[Dark Horse]

I've seen blind luck last for thousands of BETS. Not sure how many years it was though.

Exactly.

[ForgetWallStreet]

Five years of out-of-sample "dumb blind luck", gotta love these nitwits

How many years do you figure you have left before the sharper books start offering all +4.5-6 RD's at +4 +high vig?

[tomcowley]

So, instead of actually doing any investigative work, you're just inclined to say that I- the person who has actually done work- must be missing something. Thanks for sharing. I value that opinion very highly.

[Bill the cop]

One question for you Bill: If you'd also been able to also tease +3.5 and +4 RDs up to +10.5 for the same price, how would you have done on those the last 5 years?

Now we're down to hypothetical questions about 7 point teasers at even money.

Okay, just like the GodFather2 (YOU DREW ME BACK IN) to answer your post.

Over the last 5 years RD3.5 to 4 teased up 7 points went 74-29-1 for 71.8%

It's getting a little old with your dick wagging database stuff. If you really do have a database for the last 5 years (I really do, Don Best closers), show where the RD+4.5 to +6 "jumps off the page and slaps you in the face".

Here's my figures, plain and simple:

2005-2010
RD4.5 to 6, 112-36-1 for 75.7%

Show me data that conflicts with mine or STFU!

[Dark Horse]

So, instead of actually doing any investigative work, you're just inclined to say that I- the person who has actually done work- must be missing something. Thanks for sharing. I value that opinion very highly.

I'm saying 'blind luck' is the easy way out. If it was my work I wouldn't accept it. If this topic was truly interesting to me I would do more investigative work. But I'm occupied elsewhere, focusing on topics of greater fascination to me personally. Sorry.

[tomcowley]

Okay, just like the GodFather2 (YOU DREW ME BACK IN) to answer your post.

Over the last 5 years RD3.5 to 4 teased up 7 points went 74-29-1 for 71.8%

Well, the push on the 11 doesn't cover 10.5, so 74-30 71.1%. So if you could tease these at the same price, which are better teams getting more points, you would lose money. And it doesn't make you wonder, at all, that worse teams are covering a particular spread at a greater rate than better teams are. That might make some people wonder, but you just keep betting and don't give a shit.

Teasing 4.5 and 5 up 6 points to +10.5 and +11: MONEY IN THE BANK!!!!!!!!!!!!!!!
Teasing 3.5 and 4 up 7 points to +10.5 and +11 for the exact same price: MONEY DOWN THE TOILET

Move along, really nothing to see here.

Show me data that conflicts with mine or STFU!

Your data isn't bad. You just have no idea how to do anything but mindlessly mine it.

[Pancho sanza]

Okay, just like the GodFather2 (YOU DREW ME BACK IN) to answer your post.

Over the last 5 years RD3.5 to 4 teased up 7 points went 74-29-1 for 71.8%

It's getting a little old with your dick wagging database stuff. If you really do have a database for the last 5 years (I really do, Don Best closers), show where the RD+4.5 to +6 "jumps off the page and slaps you in the face".

Here's my figures, plain and simple:

2005-2010
RD4.5 to 6, 112-36-1 for 75.7%

Show me data that conflicts with mine or STFU!

My data doesn't conflict with yours, so what, no one is saying you made up the results.

Variance is your friend when you have no edge, that's whats happened here, you've had some good luck basically. You've mistaken it for an edge, thats deadly in sportsbetting if you can't tell the difference.

That subset probably hits 68-69 % going forward, its not that unusual that it would hit much higher over a short run, run a simple test and you''ll see.


http://www.stat.tamu.edu/~west/apple...omialdemo.html

[djiddish98]

So it's no surprise that this subset
also did well ATS. From 2005 to 2010 it has went 80-67-2 ATS for 54.4%.

Do you expect this subset to continue to win at 54.4% ATS going forward?

[Peregrine Stoop]

Do you expect this subset to continue to win at 54.4% ATS going forward?

If he really believed in what his database is spitting out, he should be blindly betting these teams ATS and raking in much more

[statictheory]

I've seen blind luck last for thousands of BETS. Not sure how many years it was though.

Really? Thats a big statement . lets see your numbers on that

[Bill the cop]

If he really believed in what his database is spitting out, he should be blindly betting these teams ATS and raking in much more

If you read my whole post you'd know that I bet them in teasers AND ATS (and sometimes I put them in parlays too).

[tomcowley]

If you read my whole post you'd know that I bet them in teasers AND ATS (and sometimes I put them in parlays too).

[djiddish98]

If you read my whole post you'd know that I bet them in teasers AND ATS (and sometimes I put them in parlays too).

So can I infer here that you expect them to continue to win at 54.4% pace ATS?

[wrongturn]

I think Bill already said things can change. He will be fishing all day when that happens, not now.

[ForgetWallStreet]

Really? Thats a big statement . lets see your numbers on that

There are plenty of people ITT who have run significantly above/below expectation over 5K+ bet samples. If you don't believe me, run some sims.

[Dark Horse]

It must be excruciatingly painful to know it all. Five years ago it was data mining. Today, five years later, it's blind luck. lol

I try to combine two approaches to betting. One is the sure and steady, the other is the quick and nimble. They're very different. But only the latter is awake to short term fluctuations. So I totally agree with what BTC is doing. It's far from stupidity, but rather an art in itself to be awake to whatever is going on at any given time. It's not about approval from the (self-appointed) scientific community. It's about money in the bank.

Example. Check the MLB unders for low totals (7.5 runs and lower) since the start of August. The season is over now, so it won't hurt mentioning, because it won't continue anyway. It was right there, hidden in plain sight, and without any scientific backing. Simple to the point of silly.

[tomcowley]

Clearly there are meaningful short-term fluctuations that apply to exactly the small +4.5-5.5 subset that don't exist at all in the 3.5-4 and 6-6.5 data. I mean seriously, what could logically explain the magic of 4.5-5.5 spreads besides dumb luck? Basically nothing.

If you tell me that perception of one particular team, or the entire market, is slow to adapt, then that's possible. But this subset is akin to saying that games that closed u7.5 -108 to -113 are amazing under plays while games that closed 7 to 7.5 -114 and 7.5 -107 to 8 are perfectly normal, and that I should believe that's something besides noise, and that it will be amazing going forward, GMAFB.

[trixtrix]

i try to figure out what that pic meant, took me > 1 min to figure out that's a data-miner.. i feel so ashamed

still not sure why you squibbled out his face though

[donjuan]

i try to figure out what that pic meant, took me > 1 min to figure out that's a data-miner.. i feel so ashamed

still not sure why you squibbled out his face though

Blind, IMO.

Tom Cowley's point about 3.5-4 RDs should end thread but the guppies are sure to continue with random nonsensical ramblings.

[durito]

Pinny sure seems to have a lot more 2.5's and 1.5's available this year than ever before. What's the deal? Higher totals? Are they implying that Balt is not a good leg this week?

[subs]

may i ask a question please. if the above subset is winning because of luck, which entirely possible, could that also mean the RF 7.5 - 8.5 which have been losing are also just losing because of luck? or are we just assuming the RF 7 -9 set is a loser?

i understand not capturing the whole of the 3 an 7 is not ideal. can this also be variance? not sure i'm convinced either way really.

i really had to laugh out loud there TomCowley

[donjuan]

Subs,

In short, yes. I'd play RF -7 as well.

Durito,

Pinny hangs -6.5s all the time that are obvious +ev 6 pt teaser legs so I'm really not sure what they're thinking.

[durito]

Subs,

In short, yes. I'd play RF -7 as well.

Durito,

Pinny hangs -6.5s all the time that are obvious +ev 6 pt teaser legs so I'm really not sure what they're thinking.

I've noticed that before too, but find it odd that they are suddenly taking some basic strategy legs they never took before. They do weird shit though, like how all of a sudden their football limits are 2-3x what they were since uigea passed.

[TMoney33]

Many knows how to play them...BS cross both number 3, 7 etc. RD +4.5-+6, HF of 7-9.

Hi,

As my question will obviously show I'm a complete noob, but what are BS, RD, and HF?

Thanks

[subs]

Subs,

In short, yes. I'd play RF -7 as well.

thank you sir - a few points your way

TMoney33

Hi,

As my question will obviously show I'm a complete noob, but what are BS, RD, and HF?

Thanks

BS basic strategy applies to teasers - from stanford wong's book
RD road underdog
HF home favorite

BOL (best of luck - lol laugh out loud)

[Poogs]

Thanks again Bill. I have a couple theories to "possibly" explain part of the road dog 4.5-6 teaser bias compared to that of home dogs in the same range. Good chance they are both BS The first is a kind of strength of opponent factor. When road dogs are in that range we are obviously dealing with teams that are within a couple points on a power rating basis, and when the home dog is in that range it is closer to 7-9 point difference. Is there a possible difference in expectations of teams in each scenario when the game starts to get a little out of hand? In the Minnesota +4.5 at Jets example: Are the Jets more likely to be content to be up a TD or 10 points when playing at home vs a solid team like the Vikings... compared to the home dog scenario where for example the Redskins would be up 7 or 10 points at a weak team like Buffalo? Is the mindset of the Redskins and Bills more likely to lead to a blowout if things start to get out of hand, where the Vikings and Jets are expecting a closer game from each other, and play to that level? Also, could there be somewhat of a diminishing homefield advantage as games get to double digit margins? The Bills crowd is much more likely to be a factor when the Bills are within a TD of the Redskins than if they get down 10 or 14 points in the second half. Maybe even the home crowds let up some when they are up a comfortable margin so road teams trying a backdoor cover of the +10.5 face less resistance than if they are trying to win the game. This could possibly both hurt the home teasers and help the road teasers. This could also be a part of the theory for why basic strategy teasers of road favorites -7.5 to -8.5 perform poorly. The closer the game, the more the home field edge comes into play. The road favorite is going to have a harder time coming from behind in a 17-21 game to cover the (-2) teaser line than they are to just hold onto a double digit lead and keep the -8 margin in the 4th quarter.

any comments on this?

[wrongturn]

I've noticed that before too, but find it odd that they are suddenly taking some basic strategy legs they never took before. They do weird shit though, like how all of a sudden their football limits are 2-3x what they were since uigea passed.

They could manipulate the lines to make people bait on them.

[LegitBet]

great thread....look for my response witin th eweek..i have a take on teasers i have not heard of before..
thanks

[statictheory]

There are plenty of people ITT who have run significantly above/below expectation over 5K+ bet samples. If you don't believe me, run some sims.

I believe you. im just being a smart ass. But seriously i question most sample sizes
in sports betting for exactly that reason. In poker for instance the edges are very small and sample sizes to determine true profitability are huge. Only the internet with the ability to play significant numbers of hands has helped with confidence levels , but it seems with sports betting that since the numbers everything is based on are so small, everything is just noise and randomness, but I suppose Im wrong about that.

[Highland]

Well, the push on the 11 doesn't cover 10.5, so 74-30 71.1%. So if you could tease these at the same price, which are better teams getting more points, you would lose money. And it doesn't make you wonder, at all, that worse teams are covering a particular spread at a greater rate than better teams are. That might make some people wonder, but you just keep betting and don't give a shit.

Teasing 4.5 and 5 up 6 points to +10.5 and +11: MONEY IN THE BANK!!!!!!!!!!!!!!!
Teasing 3.5 and 4 up 7 points to +10.5 and +11 for the exact same price: MONEY DOWN THE TOILET

Move along, really nothing to see here.

Your data isn't bad. You just have no idea how to do anything but mindlessly mine it.

That's definitely not good evidence for the road teasers. But maybe there is some variance involved in the 3.5's and 4's as well. Bill, what do home dogs from 3.5 and 4 teased to +10.5 points look like?

[Dark Horse]

Just want to make sure I chew on the right paradox, so if someone could confirm objectively. For 6 pt teasers:

L16 years
1) RF 7.5-8.5: 64.8%
2) HF 7.5-8.5: 73.8%
3) RD 1.5-2.5: 72.3%
4) HD 1.5-2.5: 71.8%

That's all BS or Wong.

L16 years, as pointed out by BTC:
A) HF 7 to 9: 74.4% (with above HF sample size expanded from 225 to 487 decisions; so a simple 'thank-you' might be in order just for that)
B) RD 4.5 to 6: 74.9% (sample size 435 decisions)

Based on 16 years of data. Five years ago the B) subset was criticized as data mining, but since then it has continued to produce at 75.6 % (before the start of this season). The critics, relying on push frequencies, consider this blind luck.

So the idea is that for HF's crossing the 3 and 7 is key, but for RD's crossing the 7 and 10 is key (RD's 1.5 to 2.5 are performing about 2.5% below the 4.5 to 6 subset, over 16 years). My only question is if I correctly identified the apparent paradox. I'm going to look for the answer in the game of football itself, in the hope of learning something about how it is played.

[Peregrine Stoop]

Dark Horse, not paying attention to the underlying ATS is a big mistake

look at what range the extra 6 points picks up... >21% = $$$$

[Peregrine Stoop]

The critics, relying on push frequencies, consider this blind luck.

If you looked at every single combination of road and home favorites and underdogs, what is the chance that you would find a subset that crushed ATS like +4.5 to +5.5 road dogs over that sample?

[djiddish98]

Bill - how far back does your data go? Did you arbitrarily choose 1994 or is that the beginning of your dataset?

[djiddish98]

B) RD 4.5 to 6: 74.9% (sample size 435 decisions)

I'll ask you as well DH - can you expand that data set back into the early 80s? What kind of results do you see?