Some help please....

[Mappo]

I have a theory about teams covering that I wanted to test out. Does anyone know of a good site that has quick and easy access to this year's spreads and who covered from week to week?

[JohnnyMapleLeaf]

I like covers.com....I think thats alright to post, as it doesn't have a forum...

[Mappo]

Thanks a lot, that is exactly what I am looking for. I'll let everyone know if I find out anything interesting.

[yomonte]

Whats your thoery map, I may have the answer.

[Mappo]

It's nothing particularly Earth shattering.

My thought was that many of you have commented after a team has a particularly good game that the next week's line sometimes has an "overreaction" (that is, gives that team a more favorable line than they otherwise should deserve). The same could be said, I suppose, about a team that has a particularly bad game the week before.

So, my thought was that if a team that covered the week before is playing a team that didn't cover the week before, that then might be a good time to play the team that didn't cover. Obviously, tons of other factors come into play, but, over the long run, if this could be statistically shown to be a true hypothesis, it might prove a useful data point as supporting evidence if you wanted to pick for or against a particular team.

I actually have a bit of a statistical background and own a decent statistics program...so I figured I might be able to put that to good use. I've been toying with the idea for next year to test prospectively some of the things that we often tell ourselves (e.g. betting against a team that is in a classic "let down" situation, watching for "suspicious" line movement, etc.) to separate the truisms from the falsehoods. If I ended up doing it, I'd have to set up the objective criteria a priori for each of the things I would want to test and maybe I could get the board's help with that. I imagine perhaps many of these things have already been looked at objectively, so maybe I shouldn't waste my time...but I figured it might be fun and useful if not.

[yomonte]

Kinda late in the game for that. Next year, I recommend you keep stats, unless you have nothing better to do for the next few weeks, when the season will end, to go back thru all the games and figure out what you are looking for. I have done what you are looking for in the past and just like everything else, it all evens out. eg, teams that win by 20 or more at the beginning of the season, go under the next game. But just as you see a trend, it goes the other way. Some call it trends, I call it a kaiser blade. I used to keep track of alot of ****, but like every year, it will even out. Catch it mid way and fade. Home dogs of +3 were gold, but its evened out mid way. A learned colleage of mine once said, for every action, there is an opposite and equal reaction. And for those of you that don't who that is, its Alan Einstein.

[yomonte]

Map, in all seiousness, Ithink along the same lines as you. You keep track of that from the beginning, and fade it when you see a pattern going the other way. I've been doing it for years and still can't win. But remember, don't keep playing the same trend to the end.

[homedawg]

I concur Yalbert,
When the past, shows uneven results, many interpret this as a trend, but in reality, there is always a 50/50 chance at the outcome, therefore, statistically, the advantage favors the opposite of previous uneven results.