Dogs: Play the ML or Take The Points??

[Bad_Cat]

Hmm, I think I mentioned this in a post a few days ago in Harry's thread.

http://www.therxforum.com/showthread.php?t=1073123

Post #17 in the above thread. The Eagles lost the game but buying the hook was dumb if you follow what I said in the post with betting dogs. I thought harry seemed more intelligent than that. None of the games he posted with buying hooks and paying extra came into play.

Just food for thought

 

[Mobdeeper7239]

Fair enough. I'll run those numbers. Dogs that cover but lose the game.

Dogs up to +3, that lost the game, have covered 8.3% of the time.

D and line <= 3 and L and season >= 2007
SU: 0-471-0 (-11.09, 0.0%) Teaser Records ATS: 36-396-39 (-8.72, 8.3%) avg line: 2.4 +6: 240-215-16 (52.7%) -6: 0-471-0 (0.0%) +10: 300-161-10 (65.1%) -10: 0-471-0 (0.0%) O/U: 240-223-8 (1.26, 51.8%) avg total: 44.0 +6: 154-314-3 (32.9%) -6: 327-134-10 (70.9%) +10: 106-359-6 (22.8%) -10: 374-92-5 (80.3%)

3.5 to 5.5 cover 24%

D and 3.5 <= line <= 5.5 and L and season >= 2007
SU: 0-413-0 (-11.65, 0.0%) Teaser Records ATS: 99-310-4 (-7.40, 24.2%) avg line: 4.3 +6: 222-183-8 (54.8%) -6: 0-413-0 (0.0%) +10: 280-131-2 (68.1%) -10: 0-413-0 (0.0%) O/U: 202-199-12 (0.43, 50.4%) avg total: 44.1 +6: 124-281-8 (30.6%) -6: 278-127-8 (68.6%) +10: 84-326-3 (20.5%) -10: 322-87-4 (78.7%)

6-7 cover 27.7%

D and 6 <= line <= 7 and L and season >= 2007
SU: 0-293-0 (-14.38, 0.0%) Teaser Records ATS: 78-204-11 (-7.82, 27.7%) avg line: 6.6 +6: 143-145-5 (49.7%) -6: 0-291-2 (0.0%) +10: 179-110-4 (61.9%) -10: 0-293-0 (0.0%) O/U: 150-138-5 (2.02, 52.1%) avg total: 44.6 +6: 107-183-3 (36.9%) -6: 210-80-3 (72.4%) +10: 79-211-3 (27.2%) -10: 235-54-4 (81.3%)

7.5 to 10 cover 38.5%

D and 7.5 <= line <= 10 and L and season >= 2007
SU: 0-325-0 (-14.80, 0.0%) Teaser Records ATS: 124-198-3 (-6.07, 38.5%) avg line: 8.7 +6: 181-140-4 (56.4%) -6: 22-296-7 (6.9%) +10: 215-104-6 (67.4%) -10: 0-325-0 (0.0%) O/U: 153-164-8 (-0.06, 48.3%) avg total: 44.0 +6: 90-228-7 (28.3%) -6: 225-97-3 (69.9%) +10: 55-265-5 (17.2%) -10: 262-61-2 (81.1%)

Dogs of 10.5+ cover 40.8%

D and 10.5 <= line and L and season >= 2007
SU: 0-214-0 (-16.93, 0.0%) Teaser Records ATS: 87-126-1 (-3.84, 40.8%) avg line: 13.1 +6: 124-83-7 (59.9%) -6: 50-160-4 (23.8%) +10: 152-58-4 (72.4%) -10: 15-194-5 (7.2%) O/U: 111-101-2 (0.88, 52.4%) avg total: 44.3 +6: 76-138-0 (35.5%) -6: 149-61-4 (71.0%) +10: 51-157-6 (24.5%) -10: 170-40-4 (81.0%)

Biz, the points mattering percentages you gave above include dogs who cover but don't win over the last 5 years, correct?

 

[Biz]

These numbers are taken straight from the Killer Sports data base. The numbers are the numbers, whether you agree with them or not.

+120 isn't an inflated number for +3 dogs. Go to any site and look at ML prices for this range.

Tool,

You are making a lot of assertions without any data or evidence to back up your assertions. You stated its a "known assumption" that 2.5 dogs win at 45%. The numbers show otherwise.

I'm all for a reasoned debate, but instead of assertions and assumptions, bring some actual facts or data to bolster your claims.

I'm not telling anyone to start betting ML dogs, I simply said its something to think about.

I also said to track your dog bets, and compare the results spread vs ML.

 

[Biz]

Biz, the points mattering percentages you gave above include dogs who cover but don't win over the last 5 years, correct?

Since 2007 correct.

These teams lost the game, the ATS numbers are for those teams.

 

[Biz]

6-7 pt dogs that Lose cover 27.7%

D and 6 <= line <= 7 and L and season >= 2007
SU: 0-293-0 (-14.38, 0.0%) Teaser Records ATS: 78-204-11 (-7.82, 27.7%) avg line: 6.6 +6: 143-145-5 (49.7%) -6: 0-291-2 (0.0%) +10: 179-110-4 (61.9%) -10: 0-293-0 (0.0%) O/U: 150-138-5 (2.02, 52.1%) avg total: 44.6 +6: 107-183-3 (36.9%) -6: 210-80-3 (72.4%) +10: 79-211-3 (27.2%) -10: 235-54-4 (81.3%)

6-7 pt dogs that cover, winning 55.6% at avg +260 ML

D and ats margin > 0 and season >= 2012 and 6 <= line <= 7
SU: 50-40-1 (4.05, 55.6%) Teaser Records ATS: 91-0-0 (10.60, 100.0%) avg line: 6.5 +6: 91-0-0 (100.0%) -6: 51-39-1 (56.7%) +10: 91-0-0 (100.0%) -10: 34-52-5 (39.5%) O/U: 45-43-3 (1.89, 51.1%) avg total: 45.6 +6: 31-59-1 (34.4%) -6: 66-25-0 (72.5%) +10: 25-65-1 (27.8%) -10: 71-19-1 (78.9%)

Anyone that wants to do the math knock yourself out.

I only started this thread because MCM asked about it.

These are the facts, they aren't assertions. GL

 

[Mobdeeper7239]

Since 2007 correct.

These teams lost the game, the ATS numbers are for those teams.

Cool. Thanks for clarification.

 

[joel0421]

The ending number yesterday changed the outcome of 1 out of 14 games, I believe. In other words give the pts with the favorite and take the ml with the dog when betting a side in the NFL.

 

[tool21]

I can list sources all day.

Go to Google and search for nfl spread converting to win%. A spread of +2.5 wins ~45% of the time in the NFL. Like i stated.

Here's a source: http://wizardofodds.com/games/sports-betting/nfl/
Here's a source: https://www.boydsbets.com/nfl-spread-to-moneyline-conversion/

Now convert that win % (45) to a moneyline using the links below. You will notice that these converters don't include the juice. You will add or subtract 10 from each side. So you get +110 on the one, and +112 on the other. Not +120.

Here's a source: https://www.boydsbets.com/money-line-conversion-chart/
Here's a source: http://www.covers.com/sportsbetting/money_lines.aspx

If you can get your database to tell me what the ACTUAL moneyline for those games are, I'll believe you 100%. Otherwise your just taking an average on what you think the moneyline is. I think that's your error.

 

[MotorCityMadman]

2016: 25.3%
2015: 31.1%
2014: 38.2%
2013: 35.9%
2012: 30.8%

This is the percentage of games that points made a difference regardless of spread.

There were a lot of close games this season where a 1/2 point or some can make a difference. Using this site oddsshark this is what I came up with. I went through the lines and scores one by one.



Week 1 2-12-2
Week 2 0-16
Week 3 1-15
Week 4 0-15
Week 5 2-12
Week 6 3-12
Week 7 1-13-1
Week 8 2-11
Week 9 1-11-1
Week 10 0-15
Week 11 1-12-1
Week 12 3-12-1
Week 13 1-14
Week 14 1-14

17 games the spread did matter
184 games did not
There were 7 games that would be considered a tie.

So I come up with 8.17% for 2016.

 

[MotorCityMadman]

Point Spread	Money Line
-2	-130/+110
-2.5	-140/+120
-3	-155/+135
-3.5	-175/+155
-4	-200/+170
-4.5	-220/+180
-5/-5.5	-240/+190
-6	-270/+210
-6.5	-300/+220
-7	-330/+250
-7.5/-8/-8.5	-360/+280
-9/-9.5	-400/+300
-10	-450/+325

 

[joel0421]

The bottom line to these numbers is basically pick the team that you have capped to win the game, regardless of the line, and then play the odds accordingly

 

[Mobdeeper7239]

There were a lot of close games this season where a 1/2 point or some can make a difference. Using this site oddsshark this is what I came up with. I went through the lines and scores one by one.



Week 1 2-12-2
Week 2 0-16
Week 3 1-15
Week 4 0-15
Week 5 2-12
Week 6 3-12
Week 7 1-13-1
Week 8 2-11
Week 9 1-11-1
Week 10 0-15
Week 11 1-12-1
Week 12 3-12-1
Week 13 1-14
Week 14 1-14

17 games the spread did matter
184 games did not
There were 7 games that would be considered a tie.

So I come up with 8.17% for 2016.

A tie/push is considered points coming into play.

 

[MotorCityMadman]

A tie/push is considered points coming into play.

Okay I will buy that. Make it 11.53%.

 

[Biz]

I can list sources all day.

Go to Google and search for nfl spread converting to win%. A spread of +2.5 wins ~45% of the time in the NFL. Like i stated.

Here's a source: http://wizardofodds.com/games/sports-betting/nfl/
Here's a source: https://www.boydsbets.com/nfl-spread-to-moneyline-conversion/

Now convert that win % (45) to a moneyline using the links below. You will notice that these converters don't include the juice. You will add or subtract 10 from each side. So you get +110 on the one, and +112 on the other. Not +120.

Here's a source: https://www.boydsbets.com/money-line-conversion-chart/
Here's a source: http://www.covers.com/sportsbetting/money_lines.aspx

If you can get your database to tell me what the ACTUAL moneyline for those games are, I'll believe you 100%. Otherwise your just taking an average on what you think the moneyline is. I think that's your error.

Tool,
Its impossible to get every ML for the games. I went to Heritage and looked at past ML for dogs in this range. Some were more than 120 and some were in the 140-150 range. I took a reasonable odds range of +120.

Can't open your links on my phone, will look when I get home.

 

[MotorCityMadman]

Biz if you are looking for the money line this site goes back to 2007.

http://www.footballlocks.com/nfl_odds.shtml

 

[Biz]

I can list sources all day.

Go to Google and search for nfl spread converting to win%. A spread of +2.5 wins ~45% of the time in the NFL. Like i stated.

Here's a source: http://wizardofodds.com/games/sports-betting/nfl/
Here's a source: https://www.boydsbets.com/nfl-spread-to-moneyline-conversion/

Now convert that win % (45) to a moneyline using the links below. You will notice that these converters don't include the juice. You will add or subtract 10 from each side. So you get +110 on the one, and +112 on the other. Not +120.

Here's a source: https://www.boydsbets.com/money-line-conversion-chart/
Here's a source: http://www.covers.com/sportsbetting/money_lines.aspx

If you can get your database to tell me what the ACTUAL moneyline for those games are, I'll believe you 100%. Otherwise your just taking an average on what you think the moneyline is. I think that's your error.

+2.5 does not win 45% of the time. That number is the expected win percentage, not the actual win percentage. Its won 50.4% since 2000. I've posted the results of the actual games. Lets just agree to disagree, lets try this.

The more relevant number........ How many times does the team win SU when they cover, and how many times does the team not cover??

Lets agree that +120 is too high and use +110. Here are the results of dogs up to 3.

DOGS UP TO +3

D and ats margin > 0 and season >= 2012 and line <= 3
SU: 202-22-2 (8.99, 90.2%) Teaser Records ATS: 226-0-0 (11.29, 100.0%) avg line: 2.3 +6: 226-0-0 (100.0%) -6: 147-63-16 (70.0%) +10: 226-0-0 (100.0%) -10: 95-121-10 (44.0%) O/U: 112-104-10 (0.50, 51.9%) avg total: 45.4 +6: 66-159-1 (29.3%) -6: 158-65-3 (70.9%) +10: 49-175-2 (21.9%) -10: 181-44-1 (80.4%)

So 90% wins straight up. Using an average of +110: 90 x 1.1=9.9 units won. 10 games lost x 1.1 (juice) = 11 for a net loss of 1.1. Now we need to figure out how much was saved in juice playing ML instead of losing -110 on the losses.

D and season >= 2012 and line <= 3
SU: 202-241-2 (-0.78, 45.6%) Teaser Records ATS: 226-201-18 (1.53, 52.9%) avg line: 2.3 +6: 333-103-9 (76.4%) -6: 147-282-16 (34.3%) +10: 367-72-6 (83.6%) -10: 95-340-10 (21.8%) O/U: 218-212-15 (0.51, 50.7%) avg total: 45.4 +6: 126-316-3 (28.5%) -6: 307-130-8 (70.3%) +10: 96-346-3 (21.7%) -10: 355-85-5 (80.7%)

Over this same time frame, these dogs did not cover 201 games. At -110 that is a juice loss of 20.1 units. 20.1 in juice savings less the 1.1 lost previously is a net of 19.

Is this right, or am I missing something? Serious question. My head is starting to hurt screwing around with this and I need to get on to other things.

If this is correct, then we need to add the saved juice from the prior calculations, which would make the results look better for each of the other odds ranges.

Last thing.......Great discussion with you Tool, appreciate the dialogue.

 

[Bad_Cat]

Good post. Insightful for those who don't understand the value of money line dogs.

 

[Biz]

Dogs that won outright today:
Miami -4.5/+190
Jacksonville +4/+180
San Francisco +6/+220
Arizona +8/+305
Cleveland +4/+170

Every dog that covered today also won outright.

Spread: 5-0 (+5 units)
ML: 5-0 (+10.65 Units)

 

[chonce]

Interesting post. The tricky part is there are infinitie variables one could dissect -- maybe exceptions could be proven on games with a 3-point spread (haven't tested it, just a possible variable that might matter).


I did similar math trying to figure out if I should buy points. It seems one is better off just never buying the points, but I still do it from time to time. But playing a dog on the ML is more like selling points.

Anyway, I suspect there is probably some extensive research on this somewhere -- probably purchased/commissioned by large books where they have resources to test on huge sample sizes and account for so many variables. Would love to read it.

I wonder if you could find some of this in quantitative analysis student thesis papers. I am sure more than a few have written about vegas, and tested models.

 

[chonce]

Here is a paper from a Statistics Department (dated, but it mostly just to demonstrate) where they run models on potentially finding profitable strategies. I would be willing to bet one could find many more on Google Scholar, or if we have anyone who is in a college, or teaches, or works in journalism, Lexis Nexus.

Would be cool to compile this research: http://www.ylikerroin.com/file/Complete.pdf

"In this thesis, we investigated the benefits of using a statistically acceptable model as a support of one’s decisions both from bookmaker's and punter's point of views and concluded that it would have potential to improve their performance.

The model proposed here was proven to be useful for football betting purposes. The validation indicated that it quite effectively captured many aspects of the game and...."