Results 1 to 7 of 7

Thread: Betting systems

  1. #1
    the one sun's Avatar
    Join Date
    May 2005
    Location
    Cluj ( all of the time )
    Posts
    5,832
    Reputatie
    1

    Betting systems

    System 1:

    X halftime in "Liga 1" games with no closing line favorite of less than 2.00

    Example : CFR (2.00) --- X (3.20) --- Petrolul (4.00)

    2014/2015 season : 75W-65L ( 53.57% wins )

    Lowest estimated odds for halftime X in such games : 1.95
    Estimated ROI for 2014/2015 : 6.25u ( 4.46% profit )

    2015/2016 so far : 16 wins - 9 losses

    Next steps:

    add 2 seasons of data from Liga 1 to see if the system holds
    add at least one season from 2 other leagues to see if the model can be applied leagues other than Liga 1

  2. #2
    the one sun's Avatar
    Join Date
    May 2005
    Location
    Cluj ( all of the time )
    Posts
    5,832
    Reputatie
    1
    System #1 ( Liga 1 2013 - 2014 season ) : 68W-54L ( 55.73 % wins ) --> 10.6u profit ( 8.68% ROI )

    This round ( 6th of the current season ) , there were 4 games that fit the model, all winners.
    ( botosani-iasi, astra-steaua, cfr-petrolul and concordia-pandurii )

    some quick mentions:

    1) all odds considered belong to bet365
    2) profit and ROI were calculated with average odds of 1.95 ( like in the above post )
    3) bwin has the highest odds for halftime X in Liga 1, regularly around 2.00 ( so profit and ROI can be even greater in theory)
    4) I still need to scrape one more season to reinforce the statistics
    5) Now I am really looking forward to see if the model applies to other leagues

  3. #3
    the one sun's Avatar
    Join Date
    May 2005
    Location
    Cluj ( all of the time )
    Posts
    5,832
    Reputatie
    1
    System 1:

    Liga 1 14/15 (2.29 Goals per game) 75W-65L
    Liga 1 13/14 (2.38 Goals per game) 68W-54L
    Liga 1 12/13 (2.67 Goals per game) 67W-56L

    The system went 3-0 in the 7th round of the current season of Liga 1 ( 5 out of 7 games played ) and is already 23-9 for the season

    System 1 translated to other leagues:

    EPL 14/15 : 84W-85L (2.57 Goals per game)
    Ligue 1 14/15 : 106W-106L (2.49 Goals per game)

    EPL and Ligue 1, despite of being 49% and 50% respectively, are actually profitable, since bwin average odds for X halftime are slightly above 2.00. Also, the system will be applied to two more seasons of each, to see if maybe if the 14/15 seasons were outliers

    Next Steps :

    System 1 applied to other leagues with a smaller goals per game average. I would think that a smaller goals per game number would translate into a bigger percentage of wins for the system.

  4. #4
    the one sun's Avatar
    Join Date
    May 2005
    Location
    Cluj ( all of the time )
    Posts
    5,832
    Reputatie
    1
    System 1 so far:

    Liga 1 14/15 75-65
    Liga 1 13/14 68-54
    Liga 1 12/13 67-56

    EPL 14/15 84-85

    Ligue 1 14/15 106-106
    Ligue 2 14/15 124--124

    System 2:

    System 2A : MLB Reg Season Games part of a series between division opponents who in their previous series game scored 5 or more runs each
    System 2B : MLB Reg Season Games part of a series between division opponents who in their previous series game scored 2 or fewer runs each

    For System 2A we want the over ( both teams have hot bats, we predict that will hold the next game )
    For System 2B we want the under ( both teams can't hit/score runs , we predict that holds in the next game )

    The reason we only look for division opponents is because those teams play each other 15+ times a season, they know each other inside out and is more likely that previous occurrences will hold, they also know the ballpark better, they usually compete for division standings and the series is normally played between 2-4 games in consecutive days ( no lay-offs, travels, etc. ).

    Example :

    Game 1: Cardinals - Reds 2-1 -> game 1 in St. Louis. This only serves as potential trigger, neither system applies
    Game 2: Cardinals - Reds 10-8 ( U/O line is 7 ) -> this game will be a loser in System 2B
    Game 3: Cardinals - Reds 5-6 ( U/O line is 7.5 ) -> this game will be a winner in System 2A
    Game 4: Reds - Cardinals 7-9 ( U/O line is 7 ) -> neither system applies to this game since game 4 starts a new series in Cincinatti

    (2014 season numbers)

    AL EAST Over : 11 - 6 Under : 4-6
    AL CENT Over : 16 - 10 Under : 9-4
    AL WEST Over : 10 - 9 Under : 5-3

    AL : Over : 37 - 25 & Under 18 - 13 --- > Total 55 - 38

    Since Pinnacle sports usually offers 1.95 odds for Under/Overs in MLB , System 2 ROI would be 15.32%


    Next Steps : System 2 numbers for NL (2014) , AL (2013), NL (2013) to see if the numbers hold

  5. #5
    Abomination Johnny D's Avatar
    Join Date
    Apr 2007
    Location
    Zalau
    Posts
    61,509
    Reputatie
    0
    ._.'
    The difference between a good and an awesome diplomat is the ability to reconcile the arrogant assholes with the scardy whiners, it's almost an art really!
    I know because I'M BOTH!

  6. #6
    the one sun's Avatar
    Join Date
    May 2005
    Location
    Cluj ( all of the time )
    Posts
    5,832
    Reputatie
    1
    System 2 updates (2014 season)

    AL 55 - 38
    NL 45 - 51
    ----------------
    MLB 100 - 89 ( 6 unit profit with 189 invested at avg odds of 1.95 ) --> 3.17% ROI

    of which

    2A: Over 64-54 ( 5.76% ROI )
    2B: Under 36-35 ( losing model in 2014 )

    I really dont recommend anything else except for AL + Over , which was 37-25.
    The rest of the models are too close to call...
    Anyway, I am in the process of covering 2013 numbers for System 2, to see if AL + Over makes sense in the long run. Because AL uses Designated Hitters I'm not surprised it might be.

    In the mean time ...

    System 3 ( 3 game chase MartinGale in MLB subsystems )

    All the following systems will use a 3-Game Martingale chase method ( 1u - 2u - 4u ).
    After 3 losses, we stop. After any win, we stop.

    Here we go:

    Cincinatti Reds - 2014 Season

    3 Game U chase after L with less than 3 total runs ( cold bats + good bullpen )
    13-1
    3 Game L chase after a 3 game losing streak ( bad confidence, hard to think a 3W streak follows )
    10--0
    3 Game W chase after a 3 game winning streak ( Godo confidence, hard to believe 3L streak follows )
    9--1
    3 Game L chase after a shutout loss ( this is actually a bad/not-profitable subsystem )
    12--3

    3 Game W chase after a shutout loss ( teams want to bounce back, opposite of previous one )
    14--1
    3 Game L chase after a 7 + run loss ( blow out loss, hard to believe 3W streak follows, probably bad bullpen too )
    6--0
    3 Game O chase after a 15+ run game ( hot bats + bad bullpen, hard to believe 3 U follow, this would mean bats go cold and bullpen is solid suddenly )
    6--0

    Overall : 58-3

    Short mention, because any loss is actually a 7 unit loss, and any win is a 1 unit win, the actual unit based record is 58-21. Also, juice needs to be considered, but I think hitting 73 % is more than enough to cover it.

    I will come back in the following days with the complete numbers for 2014 and more detailed explanations regarding the subsystems in System 3 ( not all are applied in the same way )

  7. #7
    the one sun's Avatar
    Join Date
    May 2005
    Location
    Cluj ( all of the time )
    Posts
    5,832
    Reputatie
    1
    MLB Systems ( continued )

    I have managed to automate the process of scraping game information ( results, home/away moneylines and under/over lines, all courtesy of Pinnacle Sports ). This makes it much more easier to create mlb systems, because i dont have to crunch numbers by hand, there are no mistakes and it's much faster.

    Just a snippet of things to come :

    System name : "L after LLL(L)W"
    System explanation : "After a team loses 3 (or more) games in a row and wins the next game, this system expects the following game to be a loss"
    System reason : "We want to fade teams that snap a 3 (or more) game losing streak, because the pressure to get a win is suddenly off and the quality/form of the team still remains poor.
    System test : 1 unit will be represented as a flat 100$ bet. Each game that fits the system will have 1 unit wagered on it.

    System results for 1st half of 2014 season:

    137 occurences , 74 wins and 63 losses
    13700$ -- > 14719$
    ROI : 7.43%

    ATL : 1 - 3 : -254 ----- > 1 - 3 -254$
    PHI : 3 - 3 : -29 ----- > 4 - 6 -283$
    WSH : 3 - 2 : 175 ----- > 7 - 8 -108$
    NYM : 5 - 2 : 209 ----- > 12 - 10 101$
    MIA : 3 - 1 : 236 ----- > 15 - 11 337$
    BAL : 1 - 2 : -100 ----- > 16 - 13 237$
    BOS : 3 - 2 : 100 ----- > 19 - 15 337$
    NYY : 1 - 3 : -176 ----- > 20 - 18 161$
    TB : 3 - 4 : 5 ----- > 23 - 22 166$
    TOR : 3 - 1 : 200 ----- > 26 - 23 366$
    STL : 2 - 1 : 103 ----- > 28 - 24 469$
    PIT : 5 - 1 : 445 ----- > 33 - 25 914$
    CHC : 6 - 1 : 480 ----- > 39 - 26 1394$
    MIL : 1 - 1 : 0 ----- > 40 - 27 1394$
    CIN : 3 - 1 : 258 ----- > 43 - 28 1652$
    DET : 3 - 2 : 140 ----- > 46 - 30 1792$
    MIN : 2 - 3 : -138 ----- > 48 - 33 1654$
    CHW : 0 - 0 : 0 ----- > 48 - 33 1654$
    KC : 1 - 3 : -228 ----- > 49 - 36 1426$
    CLE : 2 - 5 : -310 ----- > 51 - 41 1116$
    LAA : 1 - 1 : 28 ----- > 52 - 42 1144$
    HOU : 4 - 4 : -180 ----- > 56 - 46 964$
    OAK : 1 - 2 : -80 ----- > 57 - 48 884$
    TEX : 2 - 3 : -187 ----- > 59 - 51 697$
    SEA : 0 - 3 : -342 ----- > 59 - 54 355$
    LAD : 2 - 0 : 323 ----- > 61 - 54 678$
    ARI : 4 - 4 : -78 ----- > 65 - 58 600$
    COL : 3 - 1 : 207 ----- > 68 - 59 807$
    SF : 2 - 1 : 120 ----- > 70 - 60 927$
    SD : 4 - 3 : 92 ----- > 74 - 63 1019$

    The numbers above do not make the system profitable. For that, there have to be at least 3 complete profitable seasons

    I will obviously get back in the near future with more systems and hopefully a bigger database of games.


Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •