Help with Teasers

[Bill the cop]

Yeah, the super-special reason is that in the past the market was biased. Due to myriad factors, I (and many others) believe this is no longer the case. The market has become a lot more efficient in the last 10 years.

But that's not what I asked you. I asked you about the specific "sweet spot" for your RD teasers, not about their overall cover rate going forward. What makes you think a RD's distribution is significantly different than a HD's distribution?

And lastly, if you think there is a significant bias in NFL ATS going forward, you don't have any clue what you're doing. Which brings me back to my first post on this thread...

If you're looking for possible reasons that the RD's distribution is different than a HD's look at the ML cover rate for them both. The HD 4.5 to 6 covers the ML at 27.3%, the RD 4.5 to 6 ML cover rate is 35.4%.

If you think you know anything at all about NFL teasers and their associated sample sizes, give just one example of a teaser subset (other than the ones I advocate) that even comes close in cover rate and sample size to the ones I bet.

Where did I say that there's a significant bias in the NFL ATS going forward? I was merely citing the Fav/Dog ATS cover rate for the last 16 years.

Your first post was "RD+4.5 to +6 is pretty atrocious if you have any clue what you're doing". All I can say is I heard all this crap 5 years ago from other "pros". Here it is 5 years later and the subset continues to out perform all other teaser subsets (seems to me SOMEONE doesn't have a clue, and it isn't me)!

[gomiamigo]

Why would that subset do better? Past performance is not blah blah.

[TomG]

BTC, you may want to examine the split between AFC/NFC road/dogs and home/favs to find further profitable subsets.

[Bill the cop]

BTC, you may want to examine the split between AFC/NFC road/dogs and home/favs to find further profitable subsets.

Yes, I've already done that. The absolute best subset is RD4.5 to 6 between divisional opponents. It's 78.8% (but don't tell anyone, they'll think you don't have a clue)

[ouman101]

Very insightful BTC and others. Thanks for the help

[trixtrix]

hmm my data suggests if there ever was a historical bias, it was on HOME dogs, NOT road dogs

[Justin7]

hmm my data suggests if there ever was a historical bias, it was on HOME dogs, NOT road dogs

On 4.5 to 6 point dog teasers?

[trixtrix]

no, i mean just on small home dogs in general, spread <4 i believe it was when i did the analysis, my database starts from the 1983 so i think i grouped it by the decade 80/90/2000's etc and see how hd have performed ats in each of these decade, it was definitely a downward trending graph

[Justin7]

Trixtrix,

1. I would focus on recent data. The internet has made the lines more efficient. Stick to plays in the last 10 years, or even the last 5.
2. why would you look at dog rates from <=4 to analyze >=4.5? The home dog subset of 4.5-6 does especially bad. If you have other reasons to play them, fine... but at least be aware that this set has done terrible recently (although BTC points out that hd divisional rivals do much better than hd non-divisional rivals.)

[trixtrix]

jesus dude, i was just making an observation that if there were any bias in the old nfl market, i only observed it in small home dogs and not medium road dogs. now i could be wrong, but that was just my observation

[Justin7]

hmm my data suggests if there ever was a historical bias, it was on HOME dogs, NOT road dogs

Sorry trixtrix. I read this, and didn't read what you thought. No religious icons are necessary.

[trixtrix]

as far as teasable subsets go, i already stated in this thread:

that home dogs +4.5 thru +6 did horrible in my crappy data base, 67% covers when adding 6 pts

[donjuan]

If you're looking for possible reasons that the RD's distribution is different than a HD's look at the ML cover rate for them both. The HD 4.5 to 6 covers the ML at 27.3%, the RD 4.5 to 6 ML cover rate is 35.4%.

If you think you know anything at all about NFL teasers and their associated sample sizes, give just one example of a teaser subset (other than the ones I advocate) that even comes close in cover rate and sample size to the ones I bet.

Where did I say that there's a significant bias in the NFL ATS going forward? I was merely citing the Fav/Dog ATS cover rate for the last 16 years.

Your first post was "RD+4.5 to +6 is pretty atrocious if you have any clue what you're doing". All I can say is I heard all this crap 5 years ago from other "pros". Here it is 5 years later and the subset continues to out perform all other teaser subsets (seems to me SOMEONE doesn't have a clue, and it isn't me)!

Do you ever answer a question with a response that actually matches the question? I asked you why you think a HD's distribution would be significantly different than a RD's (expected distribution, not historical distribution) and you answer with this garbage? Not to mention this gets back to increasingly efficient markets (which are unbiased, of course) which you epically fail with regard to.

Let's try again.* (1)Do you think markets have become more efficient and are today unbiased? If you don't think this, and that you can keep betting certain subsets ATS in the NFL at WA prices, I can't help you and you'll lose all your money. However, if you do think this then you have to think that RDs and HDs have different distributions (it should be obvious but I mean going forward) for you not to be betting both. So again, (2) why do you believe that RDs and HDs of 4.5-6 have different distributions for the numbers 5-12 going forward? (3) Also, if ATS NFL markets are unbiased, and the 4.5-6 subset is profitable, why would you not be teasing the 10.5-12 point favorites? (4) And if not, again, why do you think they will have a significantly different push percentage aggregate on the numbers 5-12 than the 4.5-6 dogs. Lastly, (5)why do you think RDs of 4.5-6 will have higher push percentages on the numbers 7-12 than games lined within 2 points of those numbers have in the past (for example, why would a game with a spread of 5 land on the 10 more than games lined between and 8 and 12 do)?

*Questions are numbered so you can directly answer each question.

[trixtrix]

Yes, I've already done that. The absolute best subset is RD4.5 to 6 between divisional opponents. It's 78.8% (but don't tell anyone, they'll think you don't have a clue)

how many samples?

[Dark Horse]

BTC, your teaser insight, both here and in other threads, is much appreciated. Do you keep records to preceding games, such as HF-7 or HF-9 off a SU loss?

[Bill the cop]

BTC, your teaser insight, both here and in other threads, is much appreciated. Do you keep records to preceding games, such as HF-7 or HF-9 off a SU loss?

Dark Horse,
No, I don't track that. If you do, I'd be interested in your findings.

[Bill the cop]

how many samples?

It was a real pain to scrape all the data because of all the realignment that took place over the last 10 years. We came up with a SS of 204 for the RD4.5 to 6 subset (160-43-1). It will be easier now with 96 division games per year (assuming no more realignments).

[Bill the cop]

Do you ever answer a question with a response that actually matches the question? I asked you why you think a HD's distribution would be significantly different than a RD's (expected distribution, not historical distribution) and you answer with this garbage? Not to mention this gets back to increasingly efficient markets (which are unbiased, of course) which you epically fail with regard to.

Let's try again.* (1)Do you think markets have become more efficient and are today unbiased? If you don't think this, and that you can keep betting certain subsets ATS in the NFL at WA prices, I can't help you and you'll lose all your money. However, if you do think this then you have to think that RDs and HDs have different distributions (it should be obvious but I mean going forward) for you not to be betting both. So again, (2) why do you believe that RDs and HDs of 4.5-6 have different distributions for the numbers 5-12 going forward? (3) Also, if ATS NFL markets are unbiased, and the 4.5-6 subset is profitable, why would you not be teasing the 10.5-12 point favorites? (4) And if not, again, why do you think they will have a significantly different push percentage aggregate on the numbers 5-12 than the 4.5-6 dogs. Lastly, (5)why do you think RDs of 4.5-6 will have higher push percentages on the numbers 7-12 than games lined within 2 points of those numbers have in the past (for example, why would a game with a spread of 5 land on the 10 more than games lined between and 8 and 12 do)?

*Questions are numbered so you can directly answer each question.

"GET BACK TO US IN 3 OR 4 YEARS WHEN YOU'VE LOST ALL YOUR MONEY BETTING INTO THIS STUPID SUBSET AND WE'LL SEE HOW GOOD IT IS"

ANONYMOUS "PRO" CIRCA 2005


Well here it is 5 years later and anonymous pro was right, it isn't as good as it was 5 years ago, IT'S BETTER.

The problem with you is you're hung up on how to value the 6 points picked up with the teasers. I don't use push ATS percentages to determine the value of the teaser points, I use MOV. I look at a line of, say, +5.5, all I need to know is what percentage of time the Fav wins the game by 12 or more. If the Fav wins by 12 or more >30% of the time, no bet, etc.

The only criteria I think is important is "has this been a winning subset over a reasonable sample size?" . If I can answer that question to my satisfaction, I make the bets, simple as that.

The bottom line, see first statement from years ago. If I'm winning money betting this subset, why should you care? As I've said before, if it stops being profitable, I'll stop betting it. BTW, so far 2010, business as usual for the subset, 6-2 for 75%. Oh, and as far as you questions by the number, stick them up your ass.

[LegitBet]

Search posts...its out there

[Poogs]

Why would anyone possibly be mad at bill the cop? He's giving great info for free and its not like he's forcing anyone to bet. If you disagree just don't tease his subsets.

[dogman]

Right, I have heard BTC defend himself over at LVA for a long time and now here and he doesn't have to. He dishes out great information(maybe too much) and is helping many bettors win. I consider him one of the top posters and if anyone here is betting teasers I would listen when he speaks.

[tomcowley]

The only criteria I think is important is "has this been a winning subset over a reasonable sample size?" . If I can answer that question to my satisfaction, I make the bets, simple as that.

The only question that's actually important is should it win going forward. Morrison circa 2009 would have passed your test with flying colors, but nobody took it seriously because it was obviously retarded. And bet-by-bet, Morrison had you making way better bets than this crap- you actually posted, in this thread, exactly why the bets are awful, but you're too blind to see it.

[ForgetWallStreet]

Why would anyone possibly be mad at bill the cop? He's giving great info for free and its not like he's forcing anyone to bet. If you disagree just don't tease his subsets.

At least you have a chance to run good on this particular scam.

[Poogs]

What scam?? What am I missing? Is bill selling something?

[Dark Horse]

Why would anyone possibly be mad at bill the cop? He's giving great info for free and its not like he's forcing anyone to bet. If you disagree just don't tease his subsets.

The math inquisition is highly uncomfortable with new ideas. Those ideas, and their proponents, are either laughed off as ridiculous, or 'gracefully' placed on trial until proven. In their unfortunate arrogance, they usually fail to realize that their very 'kingdom' is comprised of ideas that were once new and unproven.

[djiddish98]

(1)Do you think markets have become more efficient and are today unbiased?

I'm going to go slightly OT, but run tangent to this point:

I would be inclined to think that the markets have become more efficient, due to more computing power, better / more available information, and more participants in gambling markets. Indeed, the much publicized home dog bias has disappear in the NFL.

However, it seems like spreads are still off the mark by about 10 points a game (in either direction), which has historically been the case. I would expect that number shrink down based on the above. 2008 and 2009 appear to be some of the worst years with an average around 11.

What I'm measuring here is the absolute value of favorite action points, or the difference between the favorite's margin of victory and the spread.

So the data (albeit not entirely reliable on my part) runs counter to my thoughts. Is the issue with averages (excel doesn't play nice with median) or should we still continue to expect spreads to be off by about 10 points going forward, even as markets continue to get more efficient.

Could it also be the switch to more boom-or-bust passing offenses essentially "eating up" the efficiency gained over the last 20 years? How does the average total compare over the last two decades?

Hopefully food for thought.

[Peregrine Stoop]

djiddish, the variance in the underlying event doesn't magically disappear due to more efficient spread betting

only changes in the actual game of football will effect the range of scores in the underlying event

[djiddish98]

djiddish, the variance in the underlying event doesn't magically disappear due to more efficient spread betting only changes in the actual game of football will effect the range of scores in the underlying event

Thanks for replying.

The underlying event has changed to a certain degree over the last 20 years. Whether it's material or not, I'm unsure.

[wrongturn]

Appreciate all of bill the cop's postings. It is interesting that subset RD 4.5-6 works but similar line like RD 6.5 does not work. I guess that is why some guys are suspicious. If you trust push chart, none of those should work. But push chart only counts pushes around the close line, so it may not be accurate for teasers.

[dimon]

man, what a discussion...lots of great info, I think that people have the right to speak up, and disagree....Bill you do your thing and win some money, let the people dubt, I went back and check the stats and it looks good to me to play it, I am not a max player, and sure will like to add more betting oportunities going forward...so I will go back a few posts, and since I am the one who started this thread, ask the participants again...

should we talk about the potential teaser plays for the upcomming Sunday?

I can start a new thread in NFL and we can talk about it, or here...let's hear the opinions...if this will help us all win, why not share and debate...let's just be nice ok.

[djiddish98]

Appreciate all of bill the cop's postings. It is interesting that subset RD 4.5-6 works but similar line like RD 6.5 does not work. I guess that is why some guys are suspicious. If you trust push chart, none of those should work. But push chart only counts pushes around the close line, so it may not be accurate for teasers.

If anything, the push % should be slightly lower the farther you get from the spread.

These teases have worked because the ATS is great and/or there's some property of RD 4.5-6 that increases the push probability of the numbers it crosses in a 6 point teaser.

If you ascribe to the second part of the sentence, I'd like to hear the reasoning why a RD 4.5-6 is more likely to push at 10 than a home fave at 14-15.5.

The evidence that it worked in the past doesn't necessarily convince me that it'll work in the future.

[donjuan]

"GET BACK TO US IN 3 OR 4 YEARS WHEN YOU'VE LOST ALL YOUR MONEY BETTING INTO THIS STUPID SUBSET AND WE'LL SEE HOW GOOD IT IS"

ANONYMOUS "PRO" CIRCA 2005


Well here it is 5 years later and anonymous pro was right, it isn't as good as it was 5 years ago, IT'S BETTER.

The problem with you is you're hung up on how to value the 6 points picked up with the teasers. I don't use push ATS percentages to determine the value of the teaser points, I use MOV. I look at a line of, say, +5.5, all I need to know is what percentage of time the Fav wins the game by 12 or more. If the Fav wins by 12 or more >30% of the time, no bet, etc.

The only criteria I think is important is "has this been a winning subset over a reasonable sample size?" . If I can answer that question to my satisfaction, I make the bets, simple as that.

The bottom line, see first statement from years ago. If I'm winning money betting this subset, why should you care? As I've said before, if it stops being profitable, I'll stop betting it. BTW, so far 2010, business as usual for the subset, 6-2 for 75%. Oh, and as far as you questions by the number, stick them up your ass.

Why do you refuse to answer my questions? And the problem I have is not with you betting them, it's that you're tarding up the forum by passing them off as +EV and refusing to answer accurate criticism.

[donjuan]

djiddish, the variance in the underlying event doesn't magically disappear due to more efficient spread betting

only changes in the actual game of football will effect the range of scores in the underlying event

The variance in the underlying event doesn't change, but the observed distribution should change the more efficient the market gets.

[donjuan]

I'm going to go slightly OT, but run tangent to this point:

I would be inclined to think that the markets have become more efficient, due to more computing power, better / more available information, and more participants in gambling markets. Indeed, the much publicized home dog bias has disappear in the NFL.

However, it seems like spreads are still off the mark by about 10 points a game (in either direction), which has historically been the case. I would expect that number shrink down based on the above. 2008 and 2009 appear to be some of the worst years with an average around 11.

What I'm measuring here is the absolute value of favorite action points, or the difference between the favorite's margin of victory and the spread.

So the data (albeit not entirely reliable on my part) runs counter to my thoughts. Is the issue with averages (excel doesn't play nice with median) or should we still continue to expect spreads to be off by about 10 points going forward, even as markets continue to get more efficient.

Could it also be the switch to more boom-or-bust passing offenses essentially "eating up" the efficiency gained over the last 20 years? How does the average total compare over the last two decades?

Hopefully food for thought.

Just off the top of my head, probably a combo of variance and higher totals.

[donjuan]

Why would anyone possibly be mad at bill the cop? He's giving great info for free and its not like he's forcing anyone to bet. If you disagree just don't tease his subsets.

Stick to getting making bets via PM with randoms and whining about it when you get scammed.