Journeyman said:
My two sports to handicap are definitely football and basketball....so handicapping baseball is quite NEW to me. I read through the book and love the information. It gives a lot of interesting perspectives and statistics to analyze. It still leaves a ton of room for subjectivity which will have to be learned over time.
The thing I am worried about is setting up good spreadsheets to analyze the data. I want to make sure that it can be updated easily.
Don,
I dont know how old you are but if you dont want a brain tumour early in life, lay off the heavy stats. Take it from someone who has done every research possible and every statistical angle, keep it simple and try to find a niche that the oddsmakers havent time to, that doesnt take up your life and is fundamentally sound. That way you might hit on Nirvana or have an enjoyable life betting, but you wont have that wasted feeling chasing a rainbow.
Best O Luck the rest of the year.
I dont know how old you are but if you dont want a brain tumour early in life, lay off the heavy stats. Take it from someone who has done every research possible and every statistical angle, keep it simple and try to find a niche that the oddsmakers havent time to, that doesnt take up your life and is fundamentally sound. That way you might hit on Nirvana or have an enjoyable life betting, but you wont have that wasted feeling chasing a rainbow.
Best O Luck the rest of the year.
Don,
If you're referring to my sheet, it's set up to use equalized statistics, with park factors worked in later, so that we're using Park 100 stats for both hitters and pitchers, as for our hypothetical game, both sets of players are in the same ballpark..
Be very careful using RAW stats from sites that don't account for the fact Rockies hitters play at a high elevation etc..
If you're referring to my sheet, it's set up to use equalized statistics, with park factors worked in later, so that we're using Park 100 stats for both hitters and pitchers, as for our hypothetical game, both sets of players are in the same ballpark..
Be very careful using RAW stats from sites that don't account for the fact Rockies hitters play at a high elevation etc..
They do, yes..
But my point is if you take a player's raw stats, you first have to normalize (remove the biases of their home park), then adjust to the park (if you want a reliable total).. using stats like EqOBP, EqSLG, PERA neutralises that, and you can then just do step 2 and adjust for the park the game is being played in
But my point is if you take a player's raw stats, you first have to normalize (remove the biases of their home park), then adjust to the park (if you want a reliable total).. using stats like EqOBP, EqSLG, PERA neutralises that, and you can then just do step 2 and adjust for the park the game is being played in
About a week on and off, much time was spent transferring PECOTA projections to start off with, setting lineups, depth charts, pitching rotations etc so that it's all quick to adjust once the season starts..
Still need to work out the formulae for total adjustments on park factors and umpire/weather info -- have ML's and RL's sorted now (RL adjustments w/ thanks to Cutter's posts @ FezziksPlace)
Still need to work out the formulae for total adjustments on park factors and umpire/weather info -- have ML's and RL's sorted now (RL adjustments w/ thanks to Cutter's posts @ FezziksPlace)
Now I'm getting worried about Baseball betting, when people like Santo and Labeeb start exchanging posts that a Physics Professor wouldnt understand, what chance the populus.
While not admonishing both those posters and what might be their angle, there's not much the books dont put into their prices that is not known on a statistical scale, more so with US sports.
I will tell you one thing, if the public are aware of a situation, statistic or betting pattern, so are the books, and they act accordingly. Like I keep telling everyone, find something tangible that the books either find insignificant or too hard to evaluate, more often its very simple.:drink:
While not admonishing both those posters and what might be their angle, there's not much the books dont put into their prices that is not known on a statistical scale, more so with US sports.
I will tell you one thing, if the public are aware of a situation, statistic or betting pattern, so are the books, and they act accordingly. Like I keep telling everyone, find something tangible that the books either find insignificant or too hard to evaluate, more often its very simple.:drink:
I come back to the old adage of the books trying to balance action, especially on RL's... This means that the true runline is forgotton in the mix, in reality of course the RL is:
P (Royals +1.5) = P (Royals Win) + P (Yanks by 1 | Yanks Win) -- adjusting for totals, home advantage..
Let's try a real-world example..
The White Sox are home favourites on Sunday against the Cleveland Indians. They are favourites of the magnitude -124, or are expected to win the game 55.3571% of the time...
Pinny currently have the following:
CLE +1.5 -206
CHA -1.5 +186
In theory:
P (CLE +1.5) = P (CLE WIN) + P(CHA BY 1 | CHA WIN)
= 0.446429 + (0.29*0.553571)
= 0.446 + 0.1605
= 0.6065 (or actually 0.606964 if we went to full digits)
= -154.43
Now we have to allow for Chicago being the home team, which reduces the chance of them winning by >1 run (as they (a) wouldn't bat in the bottom of the 9th, and (b) the game would end if they go ahead by 1 in the bot 9 or extra innings)..
Stats I had to hand from DrBob indicate home teams cover 68.1% of the time, and road teams 76.5% of the time, an order of magnitude of around 112%..
So we adjust the P(CHA by 1 | CHA WIN) part accordingly..
P (CLE +1.5) = P (CLE WIN) + P(CHA BY 1 | CHA WIN)
= 0.446429 + (0.325771*0.553571)
= 0.446 + 0.180377
= 0.626337 (or actually 0.626766089 if we went to full digits)
= -167
That gives us a no juice Runline of -167/+167, or roughly a 3% edge on the Pinnacle line of betting +183... is that enough to bet it? Who knows..
The total is a bit on the low side, which may account for some of it.. but even if we push the 1-run chance up a couple of percent you still only get -173ish..
There will be better examples, but I only had one game to work with atm
P (Royals +1.5) = P (Royals Win) + P (Yanks by 1 | Yanks Win) -- adjusting for totals, home advantage..
Let's try a real-world example..
The White Sox are home favourites on Sunday against the Cleveland Indians. They are favourites of the magnitude -124, or are expected to win the game 55.3571% of the time...
Pinny currently have the following:
CLE +1.5 -206
CHA -1.5 +186
In theory:
P (CLE +1.5) = P (CLE WIN) + P(CHA BY 1 | CHA WIN)
= 0.446429 + (0.29*0.553571)
= 0.446 + 0.1605
= 0.6065 (or actually 0.606964 if we went to full digits)
= -154.43
Now we have to allow for Chicago being the home team, which reduces the chance of them winning by >1 run (as they (a) wouldn't bat in the bottom of the 9th, and (b) the game would end if they go ahead by 1 in the bot 9 or extra innings)..
Stats I had to hand from DrBob indicate home teams cover 68.1% of the time, and road teams 76.5% of the time, an order of magnitude of around 112%..
So we adjust the P(CHA by 1 | CHA WIN) part accordingly..
P (CLE +1.5) = P (CLE WIN) + P(CHA BY 1 | CHA WIN)
= 0.446429 + (0.325771*0.553571)
= 0.446 + 0.180377
= 0.626337 (or actually 0.626766089 if we went to full digits)
= -167
That gives us a no juice Runline of -167/+167, or roughly a 3% edge on the Pinnacle line of betting +183... is that enough to bet it? Who knows..
The total is a bit on the low side, which may account for some of it.. but even if we push the 1-run chance up a couple of percent you still only get -173ish..
There will be better examples, but I only had one game to work with atm
