Posted this at Covers, but it fits here as well

We had quite a debate about "key" numbers and how pushes on these numbers were significant in trying to figure out betting strategies.

I stated back then that a -3 number was no where near as significant as people made it out to be. And that worrying about it, and/or buying of it was simply not a good play to make.

Of course the usual arguments and bantering went on and on. And of course last year was used as the weapon of choice since 12 games did in fact land right on that number of 3.

Well I said this preseason that I expected that to drop off severely this year. And that pushes on that spread would be very minimal. I believe I said 5 or 6 would hit.

Well we just passed week 10 and there have been two games that landed on 3 that closed at 3. In all games this year there have only been 3 pushes in total. So basically a banner year in avoiding ties.

There are a lot of factors that have gone into this. First and foremost the way dogs have been winning Su is a major reason. Another is simply because the line doesn't mean anything. It is just a way to get people to either lay off the game, or for books to rationalize gouging a little more money out of you. Since you never see -3 -110 anywhere, it is always -3 -115;-120; even -140 in some cases. So a way for the books to tweek the odds to get both more vig, and keep people from betting a dog with the pyschological half point. It is so fuuny, how a guy will play a dog +3.5 -115, but won't play the same dog +3 +110. Bascially that is a 25 cent half point, that 9 out of 10 games (on a 20 year ave) doesn't mean dick.

I haven't bothered to see how many games have closed at -3 this year. But I would say there have been 40 at least. SO 2 of 40 is not a very big percentage. That is why when we we debating this way back when, using past data as fact was tainted, because in the real world the results fluctuate.

But when seeing that the previuos year an inordinate amount of pushes happened, it is actually more correct to assume that the following year the reverse would be true.

For the most part all things try to reach a 50/50 split. But due to parity and free agency and injury, the sides are a lot more difficult to determine. So that ine of thinking doesn't always work when trying to make the conclusion that dogs have been covering steadily, the lines will be adjusted, so that the faves will "catch up".

This USED to be true. But not anymore. That is why I went into a skid the past few weeks. Simply put, the "dogs" are actually better teams and win SU these days. Even when the "better" team is favored, it hasn't always been that way this year.

Case in point...

Dogs this year in 142 game are 81-54-3 ATS, with 4 Pk em games intermixed.

So a very huge avantage for the dogs.

But when looking for a "pattern" of "bad linemaking" We see it isn't the case as there is an exact 71-71 split for over and unders. That is as good as it gets. So the lines are not wrong, just the teams have not cooperated.

Now looking at SU winners in regards to ATS covers. I have long said, if you can pickthe SU winner, you will get the cover. Historically 17% of the time the line does not even factor into a game. Well this year thus far of the 142 games played, the SU winner was 113-26-3 ATS. So that is right in line with the expected 83% rate we generaly see.

So this year has held true to a lot of things that could have been expected. But it has also broken the mold in regards to a lot of other traditonal thinking.

When you can blindly bet dogs and collect at exactly a 60% rate there is something wrong. But the books have done everything they can to get the "better" teams to win. They have posted ridiculously low numbers for the most part, but the dogs continue to win SU. And flip flopping sides would not be smart, as for example how many people would have bet Oak last night had they been a fave? As it was I thought the 6 points was very low considering all the factors going into that game. So the books cannot put up numbers that are completely off. Besides, they actullay do better when dogs cover anyways. So they are getting the best of both worlds. They are posting short numbers to keep people betting the faves, and they are still getting the dogs to win SU.

OH MY!

12% of the time, a team favored by 3 will win by 3. Aces Gold went bankrupt letting people buy through the 3.

To look at this year's data where it has hit under 12%, and try to minimize the value of 3 is silly. -3 teams are always going to win by 3 in the NFL at a 10% or higher clip.

Your lead-in comment spoils an article that otherwise is well written and useful.

Ps. NE -3.5 vs. Chicago (wins by 3). Jets -3 vs. Miami (wins by 3). PIT -3 vs. Cleveland (wins by 3). In just the past two weeks 3 games middled for sharp bettors on the hugely important 3. Not surprising, since 1 out of 9 games will hit on the number. That turns 4-4-1 into 5-4 for those getting the hook on the 3. Last time I checked 55% makes a nice profit betting.

Revinax, look at all the comments and results I have posted in other topics about this subject. Over the past 20 years, the gain in "buying off" 3 would have been I believe a profit of 2400 dollars. So in essence and average of 120 bux a year. Or in terms of vig, roughly a sinlgr bet gained. to me that is simply not a big enough gain to let it worry me.

But if this year continues its pattern, then that will cut into that "advantage" a lot. Again, we are trying to use a median/average numbers of occurances to put a "value" on this. But as I stated in the above piece, that isn't possible.

So for lack of a better statement. Go with the flow, and let the wins and losses take care of themselves. Worrying about a line should be the last thing you do.

But as to your point about those middles. had you done it on every game this year you would be behind. Again, easy to point things out after the fact. So when you already have the answers it is easy to pick and choose the games that fit the mold you are looking for. but before hand you have to use every examople to guarantee to getthe 'right" results, along with the "wrong'>

YOU are the one using short term data. I'm the one saying "3 is going to hit at 10% or more (I use 12%)". Why? Historical data shows this! It has always hit at 10% plus, and the only thing that has changed in the NFL is the two point conversion which promotes more 3 point outcomes.

Want to bet me that games lined on 3 won't land on 3 10% of the time over the next 10 years? You would lose. Anyone who thinks it will hit at under 10% going forward can be dismissed immediately since it is a "solve by inspection" easy problem. Go ask any true winning pro, or go to any database.

Having said this, the true pros know that games lined at 2.75 or 3.25 represent "plus vig" betting opportunities. EXAMPLE: Cincy/CLEV is 3.5 but may leak to 3.25. If this happens, you simply have to look to lay -3 or take +3.5. Why? Because both of them together are a profitable middle. Given, they are profitable together, it's prettty obvious that 1 of the 2 is likely very profitable.

Time to unload on the Bengals.........

I have 20 years of data. More than enough, and yes it is right around 10%, closer to 11 actually. But as I said, last year it was well over that. Thus my thinking that this year it would be less... a regression to the mean theory. As in. If 20 years it has averaged 10%, then all of a sudden there is a year with 20%, it would stand to reason that there would be a year or two, where it would be lower, to get back to that 20 year average.

Agani, this is not scientific. But niether is gambling. just another way I utilize the data I see. That is the only difference between gamblers actually. the data and stats are there for everyone to see. The INTERPRETATION and subsequent UTILIZATION is what makes some better than others.

Anyone can break a game down, but to know the times to lay off or play a game is the ultimate secret. There is obviously no fol proof reason. But I have found that for the most part what I have said is correct and has been shown.

As for your example of Cinci. They have historically been the worst team ATS.And have actually been a great blind bet. I don't have theier full 20 year history handy, but I do have their record since 1998, so nearly 5 years...

They are 18-55 SU and 26-45-2 ATS. If they are not an automatic fade I don't know what is. According to your theory the books would have figured them out by now and tweeked their numbers. but they can't, that is how bad Cinci is. In a related vein, Cinci's Totals record in terms of OVER/UNDERS is a perfect 36-36-1. So again, it isnt like the books are baffled by them, they simply stink.

And considering that they have been a dog in a vast majority of those games, you could take all that a step further and amend my original post about all dogs in general, and if you simply took out all the Cinci games, I am sure dogs would blindly be covering nearly 63%. So the fact that Cinci is an excpetion to the rule shows how you cannot blindly throw out numbers and expect them to work.

That is why I wrote this. To show that you cannot trust data that is used in its broadest terms. To truly get a good picture you need to break it down.

I will go you one further. I will take the opposite of Cinci in every game over the next ten years, and be further ahead than if I handicapped the games. Tey are that much of an automatic winner.

Also, in response to the challenge. Why bet they will or will not hit 10% over the next 10 years. The point is to play the team regardlees of the number. besides, when did having a push become a bad thing?

[This message was edited by wantitall4moi on 11-12-02 at 08:12 PM.]

Go take a stats or prob. course. You do not understand what regression to the mean is!

That is the only comeback? /infopop/emoticons/icon_smile.gif

I figured you were going to tell me I was wrong about Cinci' ATS record. Which I was. They actually had a winning season ATS a few years back. 95 I think. They were either 10-6 ATS or 9-6-1 depending on your sources. So they have actually has a winning season against the number.

But I did find their ATS record since 1992 to present. And this is the "more favorable" record I might add... 70-94-3 ATS. (another set of lines had them at 68-95-4 ATS) So while not as bad as I thought, still pretty impressive nontheless.

Math is math and you can formulate all you want. I will stick with what I know. Past results mean more than probable outcomes.

Which is directly related to your argument...where does that 10% come from that you predict?... a median of the past results. Therefore you are simply carrying what has happened in the past and using it to try and predict the future. But that is actually a false assumption. Because the "database" is constantly growing, the sample size right now is that many games smaller than it will be in 10 yrs. So in 10 years, the more "accurate" percentage may actually become apparent. Whether that percentage be 9% or 12%, you cannot automatically assume that it will be 10%. Because the data you are using is incomplete.

last week pick 5 games su perfect and went 1-4. so all statistics don't mean s&*^ on a given week; just over TIME. Forget about the hooks in the long run it's a losing proposition; whereas if I could possibly go 5-0 su, every week; I'm gonna be way ahead eventually, excluding last week!!!

You are right, 1-4 sucks, but you know that over the long time that 1-4 week will eventually be surpassed by the the "law of averages" (incorrect term, but as close as I can get).

Here is the NFL DATA from WK1 1996 thru WK8 2002 (1646 games)for winning margins and totals points.

Click WINNING MARGINS & TOTAL POINTS

Breakdown by year for winning margin of 3.0 points (playoffs included)

1996 - 12.7%
1997 - 17.1%
1998 - 12.7%
1999 - 15.8%
2000 - 16.6%
2001 - 17.4%
2002 - 17.2% (up to WK8 only)

AVG. - 15.6%

That is understood. But unless the LINE is -3, then this data is not relevant

Try looking at this probability question this way…

Line data is conditionally based on the oddsmaker or public moving the line. Looking at what spreads of 3 that actually land on a margin of 3 points is somewhat meaningless, as the variable (the line) does not have an exact defined element. The line and its outcome are partially coincidental (“12% of the time, a team favored by 3 will win by 3”).

IMHO, winning margins produce the best guess for a probable outcome. And whether or not you should buy off a spread number.


Take the Buffalo Bills vs. Chicago Bears for example, the spread for the next game is 3 points, do you buy off the 3 or not? Do you ignore the fact that 22.4% (last 6+ years) of the time the Bills had a margin of exactly 3 points and Chicago has a 20.2% occurrence of 3 points?

Or what about the Broncos, where only 12.4% of the time they end in a 3 point margin, playing against the Cincinnati with an 8.7% occurrence of 3 points?

I factor in some of what you do. But I never buy off a number ever.

When I do look at it that way, I take it even further. While overall those percentages may be accurate, as in Buff 22.4% margin of victory of 3. It doesn't matter what their win % by 3 is if the line wasn't 3. Also, are the percentages you quoted for their opponents, when they LOST by 3 or just when they were in games with 3 point score differentials.

The way I break it down would be to look at Buff, in all games they closed at -3. I have 14 games from 1992 to last year (haven't updated them this year, simply because there have been only 3 pushes.) But anyways, of the 14 games Buff was favored by -3, they won 9 SU and lost 5 SU . Of the 9 they won, 3 were by excatly 3 pts(pushes). So actually, their percentages of pushes on 3 (33% when they won by 3) is more than their overall win% by 3 (22.4%). But overall the results are similar 3 of 14 for 21.4%. But that may actaully be coincidence.

In any event, I would still not buy off that 3. Because even though they have shown a "knack" for pushing". would take my chances on getting the fullpayoff, rather than worrying about a oush. As I have mentioned before, what is so wrong with a push?

Now as for Denver They have had the same number of times closing at -3 (14) they won 8 SU, and lost 6 SU, with no pushes. So as it is, there is no advantage or disadvantage. In all actuality you would be just as well off betting the opposing team ML.

But that is me doing exactly what others do, confusing the issue in terms of numbers with things that may or may not be correct, but for the most part twisting the results in the direction that makes what I am saying correct.

This chart is the result of quite a bit of work. It does has a slight margin for error. That margin of error is based on books posting slightly different lines. All my lines are from the same book, so in that case they are uniform. Another margin for error is line moves that may have occured. But this is an in depth chart of the -3 closing pointspread.

The numbers will be slightly hard to follow. But it is listed this way.

First number: TOTAL number of times the team closed at -3

Second number: SU win

Third number: SU loss

Fourth number: push

Az 8 6 2 1
At 16 9 7 6
Balt 11 9 2 2
Buff 14 9 5 3
Car 11 7 4 1
Chi 10 5 5 0
Cin 5 5 0 0
Cle 0 0 0 0
Dal 9 4 5 0
Den 14 8 6 0
Det 7 4 3 0
GB 10 7 3 1
Ind 11 8 3 2
KC 12 7 5 2
Mia 15 10 5 2
Minn 9 4 5 0
NE 15 11 4 3
NO 6 3 3 0
NYG 10 7 3 1
Jets 7 3 4 1
Oak 14 5 9 3
Phil 11 8 3 2
Pitt 13 8 5 3
Rams 6 4 2 1
SD 11 5 6 1
SF 8 5 3 1
Sea 13 10 3 2
TB 14 8 6 0
Tenn 18 10 8 1
Wash 6 2 4 0

Overall there have been 324 games that closed at -3 from 1992 to 2001. The games that pushed were 39. Last year was a banner year, as 11 of 52 games landed on a push.

The info about SU wins and losses is also key. Since the results there are 199 wins and 125 losses. That figures out to a 61.4% win rate. that is almost exactly equal to a -160 fave on the ML. Now this is very relevant snce I have been arguing with many people about buying the half point. Well if you can buy it, it is right around -125. Now if you look at your -3 faves from time to time, their ML are anywhere from -140 to -145, depending on whether or not the books has them laying extra vig. But regardless, Even at -150 the win percentage is still high enough to actually make it profitable over the long run to play the faves ML rather than worrying about the -3 line at all.

I don't necessarily think this way, but it is definately something for those that buy off -3 should think about.

I will however amend this to say that the whole ML thing would have suffered this year as the faves have not been winning at a high enough percentage. So more than likely a short term blip in something that would seem to have long term reliabilty. But as I stated before, there is no way to know what the proper percentages are, as the database is updated with every game played.

covers sucks

Brilliant post from someone who was pimping his site on here the other day.

I'm going back to SharpSportsBetting.com. Those geeks argue and fight over stuff, but at least they have good debates (two competent people talking!)

.....I've got to deal with people here

1) Saying they would "never" buy points and
2) Saying the NFL 3 isn't important and
3) Saying it will hit under 10%

You guys don't get out much do you?

Figured I would bump this up since it is close to football season, and also since I was pretty much right on with my predictions.

Since we had this argument 2002 season, there have only been 7 pushes when the line was posted -3. That is from the beginning of the 2002 season BTW not just midway through. At the time I wrote this there were only 2, No more hit that year, and 5 hit last year. A total of 112 games have been lined at 3 during the 2002 and 2003 season. Simple math shows that that is a percentage of about 6% for the past two seasons.

So not only has it been below 10%, it is a lot closer to 5%.

Now does that mean anything for this year? Maybe maybe not. Some people would have you believe it will regress toward the mean and this year might be higher than normal. I am not so sure.

One other factor that is also lost in this is that football has come back to an era of defences, so lower scoring tighter games are the norm, yet a 3 point game is hard to come by for some reason. To me that only makes my arguments stronger. When it should be expected to be higher and isn't then that makes the buy off even more foolish.

But all this data only matters if you bet every game. Had you bought off only on the 7 that did land on seven then it is a moot point to you. But then again if you are that good to pick the only games that mattered, then you wouldn't need to bet those games in the first place.