The Ouija Board Part 3

The first two parts of this series were posted in April 2017 and they don’t contain any explanation as to how  probabilities are derived from sports book lines.  I thought I explained this once but can’t seem to find it.   Since the Cubs are once again a betting opportunity tonight a new Ouija Board will be introduced with more information.

Ouija Board

DATE 06_27_10:10_PM CHN LAN

LINEAWAY CHN [ 0.444 ] < 0.444 > +125 $125
STARTAWAY 0.80(0.540) Kyle_Hendricks_CHN TIER 3
LINEHOME LAN [ 0.592 ] < 0.574 > -135 $74
STARTHOME -0.13(0.493) Alex_Wood_LAN TIER 3
CHN 43 34 LAN 42 36
CHN Lineup 2 ==> LAN Starter 3 / Relief 3 == 0.515 CHN 4.53 runs
LAN Lineup 3 ==> CHN Starter 3 / Relief 3 == 0.485 LAN 4.38 runs

Cubs relief dropped to Tier 3 but lineup up to Tier 2 and these two teams are still very very equal.  Tier Combos simulations give the Cubs a 0.515 edge yet the markets have them a 0.444.  Since 0.515 > 0.444 + 0.07 this is considered a betting opportunity.

There are two new numbers in the LINEAWAY and LINEHOME records which show the actual line listed by the sports book and the amount of money you would win by risking $100 on each line.  This provides more context to the other line numbers.  Colors and formatting of the above is still a work in progress.

Probabilities that have been shown here for the last few years are derived from sports book lines.  The Cubs are at +125 which means you get 1.25 times whatever you bet if the Cubs win.  That’s easy.  To calculate the probability you would need to break even on this bet would be as follows:

P(break even) = 100/(100 + line) = 100/(100 + 125) = 0.444

Not too complicated.  For negative lines representing the favorites the break even probability would be as follows:

P(break even) = abs(line)/(100 + abs(line)) = 135/(100 + 135) = 0.574

If the Cubs win tonight the house takes $135 from LAN bettors and gives $125 to CHN bettors and they pocket $10.  If LAN wins they take $100 from CHN bettors and give that to LAN bettors and they pocket nothing.  In the long run the house always makes money this way.

Since we think the Cubs have a higher probability ow winning than 0.444, in the long run we’ll make money — that is if the expected probabilities crunched from these simulations are accurate.

Right now this model has taken 3 years of daily lines from every team, six snapshots a day, once every two hours.  Then there are 4 more years of once a day snapshots, possibly less accurate, before this, 2011-2014 as well.   These simulations will be tested against this dataset in coming months.

Hopefully I didn’t screw up the above math too badly.  Weirdness in Baltimore Oriole home game lines has been observed these last couple of weeks.  This probably needs a post of its own soon.  Also working on a status update for the Atlanta Braves who was a team near the bottom of our career value ranking done last April when there was nothing else to post about.  They sure turned things around.   Might have to do a Cubs status on 6/30 since things have been in flux on the Cubs roster as of late.  Until then ….