[Roscoe_Word]

Lets see if the books adjust for the "Home Team on Thur Nite" trend this year. I'm thinking it continues and goes 9-7 for a 56% clip.

[CanuckG]

divisional under trend as well people will hop on

[mrpickem]

Over the last 25 years, NFL teams that won at least 4 games more than their opponent the previous season have only covered the spread 41% in week 1

[MadSl1m]

this dumb square shit never works

[nash13]

It works if you can detect early patterns.
More so what the oddsmaker are doing.
This year it worked fine with betting favs in baseball und certain circumstances.
They've won over 60% of the time.

[nash13]

In other words a non-streaking dog made you over 100 Units if you catch the odds early on value bets.
1700 games returned 100 Units profit.

[nash13]

or something else to chew on:
taking the biggest fav of he day in the MLB has gone
612-220 giving you nearly 110 units in profit. this works even better when you take home favs on the run line and af on the ML.
139-42 this year so far with 30 units profit.

[Gaze73]

or something else to chew on:
taking the biggest fav of he day in the MLB has gone
612-220 giving you nearly 110 units in profit. this works even better when you take home favs on the run line and af on the ML.
139-42 this year so far with 30 units profit.

Very interesting stats, you'd expect that the biggest fav of the day would be a big money sink.

[nash13]

It depends on when you take the bet, because the games don't start at the same time some teams are the biggest fav of the day, when they weren't at the time they played, because the odd of the former lowest odd increased.

[nash13]

And on another note: if your trend holds some kind of logic and is not made up or artificial circumstances, it can hold up for a long time.

[Nitrogen]

If a trend is based on how teams perform in measurable situations. Ie: how does a team/coach play after a MNF game, on limited rest, after a blowout loss, after X number of road games in a row.

Those (to me) have proven valuable.

It's the ones where there's "squishy" data (this team is 8-1 when coming off a loss where the line was -2.5 to +2.5) type that give the Marc Lawrence types something to write about, but have no bearing in betting reality.

[BeatTheJerk]

There’s 29 different 3-0 trends in the Nccaf & NFL on the opening line currently for ats & totals. I like to pick out my top 10 favs & bet them individually & parlay them.

[Bsims]

On one hand it shouldn't work because that would be too easy. I assume that trends are simply calculated into the line. On the other hand, some trends look really strong and there are no mixed signals, so how could you lose money following them? Let's use NBA for example. Bucks are 17-3 ATS after coming off a loss. Would you have bet against Bucks yesterday knowing that? They won by 10. Maybe the coach gives them a good pep talk after a loss or something. There must be some reason for that because winning 17/20 coinflips seems unlikely.

Let's examine some trends coming up today:

Game 1:
The Rockets are 22-43-5 OU (-5.66 ppg) since Jan 03, 2018 on the road
The Rockets are 1-10-1 OU (-10.33 ppg) since Mar 08, 2019 as a favorite
Houston Rockets vs Western (Under) 27-20-1
Sacramento Kings as Home (Under) 22-15-1

Game 2:
The Nuggets are 4-18 OU (-9.30 ppg) since Feb 08, 2019
The Nuggets are 5-14 OU (-5.03 ppg) since Jan 08, 2019 on the road
Denver Nuggets vs Western (Under) 28-18-0
The Warriors are 4-13-1 OU (-3.33 ppg) since Feb 23, 2019
The Warriors are 1-7 OU (-3.44 ppg) since Feb 23, 2019 as a home favorite


Knowing that, will anyone here bet on the overs? Good luck, but I'll have a punt on the unders. Is there any value though? I have no idea, I've never done trend betting. But surely there must be at least 1% edge, the plays are so obvious.

My view on the original question is "NO". Unless you can spot a trend and articulate some reason there might be a relationship, then still "NO". In the 1st example, and all others cited in this thread, I don't see any reasons given as to why the trend should continue. Without any rationale, then the "spotted" trend is nothing more than falling into a data mining trap. Sorry guys, but trend betting as described in this thread is a waste of time and money.

[nash13]

I am sorry to say, but that's simply not true.
Not every trend is the same as described.

A(points@1)	query
99.50	season = 1995
96.90	season = 1996
95.57	season = 1997
91.58	season = 1998
97.47	season = 1999
94.81	season = 2000
95.48	season = 2001
95.16	season = 2002
93.05	season = 2003
97.19	season = 2004
97.06	season = 2005
98.47	season = 2006
99.59	season = 2007
99.74	season = 2008
100.26	season = 2009
99.21	season = 2010
95.99	season = 2011
97.95	season = 2012
100.91	season = 2013
100.05	season = 2014
102.49	season = 2015
105.62	season = 2016
106.22	season = 2017
110.99	season = 2018
-	season = 2019

This is a table for Average points scored by a team over a season.
You can see an increase of 10 points.

A(rest@1)	query
1.10	season = 1995
1.09	season = 1996
1.09	season = 1997
0.81	season = 1998
1.08	season = 1999
1.08	season = 2000
1.08	season = 2001
1.11	season = 2002
1.13	season = 2003
1.11	season = 2004
1.11	season = 2005
1.11	season = 2006
1.11	season = 2007
1.11	season = 2008
1.11	season = 2009
1.11	season = 2010
0.92	season = 2011
1.11	season = 2012
1.11	season = 2013
1.11	season = 2014
1.11	season = 2015
1.12	season = 2016
1.19	season = 2017
1.19	season = 2018
1.16	season = 2019

Showing 1 to 25 of 25 entries
This is the amount of Avg. Rest per team in the season.
Combine these information and look for misused Trends by the public and you will have something to depend on.
If you know how the market reacts to this trend, then you can use this information in your favor.

[nash13]

Or to make it very simple:
In the NBA since 2014 pace increased, scoring increased, players have more rest, the 3pt shooting is more important than ever.
What do other people make with this information and how is the effect in the lines?

[gojetsgomoxies]

i think some of the stuff posted works...

BUT,

i think you really need a good way to find these...... and also, a good feel for trade-off between small sample size and way too long a trend (which i'd think is many cases is better bet to reverse).......

an example of a good long term trend is army-navy UNDER. i think you should play that forever. i don't think many people understand how powerful the underlying rationale for low scoring games between them is. basically, both teams eat up the clock and play 4 down football. so the entire game is teams moving slowly down the field.

but something like gonzaga at home vs. WCC. just seems like the lines get bigger and bigger every year (adjusted for strength of gonzaga that year). i don't think there's any powerful underlying rationale for this to work. i think everyone knows G much better than WCC opponents and probably crazy HFA

baseball taking pitchers -170 and bigger favorite is interesting. i think the strength of this is somewhere in-between. and this is very price-dependent too, these lines seem to move alot

[gojetsgomoxies]

Or to make it very simple:
In the NBA since 2014 pace increased, scoring increased, players have more rest, the 3pt shooting is more important than ever.
What do other people make with this information and how is the effect in the lines?

i don't have a comment on this specifically............ but looking at ncaa scoring leaders each season, scoring averages were much much higher with no 3-point shooting and no shot clock. makes zero sense........... stats-heads seem blown away that pete maravich had a higher scoring average than his FG% for 3 years straight.

[gojetsgomoxies]

things like rest days, back-to-back games, early eastern start for west coast teams make all kinds of sense, but you'd think the market would adjust for these.

[gojetsgomoxies]

here's something i've found,

oregon beats washington in football forever. most games not close.... then washington gets better, oregon gets worse....... you'd think, well, oregon has a winning culture and they'll fix things, so go with mean-reversion........ but my experience is once washington breaks through on oregon, the effect is very powerful. releases alot of energy.

basically i think if you looked at what point spread would be last 5 years before this and compared to this year and found big difference that you want to fade that difference not tail it......... i would love a study on this.

[QuantumLeap]

Or to make it very simple:
In the NBA since 2014 pace increased, scoring increased, players have more rest, the 3pt shooting is more important than ever.
What do other people make with this information and how is the effect in the lines?

Do you think the books know this and adjust their lines accordingly?

It's funny, because I've seen some trends work for well over 10 years and they are public knowledge yet the books don't adjust for this limited circumstance. I'm thinking that trends that happen less than a dozen times a year aren't worth it for the books to adjust their lines.

[QuantumLeap]

baseball taking pitchers -170 and bigger favorite is interesting. i think the strength of this is somewhere in-between. and this is very price-dependent too, these lines seem to move alot

I've just come across this one recently. It's worked for each of the last 6 seasons.

What I wonder is why didn't it work before that? What changed that caused this to work?

A more pertinent question for me would be, how do I tell when this trend is no longer valid? How many months in a row of it resulting in -EV do I have to go through losing my money before I realize something has changed?

I think these are valid trends. They consist of situations that aren't dependent on a team and they don't limit the time period to just the last 20 games or so. They've proven themselves for years.

[TommieGunshot]

What I wonder is why didn't it work before that? What changed that caused this to work?

A more pertinent question for me would be, how do I tell when this trend is no longer valid? How many months in a row of it resulting in -EV do I have to go through losing my money before I realize something has changed?

If you have no theory about the data, there was never any validity to it. If every game was decided by a coin flip or dice roll, patterns like this would emerge all the time, but the past results would have no meaning.

[QuantumLeap]

If you have no theory about the data, there was never any validity to it. If every game was decided by a coin flip or dice roll, patterns like this would emerge all the time, but the past results would have no meaning.

Just because I don't have a theory now doesn't mean it's not valid. I may discover later why it happens the way it does and why it changed one year. My knowledge or lack of knowledge has no bearing on the validity of the system. Those are two independent entities.

Something tells me 5% ROI over 2500 games isn't a coin flip. I could be wrong. What kind of sample size is needed in order for a system to be statistically significant?