I have a guide that I received from DSI with vast details about teams' tendencies in certain situations. An example is that according to the book, St. Louis is 19-5 ATS at home off a division contest.

My question is, how strongly does this play into the decision making process? Is it comparable to Roulette, where past results have no bearing on future events (but you still see the tower with the last 30 results for obvious reason)? Or is it more like poker, where a player's past behaviors are a critical part of the equation?

In my first week picking every game, I went 9-5-2 (using WSEX lines) and I relied quite heavily on the advice in the book. When I knew next to nothing about the teams in question, my decision was almost entirely based on what I'd read. The more 'other' info I had, the less I relied on the book, but I still read it and took it's advice into account. I know my sample size is laughable, but you have to start somewhere.

Finally, I realize that the purpose of the book is just to get me to place a bet, they don't care how well informed I think I am, so long as I give action.

Any insights would be much appreciated.

I don't use these things at all

I know some people believe in them quite strongly

Lester,
I should give a fuller answer as this comes up from time to time on the forum. Basically, is there a good reason for the stat they are giving? If there is a good theory behind it, it might be worth something. Otherwise, it's just statistical noise-data mining. Given enough numbers on anything, you'll find a few weird subsets that work for no apparent reason.

Here's an example of theory...IIRC, in NCAAF bowl games, teams favored by over 7.5 points are only 38% ATS since 1980. The idea is that these teams are usually in bowl games that they are less motivated to play. The other team better prepares and is more dedicated to the game over the 3 week pre-bowl practice span. The favorite is usually quite happy with their season and has nothing left 'to prove.' The underdog is still fighting for respect.

Now, obviously, the theory is just general as not everygame will fit that case (for instance, a big favorite for a national title game should be pretty focused). However, it does make some intuitive sense and is a theory that one could have before looking at the data to confirm it.


I guess that is the test in my eyes. Is this a relation that one could theoretically expect before looking at the data? If so, it might have some merit. If not, it's probably just a result of data mining and searching for patterns---such as Sklansky's book talking about Super Bowl Champs poor ATS record in week 2 the following season.


Load up those bets against the Colts

Thanks for taking the time to write that.

What you say adds up. Some of the tidbits they offer make sense and I feel compelled to act on the information. But, other times the information makes almost no sense, like this gem:

"New Orleans is 18-3 ATS against division foes."

So this data goes back four years. But this doesn't make sense. Last year NO was good, but before that, they sucked supreme ass. With Atlanta and Carolina in their division, why would they be so great against the spread?

Here's another lovely one:

"Det is 12-4 ATS at home after scoring 28 or more points."

How far back did they have to go to gather that data? 1962? Detroit only scores that many points in a game once a season.

Detroit is home this week and at -3 that has to be a lock, or at least a 3-1 lock.

Well, I was planning on picking Detroit this weekend anyhow. Now I have someone other than myself to blame if I am wrong.

Thanks for the replies.

Later

Actually this one COULD make sense if for some reason NO is more focused(or matches up better)against division foes as opposed to the rest of the league. Or if the oddsmakers perception of their strength against these foes is incorrect.

You are correct here,this one is totally useless.

Stats that relate to same coaches/schemes and motivation would have some applicability, even if very small.

If Martz' home games on turf tend to go Over more than normal over a stat valid sample, that is understandable if other teams D has trouble stopping his O, due to Martz' "genius," to make up an example, or a coach like, say, Vermeil tends to go for it on 4th in his own side of the field, that will lead to more scoring than you would normally expect. Play-it-safe coaches affect outcomes similarly, 'run-it-up' college coaches, etc.
Vegas can be quite slow to notice these changes, mt2r can comment on a CBB team that switched to the Loyola Marymount offense last year and it played utter havoc with the totals, Vegas had NO clue where to open the line, like 0.1% clue.

Motivation can be a very big factor and I think mt2r nailed it on the head similarly. Not all teams are equally motivated every game. Yeah, USC and Texas were both up for the BCS Title, but don't you think Texas took the game a tad more seriously given they were +260 underdogs at one point and USC was supposedly 'the bestest football team ever zomg?' I bet they did. [and made a lot of money doing so.]
JMHO.

The data mining stuff is for squares. That should be obvious. Pulling data on Martz totals turf v grass may lead to an interesting result, that's a difference that would make sense -- even if the result shows no difference, you move on. Etc.