[Justin7]

Care to send me your raw data, and i'll verify or refute your hypothesis?

[RickySteve]

i would be sick, if i were to lose alot doing this.

hoping for the best.

Mr. Progression

I hope you have health insurance.

[reno cool]

what happens if you verify his hypothesis?

I tend to agree with Mudcat.

[Justin7]

I'll be impressed, and study it further before betting.

[Kingctb27]

Got a quick question here... Let's say that a LeBron James type in the NBA is out for the first 20 games of the year and then he comes back to the team. Would you still have the Cavs in your criteria or would you just exclude them because the superstar player obviously skews your system... ?

[Ganchrow]

So I finally back tested the OP's "system" for regular season MLB 1999 through 2008 and regular season NHL 2005/6 through 2007/8 using closing ML prices from covers.com.

Flat-betting results follow:

MLB:

Season	Tot Bet	Tot Res	% Res
1999	 1,796	-64.97	-3.62%
2000	 1,804	+57.39	+3.18%
2001	 1,804	-45.04	-2.50%
2002	 1,787	-21.48	-1.20%
2003	 1,789	-18.14	-1.01%
2004	 1,777	-69.54	-3.91%
2005	 1,787	-19.78	-1.11%
2006	 1,716	 +9.82	+0.57%
2007	 1,731	+32.78	+1.89%
2008	 1,706	-36.44	-2.14%
TOTAL	17,697	-175.4	-0.99%

NHL:

Season	Tot Bet	Tot Res	% Res
2005/6	  882	-28.27	-3.21%
2006/7	  877	 +4.95	+0.56%
2007/8	  861	+57.30	+6.66%
TOTAL	2,620	+33.98	+1.30%

[reno cool]

Would you say that these # are significantly better than random given the vig? Or likely due to luck, or some sort of "data mining"?

[Stumpage]

Ganchrow, im not doubting your numbers but are you sure that you tested properly?

Well there's a sentence I never thought I'd see in my lifetime.....

[20Four7]

Well Stumpage I'm sure Ganch's bad numbers are better than my good ones.

I agree that that's not a sentence I would have thought ever would show up on this forum. I'd be interested in the other sports to see where they stack up.

[Peep]

Well there's a sentence I never thought I'd see in my lifetime.....

[Ganchrow]

Ganchrow, ... are you sure that you tested properly. its very difficult to do this... you have to know the exact records every day in each matchup

Quite sure. I spent the better part of 10 minutes writing a simple script to perform the back testing. It's very straightforward. Let me know if you'd like me to post the Perl source code.

For which years is the discrepancy with your data most pronounced?

[Munson15]

I figured it was too good to be true. Would like to know if perhaps a couple of seasons were profitable and then the curtain came down, or if line shopping could make that much of a difference.

[durito]

Let me know if you'd like me to post the Perl source code.

I'd be interested in seeing this if you have a chance. Thanks.

[Ganchrow]

I'd be interested in seeing this if you have a chance. Thanks.

Input data (defined in the source) needs to be sorted ASC by date.
Perl source code follows:

#!perl -w

use strict;

use constant LEAGUE          =>    uc 'MLB';
use constant MIN_GAMES       =>    ( LEAGUE eq 'MLB' ? 162/4 : (LEAGUE eq 'NHL' ? 82/4 : ''));
use constant MIN_DOG_ODDS    =>    +100;
use constant TOTAL_YEAR      =>     99999999;

my ($wins_r, $losses_r, $tot_bet_r, $tot_win_r, $tot_var_r, $count_r, $tot_util_r, );
my ($tot_bet, $tot_win, ) = (0,0);
my ($date, $vt, $ht, $vml, $hml, $tot, $tadj, $oadj, $uadj, $vsc, $hsc,$year,$month);
while (<>) **
    chomp;
    if (LEAGUE eq 'NHL') **
        ($date, $vt, $ht, $vml, $hml, $tot, $oadj, $uadj, $vsc, $hsc) = split/\t/;    # nhl
        # fmt (tab sep)=> date, visit team, home team, visit ml, home ml, total, over adj, under adj, visit score, home score    
    ** elsif (LEAGUE eq 'MLB') **
        ($date, $vt, $ht, $vml, $hml, $tot, $tadj, $vsc, $hsc) = split/\t/;        # mlb
        # fmt (tab sep)=> date, visit team, home team, visit ml, home ml, total, total adj, visit score, home score
    **
    next unless $date > 0;
    ($year,$month) = $date =~ /^(....)(..)/;
    $year++ if (LEAGUE eq 'NHL' && $month+0 >= 8);
    if (     
        ( ($wins_r->**$year**->**$vt** + $losses_r->**$year**->**$vt** < MIN_GAMES) &&
          ($wins_r->**$year**->**$ht** + $losses_r->**$year**->**$ht** < MIN_GAMES)
        )
    ) **
        # not enough games played
        next;
    **

    my ($dog, $fave, $dog_score, $fave_score, $dog_ml, $fave_ml,);
    if ($vml > $hml) **
        # visitor is dog
        $dog = $vt;
        $fave = $ht;
        $dog_score = $vsc;
        $fave_score = $hsc;
        $dog_ml = $vml;
        $fave_ml = $hml;
    ** elsif ($hml > $vml) **
        # home is dog
        $dog = $ht;
        $fave = $vt;
        $dog_score = $hsc;
        $fave_score = $vsc;
        $dog_ml = $hml;
        $fave_ml = $vml;
    ** else **
        # no dog
        next;
    **

    my ($fave_rec,$dog_rec) = (0,0);
    if ($fave_rec + $dog_rec > 0) **
        $dog_rec = $wins_r->**$year**->**$dog** / ($wins_r->**$year**->**$dog** + $losses_r->**$year**->**$dog**);
        $fave_rec = $wins_r->**$year**->**$fave** / ($wins_r->**$year**->**$fave** + $losses_r->**$year**->**$fave**);
    **

    if ($fave_rec > $dog_rec) **
        # fave record better than dog
        next;
    ** elsif ($dog_ml < MIN_DOG_ODDS) **
        # not a big enough dog
        next;
    **

    my $dog_win = ($dog_score > $fave_score ? 1 : 0);
    my $dec_dog_odds = &us2dec($dog_ml);
    my $edge = 0.01;
#    my $bet = $edge/($dec_dog_odds-1);
    my $bet = 1;
    my $win_amt = $bet * ( $dog_win * $dec_dog_odds - 1 );
    my $var = $bet*$bet*($dec_dog_odds-1);
    my $util = $bet < 1 ? log(1+$win_amt) : 0;

    print "$date\t$dog\t$fave\t$dog_ml\t$fave_ml\t$dog_score\t$fave_score\t$dog_rec\t$fave_rec\t$bet\t$win_amt\t$var";
#    print "\t$util";
    print "\n";

    $tot_var_r->**+TOTAL_YEAR** += $var;
    $tot_bet_r->**+TOTAL_YEAR** += $bet;
    $tot_win_r->**+TOTAL_YEAR** += $win_amt;
    $count_r->**+TOTAL_YEAR**++;
    $tot_util_r->**+TOTAL_YEAR** += $util;
    $tot_var_r->**$year** += $var;
    $tot_bet_r->**$year** += $bet;
    $tot_win_r->**$year** += $win_amt;
    $tot_util_r->**$year** += $util;
    $count_r->**$year**++;
** continue **
    if ($vsc > $hsc) **
        $wins_r->**$year**->**$vt**++;
        $losses_r->**$year**->**$ht**++;
    ** elsif ($vsc < $hsc) **
        $wins_r->**$year**->**$ht**++;
        $losses_r->**$year**->**$vt**++;
    **
**
foreach my $year (sort keys %$tot_bet_r) **
    my $disp_year = $year;
    $disp_year = 'TOTAL' if $disp_year == TOTAL_YEAR;
    printf STDERR "$disp_year\t$count_r->**$year**\t%0.4f\t%0.4f\t", $tot_win_r->**$year**,$tot_bet_r->**$year**;
    printf STDERR "%0.4f%%\t", 100*$tot_win_r->**$year**/$tot_bet_r->**$year**;
    printf STDERR "%0.6f", sqrt($tot_var_r->**$year**)/$tot_bet_r->**$year**;
#    printf STDERR "%0.3f\t", $tot_win_r->**$year** / sqrt($tot_var_r->**$year**);
#    printf STDERR "%0.4f%%\t", 100*exp($tot_util_r->**$year**)-1 if $tot_util_r->**$year**;
    printf STDERR "\n";
**

sub us2dec($) **
    my $us = 0 + shift;
    return (
        $us >= 100
        ? 1 + $us / 100
        : ( $us <= -100 ? 1 - 100 / $us : 0 )
    );
** 

[Ganchrow]

Sample data for NHL 2008 season though yesterday:

DATE	V TEAM	H TEAM	V SIDE	H SIDE	TOTAL	O_ML	U_ML	V SCORE	H SCORE
20081004	NY RANGERS	TAMPA BAY	-114	104	5.5	-115	-105	2	1
20081004	PITTSBURGH	OTTAWA	-105	-105	6	107	-127	4	3
20081005	OTTAWA	PITTSBURGH	-105	-105	6	-111	-109	3	1
20081005	TAMPA BAY	NY RANGERS	111	-121	5.5	102	-122	1	2
20081009	ANAHEIM	SAN JOSE	130	-140	5	-110	-110	1	4
20081009	BOSTON	COLORADO	136	-146	5.5	107	-127	5	4
20081009	CALGARY	VANCOUVER	117	-127	5	-129	109	0	6
20081009	TORONTO	DETROIT	272	-302	5.5	-112	-108	3	2
20081010	CHICAGO	NY RANGERS	129	-139	5.5	122	-142	2	4
20081010	COLUMBUS	DALLAS	156	-166	5	-115	-105	5	4
20081010	FLORIDA	CAROLINA	134	-144	6	133	-153	4	6
20081010	MONTREAL	BUFFALO	-111	101	6	107	-127	1	2
20081010	NASHVILLE	ST. LOUIS	-105	-105	5	-132	112	2	5
20081010	NY ISLANDERS	NEW JERSEY	188	-208	5	-108	-112	1	2
20081010	WASHINGTON	ATLANTA	-135	125	6	122	-142	4	7
20081011	ATLANTA	FLORIDA	135	-145	5.5	-104	-116	2	3
20081011	BOSTON	MINNESOTA	117	-127	5	-104	-116	3	4
20081011	CAROLINA	TAMPA BAY	-101	-109	5.5	-109	-111	4	3
20081011	CHICAGO	WASHINGTON	120	-130	5.5	-106	-114	2	4
20081011	COLUMBUS	PHOENIX	124	-134	5.5	116	-136	1	3
20081011	DALLAS	NASHVILLE	-101	-109	5	-133	113	1	3
20081011	DETROIT	OTTAWA	-140	130	5.5	106	-126	3	2
20081011	LOS ANGELES	SAN JOSE	200	-220	5.5	101	-121	1	3
20081011	MONTREAL	TORONTO	-126	116	5.5	-115	-105	6	1
20081011	NEW JERSEY	PITTSBURGH	122	-132	5	-121	101	2	1
20081011	NY RANGERS	PHILADELPHIA	124	-134	5.5	115	-135	4	3
20081011	ST. LOUIS	NY ISLANDERS	-101	-109	5	-110	-110	2	5
20081011	VANCOUVER	CALGARY	117	-127	5	-127	107	5	4
20081012	COLORADO	EDMONTON	117	-127	5.5	-138	118	2	3
20081012	PHOENIX	ANAHEIM	165	-175	5	-113	-107	4	2
20081012	SAN JOSE	LOS ANGELES	-155	145	5.5	116	-136	1	0
20081013	BUFFALO	NY ISLANDERS	-128	118	5.5	116	-136	7	1
20081013	DETROIT	CAROLINA	-136	126	5.5	-117	-103	3	1
20081013	MONTREAL	PHILADELPHIA	-108	-102	5.5	-114	-106	5	3
20081013	NASHVILLE	CHICAGO	130	-140	5.5	-116	-104	3	2
20081013	NEW JERSEY	NY RANGERS	111	-121	5	-110	-110	1	4
20081013	ST. LOUIS	TORONTO	105	-115	5.5	108	-128	5	4
20081013	VANCOUVER	WASHINGTON	102	-112	5.5	104	-124	1	5
20081014	ANAHEIM	LOS ANGELES	-158	148	5	-113	-107	3	6
20081014	COLORADO	CALGARY	135	-145	5.5	-107	-113	4	5
20081014	COLUMBUS	SAN JOSE	195	-215	5	-107	-113	2	5
20081014	MINNESOTA	ATLANTA	-112	102	5.5	107	-127	4	2
20081014	PHILADELPHIA	PITTSBURGH	157	-167	5.5	-123	103	2	3
20081015	BOSTON	MONTREAL	162	-172	5.5	-102	-118	3	4
20081015	BUFFALO	NY RANGERS	132	-142	5.5	122	-142	3	1
20081015	EDMONTON	ANAHEIM	130	-140	5.5	120	-140	3	2
20081015	NASHVILLE	DALLAS	165	-175	5	-128	108	4	6
20081015	PHOENIX	CHICAGO	118	-128	5.5	-112	-108	1	4
20081016	DALLAS	ST. LOUIS	103	-113	5	-151	131	1	6
20081016	MINNESOTA	FLORIDA	105	-115	5	-110	-110	6	2
20081016	NEW JERSEY	ATLANTA	-142	132	5	-136	116	1	0
20081016	NY ISLANDERS	TAMPA BAY	174	-184	5.5	-110	-110	4	3
20081016	PHILADELPHIA	COLORADO	116	-126	5.5	-134	114	2	5
20081016	VANCOUVER	DETROIT	175	-185	5.5	133	-153	4	3
20081016	WASHINGTON	PITTSBURGH	118	-128	5.5	-122	102	4	3
20081017	CAROLINA	LOS ANGELES	-123	113	5.5	-128	108	3	4
20081017	EDMONTON	CALGARY	124	-134	5.5	-105	-115	4	3
20081017	NASHVILLE	COLUMBUS	135	-145	5	-146	126	3	5
20081017	PHOENIX	OTTAWA	140	-150	5.5	103	-123	3	6
20081017	SAN JOSE	ANAHEIM	100	-110	5	-105	-115	0	4
20081017	TORONTO	NY RANGERS	184	-197	5.5	-116	-104	0	1
20081017	VANCOUVER	BUFFALO	168	-178	5.5	114	-134	2	5
20081018	BOSTON	OTTAWA	132	-142	5.5	-108	-112	4	2
20081018	BUFFALO	ATLANTA	-125	115	5.5	-137	117	2	3
20081018	CALGARY	EDMONTON	121	-131	5.5	-125	105	2	3
20081018	CHICAGO	ST. LOUIS	102	-112	5.5	103	-123	3	4
20081018	COLORADO	DALLAS	127	-137	5.5	-127	107	5	4
20081018	COLUMBUS	NASHVILLE	143	-153	5.5	123	-143	3	6
20081018	MINNESOTA	TAMPA BAY	114	-124	5.5	105	-125	1	0
20081018	NEW JERSEY	WASHINGTON	118	-128	5	-120	100	4	3
20081018	NY ISLANDERS	FLORIDA	150	-160	5.5	102	-122	0	2
20081018	NY RANGERS	DETROIT	170	-180	5	109	-129	4	5
20081018	PHILADELPHIA	SAN JOSE	175	-185	5.5	112	-132	4	5
20081018	PHOENIX	MONTREAL	220	-240	5.5	-131	111	1	4
20081018	TORONTO	PITTSBURGH	222	-242	5.5	-141	121	1	4
20081019	CAROLINA	ANAHEIM	152	-162	5.5	-102	-118	3	1
20081019	VANCOUVER	CHICAGO	-115	105	5.5	111	-131	2	4
20081020	DALLAS	NY RANGERS	140	-150	5.5	110	-130	2	1
20081020	FLORIDA	MONTREAL	204	-224	5.5	-105	-115	1	3
20081020	PITTSBURGH	BOSTON	130	-140	5.5	-131	111	2	1
20081021	ANAHEIM	TORONTO	-140	130	5.5	110	-130	3	2
20081021	ATLANTA	TAMPA BAY	145	-155	5.5	-115	-105	2	3
20081021	BOSTON	BUFFALO	178	-191	5.5	-111	-109	2	3
20081021	VANCOUVER	COLUMBUS	127	-137	5.5	115	-135	2	4
20081021	WASHINGTON	CALGARY	117	-127	6	-132	112	1	2
20081022	DALLAS	NEW JERSEY	130	-140	5	-110	-110	0	5
20081022	DETROIT	ST. LOUIS	-142	132	5.5	110	-130	4	3
20081022	EDMONTON	CHICAGO	136	-146	5.5	-121	101	0	3
20081022	FLORIDA	OTTAWA	197	-217	5.5	-122	102	3	1
20081022	SAN JOSE	PHILADELPHIA	-116	106	5.5	107	-127	7	6
20081023	BUFFALO	MINNESOTA	130	-140	5.5	145	-165	4	3
20081023	CALGARY	NASHVILLE	113	-123	5.5	-104	-116	5	3
20081023	CAROLINA	PITTSBURGH	142	-152	5.5	-132	112	1	4
20081023	DALLAS	NY ISLANDERS	-120	110	5.5	123	-143	5	3
20081023	EDMONTON	COLORADO	141	-151	5.5	-146	126	1	4
20081023	TORONTO	BOSTON	173	-183	5.5	103	-123	4	2
20081023	WASHINGTON	PHOENIX	-113	103	5.5	-128	108	1	2
20081024	ANAHEIM	OTTAWA	135	-145	5.5	125	-145	4	3
20081024	ATLANTA	DETROIT	245	-265	5.5	117	-137	3	5
20081024	LOS ANGELES	ST. LOUIS	157	-167	5.5	-147	127	4	0
20081024	NY RANGERS	COLUMBUS	-105	-105	5	-142	122	3	1
20081024	PHILADELPHIA	NEW JERSEY	155	-165	5.5	119	-139	6	3
20081024	SAN JOSE	FLORIDA	-152	142	5	-126	106	3	4
20081025	ANAHEIM	MONTREAL	165	-175	5.5	125	-145	6	4
20081025	ATLANTA	BOSTON	172	-182	5.5	104	-124	4	5
20081025	BUFFALO	COLORADO	120	-130	5.5	-137	117	1	2
20081025	CALGARY	PHOENIX	-106	-104	5.5	-115	-105	4	1
20081025	CAROLINA	NY ISLANDERS	-136	126	5.5	102	-122	4	3
20081025	COLUMBUS	MINNESOTA	170	-180	5	-122	102	1	2
20081025	DETROIT	CHICAGO	-113	103	5.5	-101	-119	6	5
20081025	EDMONTON	VANCOUVER	116	-126	5.5	-102	-118	3	6
20081025	FLORIDA	ST. LOUIS	123	-133	5.5	-102	-118	0	4
20081025	LOS ANGELES	NASHVILLE	162	-172	5.5	-116	-104	4	5
20081025	NEW JERSEY	PHILADELPHIA	-105	-105	5.5	102	-122	2	3
20081025	OTTAWA	TORONTO	-116	106	5.5	-126	106	2	3
20081025	PITTSBURGH	NY RANGERS	122	-132	5.5	125	-145	2	3
20081025	SAN JOSE	TAMPA BAY	-120	110	5.5	-104	-116	3	0
20081025	WASHINGTON	DALLAS	116	-126	5.5	-111	-109	6	5
20081027	ANAHEIM	COLUMBUS	-114	104	5	-110	-110	3	2
20081027	BOSTON	EDMONTON	106	-116	5.5	-114	-106	1	0
20081027	CHICAGO	MINNESOTA	130	-140	5	-110	-110	2	3
20081027	DETROIT	LOS ANGELES	-162	152	6	120	-140	4	3
20081027	NY RANGERS	NY ISLANDERS	-182	172	5.5	133	-153	4	2
20081027	OTTAWA	BUFFALO	127	-137	6	119	-139	5	2
20081028	BOSTON	VANCOUVER	166	-176	5.5	101	-121	1	0
20081028	CAROLINA	MONTREAL	172	-182	5.5	-125	105	2	3
20081028	COLORADO	CALGARY	125	-135	5.5	-135	115	0	3
20081028	NASHVILLE	WASHINGTON	105	-115	5.5	-110	-110	3	4
20081028	PHILADELPHIA	ATLANTA	-125	115	5.5	-146	126	7	0
20081028	PITTSBURGH	SAN JOSE	152	-162	5.5	120	-140	1	2
20081028	TAMPA BAY	TORONTO	116	-126	5.5	-104	-116	3	2

Tab separated.

[durito]

Thank you!

[BigCap]

NBA data for 2003-2007, min 21 games played, bet on dog if win % better than opponent, using no-vig money line:

2003:
wins: 45, +66.7u
losses: 89, -89u

2004:
wins: 87, +129.3u
losses: 151, -151u

2005:
wins: 87, +137.8u
losses: 155, -155u

2006:
wins: 93, +140.8u
losses: 155, -155u

2007:
wins: 99, +154.5u
losses: 149, -149u

[wannabet]

Very nice. Basketball would turn out well over the past couple of years I'm thinking, don't have any data, but it does sound very good. Also very easy to play, no work or research involved, but you have to find the best lines, which isn't easy. Well, I wanted to comment, thanks.

[Munson15]

NBA data for 2003-2007, min 21 games played, bet on dog if win % better than opponent, using no-vig money line:

2003:
wins: 45, +66.7u
losses: 89, -89u

2004:
wins: 87, +129.3u
losses: 151, -151u

2005:
wins: 87, +137.8u
losses: 155, -155u

2006:
wins: 93, +140.8u
losses: 155, -155u

2007:
wins: 99, +154.5u
losses: 149, -149u

Thanks for doing the research, there is an NFL system I read years ago which advocates taking the points in the same scenario, but I never tracked it much above 50% either.

[Igetp2s]

As mentioned in a prior post, how do injuries affect your system?

If a star player or starting QB is out, of course the team will be an underdog even with a better record. How is that finding value?

If you don't take significant injuries into account, it seems you're comparing apples to oranges.

[Peep]

New England tonight +6.5 should fit the system perfectly.

[rerics]

How do you handle NHL records? For example, Columbus (7-6-2, 16 pts) is playing Phoenix (7-6-0, 14 pts) on Wednesday. Phoenix is the underdog, but are they equal to Columbus with identical W-L records, or does the number of points/overtime losses make Columbus the team with the better/worse record? I know the season isn't 25% old yet, but I'm asking just as a hypothetical.

[Munson15]

Did this thing die a quick death? Does anyone have any data from this year, or can I assume that the results were bad?

[hajune]

here it is..

1. Play only on Professional Sports.

2. Play every dog with equal or better record than opponent.

3. Begin play only after 25% of the season has been played.

4. risk the same amount on each wager.

over thousands of wagers, here are the results.

an average of (+1.37) payback on winners.
an average winning rate of (45.5%)

system breaks even at just over 42%.

should net you a positive (+7 units) for every 100 wagers.

doubt me. look it up...

MLB Playoffs: 3-3
NFL week 5: 3-2

remember, 25% of the season must be played. only use for the 4 major pro-sports.

track it and let me know how you do.......... play all underdogs with equal or better records.

Mr. Progression

It works. You are on the right track.

[Munson15]

It works. You are on the right track.

What makes you say it works? Do you have any data? I don't know what to make of this and I'm sure there would be many twists and turns in a system involving this many plays. Elaborate, please. The post above regarding NBA indicates a very poor performance over several years' time.