...............

a bettor might have an edge betting dogs at the very beginning of the season
most of the teams have changed - some a great deal
there is less predictability
that unpredictability tends to favor the dog
even teams that haven't changed much - the 29 year olds are now 30
such players may have lost half a step

How about you download some data, backtest this, then let us know. Thx!

sorry, I'm not real geekay - I would have to do it most by hand and it would just be too much for me

I am working on something re road dogs which I will post a few months later if it's useful - I am doing this by hand mostly but its not to tough going day by day

but thanks anyway for the suggestion

This thread has determined that was a lie.....

This isn't a theory. It's not even really a hypothesis. Do you just want someone to agree so that you can bet on it?

Like it's been said in here before, do a bit of work and backtest it.

FYI, if there's unpredictability, it can't favour one side or the other.

Bro, get Google Sheets, go on Coursera or about a million other free resources that you can learn the skills to actually find an edge.

By hand is pointless and will take forever to even datamine a month of games, which is a nothing sample.

....................

I didn't post this earlier because I didn't think there was much interest



I had an idea that the dog might be a good bet very early in the NBA season because the abilities of the teams are not well known. many teams add or subtract players and even those who don't do much of that - the 28 year olds are now 29.

it seemed to me that there would be significantly greater unpredictability and that unpredictability would favor the dogs

using covers.com I tracked the against the spread results of every single game of the first 21 days of the 2020/2021 NBA season



results:



NBA dogs December 22 2020 thru January 11 2021




92-61..............60% winners..............R.O.I.....................14. 8%










anybody who wants to double check my stats can do this at covers.com
covers.com is the only site I know of that provides this type of historical data




when I feel like I have more energy I will work on prior years and report back


Note: post #5 - VeniceMerchant wrote: "if there's unpredictability it can't favor one side or the other"



that is not correct.............because................if there is total unpredicatability about the result...................the dogs are getting points


and yes, I'm not a naive person - so I recognize that the sample size is not large enough to come to any conclusions


I will continue on looking at this in various sports in my own non - geekay plodding way - as inefficient as it may be




Note - I believe the historical line quoted by covers.com is the last available line before gametime and the most common line as they survey multiple books

you have nothing to worry about buddy
I would never ever want somebody on this site to agree with me or do anything else for me


in my OP I asked if they would agree or disagree
but that in no way means that I wanted them to agree

Your theory makes sense, which is a good start.

Most systems are based on past results which look ok, but what happened in the past doesnt mean it will always happen again in the future!

In the main odds layers are better informed than punters and consequently the odds offered are seldom far off the mark.

......................

the image below is the results of statistician Mike Shackleford's tracking of the point spread results of over 2,000 NFL games. He runs the Wizard of Odds website and is a very sharp guy


as you can see - the away dog won at a 53.75% clip with a 2.57% R.O.I.






https://wizardofodds.com/games/sports-betting/nfl/

Your sample of Dec 22 to Jan 11 is way too small to even take seriously. Before I created a successful NBA model, I had a bunch fail. The first and worst one had a winning November in NBA ML. It then proceeded to get smoked by the books most months over a 5 year sample.

Fortunately I have scripts that run this, or doing it manually would suck.

FYI, sportsbookreview.com has historical opening and closing lines for most of the major books online.

That's where I'd start.




As far as a theory "making sense" in my experience, most people's intuition is just confirmation bias.

If this was all there was to making money at sportsbooks, my guess is that books would figure out a way to adjust long before you find out.

incorrect

you're giving books way too much credit
you're making them look like geniuses
they're not geniuses

they won't adjust just because there may be a profitable angle

they only adjust if there is too much money on that angle causing a reduction in their R.O.I.

So you’ve been able to successfully beat the books long term using this strategy?

If not, you have no basis to tell me I’m wrong.

A link on wizard of odds said it’s so, so it must be true.
-they don’t even post the games & lines and you trust it as gospel? Jfc

Why are you giving away your secrets for free? then arguing with people who disagree with you lol

OP, your theory does make sense as to something to test.

new season = new info... market may be slow to adjust.... a great spot is a new coach. i remember when an NBA team when from a really low scoring coach to a high scoring coach. took the market a few games to catch on. or a new coach whose players don't fit the coach's plan.

someone on another site had an NBA superstar theory: bet a team first game without superstar, fade a team first game with superstar back... i tend to agree. it makes sense behaviourly to me.

anyway, as to your theory, what leagues? how many games into new season? how big an underdog?........ base it on big favorite/underdog or terrible/outstanding teams from last season (will yield similar picks for sure)

i ran some basic early season numbers for the 4 leagues... generally do 5 seasons and i think the database doesn't do 2020 (abbreviated season) for these. not sure if it's my coding or the defaults.

anyway, first month of new season... backing outstanding or terrible teams (straight-up) from last season:

2 lines looked interesting,

1) fading NHL losing teams from previous season.. i think it was fading. i'll check

2) backing outstanding NFL teams from previous season... outstanding regular season, nothing to do with how they did in playoffs previous year.

you might be able to poke around and data mine some reasonable stuff for the other 6 lines......... my two trends are moderately data-mined. not bad.