[Consigliere]

HF and division=o: division and day=Thursday and total <45 Something for the under tonight

[Mako-SBR]

FortySix, your posts continue to have smilies in the code despite JMon posting about how to correct that multiple times. Please edit and clean your posts after you make them, thanks.

For those who haven't seen it:

INSTRUCTIONS TO ELIMINATE SMILIES (thx J)

1. Make your post as normal
2. Edit your post.
3. Click disable smilies from text
4. Save.

[Consigliere]

HF and division=o: division and day=Thursday and total <45 Something for the under tonight

Big under hit. Any promising trends for this weekend?

[Mako-SBR]

Big under hit. Any promising trends for this weekend?

That was a good one, nice Consig!

I'm on three for the weekend, do not like all three picks, but that's the game with SDQL...have to bet the games your queries point to if you believe your scenarios are sound (thanks to JMon and others who contributed to these):

A and P:L and p:AL and op:L and 4<=game number<=11 and season>=2008

H and 4< tS(ats margin>0,N=7)< 7 and o:rest<9 and week<=15 and week!=14 and week!=13 and season>=2008

H and p:AW and op:HW and (5.5 >= line >= 1.5 or line >= 8.5) and division!=NFC East and season>=2008

[Consigliere]

That was a good one, nice Consig!

I'm on three for the weekend, do not like all three picks, but that's the game with SDQL...have to bet the games your queries point to if you believe your scenarios are sound (thanks to JMon and others who contributed to these):

A and P:L and p:AL and op:L and 4<=game number<=11 and season>=2008

H and 4< tS(ats margin>0,N=7)< 7 and o:rest<9 and week<=15 and week!=14 and week!=13 and season>=2008

H and p:AW and op:HW and (5.5 >= line >= 1.5 or line >= 8.5) and division!=NFC East and season>=2008

Awesome. Will check these out. I have nothing in my bank for NFL trends but will try and do some legwork this weekend and report back anything promising.

[FortySix]

Hey Mako,

With H and p:AW and op:HW and (5.5 >= line >= 1.5 or line >= 8.5) and division!=NFC East and season>=2008, whats the play? When I put that into SDQL there isn't any active queries..

[JMon]

Hey Mako,

With H and p:AW and op:HW and (5.5 >= line >= 1.5 or line >= 8.5) and division!=NFC East and season>=2008, whats the play? When I put that into SDQL there isn't any active queries..

It moved out of the line parameter. Easy check is too simply remove it from the query

[JMon]

That was a good one, nice Consig!

I'm on three for the weekend, do not like all three picks, but that's the game with SDQL...have to bet the games your queries point to if you believe your scenarios are sound (thanks to JMon and others who contributed to these):

A and P:L and p:AL and op:L and 4<=game number<=11 and season>=2008

H and 4< tS(ats margin>0,N=7)< 7 and o:rest<9 and week<=15 and week!=14 and week!=13 and season>=2008

H and p:AW and op:HW and (5.5 >= line >= 1.5 or line >= 8.5) and division!=NFC East and season>=2008

Yes, I was going to post this. I will be on the Bears...love the ML/margin of victory in this revenge angle! May put a little on the ML.

[JMon]

Over - Mia/Det

SDQL- Play over in a non-conference matchup; on a road team with a winning streak of three more.

A and streak>=3 and conference != o:conference and 32= line>=-10

[emceeaye]

Can't remember where I got this but although it favors Panthers ATS, I'm reticent to act on team-specific queries of low sample sizes across multiple seasons...just in case, here it is:

(team=Cowboys or team=Eagles) and F and p:F and season>=2011

[Consigliere]

Tough week on the trend bets this week. Luckily I went with GB not CHI and stayed away from MNF. The OU trend I posted is active again this Thursday with Bills vs Dolphins. Like this pick a lot, two solid defences and two struggling offences. HF and division=o:division and day=Thursday and total <=45

[Mako-SBR]

NFL is just tough for situational analysis because of a variety of factors...it makes the NBA seem like child's play in comparison.

Things like Special Team play being so variable week to week (fluke muffed punts, runbacks for TDs, etc), prime-time games versus non-prime-time (tough to quantify whether a scenario is still valid if it occurs on SNF or MNF where stars tend to shine brighter---see GB/CHI this past MNF, sigh), the smaller sample sizes due to limited games (try creating good TNF trends, nearly impossible due to low volume and Thanksgiving screwing it up), etc, etc, etc.

Not the easiest thing to do, but definitely looking forward to seeing how the NFL tab of nash's master sheet grows.

[Consigliere]

Yeah NFL has been tough the last few years. Like you said, crazy things happen. Also the trend to more passing seems to make results more variable and difficult to predict. Will work on NFL some more this week to see if anything. Feels like this time in the year is prime time to tail the big spreads...going to do some work on this angle.

[mdunlap3]

Query I just started fooling with...small N, but looks pretty solid, worth a shot I think. FINS SU

The Bills playing an away game against a divisional opponent coming off a loss where they scored less than 14 points since 2005.

team=Bills and A and p:L and DIV and points<14 and season>=2005

[Consigliere]

HF and division=o: division and day=Thursday and total <45 Something for the under tonight

Another Thursday night division game under hit big time...keep on rolling! I will post any compelling trends I find but probably not too much time this weekend to look at.

[Mako-SBR]

Another Thursday night division game under hit big time...keep on rolling! I will post any compelling trends I find but probably not too much time this weekend to look at.

Nice Consig!

That game was a perfect example of what we were talking about this week in terms of difficulty with SDQL for NFL forecasting...from the crazy "safety" call, to the game-changing pass interference call shortly after, to the Buffalo kicker missing by a few inches...the variables in a typical NFL game are just insane to try and get a handle on.

[b1slickguy]

HF and rest>10 and po:TY - p:TY >=150 and season>=1996

[moshi]

A couple I am looking at this week:

50>=Andrew Luck:p:passes>=35 and Andrew Luck:p:completions/Andrew Luck:p:passes>.45

ATS 20-3-1 (87.0%) Have been riding this all year. 5-1 in 2014.

team=Packers and p:margin>=3 and po:PENY>=95 and NB and 19971201<=date

ATS 11-0-0 (100%) but there's also...

4< tS(ats margin>0,N=7)< 7 and H and month=11 and 2008<=season and o:rest<9

...which is on the other side, so may lay off this one.

Marshawn Lynch:p:rushing yards>=113 and NB and season>=2011

ATS 12-0-0 (100%)

team=Broncos and F and p:AFW and date>=20061029

OU 13-3-0 (81.2%)

I know there's some weaknesses around player/team-based queries so play at your own risk !

[JMon]

Consider the Packers today!

9>=line >= 3.5 and 20>=week >= 7 and WP >= 77 and 1993<=season

contrarian view where we want to fade a winning dog in latter half of the season...a on heater.

[JMon]

Consider the Broncos today!

Basic premise behind this sdql is playing a favorite where their run defense is performing good the last few games and conversely their oppt's rushing offense is not.

SDQL

-9.5<=line<=-5 and po:rushing yards<=99 and ppo:rushing yards<=99 and op:rushing yards<=99 and opp:rushing yards<=99 and oppp:rushing yards<=99

[JMon]

HF and rest>10 and po:TY - p:TY >=150 and season>=1996

week > 8 and wins = 0 and D and conference = o:conference and losses > 6 and o:wins > 2

Winless conference dogs (seven or more losses) seem to cover in the latter part of the season. It tightens up as divisional game.

[moshi]

Something on the Bears/Bucs game in week 12:

team=Bears and NDIV and n:DIV and total<47 and season>=2010

OU 16-1 (94.1%) since 2010.

[Jay Price]

I cherry picked this one, but tossing it out there for those that believe:

po:division=division and ppo:division=division and o:division=division and HD and total>40 and total<44

Raiders

ATS:	16-3-1 (7.70, 84.2%)

[Cutler'sThumb]

Hi guys, maybe someone can come up with the correct query for this trend I manually back tested last summer. I've been fiddling around with it but the language is really still all greek to me.
Here's what I got manually: Inter conference dome games OVER is 60-32 since '08. I think that result is with me eliminating the game if both teams are dome teams. The idea is unfamiliar teams playing each other on a fast surface = more scoring.

[JMon]

Hi guys, maybe someone can come up with the correct query for this trend I manually back tested last summer. I've been fiddling around with it but the language is really still all greek to me.
Here's what I got manually: Inter conference dome games OVER is 60-32 since '08. I think that result is with me eliminating the game if both teams are dome teams. The idea is unfamiliar teams playing each other on a fast surface = more scoring.

something like this???? non conference on artificial turf, but only back 2010 - non sure how to eliminate games if both are dome teams.

conference!=o:conference and surface=artificial and 2010<=season and H

[JMon]

Consider a small favorite (ATL) in second half the season, good offense against a average defense, coming off win by 6 pts or less

F and 27>=tA(points)>=23 and 23>=oA(o:points)>=18 and game number>=8 and p:W and p:margin<=6 and 2005<=season and line>-5.5

[mdunlap3]

Dr.M is really struggling as of late....I believe he is 1-4 this week. He was horrible the last two weeks too. Very frustrating to say the least...

[JMon]

NFL and sdql is tough..least favorite sport to use it. I think mako warned of the cherry picker

[nash13]

NFL style of sport is no easy field to use a tool like SDQL. Few Games, few situations that work long term.
Best sport to use for me is MLB. Tons of games, tons of situations. Simple as that.

[JMon]

NFL style of sport is no easy field to use a tool like SDQL. Few Games, few situations that work long term.
Best sport to use for me is MLB. Tons of games, tons of situations. Simple as that.

agree...luv NBA as well.

[mdunlap3]

Can't disagree with you gentlemen, just frustrating ya know. You guys mess with college basketball at all? I've been looking at this website kenpom.com, he has some sick stats. Been doing pretty well so far this season, you guys should check it out.

[Cutler'sThumb]

conference!=o:conference and surface=artificial and 2010<=season and H

Thanks JMon! This is pretty close...it gives us any game played on an artificial surface, which would seem to fit the overall logic. Bigger sample size that still hits at 60% (although only 11-12 this year). Maybe worth rolling with, but this season's results are a little troubling.

[JMon]

Consider the Jets ATS, possible ML

AFC divisional revenge game

P:DIV and P:L and DIV and P:season = season and AD and conference = AFC and 18>game number>6 and line<7 and 1992<=season and losses<11 and o:wins<11

[parlayin]

what's a good way to include high-scoring and low-scoring teams in a query? Like average ppg

[JMon]

what's a good way to include high-scoring and low-scoring teams in a query? Like average ppg

Using the summative method with the points parameter would probably be the best.

For example....below you will find teams that averages 30 or more points a game in 2014.

tA(points)>=30 and 2014=season