Also a general question which I personally haven't looked into: how do books/public value previous meetings between two teams the same season, and, more importantly are there any known research / analysis on the actual correlation between same match up results?
0
Also a general question which I personally haven't looked into: how do books/public value previous meetings between two teams the same season, and, more importantly are there any known research / analysis on the actual correlation between same match up results?
I appreciate all the work you've done with this. I'm looking into doing some of my own modeling, but I'm not too strong with probabilities as of yet. Still need to do more learning.
0
I appreciate all the work you've done with this. I'm looking into doing some of my own modeling, but I'm not too strong with probabilities as of yet. Still need to do more learning.
Love the Bucks tonight. But the Vegas line seems accurate for the best team in the East at home against a struggling team. Bucks actually playing very well if you look deeper though.
0
Love the Bucks tonight. But the Vegas line seems accurate for the best team in the East at home against a struggling team. Bucks actually playing very well if you look deeper though.
Limit starters with primarily defensive strengths? Or are there no names out?
From Bucks side there are no names out yet, Atlanta will be most likely ack to the starting lineup, more or less. Maybe books expecting better defence?
0
Quote Originally Posted by Bermax:
Limit starters with primarily defensive strengths? Or are there no names out?
From Bucks side there are no names out yet, Atlanta will be most likely ack to the starting lineup, more or less. Maybe books expecting better defence?
Also a general question which I personally haven't looked into: how do books/public value previous meetings between two teams the same season, and, more importantly are there any known research / analysis on the actual correlation between same match up results?
I like to take the team that lost the prior meeting if they are a dog. I believe it has been a profitable angle, at least in college. I think the books may account for this with maybe a 1 point adjustment which I don't think is enough. Is this something you can run- what is spread record of teams seeking revenge that are dogs?
0
Quote Originally Posted by Bermax:
Also a general question which I personally haven't looked into: how do books/public value previous meetings between two teams the same season, and, more importantly are there any known research / analysis on the actual correlation between same match up results?
I like to take the team that lost the prior meeting if they are a dog. I believe it has been a profitable angle, at least in college. I think the books may account for this with maybe a 1 point adjustment which I don't think is enough. Is this something you can run- what is spread record of teams seeking revenge that are dogs?
Just read this thread from top to bottom. Covers needs more stat/model greeks like you on this forum. From what I took from your opening post...you are completely unbiased and aren't familiar with the NBA. And IMO that is a good thing. If you eventually hammer down your system to get to a 55-60% succes rate you will be in plus money as long as you stick to your guns and don't let bias or "hunches" effect your wagers.
With that being said, here are a couple things that may help your system.
Turn over margin is important IMO. If team A has a -.5 t.o.m. and team B has a +2.5 t.o.m. that's theoretically 3 extra possessions for team B. Which is a possible 6 to 9 more points if team B were to score on those 3 extra possessions. From there you could calculate team B's shooting percentage/adj offense vs team A's adj defense and get a "ball park" # of the "probable" extra points off of turnovers for team B.
Another good stat to look at is SOS. How you would incorporate that into your system is up to you and if it fits into your system.
Personally, I always liked the theory that shooting% and offense is never guaranteed. A team can be off shooting the ball for many reasons, and visa versa(sometimes they just shoot really well) 2 things that are a constant are REBOUNDING and DEFENSE. I myself take into account rebounding always.
Good luck w/ your system.
**love the Boston/Charlotte #'s you posted. I locked into Boston +4 last night, and just hit the ML**
0
BERMAX,
Just read this thread from top to bottom. Covers needs more stat/model greeks like you on this forum. From what I took from your opening post...you are completely unbiased and aren't familiar with the NBA. And IMO that is a good thing. If you eventually hammer down your system to get to a 55-60% succes rate you will be in plus money as long as you stick to your guns and don't let bias or "hunches" effect your wagers.
With that being said, here are a couple things that may help your system.
Turn over margin is important IMO. If team A has a -.5 t.o.m. and team B has a +2.5 t.o.m. that's theoretically 3 extra possessions for team B. Which is a possible 6 to 9 more points if team B were to score on those 3 extra possessions. From there you could calculate team B's shooting percentage/adj offense vs team A's adj defense and get a "ball park" # of the "probable" extra points off of turnovers for team B.
Another good stat to look at is SOS. How you would incorporate that into your system is up to you and if it fits into your system.
Personally, I always liked the theory that shooting% and offense is never guaranteed. A team can be off shooting the ball for many reasons, and visa versa(sometimes they just shoot really well) 2 things that are a constant are REBOUNDING and DEFENSE. I myself take into account rebounding always.
Good luck w/ your system.
**love the Boston/Charlotte #'s you posted. I locked into Boston +4 last night, and just hit the ML**
Ok bump for input. Under the assumption that model is a pretty good raw stats line - What new known facts are out on the Charlotte game that would see their average offensive performance / Bostons defensive performance increase / decrease with 5pts a pretty big shift on a team total <100
8 games ago Charlotte only scored 66 points, which is a severe outlier probably skewing the model.
0
Quote Originally Posted by Bermax:
Ok bump for input. Under the assumption that model is a pretty good raw stats line - What new known facts are out on the Charlotte game that would see their average offensive performance / Bostons defensive performance increase / decrease with 5pts a pretty big shift on a team total <100
8 games ago Charlotte only scored 66 points, which is a severe outlier probably skewing the model.
8 games ago Charlotte only scored 66 points, which is a severe outlier probably skewing the model.
Valuable notion. "correcting" this extreme outlier to a more normal score adds a full point to cho's team total. Had the game been within the last 6 it would have been three points.
0
Quote Originally Posted by tsw:
8 games ago Charlotte only scored 66 points, which is a severe outlier probably skewing the model.
Valuable notion. "correcting" this extreme outlier to a more normal score adds a full point to cho's team total. Had the game been within the last 6 it would have been three points.
Bermax so in your opinion what's the best bet of the night im having trouble getting the actual best play. Overall in spreads or totals rather not TT.
I'm not doing any picking, though I realise this thread maybe should've been posted in the general gaming forum. Read my first post from start to end and especially the DISCLAIMER.
0
Quote Originally Posted by Crack5511:
Bermax so in your opinion what's the best bet of the night im having trouble getting the actual best play. Overall in spreads or totals rather not TT.
I'm not doing any picking, though I realise this thread maybe should've been posted in the general gaming forum. Read my first post from start to end and especially the DISCLAIMER.
Can you share us the excel files? I can try to at least make it more user friendly and may be account into the "bench" numbers to make it better.
I will make a complete write up on the model when I get the time later this week.
My spreadsheet's perfectly user friendly as it is, and with all respect I don't think that the excel (libreoffice actually) file in itself is that interesting, I'd rather keep that to myself since I'm not a big fan of personal files floating around with formulas unexplained.
But to reiterate, I'll describe the variables and their weighing in perfect detail later so that anyone who feel the need could build their own sheet to their liking. It is not a complicated model.
Regarding the bench numbers please share your thoughts in the thread on how to quantify that. I'm sure many would love to hear your input.
0
Quote Originally Posted by TimeMachine:
Can you share us the excel files? I can try to at least make it more user friendly and may be account into the "bench" numbers to make it better.
I will make a complete write up on the model when I get the time later this week.
My spreadsheet's perfectly user friendly as it is, and with all respect I don't think that the excel (libreoffice actually) file in itself is that interesting, I'd rather keep that to myself since I'm not a big fan of personal files floating around with formulas unexplained.
But to reiterate, I'll describe the variables and their weighing in perfect detail later so that anyone who feel the need could build their own sheet to their liking. It is not a complicated model.
Regarding the bench numbers please share your thoughts in the thread on how to quantify that. I'm sure many would love to hear your input.
Just read this thread from top to bottom. Covers needs more stat/model greeks like you on this forum. From what I took from your opening post...you are completely unbiased and aren't familiar with the NBA. And IMO that is a good thing. If you eventually hammer down your system to get to a 55-60% succes rate you will be in plus money as long as you stick to your guns and don't let bias or "hunches" effect your wagers.
With that being said, here are a couple things that may help your system.
Turn over margin is important IMO. If team A has a -.5 t.o.m. and team B has a +2.5 t.o.m. that's theoretically 3 extra possessions for team B. Which is a possible 6 to 9 more points if team B were to score on those 3 extra possessions. From there you could calculate team B's shooting percentage/adj offense vs team A's adj defense and get a "ball park" # of the "probable" extra points off of turnovers for team B.
Another good stat to look at is SOS. How you would incorporate that into your system is up to you and if it fits into your system.
Personally, I always liked the theory that shooting% and offense is never guaranteed. A team can be off shooting the ball for many reasons, and visa versa(sometimes they just shoot really well) 2 things that are a constant are REBOUNDING and DEFENSE. I myself take into account rebounding always.
Good luck w/ your system.
**love the Boston/Charlotte #'s you posted. I locked into Boston +4 last night, and just hit the ML**
Exactly. One big reason the average Joe loses is they watch games on TV and become biased toward a team. Because they don't understand what is sustainable, repeatable skill and what part of what they saw is pure luck.
Luck eventually shows up in.....regression.
Field goal % does have the highest correlation to winning NBA games of any stat, the team that shoots the better % wins about 79% of all games. Of coarse the team shooting the better % is based on playing good defense when your shots don't go in.
The team with the most rebounds wins about 69% of all games.
0
Quote Originally Posted by undermysac:
BERMAX,
Just read this thread from top to bottom. Covers needs more stat/model greeks like you on this forum. From what I took from your opening post...you are completely unbiased and aren't familiar with the NBA. And IMO that is a good thing. If you eventually hammer down your system to get to a 55-60% succes rate you will be in plus money as long as you stick to your guns and don't let bias or "hunches" effect your wagers.
With that being said, here are a couple things that may help your system.
Turn over margin is important IMO. If team A has a -.5 t.o.m. and team B has a +2.5 t.o.m. that's theoretically 3 extra possessions for team B. Which is a possible 6 to 9 more points if team B were to score on those 3 extra possessions. From there you could calculate team B's shooting percentage/adj offense vs team A's adj defense and get a "ball park" # of the "probable" extra points off of turnovers for team B.
Another good stat to look at is SOS. How you would incorporate that into your system is up to you and if it fits into your system.
Personally, I always liked the theory that shooting% and offense is never guaranteed. A team can be off shooting the ball for many reasons, and visa versa(sometimes they just shoot really well) 2 things that are a constant are REBOUNDING and DEFENSE. I myself take into account rebounding always.
Good luck w/ your system.
**love the Boston/Charlotte #'s you posted. I locked into Boston +4 last night, and just hit the ML**
Exactly. One big reason the average Joe loses is they watch games on TV and become biased toward a team. Because they don't understand what is sustainable, repeatable skill and what part of what they saw is pure luck.
Luck eventually shows up in.....regression.
Field goal % does have the highest correlation to winning NBA games of any stat, the team that shoots the better % wins about 79% of all games. Of coarse the team shooting the better % is based on playing good defense when your shots don't go in.
The team with the most rebounds wins about 69% of all games.
Hot Forum Topics
