Which line is the better value?

**texhooper** · 02-10-23, 12:05 AM

Well if you’re taking the dog you take 4 if those are your two options. I think that is fairly obvious and I don’t really think that’s what you’re asking.

I think what you’re really asking is between -3.5 and -4 what’s better, and while I am no scientist, I’d imagine that over the course of a bazillion plays that -4 at a lower price than -3.5 is gonna work in your favor in the long run.

All things being equal, that is. The sport is gonna matter too because -4 is gonna have different push rates in NFL, NCAAF, NBA, NCAAB, etc, and that half point will have different values depending on what league you’re talking about. I’d venture to say though regardless of league in the long run you would benefit more from taking -4 at -105 1000 times rather than -3.5 at -110 1000 times. You can hit a lower strike rate at -105 which will likely overcome the push rate, but that’s just a guess, and this is a pretty good question honestly. Someone will probably clarify this and tell you I’m wrong

**Nate rasta** · 02-10-23, 12:06 AM

Welcome to the forum

**turbobets** · 02-10-23, 12:23 AM

I don't know if this is updated but it will get you started.

https://www.sportsbookreview.com/betting-calculators/half-point-calculator/

**DownEastAaron** · 02-10-23, 12:28 AM

Appreciate the quick response and it makes perfect sense. However, I am more concerned with the actual math and percentages. It appears that turbobets may have directed me to a calculator that does the math. Hopefully, it has the accompanying formula as well. Thank you again!

**ChuckyTheGoat** · 02-10-23, 12:32 AM

Originally posted by DownEastAaron

Appreciate the quick response and it makes perfect sense. However, I am more concerned with the actual math and percentages. It appears that turbobets may have directed me to a calculator that does the math. Hopefully, it has the accompanying formula as well. Thank you again!

Big thing in sports-betting is to have access to estimated value of each landing point.

It won't be perfect but needs to be approximate. From there, one can calculate estimated ROI.

Really, calculating the value across points is not too tough. The hard part is generating an algorithm that gets you to a projection that is superior to the market line. Very tough.

**pologq** · 02-10-23, 07:17 AM

if you have a chance to get more points at a lower price do it. not sure why you would not.

now if it was the other way and you were giving 4 or 3.5 i would give the 4. 3.5 is basically 4 and i would rather the -105 price.

**milwaukee mike** · 02-10-23, 08:09 AM

Originally posted by ChuckyTheGoat

Big thing in sports-betting is to have access to estimated value of each landing point.

It won't be perfect but needs to be approximate. From there, one can calculate estimated ROI.

Really, calculating the value across points is not too tough. The hard part is generating an algorithm that gets you to a projection that is superior to the market line. Very tough.

good post

it's not just as simple as "that 1/2 point is worth 5 cents"... different sports mean different calculations

**DownEastAaron** · 02-10-23, 07:09 PM

This will be a challenge to incorporate into an algorithm. I am an R developer and am looking to incorporate this into a function. I know it's possible because there are several websites, including this one, that have done so. I have functions that calculate the push frequencies as far back as 2015.

**DownEastAaron** · 02-10-23, 07:30 PM

Chucky, do you have a formula that you would share for calculating the values?

**DownEastAaron** · 02-10-23, 07:31 PM

Originally posted by ChuckyTheGoat

Big thing in sports-betting is to have access to estimated value of each landing point.

It won't be perfect but needs to be approximate. From there, one can calculate estimated ROI.

Really, calculating the value across points is not too tough. The hard part is generating an algorithm that gets you to a projection that is superior to the market line. Very tough.

Chucky, do you have a formula that you would share for calculating the values?

**Optional** · 02-10-23, 07:44 PM

A user named "Ganchrow" created the half point calculator here, many years ago.

He would have been asked about this many times and I think he would have given out all his math for it too. He was good like that.

But it might take you a fair bit of searching to find it. I could not with a quick search. This post of his might be a good spot to start looking https://www.sportsbookreview.com/for...equencies.html

**ChuckyTheGoat** · 02-10-23, 08:02 PM

Originally posted by DownEastAaron

Chucky, do you have a formula that you would share for calculating the values?

See post 12 by Opti.

I'm saying that the exact values vary a little. Depends on the years you use.

IMHO, the best way to do it is to estimate F(x+o.5) and
F(x-0.5). Divide by 2 and you have your half-pt value.

**KVB** · 02-10-23, 08:05 PM

Originally posted by DownEastAaron

Hi All. This is my first post so forgive any code of ethics violations. I am struggling to come up with a way to compare lines/odds to determine which is better. Let's say I have the following

-3.5/-110
3.5/-110

-4/-105
4/105

Is there a mathematical formula that will compare these lines to determine which is the better value? Thank you in advance.

Yes, there is math here as it's a conversion. But I think we may need to rework your question a bit.

For the sake of the math, we have to figure out, first, what is fair in the market. Which one of these bets are we going to consider a fair market bet?

Let's say, as most would, that the -110's are fair. So we have Eagles -3.5 (-110) vs Eagles -4(-105). That reasons out. You are selling a 1/2 point for "5 cents".

But how does that look on the other side, the dog? I'm getting +3.5 for -110 and can buy 1/2 point at +105? (I am assuming you meant +105 there, but it's the same at -105). If -3.5 (-110) is "fair" and -4 is worth -105 then +4 here must be -115 or something with a higher price. If the favorite got cheaper a half point higher, then the dog must get more expensive.

So, let's say you meant to write this...

-3.5 (-110)
+3.5 (-110)

-4 (-105)
+4 (-115)

While we could use the +105 in our conversion, I think it's best we look at it as though we are realistically comparing two offerings. Regardless, we can do the math.

**KVB** · 02-10-23, 08:06 PM

The math we're looking for here is the Expected Value of the play.

Expected value is what you can gain minus what you can lose. That's the odds of winning times what you can win minus the odds of losing times what you can lose. We express it, often, in percentage by dividing by what is risked.

Let's say we have an expectation of 50% on that +3.5(-110) spread. 50% of the time you win, 50% of the time you lose.

Let's say you risk $110 to win $100. 50% of the time you win $100, 50% of the time you lose $110. Divide by the $110 Risk.

(.50) (100) - (.50) (110) =

50-55 = -5

Divided by the risk and you get -5/110 or an expected value of -4.55%.

So let's look at +4 (-115).

Here, we have to consider the push. Say we are dealing with the NFL, and we figure that the current push rate for 4 points in the NFL is 5%. That means we determine, in this case that 5% is the chances we tie on our +4 (-115) bet.

If we are still dealing with a 50% expectation, then we win 47.5% of the time, lose 47.5% of the time, we tie the other 5%.

So now we use our EV formula and get...

(.475)(100) - (.475)(115) =

47.5 - 54.6 = -7.1

Divide by the risk and you get -7.1/115 or an EV of -6.17%

With these numbers it's close, and both EV's are negative, but taking the +4 (-115) is a worse bet at -6.17% EV as opposed to the +3.5 (-110) that is only -4.55%

**Bostongambler** · 02-10-23, 08:08 PM

Leave it to KVB to bring brains 🧠 into answer.

**KVB** · 02-10-23, 08:13 PM

You can plug in whatever numbers you like there, and even change your expectation of winning from 50%.

But one thing I've noted, and have mentioned that push rates can be nebulus in some videos, is that most want to figure push rates based just on historical margins of victory and the chances of games landing on specific numbers.

I'm here to say, and haven't heard anyone at SBR say it yet, that this is only part of the story. There is much more to it.

First, we can take variations of margins of victory and play with recency to help hone the number. Some posters know this.

What what many fail to mention is the effect of the Total line on these push rates.

A push rate calculator that does not allow you to input a Total, especailly in the NFL, will be lacking. Likewise, anyone assessing push rates without Totals, could be lacking too.

How much those methods are lacking depends.

There, I said it. Consider your Totals in push rates, folks. Sometimes I wonder just how many readers will be able to truly appreciate that piece of advice.

Good Luck.

**KVB** · 02-10-23, 08:22 PM

Originally posted by Bostongambler

Leave it to KVB to bring brains 易 into answer.

**pavyracer** · 02-10-23, 08:24 PM

The better value will always be the one that gives you the best chance to win your bet.

**KVB** · 02-10-23, 08:59 PM

Originally posted by pavyracer

The better value will always be the one that gives you the best chance to win your bet.

This makes no sense.

If your not considering the cost and benefit of those probabilites, then you aren't concering yourself with value.

I can increase my chances of winning a bet, significantly, but if the cost of that increase is too high, it lack any value. I can get KC spread bet at +1.5 for -105, basically the currently fair line of +1.5. I can increase my chances of winning my KC spread bet by taking KC +3.5. But if I have to pay -300 for it, then I increased my chances of winning, but lessened my value by an awful lot.

**Optional** · 02-10-23, 09:17 PM

Originally posted by pavyracer

The better value will always be the one that gives you the best chance to win your bet.

Does that mean you would auto take -3.5 ahead of a -4?

**BuckyOne** · 02-10-23, 09:20 PM

Originally posted by KVB

You can plug in whatever numbers you like there, and even change your expectation of winning from 50%.

But one thing I've noted, and have mentioned that push rates can be nebulus in some videos, is that most want to figure push rates based just on historical margins of victory and the chances of games landing on specific numbers.

I'm here to say, and haven't heard anyone at SBR say it yet, that this is only part of the story. There is much more to it.

First, we can take variations of margins of victory and play with recency to help hone the number. Some posters know this.

What what many fail to mention is the effect of the Total line on these push rates.

A push rate calculator that does not allow you to input a Total, especailly in the NFL, will be lacking. Likewise, anyone assessing push rates without Totals, could be lacking too.

How much those methods are lacking depends.

There, I said it. Consider your Totals in push rates, folks. Sometimes I wonder just how many readers will be able to truly appreciate that piece of advice.

Good Luck.

Entertaining numbers. What is the numbers quoted were really available and there was no mistake? +4 -105 is the best number as far as least negative expectation???

(.475)(100) - (.475)(105)

47.5 - 49.875 = -2.375

-2.375/105 = - 2.26%

So, in my opinion, we found the best number but we still do not have positive expectations and we do not have anything of value????

**KVB** · 02-10-23, 09:24 PM

Originally posted by Optional

Does that mean you would auto take -3.5 ahead of a -4?

Let's do it for the favorite in the example.

Same numbers for -3.5 (-110) as +3.5 (-110).... -4.55% EV.

Now let's do -4 (-105)

(.475)(100) - (.475)(105) = -2.375/105 = -2.3% ev

Again both negative, but texhooper's instinct above was right in this case.

**KVB** · 02-10-23, 09:35 PM

Originally posted by BuckyOne

Entertaining numbers. What is the numbers quoted were really available and there was no mistake? +4 -105 is the best number as far as least negative expectation???

(.475)(100) - (.475)(105)

47.5 - 49.875 = -2.375

-2.375/105 = - 2.26%

So, in my opinion, we found the best number but we still do not have positive expectations and we do not have anything of value????

Yup, same formula.

But the original post actually does have +4 (+105), even if I suggested he meant -105.

Now lets bet that $100 to win 105, and reverse the actual EV...

(.475)(105) - (.475)(100) = 2.375/100 = 2.375%

Or even stick with the same $110 dollar risk....

(.475)(115.5) - (.475)(110) = 2.6125/110= 2.375%

Again, to show that you can interchange values, but that is the formula the OP is seeking.

And yes, if we consider -/+3.5 (-110) the fair line here then of course the +4 (+105) is actually not just the best value, but profitable as well.

**pavyracer** · 02-10-23, 09:53 PM

Originally posted by KVB

This makes no sense.

If your not considering the cost and benefit of those probabilites, then you aren't concering yourself with value.

I can increase my chances of winning a bet, significantly, but if the cost of that increase is too high, it lack any value. I can get KC spread bet at +1.5 for -105, basically the currently fair line of +1.5. I can increase my chances of winning my KC spread bet by taking KC +3.5. But if I have to pay -300 for it, then I increased my chances of winning, but lessened my value by an awful lot.

It makes sense. Look at Drake. He bet 1 million on Argentina to win the World Cup in 90 min.

If he bet Argentina pk he would have pushed his bet and not be pourer by 1 million now.

**pavyracer** · 02-10-23, 09:55 PM

Originally posted by Optional

Does that mean you would auto take -3.5 ahead of a -4?

If I capped the game with my lines and figured out that the game will land at 4 points differential I would have bet the -3.5.

It all depends on what you think the line should be and not what line the book is offering you.

**BuckyOne** · 02-10-23, 10:27 PM

Originally posted by KVB

Yup, same formula.

But the original post actually does have +4 (+105), even if I suggested he meant -105.

Now lets bet that $100 to win 105, and reverse the actual EV...

(.475)(105) - (.475)(100) = 2.375/100 = 2.375%

Or even stick with the same $110 dollar risk....

(.475)(115.5) - (.475)(110) = 2.6125/110= 2.375%

Again, to show that you can interchange values, but that is the formula the OP is seeking.

And yes, if we consider -/+3.5 (-110) the fair line here then of course the +4 (+105) is actually not just the best value, but profitable as well.

Yes, well done. That is what we are all looking for. How to define value. Then, the next step is to find it.

**KVB** · 02-10-23, 10:33 PM

Originally posted by pavyracer

If I capped the game with my lines and figured out that the game will land at 4 points differential I would have bet the -3.5.

It all depends on what you think the line should be and not what line the book is offering you.

Curiously, do you even recognize any contradiction between the two sentences in your post?

Maybe any at all? Maybe something? Maybe just some contradiction? Perhaps an inkling?

Do you?

It's ok to say "no" I was just curious what you thought there.

In my opinion, it depends both on what I think the line should be and what is offered in the market. In fact, I have many different levels of lines, but they are useless if the book doesn't offer the right kind of bet.

Of course, there's more than one way to skin that cat.

**d2bets** · 02-10-23, 10:43 PM

Why all the discussion? Any basketball line I'll take -3.5-110 over -4-105. Doesn't matter what you capped the game at. This is not a difficult decision.

**Nate rasta** · 02-10-23, 10:43 PM

This is beyond my pay grade

**KVB** · 02-10-23, 11:01 PM

Originally posted by d2bets

Why all the discussion? Any basketball line I'll take -3.5-110 over -4-105. Doesn't matter what you capped the game at. This is not a difficult decision.

Our discussion outlines why one would make that decision. The OP wanted to understand the math behind it, or at least some formula to add to his program. I assumed football here, in the example.

**Nate rasta** · 02-10-23, 11:03 PM

Personally in football I'll never take a favorite at -3.5 or -4. Has to be 3 or less. I'd rather lay the juice on the money line. Basketball is different of course

**Covering the #** · 02-10-23, 11:42 PM

I rarely make any single bets to make less than the bet itself. Especially if it's a bet giving away points, whether it's -3 or -10. Your best long term gain and avoiding the vig is to find a correlating play within that game based on why you are on that side of the number. For example, if you like the under in the Super Bowl, you will expect and need turnovers. You can bet the under and a fumble lost at +160 as a parlay, instead of just betting the under at -110 or whatever. It's best to use your noggin once in awhile and eliminate unnecessary juice.

**texhooper** · 02-10-23, 11:42 PM

Looks like I was wrong after all, which I predicted, meaning I am still right

**KVB** · 02-10-23, 11:50 PM

Originally posted by texhooper

Looks like I was wrong after all, which I predicted, meaning I am still right

I think you were right. Maybe not about overcoming the push rate, but it's true that in this example taking the -4 (-105) was better in the long run that the -3.5 (-110), at least in this example where I used 5% as push rate for a 4 point margin in the NFL.

As you said, it still works out in the bettor's favor, even if it just means losing less...

Originally posted by KVB

Let's do it for the favorite in the example.

Same numbers for -3.5 (-110) as +3.5 (-110).... -4.55% EV.

Now let's do -4 (-105)

(.475)(100) - (.475)(105) = -2.375/105 = -2.3% ev..