Inefficient NFL markets?

[Justin7]

At about 1:50 am EST (before time change), Matchbook has:

New Orleans -6.5 -101 / +6.5 +100

and NO moneyline at -268/+250. The ml will obviously tighten as we get closer to game time.

I contend this is due to a somewhat rare NFL inefficiency. I speculate that these lines will move apart -- either NO spread will move up (in the direction of -6.5 -105 or -7 +105), or that the dog ML will move down below +250.

I'd love to get Durito's take on this set of lines, with an explantion.

[donjuan]

Crush your stat arb Daringly.

[JustinBieber]

What was the comparison to pinnacle at that moment in time?

[Peregrine Stoop]

and what was the limit at pinnacle at the time?

[Justin7]

9:09 am EST. MB has the two same prices.

Pinny has +239/-270, and -6.5 -101 / +6.5 -107

[wrongturn]

Because road fav -6.5 win straight less than 70.7% of time, so the fair money line should be less than -242.

[Shonner]

Because road fav -6.5 win straight less than 70.7% of time, so the fair money line should be less than -242.

Wrong, you have to take into account what the total is.

Spread to ML conversions without factoring the Total are practically useless.

[Justin7]

Wrong, you have to take into account what the total is.

Spread to ML conversions without factoring the Total are practically useless.

Totals have an effect, but less than I would have guessed.

[Shonner]

Totals have an effect, but less than I would have guessed.

If you look at the Spread/ML converter (which I realize isn't perfect because it doesn't take into account totals) the fair price of a -6.5 (-101/100) is +262 / -266.

Considering the +250 was less than the +262 per the ML calculator, how did you realize/figure the line was off? The ML calculator would suggest it should have moved up to +260, not -245.

You were right as it's gone down to -246 / +244

[wrongturn]

I was wrong about 70.7%. It is actually 74.20%, although in sample of only 62 games since 1989.

[djiddish98]

I think the impact of totals are over-rated. Otherwise, wouldn't it cost more to buy off the 3 in a game where the total is 50 vs. 40?

[Shonner]

I was wrong about 70.7%. It is actually 74.20%, although in sample of only 62 games since 1989.

That's why I am wondering how the line of +250 can be considered off...and why it dropped to +245

ML on a team with an estimated win % of 25% is +280 or so.

Justin, can you please comment on this - how did you conclude that the +250 was too high and it should be lower considering these facts?

Thx

[RickySteve]

I think the impact of totals are over-rated. Otherwise, wouldn't it cost more to buy off the 3 in a game where the total is 50 vs. 40?

It should cost less.

[Michaelmakesit]

At about 1:50 am EST (before time change), Matchbook has:

New Orleans -6.5 -101 / +6.5 +100

and NO moneyline at -268/+250. The ml will obviously tighten as we get closer to game time.

I contend this is due to a somewhat rare NFL inefficiency. I speculate that these lines will move apart -- either NO spread will move up (in the direction of -6.5 -105 or -7 +105), or that the dog ML will move down below +250.

I'd love to get Durito's take on this set of lines, with an explantion.

is this an indication that NO was the bet of the day JUstin ?
did u put $ on this game Justin and pound it hard!?

[MonkeyF0cker]

So the Matchbook ML market was slow to move. What's the point?

[Pancho sanza]

Either you believe they are efficient or you don't, can't pick and choose pockets and say this one is, this one isn't.

[brumbies]

Final score: New Orleans win 34:3

Does the spread of -6/-6.5/-7 matter at all?

[MonkeyF0cker]

Either you believe they are efficient or you don't, can't pick and choose pockets and say this one is, this one isn't.

Behavioral economists would actually argue otherwise. However, for those who subscribe to the EMH (which Justin knowingly or not seems to) this would be true.

[jgilmartin]

Final score: New Orleans win 34:3
Does the spread of -6/-6.5/-7 matter at all?

Over the long term, yes, most certainly.

[brumbies]

Over the long term, yes, most certainly.

Really? Look at today's games or the weeks before and tell me how many of them ended up dead close to the spread, be it opening or closing lines.

[donjuan]

Really? Look at today's games or the weeks before and tell me how many of them ended up dead close to the spread, be it opening or closing lines.

http://www.sbrforum.com/betting-tools/half-point-calculator.aspx

This has all been hashed out before. You are wrong.

[Dark Horse]

Really? Look at today's games or the weeks before and tell me how many of them ended up dead close to the spread, be it opening or closing lines.

Off the top of my head for today:
KC opened +3, closed OAK +1. Middle on dogs won/pushed
BUF +3 pushed
DET +4 lost by 3
IND +3 lost by 2
HOU +3 lost by 6, after just missing late TD and getting FG instead (4 pts)
TB +10, lost by 6
ARI +7, lost by 3


And that for a schedule with just 12 games.

[Michaelmakesit]

Off the top of my head for today:
KC opened +3, closed OAK +1. Middle on dogs won/pushed
BUF +3 pushed
DET +4 lost by 3
IND +3 lost by 2
HOU +3 lost by 6, after just missing late TD and getting FG instead (4 pts)
TB +10, lost by 6
ARI +7, lost by 3


And that for a schedule with just 12 games.

[brumbies]

Off the top of my head for today: KC opened +3, closed OAK +1. Middle on dogs won/pushed BUF +3 pushed DET +4 lost by 3 IND +3 lost by 2 HOU +3 lost by 6, after just missing late TD and getting FG instead (4 pts) TB +10, lost by 6 ARI +7, lost by 3 And that for a schedule with just 12 games.

It does not matter whether you had the best number.

For example, whether you had -8.5 or -10 Hawks, you still lose. And whether you had +8.5 or +10 Hawks, you still win.

Whether you had -3.5 or -6 Jets, you still lose. And whether you had +3.5 or +6 Lions, you still win.

Etc etc.

[u21c3f6]

My answer is: The markets are very efficient overall. A season of ATS or odds will have an overall win rate very close to 50% or the win rate for those particular odds. The individual games however do not always have to be (and IMO are not) very efficient. This inefficiency can be exploited further through the use of correlated wagers.

If the closing lines were perfectly efficient, then the only way you could win would be to BTCL and that is just not the case. You can be profitable without regards to BTCL. Using validated criteria that produce a profit, some of your wagers will and some will not end up BTCL. Your wager does not become better or worse just because it does or does not BTCL, it just means that you got a better or worse price.

This is not to say that you cannot be profitable using BTCL. BTCL just does not have the same significance for someone using profitable validated criteria to select their wagers.

Joe.

[Wrecktangle]

The efficient market guys (in the real market, i.e. the stock market, not this thinly traded, easily manipulated sports thing we all peck at) state that you can't really make a profit by trading (over the cost of trading, i.e the vig) because the best price is baked into the current price.

So, if you bet sports you 1) cannot be an efficient market person or 2) have a gambling problem.

I happen to believe in a semi-efficient market: if you naively bet it, you will get eaten by the vig. on the other hand if you devote enough numerical resources, you can win.

[Thremp]

I don't think anyone believes in strong form efficiency in any market.

[Indecent]

It does not matter whether you had the best number.

For example, whether you had -8.5 or -10 Hawks, you still lose. And whether you had +8.5 or +10 Hawks, you still win.

Whether you had -3.5 or -6 Jets, you still lose. And whether you had +3.5 or +6 Lions, you still win.

Etc etc.

So we can assume games won't end with a final score difference of 9, or 4-6, etc? This week it didn't happen so we don't need to worry about it.... wonderful logic.

You're are looking at one week and trying to draw conclusions from what happened, try thinking about about the big picture. Take the lions/jets example. How often is the final score difference 3, 4, or 5? How much value did you lose by betting -6 instead of -3.5?

These are fairly simple concepts. If you are resistant to using this line of thinking in your handicapping, you should give in now.

[djiddish98]

It should cost less.

Correct - buying off the 3 with a total of 50 should cost less in theory (I got my numbers and words mixed up.)

However, it costs the same, doesn't it? (looking at pinny).

[Peregrine Stoop]

It does not matter whether you had the best number. For example, whether you had -8.5 or -10 Hawks, you still lose. And whether you had +8.5 or +10 Hawks, you still win. Whether you had -3.5 or -6 Jets, you still lose. And whether you had +3.5 or +6 Lions, you still win. Etc etc.

wrong... having +3 mattered a whole ton on the Chiefs
In theory, you wouldn't be betting every side. However, even if you did, the 1 out of 14 push instead of loss is a huge difference in ROI.

[Peregrine Stoop]

Correct - buying off the 3 with a total of 50 should cost less in theory (I got my numbers and words mixed up.) However, it costs the same, doesn't it? (looking at pinny).

Justin7 hinted at this. I've looked at it in a db as well. Totals make some difference, but not as much as you would think in the value of a 3 or 7, especially when going up from 40 to 50. It makes more difference in going down from 40 to 30, which leads to some interesting NFLX spots.

[Justin7]

Justin7 hinted at this. I've looked at it in a db as well. Totals make some difference, but not as much as you would think in the value of a 3 or 7, especially when going up from 40 to 50. It makes more difference in going down from 40 to 30, which leads to some interesting NFLX spots.

NFLx has a very different distribution than NFL, even for games with identical spreads and totals.

[brumbies]

So we can assume games won't end with a final score difference of 9, or 4-6, etc? This week it didn't happen so we don't need to worry about it.... wonderful logic. You're are looking at one week and trying to draw conclusions from what happened, try thinking about about the big picture. Take the lions/jets example. How often is the final score difference 3, 4, or 5? How much value did you lose by betting -6 instead of -3.5? These are fairly simple concepts. If you are resistant to using this line of thinking in your handicapping, you should give in now.

If you scroll up, you would have seen my reply to jgilmartin. I asked him to list me games over the past few weeks where the spread truly mattered at all.

Of course once in a blue moon you will have games with spreads that matter, but for the most part, it all boils down to whether you can pick a winner. You can have all the best numbers, but what's the point if they all lose?

I remembered I had Denver+3.5 against the Jets the other week. I made sure to get the best price and number. I ended up losing by the hook.

If getting the best numbers truly matter, why not simply bet on whichever 2nd half NBA team that has a better spread as compared to pregame? For eg, Lakers -10. They were down 15 at the half. 2nd half spread is -7. That makes the Lakers as +8 dogs. So isn't that value? Why can't I win simply by blind betting on such scenarios? I clearly had the best numbers.

[donjuan]

if you scroll up, you would have seen my reply to jgilmartin. i asked him to list me games over the past few weeks where the spread truly mattered at all.

Of course once in a blue moon you will have games with spreads that matter, but for the most part, it all boils down to whether you can pick a winner. You can have all the best numbers, but what's the point if they all lose?

I remembered i had denver+3.5 against the jets the other week. I made sure to get the best price and number. I ended up losing by the hook.

If getting the best numbers truly matter, why not simply bet on whichever 2nd half nba team that has a better spread as compared to pregame? For eg, lakers -10. They were down 15 at the half. 2nd half spread is -7. That makes the lakers as +8 dogs. So isn't that value? Why can't i win simply by blind betting on such scenarios? I clearly had the best numbers.

gtfo

[Data]

Of course once in a blue moon you will have games with spreads that matter

Points don't matter... I am sure I heard this before. I recall it was on Wed/day...