Google
LasVegasAdvisor.com
Web

 

Navigation:


 
FORUMS > LVASports < refresh >
Topic Title: I don't get it
Created On 1/31/08 10:05 AM
Topic View:

Pages: [ 1 2 3 4 >> Next ]
View thread in raw text format


trotman

Posts: 898
Joined: Feb 2007

1/31/08 10:05 AM
User is offline

During baseball season, if people listed a bet at -200 they were considered a square, not once do I recall, Peavy, Sabathia, or Beckett, listed a -400 favorite....

Now we have folks betting NE - 400 or more. Even Fezz on last night's call talked about laying 9-1 on no safety. Is not this the same type of thinking as a horse player (bridge jumper) wagering a great deal of money to show on a solid favorite hoping for a return of a 2.10 or 2.20 pay off....

The Giants have won 10 on the road...something even NE has not done.

What am I missing, are not these square type plays, risking a great deal for a small return?

Edited: 1/31/08 at 10:07 AM by trotman
 
Reply
   
Quote
   
Top
   
Bottom
     



Ramthebuffs

Posts: 4029
Joined: Oct 2006

1/31/08 10:07 AM
User is offline

What your missing is a great baseball team wins 60% of its games.

-------------------------
Be Champions!
lines411.com

Contact the pedophile watch groups!!! I have an 18 year old as my avatar.

Edited: 1/16/08 at 6:30 PM by TA/SPV
 
Reply
   
Quote
   
Top
   
Bottom
     



nhreapa

Posts: 861
Joined: May 2007

1/31/08 10:10 AM
User is offline

I particularly love ignoring the 6,000 things the Patriots have done more impressively than NY to say that NY has won 10 games on the road, something not even the Patriots have done. Of course, the Patriots denied themselves this opportunity by having a good regular season and finished a mere 8-0 on the road.

In my opinion, there is a whole lot you are missing.


Edited: 1/31/08 at 10:10 AM by nhreapa
 
Reply
   
Quote
   
Top
   
Bottom
     



jakem921

Posts: 1301
Joined: Feb 2006

1/31/08 10:14 AM
User is offline

What you are missing is that the ML for a 12 point fav is around -775. We could possibly get -400. That's a good bet.

Same thing for the no OT and no Saftey props. Fezz said fair price on OT is around 25:1. When you can do better, bet it.

-------------------------
RIP Marcos
 
Reply
   
Quote
   
Top
   
Bottom
     



trotman

Posts: 898
Joined: Feb 2007

1/31/08 10:16 AM
User is offline

Ram,

That is true....one other stat worth mentioning that someone else posted earlier this month...the QB with the greatest number of passing yards during the regular season has never won the Super Bowl...we all know who the leading passer was this year....

I am not saying NE won't win....but this game is starting to take on a look of like the folks who went and mortgage themselves to the hilt to buy a a second or third condo, home etc hoping to flip it...
 
Reply
   
Quote
   
Top
   
Bottom
     



ddss6_99

Posts: 14
Joined: Oct 2006

1/31/08 10:25 AM
User is offline

Quote

Originally posted by: trotman

The Giants have won 10 on the road...something even NE has not done.


...That's because NE has only had to play 8 games on the road
 
Reply
   
Quote
   
Top
   
Bottom
     



EastCoastBias

Posts: 1194
Joined: Aug 2005

1/31/08 11:24 AM
User is offline

Basically what you are missing is VALUE. Its ok to bet the 4:1's if they have value, meaning that they should be 7:1.
 
Reply
   
Quote
   
Top
   
Bottom
     



dcpdcp

Posts: 501
Joined: Nov 2006

1/31/08 11:43 AM
User is offline

Quote

Originally posted by: jakem921
What you are missing is that the ML for a 12 point fav is around -775. We could possibly get -400. That's a good bet.

Same thing for the no OT and no Saftey props. Fezz said fair price on OT is around 25:1. When you can do better, bet it.


Yes, value is always relative.

 
Reply
   
Quote
   
Top
   
Bottom
     



blogguy

Posts: 1111
Joined: Jul 2007

1/31/08 11:59 AM
User is offline

He's not missing out on VALUE, he's wondering why people think that a 4-1 favorite offers value on a neutral field against another playoff caliber team when we've all spent our whole lives seeing that there's not as much difference as people think between elite teams and the next rung when they play on neutral fields in the postseason or in championship formats in other sports.

Sharps have spent decades taking advantage of that. In this case, they're going the other way and laying the big number...basically saying that the Patriots straight up are a stronger bet than pocket aces over another pocket pair pre-flop in holdem.

The differences between the ML and pointspread could just as easily be suggesting that the PS is wrong rather than the ML is wrong. There's also a chance both are wrong, as we've seen often in neutral field games matching playoff caliber teams in football (college and pro) and in March Madness.

Unfortunately, they're only going to play the game once...and that one result won't really shed light on whether NE should have been an 88-12 favorite (-730 on the moneyline equivalent), an 85-15 favorite (-566) an 80-20 favorite (-400) or a 75-25 favorite (-300).

Would be cool if we could come up with a way to properly assess the issue without just looking at a conversion chart that consistently errs in neutral field postseason contests...and would be even further off if the current line of 12 or 12.5 is inflated on the team that's 1-7 its last eight games against expectations.

Sample size issues are always going to be an issue in this kind of discussion. Is there a way to get around that?

Any statheads want to try the Pythagorean approach?

New England: 589-274 for the regular season
NY Giants: 373-351 for the regular season
If we only look at regular season differentials, how often should a team with NE's ratio beat a team with NYG's ratio?

New England: 641-306 in all games
NY Giants: 441-402 in all games
Same question for all games

New England: 310-183 since start of November
NY Giants: 241-241 since start of November
Same question for games since start of November, which ironically hurt NYG's differential because they won a lot of close games but lost big to Minnesota

Might help us get closer to a reasonable answer that would confirm either side of the debate...or confirm the current moneyline.

-------------------------
My blog: http://statintelligence.blogspot.com
 
Reply
   
Quote
   
Top
   
Bottom
     



King Yao

Posts: 849
Joined: Oct 2004

1/31/08 12:02 PM
User is offline

Quote

Originally posted by: blogguy
He's not missing out on VALUE, he's wondering why people think that a 4-1 favorite offers value on a neutral field


Instead of focusing on the money line, the focus should be on the point spread. The fact is that -400 in the money line is a fantastic bet when the point spread is -12. Taking +12 and -400 is a nice combination. If -400 is too high, then -12 is too high as well, to a larger relative degree.

So, I suggest let's forget about focusing on a -400 favorite, but just focus on whether the -12 is right. Even if the -12 is wrong and the true line should be -10, -400 is still a good number on the fave compared to laying -10 -110.
 
Reply
   
Quote
   
Top
   
Bottom
     



DanielSong39

Posts: 3552
Joined: May 2004

1/31/08 12:20 PM
User is offline

It's all about value.

It's not square to make a bet at -200, -400, or -10000 if you're still getting bang for your buck.

But it IS square to assert that -200, -400, or -10000 can NEVER have value. That is simply not true and one can dig up a long list of counterexamples.

This is starting to sound like that "sour grapes" fable. Not everyone can win with bridgejumpers but it doesn't mean it's impossible. Opportunities do pop up at every possible range in the spectrum, and it's the job of a good gambler to take advantage of them.

-------------------------
"Betting offshore is illegal, Jon. No one does that anymore." - Baker, 12/6/07 conference call


Edited: 1/31/08 at 5:01 PM by DanielSong39
 
Reply
   
Quote
   
Top
   
Bottom
     



blogguy

Posts: 1111
Joined: Jul 2007

1/31/08 12:45 PM
User is offline

Sorry King, have to respectfully disagree...particularly coming from a starting point where the community PR made the line 7, as have other math models I've seen.

To me the value on the two things can (and should) be judged independently. I don't think there's necessarily a strong correlation between the pointspreads and the moneylines in these kinds of games. It's either a blowout or it isn't. Some scattergrams are bell curves with a strong correlation to the traditional relationship, but others are dumbells where results cluster in a way that makes outright upsets more common than people realize.

What the "right" number is and how often does the favorite wins are separate issues. I think the arbitrary link between the two has been exposed as faulty in neutral site games matching quality teams in the sports that have that characteristic.

And, in the case of the Pats, we're also talking about:
*a 24-point favorite that won 31-28 with a 410-391 edge
*a 19-point favorite that won 27-24 with a 327-376 yardage deficit
*a 21-point favorite that won 20-14 with a 265-236 yardage edge
*a 14-point favorite that won 38-35 with a 390-316 yardage edge
*a 13-point favorite that won 31-20 with a 403-350 yardage edge
*a 14-point favorite that won 21-12 with a 347-311 yardage edge

This is a team that the market has consistently been overrating the past several weeks, who now has a quarterback playing with a limp. They have definitely done a great job of finding ways to win when they don't have monster stat edges. To me there's a fine line between suggesting they're a 78-22 favorite (which would sound about right for a 7-8 point favorite give or take some change), or an 80-20 favorite, or an 85-15 favorite in terms of what's going to happen in just one given game. I don't see the evidence that -400 is a "great" play given the evidence. And, I don't think the fact that a market that has been badly missing Patriots games for two months opened the line at 14 and currently them at 12 is a defense for it being a great play.

It's kind of a futile debate, because we all think the Pats are going to win...and it's basically whether they should be a -350 or -500 favorite as best I can tell amongst the participants. I think, in general, the handicapping world should be dealing with ML and PS as different issues rather than something that's correlated in a causative way.

Basketball teams who live by the three and die by the three don't have the same correlation to ML results as medium favorites in neutral site games that teams with strong defenses do. In football, it's the same with high octane offenses. Once they're in a situation that nullifies some of the octane, they don't win as often as the market expects. I know a 40-year sample size still isn't enough to confidently "prove" that. There's no sample size that says it's not true in these kinds of games either. New England's upset of St. Louis wasn't very abberant in terms of what we've seen in our lifetimes when a team that can play ball control is matched up against a team that ran up impressive scoring margins much of the season. Upsets have happen more than 20% of the time in that kind of matchup, even if you include home field games...and particularly if you're looking at 7-10 as the key margin strike range rather than 12-14.

That's my 50th version of the longwinded take anyway, lol...

What's the best way to determine what percentage of time a team will win a game? Why would trusting a market that's been consistently wrong about a team be the right way to do that?



-------------------------
My blog: http://statintelligence.blogspot.com
 
Reply
   
Quote
   
Top
   
Bottom
     



King Yao

Posts: 849
Joined: Oct 2004

1/31/08 1:01 PM
User is offline

Jeff, good stuff as always, even if I think you are wrong, and you think I am wrong.

Follow-up stuff:

"To me the value on the two things can (and should) be judged independently. I don't think there's necessarily a strong correlation between the pointspreads and the moneylines in these kinds of games. It's either a blowout or it isn't. Some scattergrams are bell curves with a strong correlation to the traditional relationship, but others are dumbells where results cluster in a way that makes outright upsets more common than people realize. "

Show me why this game is a dumbbell (dumbell or dumbbell?) and why the traditional relationship of PS/ML doesn't hold...without saying anything about why you think the Patriots are over-rated by fans and linemakers. I think this is the crux of the disagreement and where it gets interesting...well interesting for me because I haven't heard that argument until you mentioned it last night in the conference call.

I don't mean to pick on wording...but the part where you write "It's either a blowout or it isn't" -- what do you mean by that? That's the case in every game, given that you've defined what a blowout is. I mean, can a game be both a blow out and not a blowout at the same time?

Edited: 1/31/08 at 1:03 PM by King Yao
 
Reply
   
Quote
   
Top
   
Bottom
     



EastCoastBias

Posts: 1194
Joined: Aug 2005

1/31/08 1:10 PM
User is offline

But the odds off a 4:1 dog in baseball winning are much greater than the Superbowl going into OT, because that has NEVER happened, and baseball teams consisntly lose. So you are paying a little more than 2x the amount for something that has never happened, and should be 25:1, rather than betting 4:1 on something that may happen 25% of the time.
 
Reply
   
Quote
   
Top
   
Bottom
     



blogguy

Posts: 1111
Joined: Jul 2007

1/31/08 1:10 PM
User is offline

Nothing from the Pythagorean contingent yet. Let's assume the Giants are a generic opponent that don't represent anything positive themselves, and just look at the Pats.

A team with their scoring totals should win (based on points scored squared---try saying THAT out loud!--- over points scored squared plus points allowed squared)
Regular Season: 82%
Regular Season plus Playoffs: 77%
Games Since November 1st: 74%

That's why we're talking in the range we're talking in. Full season numbers would show the Pats barely on the right side of -400 on the assumption that the Giants are a typical opponent (rather than a playoff caliber opponent). If you include the playoffs in the mix, they drop below the -400 threshold. If you only look at games since November first, they drop barely below the -300 threshold.

Now, I 'm all for giving them bonus points because of Bellicheck/Brady vs. Coughlin/Eli. But, the Giants are also representing themselves as a better than "non-entity" team in terms of the math. They played the 6th toughest schedule in the regular season, then won three road games against divisional winners. To me, that all cancels out...the Patriots will surpass their own Pythagorean numbers to a degree because of their brains...but the Giants are better than they're typical opponent.

If you focus on the full season, and assume Brady's injury is irrelevant...then 82% would be fine.
If you focus on what's happened since the defense started wearing down, it becomes harder to justify -400 by a good bit.
If Brady's injury is going to be a factor...

I don't see how you can mathematically get to New England -400 is a smart bet. It's still probably going to win anyway. That's where the value lies? Only if you accept that a market that's been mispricing the Patriots for two months at home with a healthy quarterback suddenly got them right on a neutral field with a potentially hobbled quarterback.

Debatable to me, (obviously, lol). They're still very likely to win...

-------------------------
My blog: http://statintelligence.blogspot.com


Edited: 1/31/08 at 1:13 PM by blogguy
 
Reply
   
Quote
   
Top
   
Bottom
     



King Yao

Posts: 849
Joined: Oct 2004

1/31/08 1:22 PM
User is offline

Quote

Originally posted by: blogguy
I don't see how you can mathematically get to New England -400 is a smart bet.


I'm not sure if you are asking me directly, but I'll answer anyway.

-400 is a good bet versus -12. That's the rationale for why -400 is a good bet. The pointspread and moneyline are related in my opinion. But you said in your opinion, they may not be in this game (dumbell curve), so I'll reiterate my question: why do you think this game is a dumbell curve? IMO, it is a much larger stretch to say this one game is different from all other 5,000 NFL games played in the last 20 years than to say that this one game is still somewhat similar to the other 5,000 NFL games played in the last 20 years. Are they playing with a rugby ball in this game? Are they not playing with pads in the Super Bowl? OK, I admit the last two questions were lame, just trying to lighten the thread. Its not all that serious since I'm not betting your money and your not betting mine.


Quote

Originally posted by: blogguy
It's still probably going to win anyway. That's where the value lies?


Of course not.
 
Reply
   
Quote
   
Top
   
Bottom
     



blogguy

Posts: 1111
Joined: Jul 2007

1/31/08 1:26 PM
User is offline

King, let me try it this way.

Favorites in really the 7-14 range in football that match up a high octane offense versus a ball control defense don't show a bell curve scattergram. By, "it's either a blowout or it isn't," I mean:

*Either the ball control team keeps it close and turns the game into a nailbiter, or grabs an even more comfortable win when the favored team panics and make mistakes (TB/Oakland would be an example of this but the line wasn't really in the strike zone...but that kind of game...TB ball control, Oakland high octane team posting big offensive numbers)

*Or, the ball control team fails to do what it hoped, and implodes in an effort to catch up.

So, the scattergram clusters in the +3 to -3 range on one end, and anything 21-plus on the other.

It's not a bell curve with most of the results landing near the spread...then a mountain going down both sides to the extremes. Relatively few games land near the middle, and the results cluster at the extremes.

Why are the Patriots that kind of team? Regular season results from high to low this year:

3, 3, 3, 4, 6, 17, 21, 21, 21, 21, 21, 24, 24, 31, 45, 46

Five results from 0-6
One result fro 7-20
10 results from 21-up

The playoff games were closer to the vest...both were played in cold weather, and saw the Patriots successfully run clock with a lead. Those games landed on 11 and 9...though the nine came against a team that was missing stars on offense and had a one-legged quarterback who was even more hobbled than Brady was in that game. Adjusting for home field, you get 8 and 6 as neutral equivalents.

If you want lists of other big games that had barbell potential...many of the 12 double digit SB spreads probably fit that category...though I think SF would have killed SD a million times out of a million that year (lol). Chicago had a nice edge over New England because it would have been very difficult for the Pats to own the clock with Eason or a banged up Grogan against the outstanding Bears defense. It's vulnerable defenses for the favorite that typically trigger the bar bell. That allows the opponent to run clock or put points on the board. But, if the lesser team makes a bunch of mistakes, the victory margin blows way up.

Does that explain it yet? Sometimes it's hard to get across what I'm driving at with something like this...

-------------------------
My blog: http://statintelligence.blogspot.com


Edited: 1/31/08 at 1:27 PM by blogguy
 
Reply
   
Quote
   
Top
   
Bottom
     



King Yao

Posts: 849
Joined: Oct 2004

1/31/08 1:30 PM
User is offline

Jeff, very interesting, thanks. Other than the Patriots, do you have other data?
 
Reply
   
Quote
   
Top
   
Bottom
     



blogguy

Posts: 1111
Joined: Jul 2007

1/31/08 1:32 PM
User is offline

"IMO, it is a much larger stretch to say this one game is different from all other 5,000 NFL games played in the last 20 years than to say that this one game is still somewhat similar to the other 5,000 NFL games played in the last 20 years"

Certainly not different than all other pro or college games. There have been many samples of high octane offenses running into teams that can control the ball against them...these are the types to show the barbell. On a neutral field in particular the high octane teams are generally overpriced by a good bit. Expectations are high because of a history of big margins. The things that lead to big margins are much less in place. Pro and college football...for years and years. I'd probably be happier if we split those 5,000 games into

"what happens when the best team has a dominating defense" and "what happens when they have a vulnerable defense"

Different animals that create different dynamics.

But, even without that dynamic in place...New England's Pythagorean stuff only shows them at 74% and 77% in the latter two projections. Wish we had more statheads posting here! Pythagorean stuff all over the kenpom site and knickerblogger.net for the nba...

-------------------------
My blog: http://statintelligence.blogspot.com
 
Reply
   
Quote
   
Top
   
Bottom
     



King Yao

Posts: 849
Joined: Oct 2004

1/31/08 1:36 PM
User is offline

Quote

Originally posted by: King Yao
Jeff, very interesting, thanks. Other than the Patriots, do you have other data?



Better yet, let's do this: assuming you haven't looked for more data yet, name me all the teams from 1989 to 2007 that you think have high-powered offenses. I can quickly go through my database and spit out the games they played in the three ranges you listed. I don't like sticking with one team like the Patriots...if we can come up with 10 teams, and the data reflects similarities...then I think you'd have something.
 
Reply
   
Quote
   
Top
   
Bottom
     

Pages: [ 1 2 3 4 >> Next ]
View thread in raw text format
FORUMS > LVASports < refresh >

Navigation: