Better Lines vs Better Price

[austintx05]

Wanted to create a discussion here. When I place my bets I always try to get the best line with the best price, but this can be very hard to do. With the NBA/NCAAb season just underway, I am at the point where I am evaluating what my goal should be. In doing so, I try to set a limit of juice that I would pay. I try never to pay more than 3% juice. I know that sounds very stingy, but since I use Matchbook & Mansion Betting Exchanges it really helps and if you don't already use them, you might be suprised at what good prices you can get.

I want to use this week's NFL Game as an example.

Carolina -4 vs Washington

I like Washington in this game and at Pinnacle you can get Washington +4 (-103) or at Bowman's you can get Washington +4.5 (-110).

Personally I prefer to take the low juice and hope the half point will not matter, but sometimes its that half point that can save you. I have heard from several bettors say that if you win , you don't pay juice. While this is true, no one knows what is going to happen in any game and I just know when it comes to me setting out my goals for any given season that money management and keeping my juice as minimal as possible is key to turning a profit.

So what is your stand on this matter? Do you take a line at half a point less for minimal juice, or do you pay full price for a 1/2 point more?

[Ganchrow]

To evaluate the relative importance of various points, check out the following charts:

http://forum.sbrforum.com/pro-football/18502-estimated-nfl-ats-push-frequencies.html

http://forum.sbrforum.com/college-football/18521-estimated-ncaa-football-ats-push-frequencies.html

http://forum.sbrforum.com/hoops/20924-estimated-nba-ats-over-under-push-frequencies.html

[austintx05]

Thanks for the info Ganchrow.

[pags11]

austin,

excellent thread you have going here...we had a big discussion on this last year...I think you want to evaluate each circumstance individually based on what sport it is (basically how key the number really is) and what the juice/ line trade-off is...I'd say about 75% I take the lower juice, 25% line the better number...that's just me though...

[Dark Horse]

Personally I prefer to take the low juice and hope the half point will not matter, but sometimes its that half point that can save you. I have heard from several bettors say that if you win , you don't pay juice. While this is true, no one knows what is going to happen in any game and I just know when it comes to me setting out my goals for any given season that money management and keeping my juice as minimal as possible is key to turning a profit.

So what is your stand on this matter? Do you take a line at half a point less for minimal juice, or do you pay full price for a 1/2 point more?

If you win, you do pay juice. Losing bets are 'free'. The argument isn't very important, but if you and I make a $100 bet and I win, you're out of a 100 bucks. If a middle man is involved, you will still be out of a 100 bucks, but I only get 90 something. So the winner pays the juice.

In basketball, don't mess with half pts in favor of better price. 1/2 pts in basketball are huge.

In football it all depends on the key spreads.

If you're going to be overly rigid about juice you'll miss many good opportunities. When a book like Pinny moves to -112, that means they don't want action on that side. The question is why.

[Kaps]

It all comes down to value. In your situation when you have a +4-103 versus +4.5-110 the better value is +4.5-110.
In order for the +4 to be a better value than the +4.5 then your +4 should either be +4even or +4+101 or better.
3 and 7 are your biggest key numbers in football but 4,6,10 are somewhat keys.
If you had the same situation with dead numbers then always take the lower juice. Like 1,2,5,8,9,11,12,13,

[Ganchrow]

If you win, you do pay juice. Losing bets are 'free'. The argument isn't very important, but if you and I make a $100 bet and I win, you're out of a 100 bucks. If a middle man is involved, you will still be out of a 100 bucks, but I only get 90 something. So the winner pays the juice.

It's really just two sides of the same coin.

Before the bet is placed, your expectation of juice paid is of course the theoretical hold.

Once the bet has been decided, however, it doesn't really make sense to say that only one side or the other is paying the juice. It all depends on how you look at it -- do you place a bet to risk a specified amount or do you place a bet to win a specified amount?

If the former, then you can nominally say that the winner is "paying" the juice in that he wins less on any given unit bet won.

If the latter, then you can nominally say that the loser is "paying" the juice in that he loses more on any given unit bet lost.

Two sides of the same coin.

[TLD]

Ganchrow is exactly correct and saved me the trouble.

Tout J.R. Miller is to blame for spreading this notion that you only pay juice on a winning bet.

My challenge to those who believe that is this: Let’s say you decided to open a book that charges the loser juice and not the winner (i.e., reverses the way books supposedly do it now). Explain to me, with examples, how your book would differ from current practice.

(And the answer to the original question is, of course, “it depends.” I’ll gladly bet +3, -112 instead of +2.5, –110. I’ll also gladly bet +5, –105 instead of +5.5, –115. There can’t be some general rule about whether it’s wiser to take the better line or the better odds. Case-by-case.)

[Dark Horse]

Too theoretical for my taste. The middle man solves the riddle for me. Bet without a middle man and nobody pays any juice. Add the middle man and the winner ends up with less than without the middle man. It makes no difference for the loser.

In other words, you want to pick up the tab! Just as it's not necessarily a bad thing to pay a lot of taxes, as this reflects your income.

(but like I said, the argument is unimportant. TLD, I had the same discussion with J.R. Miller once, with me taking your position. lol.)

[Ganchrow]

Bet without a middle man and nobody pays any juice. Add the middle man and the winner ends up with less than without the middle man. It makes no difference for the loser.

Or conversely, the loser ends up paying more than without the middle man. Just two sides of the same coin.

TLD, I had the same discussion with J.R. Miller once, with me taking your position. lol.

And J.R. Miller was wrong. How surprising.

[Ganchrow]

I posted the following in another thread regarding how to use push probabilities to determine the fair price of buying or selling a half-point:

How can use this data to see the value of buying (or maybe selling, if one is using pinnacle) a half point or more at certain numbers?

The results will be slightly different whether you're moving on to or off of an integer line.

Let's assume you're moving on to an integer line that occurs with probability p, and further assume that you're buying a half point, and that the theoretical hold on the unadorned line is given by H.

If the probability of the buy team winning the bet before the half point was purchased is given by b, then the win probability post-purchase would still be b, the push prob would be p, and the loss prob would be 1-p-b. Hence the relative win prob would be given by b/(1-p), and the fair decimal odds, figuring in a theoretical hold of H, would just be:

(1 - H) × (1 - p)
----------------   (buying a half point on to an integer)
      &nbs p;    b

And using similar logic, in the case of selling a half point on to an integer spread (relative win prob post-purchase of [b-p]/[1-p]), fair decimal odds would be given by:

(1 - H) × (1 -p)
----------------   (selling a half point on to an integer)
      &nbs p; b - p

<hr>
Using a specific example:

Purdue -7½ -102
Northwestern +7½ -108

This implies a Purdue ATS win prob of 49.30%, a Northwestern ATS win probability of 50.70%, and a theoretical hold, H, of 2.36%.

So let's first consider buying a half point on Purdue (b=49.30%). From the chart above we see that the spread 7 has an associated probability, p, of 5.45%. Using the buy-on-to-an-integer-formula we find that fair decimal odds (assuming zero book profit other than the original hold) for the purchase of the half point should be: (1-2.36%)*(1-5.45%) / 49.30% ≈ 1.8726, corresponding to US-style odds of about -114.60. Pinnacle is offering the line at -117, implying (if you trust the predictive power of the push frequencies above) that it's charging a player an extra 0.93% worth of hold make the half point purchase.

In the case of the sale of a half point on Northwestern (b = 50.70%) fair decimal odds would be given by (1-2.36%)*(1-5.45%) / (50.70% - 5.45%) ≈ 2.0402, or US-style odds of +104.02. Compare this to Pinnacle's line of +103 on the offered half-point, and you see that you're giving up an extra 0.49% worth of hold to make the sale.
<hr>
Anyway, the logic is nearly identical in the case of buying or selling off of an integer spread and I leave the following as an exercise for the interested reader:

Assuming no change in theoretical hold, and taking the push probabilities above as given, what money line should be associated with a side of Fave -6½ given a market of:

NCAA FB Fave -7 -102
NCAA FB Dog +7 -108

[Dark Horse]

Ganch, how does the loser pay more with a bookmaker?

a) 100 bet between you and me. Loser loses 100, winner wins 100.

b) Same bet at bookmaker with -110 juice. Loser loses 100, winner takes 91.

Occam's Razor (wikipedia): ... when given two equally valid explanations for a phenomenon, one should embrace the less complicated formulation.

[Yoshi]

If you win, you do pay juice. Losing bets are 'free'.

Thats hilarious

[Ganchrow]

Ganch, how does the loser pay more with a bookmaker?

a) 100 bet between you and me. Loser loses 100, winner wins 100.

b) Same bet at bookmaker with -110 juice. Loser loses 100, winner takes 91.

b') Bet to win the same amount at bookmaker with -110 juice. Loser loses 110, winner takes 100.

So is a) the analogue of b) or is it the analogue of b')? There's really no strictly "correct" answer -- it all just comes down to convention and/or betting style.

In <a href=http://www.amazon.com/Fixed-Odds-Sports-Betting-Statistical/dp/1843440199/sr=8-1/qid=1164261995/ref=pd_bbs_sr_1/103-6849374-8616645?ie=UTF8&s=books TARGET=_blank>Fixed Odds Sports Betting: Statistical Forecasting and Risk Management</a>, for example, Joseph Buchdahl demonstrates how a bettor engaging in "fixed profits" staking can reduce both his standard deviation and risk-of-ruin versus a flat bettor with the same average bet size. In the case of such a wagering structure, b') clearly provides a more appropriate description of that bettor's reality than b).

As I wrote before:

Do you place a bet to risk a specified amount or do you place a bet to win a specified amount?

If the former, then you can nominally say that the winner is "paying" the juice in that he wins less on any given unit bet won.

If the latter, then you can nominally say that the loser is "paying" the juice in that he loses more on any given unit bet lost.

Just two sides of the same coin.

And of course, TLD posed a rather apt question:

Let’s say you decided to open a book that charges the loser juice and not the winner (i.e., reverses the way books supposedly do it now). Explain to me, with examples, how your book would differ from current practice.