3.5 - If you can get 3.5 probably not a bad pick. Its a fairly even game, and with 3.5 I'd say Chicago wins over 55% of the games on same circumstances (so there's value). Seems solid. I'm a huge Bears fan though so not betting.

Denver +9 - Which Denver team? That's the problem that I see. I don't know who will show up. The lines for this game range from 7.5 - 9. I handicapped the Jets as +6, If I can get 9 I might take Denver.

NE -1 - No bet from me. Tough game to predict. I don't find any value in it.

Oakland -3 - I have no opinion (read: no line) on this game. Handicapping it was tough. I can't figure out which teams will show up. Will KC come with a solid enough defense, or will they let the opposition roll over them? Will Oakland actually play tough, or is the situation with that team still so chaotic that the team might just lay down early if their defense falters early to a good Thigpen led offense?

SD -4.5 - I don't get this line. I am a Falcons fan, but even ignoring that, I handicapped them as 7 point favorites. I need to relook at things, but I think the public is giving SD at home too much credit. Matt Ryan could light up the struggling SD defense, similar to what he did to Carolina. Atlanta's defense has flat out been playing better. Atlanta hungers for a playoff spot and has a better chance at getting one.

[Quote: Originally Posted by CaseyJonesWW]

Cincy +7 -- I don't see it happening. Baltimore is putting up a lot of points and keeping teams to low scoring. I don't think that Cincy can put up too many against Baltimore, but the Ravens will put up 30+. I love the Bengal's head coach but the team is in such turmoil (mostly due to factors not in his control) that I think they'll get smacked.

Chicago +3.5 - If you can get 3.5 probably not a bad pick. Its a fairly even game, and with 3.5 I'd say Chicago wins over 55% of the games on same circumstances (so there's value). Seems solid. I'm a huge Bears fan though so not betting.

Denver +9 - Which Denver team? That's the problem that I see. I don't know who will show up. The lines for this game range from 7.5 - 9. I handicapped the Jets as +6, If I can get 9 I might take Denver.

NE -1 - No bet from me. Tough game to predict. I don't find any value in it.

Oakland -3 - I have no opinion (read: no line) on this game. Handicapping it was tough. I can't figure out which teams will show up. Will KC come with a solid enough defense, or will they let the opposition roll over them? Will Oakland actually play tough, or is the situation with that team still so chaotic that the team might just lay down early if their defense falters early to a good Thigpen led offense?

SD -4.5 - I don't get this line. I am a Falcons fan, but even ignoring that, I handicapped them as 7 point favorites. I need to relook at things, but I think the public is giving SD at home too much credit. Matt Ryan could light up the struggling SD defense, similar to what he did to Carolina. Atlanta's defense has flat out been playing better. Atlanta hungers for a playoff spot and has a better chance at getting one.




Well said..... he picked every game that we had power ranked within a few points of the line.

I get the lines after I place my wager, either BetOnline or Sportsbook.  I checked the lines this morning and they're unchanged.

Sdiinc, take a look at matchbook if you haven't already.  Best US book

...ouch...1-5 this week with Jaccksonville left to play

 

what is you w-l record this season?

Actually, 3-7 this week including Thanksgiving.  And, let me tell you it takes alot of work to perform at that level.

Week 1 through 13, I'm at 50-64.  It's hard to explain why a formula can work so well one season and not the next, but I'll continue to work on it for the rest of the year.  Meanwhile, I think we've hit our stride with NBA and NHL.  Check those threads out and give me some feedback.

I was thinking about the statistics of betting football and it would seem to me that this is the hardest sport to handicap.  I am not a statistician, but I did stay a a Holiday Inn a few weeks ago...

[Quote: Originally Posted by Sumtingwong] I was thinking about the statistics of betting football and it would seem to me that this is the hardest sport to handicap.  I am not a statistician, but I did stay a a Holiday Inn a few weeks ago...

Dec 4:

SD -8.5

Dabar, sdiinc,