Help with Teasers

[wrongturn]

bill will have a perfect week if hou covers +10.5 today.

[Peregrine Stoop]

bill will have a perfect week if hou covers +10.5 today.

and would be quite dumb to put those in a teaser

[rfr3sh]

why?
RD4.5 to 6, 326-109-2, for 74.9%

[dimon]

teasers done bad, and looking forward for next week now

[tomcowley]

why?
RD4.5 to 6, 326-109-2, for 74.9%

Teams with odd numbers of both vowels and consonants against teams with an animal mascot who ran an odd number of running plays the previous week 320-265 54.7%. (Teams with an even number of either vowels or consonants, 600-600 50%, against teams with non-animal mascots, 300-300 50%, against teams who ran an even number of running plays the previous week, 300-300 50%)

[rfr3sh]

Teams with odd numbers of both vowels and consonants against teams with an animal mascot who ran an odd number of running plays the previous week 320-265 54.7%. (Teams with an even number of either vowels or consonants, 600-600 50%, against teams with non-animal mascots, 300-300 50%, against teams who ran an even number of running plays the previous week, 300-300 50%)

i posted Bill the Cops teaser data he posted

[donjuan]

why?
RD4.5 to 6, 326-109-2, for 74.9%

For starters, this:

Tenn +10.5 -234
Minn +10.5 -220

Beyond that, if you re-read the thread and can't figure out why BTC teasers (RD 4.5-6) are only for droolers, you probably can't be helped.

[durito]

Greek has texans +10.5 -220. Should probably tease instead though so you can pick up the 11 for 40 cents.

[wrongturn]

also these two teams are currently tied at top, so a tight game is expected.

[rfr3sh]

Guess I should read teh whole thread before chipping in
But I now see how playing the ALT lines can produce small edges rather then betting the teasers
but is it fair to assume most sites Alt lines are -EV because they can't offer pinny prices

[rfr3sh]

Greek has texans +10.5 -220. Should probably tease instead though so you can pick up the 11 for 40 cents.

5Dimes also has this price

[wrongturn]

Interesting stats: Road dog +5.5 covers +10.5 71.85% of time. -220 implies 68.75% break-even. So it has some value.

[durito]

Interesting stats: Road dog +5.5 covers +10.5 71.85% of time. -220 implies 68.75% break-even. So it has some value.

yes pinny and greek just give away free money for everyone

[wrongturn]

alright, no free money after all

[LegitBet]

Is there a diffinative source for teasers
Book?

[terpkeg]

I just read this entire thread and have to agree with the Math guys here. Unless, is it actually possible that there is a reason why +4.5 to +6 rd win at a higher rate than +3-+4 rd at the same +10.5-+12?

I am not saying that a move from +4.5 to +4 or vice versa will effect a teams effort. However, what if +4.5 - +6 RD consist of teams with the perfect pyscological or motivational recipes for games to fall in the teaser window.

In this range maybe you will have teams that will be behind in the power rankings, but are playing with a high motivation. Maybe because they are still in the playoff chase. However, you have favorites who go into a home game, believing they are the better team. This recipe may lead to mental distractions for the favorite if they go up by two scores against a motivated opponent who can take advantage and cut into the lead late.

Whereas, maybe games that are generally +3 to +4 pit teams that are closer together in strength causing the favorite to stay mentally focused for the duration of the game. And, teams that are +4.5 to +6 are weak for one of many reasons and they may either emotionally fold when they fall behind by two scores or be playing a team that is actually far superior.

Chances are this is just variance and the numbers will work themselves out, but maybe there is a reason this happens and Pinny is still hanging soft numbers because people are failing to realize it.

[Peregrine Stoop]

Is there a diffinative source for teasers Book?

nope... that is why these threads exist... to debate the merits

what many of us have a problem with is the BillTheCop road dog subset appears just to be datamining. Why?
1) It has done ridiculously well Against the spread. Anyone who believed in the subset would have made much much more money just doing spread bets. From an estimated growth standpoint, it would be silly to bet them in teasers rather than ATS.
2) The subset of 'better' road dogs has done worse covering the +10.5 to +11.5. What about the nature of football would lead to a road dog of 3.5 covering +10.5 less often than a road dog of 5? See, there is no theory.
3) Pinnacle, 5dimes, thegreek would not all be laying such a positive alternative spread line if these things really covered that much. OK, I guess they all could be wrong, but it sure as hell not likely.
4) If you blindly looked at all the spread subsets that existed over that period, it's not a remote chance that one just happened to run very hot.


Always keep in mind that teaser bets are a bet on the range. You are gaining an extra X pts in exchange for Y payout. Does X justify the Y payout compared to other available lines?

[Peeig]

nope... that is why these threads exist... to debate the merits

what many of us have a problem with is the BillTheCop road dog subset appears just to be datamining. Why?
1) It has done ridiculously well Against the spread. Anyone who believed in the subset would have made much much more money just doing spread bets. From an estimated growth standpoint, it would be silly to bet them in teasers rather than ATS.
2) The subset of 'better' road dogs has done worse covering the +10.5 to +11.5. What about the nature of football would lead to a road dog of 3.5 covering +10.5 less often than a road dog of 5? See, there is no theory.
3) Pinnacle, 5dimes, thegreek would not all be laying such a positive alternative spread line if these things really covered that much. OK, I guess they all could be wrong, but it sure as hell not likely.
4) If you blindly looked at all the spread subsets that existed over that period, it's not a remote chance that one just happened to run very hot.


Always keep in mind that teaser bets are a bet on the range. You are gaining an extra X pts in exchange for Y payout. Does X justify the Y payout compared to other available lines?

Maybe, but if they were, it wouldn't be widely available near close...........someone would surely hammer that line into shape.

[terpkeg]

Maybe, but if they were, it wouldn't be widely available near close...........someone would surely hammer that line into shape.

But what if all the sharpest cappers, who are the types that would be able to dycifer whether the line was off and would be betting alt lines, are under the impression that it is not off since they think the results of BTC's sample are merely variance and there is not an underlying factor that +4.5-+6 rd OR -4.5 to -6 HF have that other samples do not.

It does not have to be a big difference, merely 1 out of 10 games having a specific set of motivational or emotional factors (saying that maybe 1/2 of those 10% actually fall in the teaser range at a higher percent than they should) that fall into this category could conceivably make this go from small loser to profitable, no?

What if you narrow BTC sample size down. Maybe eliminate teams that are atleast 4 games under .500 since they are more apt to a blow out. Or, maybe I am completely off and it is actually those teams you want to keep in.

I feel like the sample size is given too much consideration. Yes, if a series of events actually was 50/50, then the sample size would be of extreme importance. But, sporting events, despite having two possible outcomes, are not 50/50 propositions, even against the spread.

For example, just say a play in a football league and my team over the past 10 years is 50 wins, 50 losses. However, when the team goes out drinking the night before, we are 3 and 13. Yes, both of these sample sizes are too small for any sttistical data to have meaning if going out drinking did not have any effect on our play.

But, it does. And, it shows in our 3-13 record. Why cant BTC sample have a factor effecting game outcomes that is subtle and we fail to see?

[Dark Horse]

As mentioned at the outset, the subset +4.5 to +6 doesn't interest me for teasers, and Durito showed why.

But the question why that subset has been so successful certainly does interest me. There is no way that I will throw 16 years of data on a 'blind luck' pile. My own definition for good and back luck involves that these manifest as short term fluctuations. Sixteen years is not short term, and nor is the five years going forward since the 'data mining' in this case.

The question revolved around two separate questions. The first, as I suggested earlier, is the ease with which '+10.5' (which is practically the same as +13...) opens the game to backdoor covers. This point, not considered by the math mafia, was more or less conceded.

The second question revolves around +4. As tomcowley pointed out, the +3.5 and +4 subset doesn't come close to the +4.5 to +6 subset. There may be another football reason, and if I do think of it I will make sure not to post it here. Suffice it to say, for now, that tomcowley for his subset wants a TD instead of 6 pts. The math boys give him this 16% difference...

I'll bow out of this thread with this question, again reasoning from a perspective of prevent defense and backdoor covers (for the teaser): who is more likely to be given a free TD at the end of a game, when being behind (up to) 17 pts? A 3.5 pt dog or a 6 pt dog? Or should we believe tomcowley, who throws these lines on one pile?

The question was always about 6 teased points, not 7. But tomcowley & his math boys, in their quick sprint to their predetermined finish line, give themselves an extra point (or 16% difference). I suppose it is only fitting that they give themselves 16% in order to throw 16 years of data on the blind luck pile.

Note that they have all the answers. I certainly don't. I like questions that open up new possibilities. They certainly don't.

[tomcowley]

As mentioned, the subset +4.5 to +6 doesn't interest me for teasers, and Durito showed why.

Seriously? I posted that 2 weeks before he did.

There is no way that I will throw 16 years of data on a 'blind luck' pile.

YEARS ROFLROFLROFLROFL. Please link to a definition of standard deviation that uses the number of years the data was collected over and not the number of data points.

The question revolved around two separate questions. The first, as I suggested earlier, is the ease with which '+10.5' (which is practically the same as +13...) opens the game to backdoor covers. This point, not considered by the math mafia, was more or less conceded.

It's completely irrelevant because that number opens the door to backdoor covers NO MATTER WHAT THE GAME SPREAD IS.

The second question revolves around +4. As tomcowley pointed out, the +3.5 and +4 subset doesn't come close to the +4.5 to +6 subset. There may be another football reason, and if I do think of it I will make sure not to post it here. Suffice it to say, for now, that tomcowley for his subset wants a TD instead of 6 pts. The math boys give him this 16% difference...

I don't "want a TD instead of 6 points". I want to see how often teams cover a particular spread. A -14 fave should cover +10.5 an awful lot. A -3 team less, a +4 team even less, and a +5.5 team less still.

Simply note that 6 teased points for 4.5 buys 10.5, and that 10.5 roughly extends to 13 (because of the rarity of 11 and 12). The question was always about 6 teased points, not 7.

What the **** nonsense is this? Only a true mongoloid would, ceteris paribus, take a worse team teased 6 points than a better team teased 7 points for the same price and final spread. What does the number of points teased have to do with any point I made?

But tomcowley & his math boys, in their quick sprint to their predetermined finish line, give themselves an extra point (or 16% difference). I suppose it is only fitting that they give themselves 16% in order to throw 16 years of data on the blind luck pile.

This is WRFF-level nonsensical rambling. If you read it again and don't see the need to edit it to something remotely coherent, I suggest getting help, because in a couple of months, you'll be calling me a liberal scumbag.

[Peregrine Stoop]

dark horse,
what is the 4 (5) worth in push pcts for a road dog in the NFL?
How does this make a magic cutoff between 4.5 and 4 point dogs (5 and 4, 5.5 and 3.5 even) in their characteristics and how a game is played?

retract your terrible narrative story that is just an example of hindsight bias par excellence to fit the data mined numbers. That cutesy story is bound to make some poor schmuck who reads things here much dumber.

[Dark Horse]

Take your pick. Sixteen years of blind luck or lmao blinders.

Yes tom, years. Luck is subject to fluctuations, as you may have noticed. No time, no fluctuations. Sample size without time is not the same as sample size with time. But that is obviously beyond you.

[tomcowley]

Yes tom, years. Luck is subject to fluctuations, as you may have noticed. No time, no fluctuations. Sample size without time is not the same as sample size with time. But that is obviously beyond you.

This is what scientologists actually believe.

I could datamine some small-sample trend all the way back to 1930s college football and it would be OMFG EIGHTY YEARS OF PURE LUCK NO WAY THAT CAN BE TRUE THIS SYSTEM IS LIKE PRINTING MONEY!!!!!!!!!!!!!

[djiddish98]

let's say that we had 30 different groupings (I'm simplifying as the group sizes differ)
RD 14.5+
RD 14.5-13.5
RD 13.5-12.5
RD 12.5-11.5
RD 11.5-10.5
RD 10.5-9.5
RD 9.5-8.5
RD 8.5-7.5
RD 7.5-6.5
RD 6.5-5.5
RD 5.5-4.5
RD 4.5-3.5
RD 3.5-2.5
RD 2.5-1.5
RD 1.5-PK
HD 14.5+
HD 14.5-13.5
HD 13.5-12.5
HD 12.5-11.5
HD 11.5-10.5
HD 10.5-9.5
HD 9.5-8.5
HD 8.5-7.5
HD 7.5-6.5
HD 6.5-5.5
HD 5.5-4.5
HD 4.5-3.5
HD 3.5-2.5
HD 2.5-1.5
HD 1.5-PK

what do you think the chances are that one of the groupings will have some amazing underlying ATS in either the dog or fav direction just by blind luck? It's more than you think.

This is how a subset like that can perform so well.

You're never going to have a sufficient sample size based on sub-setting the NFL's 14 and 16 game seasons to have any sort of confidence that the results aren't random.

[Pancho sanza]

As mentioned at the outset, the subset +4.5 to +6 doesn't interest me for teasers, and Durito showed why.

But the question why that subset has been so successful certainly does interest me. There is no way that I will throw 16 years of data on a 'blind luck' pile. My own definition for good and back luck involves that these manifest as short term fluctuations. Sixteen years is not short term, and nor is the five years going forward since the 'data mining' in this case.

The question revolved around two separate questions. The first, as I suggested earlier, is the ease with which '+10.5' (which is practically the same as +13...) opens the game to backdoor covers. This point, not considered by the math mafia, was more or less conceded.

The second question revolves around +4. As tomcowley pointed out, the +3.5 and +4 subset doesn't come close to the +4.5 to +6 subset. There may be another football reason, and if I do think of it I will make sure not to post it here. Suffice it to say, for now, that tomcowley for his subset wants a TD instead of 6 pts. The math boys give him this 16% difference...

I'll bow out of this thread with this question, again reasoning from a perspective of prevent defense and backdoor covers (for the teaser): who is more likely to be given a free TD at the end of a game, when being behind (up to) 17 pts? A 3.5 pt dog or a 6 pt dog? Or should we believe tomcowley, who throws these lines on one pile?

The question was always about 6 teased points, not 7. But tomcowley & his math boys, in their quick sprint to their predetermined finish line, give themselves an extra point (or 16% difference). I suppose it is only fitting that they give themselves 16% in order to throw 16 years of data on the blind luck pile.

Note that they have all the answers. I certainly don't. I like questions that open up new possibilities. They certainly don't.

Complete hogwash.

+10.5 about the same as +13?

Try again.

If you really believe in some backdoor cover phenomenom, you can make a killing betting 2nd halfs, just look for the right situation that puts you at +10.5 on a net basis.

Let us know how it works out.

[Dark Horse]

dark horse,
what is the 4 (5) worth in push pcts for a road dog in the NFL?
How does this make a magic cutoff between 4.5 and 4 point dogs (5 and 4, 5.5 and 3.5 even) in their characteristics and how a game is played?

If I remember correctly I introduced the prevent defense and backdoor cover into this 'cutesy story'. This element was not disputed by the math mafia.

tomcowley brought up the 3.5 and 4 pts, in response to that. To him a backdoor cover for a teased 10.5 RD should be the same for all teams. Is that a fair argument?

No.

Why not? Because a backdoor is a psychological/motivational angle.

Do you honestly think math guys have a clue about motivational angles? Believe me, psychology is their weak spot. It's all math, it works for them, and outside of it they know little.

What they fail to see is that backdoor covers can easily be a combination of perceived strength AND a big lead. They only see the big lead. They do not bring in the other element. I pointed out that a favorite of 21 pts is unlikely to be as motivated as a 3 pt fave. If such a motivational difference exists, then it exists on a smaller scale as well. How small? I don't know. I'm not the one claiming to know it all. I'm comfortable asking questions, and understand that is a process. Do you think they're comfortable not knowing?

Think tank? I don't think so.


Sweet sixteen: math guys giving themselves 16% so they can maintain their belief in 16 years of blind luck.

[tomcowley]

If I remember correctly I introduced the prevent defense and backdoor cover into this 'cutesy story'. This element was not disputed by the math mafia.

BtC had been peddling the same nonsense years before you brilliantly came up with it as well. Birds of a feather, or something.

http://www.lasvegasadvisor.com/forum/messageview.cfm?catid=36&threadid=284484&FTVAR_MSGDBTABLE=&STARTPAGE=2

His last post on page 2.

Sweet sixteen: math guys giving themselves 16% so they can maintain their belief in 16 years of blind luck.

This incoherent nonsense again? And the out-of-sample result (since "discovery" in 2003), is 168-73-1 69.6%, even with a >>50% ATS. But that's ok, they haven't performed worth a shit since discovery, so what do you do?

BtC just pretends the two out-of-sample losing years didn't exist and starts touting his new sample from 2005 forward.

You and BtC both mix your datamining years with your out-of-sample numbers and try to convince people that these are, or ever were, any good.

This methodology is what scientologists actually believe.

[terpkeg]

This is how a subset like that can perform so well.

You're never going to have a sufficient sample size based on sub-setting the NFL's 14 and 16 game seasons to have any sort of confidence that the results aren't random.

But isnt this true only if the results in the subset were random? However, since Bookmakers take stances when setting lines, isnt it possible that a particular subset may be populated by a group of emotional and motivational factors that effect the outcome of a game?

[djiddish98]

But isnt this true only if the results in the subset were random? However, since Bookmakers take stances when setting lines, isnt it possible that a particular subset may be populated by a group of emotional and motivational factors that effect the outcome of a game?

Why do -14 games exist then if every bettor assumes that this team is not as "motivated" at as a -3? Wouldn't that spread be adjusted to account for these motivational issues?

I find it far fetched that anyone could make a blanket statement about a specific spread with regards motivation and psychological factors. And even if this was true, it doesn't take long for the market to adjust (see: underdog bias)

[djiddish98]

Better yet, why don't we just subset this to death.

Road dogs - 1994 - 2009. Sunshine forecast data.

Spread Count Fave Win Fave Cover 6 PT Tease cover

4 127 0.67 0.547619048 0.636363636
4.5 112 0.60 0.410714286 0.75
5 71 0.70 0.485294118 0.75
5.5 111 0.68 0.468468468 0.747747748
6 134 0.69 0.523809524 0.736842105
6.5 141 0.63 0.439716312 0.709219858

16 years of data where the 4.5 has covered almost 60% of the time. Keep betting those ATS, they're gold!

[Dark Horse]

Complete hogwash.

+10.5 about the same as +13?

Try again.

If you really believe in some backdoor cover phenomenom, you can make a killing betting 2nd halfs, just look for the right situation that puts you at +10.5 on a net basis.

Let us know how it works out.

Dumb and dumber.

[terpkeg]

Why do -14 games exist then if every bettor assumes that this team is not as "motivated" at as a -3? Wouldn't that spread be adjusted to account for these motivational issues?

Well even if they are not as motivated, maybe they are still that much better. Or, the underdog is equaly as unmotivated. Or maybe some -14 point favorites are as motivated. Or, maybe a significant amount of nfl money somes from unsofisticated bettors who do not even consider motivation.

I find it far fetched that anyone could make a blanket statement about a specific spread with regards motivation and psychological factors. And even if this was true, it doesn't take long for the market to adjust (see: underdog bias)

I do not necessarily believe this is the case. I am just trying to explore the possibility. I am not trying to make a blanket statement about a specific spread. Obviosly not all games that are set at +4.5 to +6 would create a motivational or emotional set of factors for the final score to beome more likely to fall between +4.5 and +12. But, maybe these factors exist in enough games to make this occurance more likely.

I am not trying to justify BTC's teasers. But, I do not think the possibility can just be dismissed with statistical analysis that applies randomness to each event since sports are not random events.

Maybe I am wrong, I have not studied any Mathematics, realistically, since I was a Junior in H.S. However, I feel like the people who are the most adament about this subset not being a true indicator of future results justify it on two basis: 1) that the spread or money line makes the sporting event a random event; and 2) that the market is efficient.

Maybe that is only one justification, that the market is efficient.

But, dont the books take stances? Also, is it possible that a obscure alt line like mid range RD at double digit point spreads can remain somewhat soft for a period of time. Especially, when the sharpest cappers, do not believe the are soft.

[terpkeg]

Complete hogwash.

+10.5 about the same as +13?

Try again.

If you really believe in some backdoor cover phenomenom, you can make a killing betting 2nd halfs, just look for the right situation that puts you at +10.5 on a net basis.

Let us know how it works out.

Using this logic, you make an argument that beating the closing line has no significance. Just take the better number at halftime. But, this is not an accurate statement.

However, you are right, you cannot say +10.5 is the same as +13 when the difference between winner and loser in Sports betting is so fine.

[TheJewBear]

good post!