What is the value of a teaser?

[ronaldn]

As a new poster here but a longtime player I would like to submit the following.




<?xml:namespace prefix = o ns = "urn:schemas-microsoft-comfficeffice" /><o></o>

I have never seen anyone try to quantify the value of the extra 4.5 points in playing a teaser so I decided to do a little research myself. I wanted to find out if they would be a suitable handicapping tool to add to my armamentarium. Heaven knows, we need all the firepower we can muster when we do battle with the “books”. I am well aware that teasers are considered “sucker bets” by most people in the industry.

<o></o>

Before I begin I would like to let you know that I get all my figures from a computer program that I wrote in Basic Programming Language that I maintain and update daily. To get the results I just have to program in the parameters that I am interested in and then let the computer do the work.

<o></o>

My methodology is this. I first determined what the percentages were for certain criteria and then what the percentages would be if I teased them. Like, what are the win percentages ATS for home teams this year. I then found out what the percentages would be if I teased every home team this year and subtracted the actual ATS percentage from the teased percentage. While the theoretical ATS percentage is 50%, the actual percentage may be slightly different. And I do mean slightly as the linemaker does an amazing job of keeping everything even. I have tracked this over many years and those guys are GOOD, really GOOD.

<o></o>

This is for a 4.5 point 2 team teaser for sides and 6.5 for totals paying even odds at the sportbook I use.

<o></o>

My results are as follows for games up to and including 2/26/05:

<o></o>

There will be slight differences in the amount of plays because of pushes.

<o></o>

Home teams have covered 399 times and failed 396 plus pushes which I just ignore. (Didn’t I tell you how good the linemaker is?). Or 50.2%

If you teased every home team….534 – 273 or 66.2% for a net increase of 16%.

If you teased both sides in the same game you would have won both 31.8% of the time. I found that interesting but not particularly germane to the research. I don’t even know if the books will let you do that! LOL.

<o></o>

Home favorites were 281 – 283 or 49.8%. I told you they are good.

If teased 370 – 200 or 64.9%. An increase of 15.1%

<o></o>

Overs won 414 times and Unders 383. Wow, the linemaker is slipping here. That’s 51.9%. By teasing the over….547 – 262 or 67.6 %. An increase of 15.7%

<o></o>

I also teased a certain situation that I have been using for years and it increased the win% from 56.3% to 71.85%. An increase of 15.55%. I did this with several other situations as well, with similar results with a high of 19% and a low of 13%.

<o></o>

As you can see from these results the value of the 4.5 points is approximately a 15 to 16 percent increased chance to win on each leg of the teaser. Being you need two wins to win a teaser you need approximately a 71% win rate on each leg to theoretically break even. (.71 x .71 = .504) I have been experimenting with this over the last month and a half and have won 19 teasers, lost 15 with 1 push. Being they pay at even odds that is a slight profit. I combine it with a modest betting system that has improved on that figure somewhat. Most wins in a row are 5. Most consecutive losses so far are 4. I am hopeful that the teaser element will moderate losing streaks. So far it has but that may be just wishful thinking over the long run.

<o></o>

With the small number of outcomes to date this is certainly not evidence of a successful system. Could be just a mild hot streak. If these results can be maintained over two seasons, I think I may have something. I will continue working on it.

<o></o>

Please note. I am not advising anyone to play teasers but just wanted to share my research with you. This is definitely experimental.

<o></o>

If anyone sees any errors or inconsistencies in my methodology, please bring them to my attention. Insights, questions and comments are welcome.

<o></o>

Ron

 

[Baker]

Interesting study, I think the slight predilection towards the over may be caused by the ~6% of games that go into overtime.


There is some data I am looking for and from the sounds if it you may have it readily available. Would your program be able to spit out the average historical difference in each games opening line and the final score, with all differences treated as a positive number diverging from zero? This could help me in some research I am doing.


Baker

 

[ronaldn]

There is some data I am looking for and from the sounds if it you may have it readily available. Would your program be able to spit out the average historical difference in each games opening line and the final score, with all differences treated as a positive number diverging from zero? This could help me in some research I am doing.


Baker

Baker, I am not quite sure I understand your question. If say the line is -1, do you want to know the average difference in the final score between the home team and the visitor for that pointspread? I can do that for each pointspread. Please clarify. My program however only has one pointspread entered into it for each game and that was the ptsprd available at the time I made my plays or else the closing line.

Ron

 

[Baker]

I apologize I wasn't clear with my question. I am referring to totals.

Examples:
Ex 1-opening line is 190 with a final score of 202 equates to a difference of 12.
Ex 2-opening line is 195 with a final score of 180 equates to a difference of 15.

Average would equal 13.5 points.

I am basically seeking to find out how close the opening line on totals are to the actual final score, preferably with season by season statistics with playoffs calculated seperately.

 

[Zapster]

Baker........

I have keep track of these things. I have the spreadsheet at home for 2005. Meanwhile, here are some numbers from last season:

These are only for games in which I had a posted totals play:

11/03: 12.242
12/03: 12.341
01/04: 12.671

The above figures are cumulative, as in 01/04 includes 11/03 and 12/03. There are about 400 games in the sample.

The above numbers reflect the average of the absolute values of the difference between the line and the actual score. The sampling may be a bit "skewed" because it doesn't include the games I regarded to be "no play"

For the games since 1/1/05 I have kept track of every single game (not just the games I posted plays for) and I will post that number when I get home.

 

[Two-Niner]

Interesting thread. Can anyone tell me what percentage of games land within 5 points of the total in either direction?

 

[Zapster]

Ok.....starting January 1st and through last night I have 403 games on my sheet. The difference between the posted line and the actual score has averaged 13.5 through those games. 103 of the 403 games were within 5.....and I counted the ones that were exactly 5 as being within 5.

 

[ronaldn]

Interesting thread. Can anyone tell me what percentage of games land within 5 points of the total in either direction?

I can get that info out of my program. Give me a day or 2.
Ron

 

[Two-Niner]

Thanks alot. Around 25% is what I was thinking just skimming over some team logs. I'd assume as the total increases (say above 200) the % of games within 5 of the total decreases?

 

[ronaldn]

Zapster, thanks for jumping in with this. My figures come out with identical percentages to yours. Out of 821 games, 207 fell within 5 points of the line either way. That figure includes exactly 5 points. Almost exactly 25%

 

[ronaldn]

Thanks alot. Around 25% is what I was thinking just skimming over some team logs. I'd assume as the total increases (say above 200) the % of games within 5 of the total decreases?

When the totals number was equal to or greater than 200 then 53 out of 207 games were within 5 pts. or 25.6%
Number that fell within 6.5 pts (the teaser number) was 63 out of 207, or 30.4% which would correlate with my earlier teaser figures for totals. 30.4 / 2 = 15.2%.

 

[Zapster]

RonaldN........

This is a great thread. Thanks for starting it up.

Personally I have never put much stock in teasers that involve totals, but I always say I am very much in favor of whatever works.....

Please keep the posts coming. You will be a great addition to this forum !!

 

[Baker]

Thanks for the information. The 25% falling within 0-5 of the line falls in line with the limited data I have captured. This leads in to my next question.

Next question(s):

In relation to how the final score relates to the totals line, according to the data does this value fall within a normal distribution?

If so is the mean 0?

If the mean isn't 0 what is it?

In the limited amount of data I have the difference of the line and the final score does fall within a normal distribution (with a mean of 0) with the exception of one abnormailty. I would be very interested to know if anyone else sees the same abnormality.

 

[Kruser6]

You guys sat in the front row during class and raised your hands often didnt you!!!!???


:monsters- This is how I feel. :lolBIG:

 

[ronaldn]

RonaldN........

This is a great thread. Thanks for starting it up.

Personally I have never put much stock in teasers that involve totals, but I always say I am very much in favor of whatever works.....

Please keep the posts coming. You will be a great addition to this forum !!

Zapster, thanks for the welcome and the kind words. I too do not like playing teasers on totals. In fact, I rarely if ever play regular totals. I have been quite successful over the years in capping sides but I have never been able to get a handle on totals. In short, I dont have a clue as to how to predict totals. I keep looking but I have never seen anyone post something that works. I'm sure it's possible, I just dont know how.

 

[ronaldn]

Thanks for the information. The 25% falling within 0-5 of the line falls in line with the limited data I have captured. This leads in to my next question.

Next question(s):

In relation to how the final score relates to the totals line, according to the data does this value fall within a normal distribution?

If so is the mean 0?

If the mean isn't 0 what is it?

In the limited amount of data I have the difference of the line and the final score does fall within a normal distribution (with a mean of 0) with the exception of one abnormailty. I would be very interested to know if anyone else sees the same abnormality.

Baker, I will try to do some more research on this. I'm sure the 25% will fall within a normal distribution. I think maybe the question should be, how many points above or below the line will fall within a Standard Deviation of one. It's been a long time since taking Statistics so please forgive me if I dont grasp things right away. Might not get it done today as I'm supposed to go babysit my grandchildren but maybe with all the snow we just got my daughter will call it off. I think the Mean may be different for each total number. While theoretically the Mean should be zero I'm sure that in actuality it may show a bias the further away from the average score of all games the totals line is. I'll see if I can determine this but it may be a bigger job than I can handle. This year the average score per game is approx 193.7, up over 7.5 points from last year.

 

[Baker]

Ronald-

Thanks for your help. If you (or Zapster) have a chance could you post the totals lines and final scores for the current season?

 

[dolphin]

Football is quite different, because of the specific value of certain points - 3,4,6,7 especially - so that teasers which capture those points up or down are virtually the only ones worth considering (although not profitable without considerable screening).

Although people go on streaks with basketball teasers at times, I would contend that during the vast majority of good teaser runs, one would have done better with a $ comparable straight wager approach. Teasers have a propensity to make a handicapper lazy, as well.

Dolphin

 

[ronaldn]

Football is quite different, because of the specific value of certain points - 3,4,6,7 especially - so that teasers which capture those points up or down are virtually the only ones worth considering (although not profitable without considerable screening).

Although people go on streaks with basketball teasers at times, I would contend that during the vast majority of good teaser runs, one would have done better with a $ comparable straight wager approach. Teasers have a propensity to make a handicapper lazy, as well.

Dolphin

Thanks for your input. What you say makes a lot of sense.
Many years ago, the fellow that cut my hair was also a bookie, a middle man, before he became so rich that he no longer cuts hair. LOL. He told me that whenever anyone placed a teaser wager with him he would place it with his contact as a parlay bet. Said that most of the time if the teaser won, so would the parlay and he would pocket the difference. In my opening post I said that I was 19 - 15 betting teasers. 11 of them I needed the extra 4.5 points to be winners and 8 of them would be parlay winners.
For the bookie. The 15 outright losers was just a pass thru of money for him. Of the 8 parlay winners he made 1.6 units. (2.6 - 1 unit he paid out to the bettor). The 11 that needed the teaser points to win he had to pay out 11 units. 8 x 1.6 = 12.8 minus the 11 he paid out for a profit of 1.8 units. With practically no risk on his part. We bettors cant do that tho because we dont have anyone paying for our straight thru losses. If I was playing parlays the whole way I would have won only 8 out of 34 which would have been a loser.

 

[dolphin]

beware of small samples...nuances appear which have little to do with logic/what you can expect over the long run...I bascally propose straight wagers as opposed to the teasers...ESPECIALLY in college, where the emotions run so high that deviations from the norm are greater, despite the lower scoring...