Series Price Calculations

[d2bets]

Assuming you knew what the ML would be for each game in a best of 7 series, is there a formula for calculating the correct series price? Hypothetically speaking, let's say the "no-vig" line for a Bulls/Heat series per game is:

Game 1: Bulls -150
Game 2: Bulls -140
Game 3: Heat -220
Game 4: Heat -200
Game 5: Bulls -150
Game 6: Heat -220
Game 7: Bulls -150

What would the correct no-vig series price be, and how do you arrive at that?

[mebaran]

How do you even arrive at anywhere close to accurate numbers for the last few games? There is so much uncertainty with the last few.

I would say simply figure out the odds of each team winning at each arena, then run a simulation based on those two numbers. You can start adding other factors in (-140 for 2nd Bulls home games instead of -150, etc.), but I assume with series pricing, you're not really gaining that much by trying to be uber accurate with every game price.

[d2bets]

How do you even arrive at anywhere close to accurate numbers for the last few games? There is so much uncertainty with the last few.

I would say simply figure out the odds of each team winning at each arena, then run a simulation based on those two numbers. You can start adding other factors in (-140 for 2nd Bulls home games instead of -150, etc.), but I assume with series pricing, you're not really gaining that much by trying to be uber accurate with every game price.

You didn't read my question. That was just a hypothetical. I'm not saying those game numbers are accurate. That's not important. My question is let's assume here are your numbers to work with, whatever they are. Based on those numbers, then how do you mathematically calculate a correct series price. This is a math question, not a capping question.

[Blax0r]

I'll give this a shot.

First let's convert those ML's to win probabilities. I'll do this for the Bull's series price:

Game 1: 150/250 = .6
Game 1: 140/240 = .583
Game 1: 100/320 = .3125
Game 1: 100/300 = .33
Game 1: 150/250 = .6
Game 1: 100/320 = .3125
Game 1: 150/250 = .6

So we need to find the p(bulls win 4 games). Showing the work to get the series price is too much; I'll start the pattern, and probably you'll see what I'm doing. Let's break this down by possible exact series scores:

4 - 0 = p(game 1,2,3,4 win) = .03646
4 - 1 = p(game 1,2,3 win and game 4 loss) + p(game 1,2,4 win and game 3 loss) + and so on. Multiple the resulting sum of this by p(game 5 win) + p(game 5 loss and game 6 win) + p(game 5,6 loss and game 7 win).
... and so on for 4-2 and 4-3.

Add up all of the percentages.

I got ~45%, which is about +122. Could others double-check?

[d2bets]

I'll give this a shot.

First let's convert those ML's to win probabilities. I'll do this for the Bull's series price:

Game 1: 150/250 = .6
Game 1: 140/240 = .583
Game 1: 100/320 = .3125
Game 1: 100/300 = .33
Game 1: 150/250 = .6
Game 1: 100/320 = .3125
Game 1: 150/250 = .6

So we need to find the p(bulls win 4 games). Showing the work to get the series price is too much; I'll start the pattern, and probably you'll see what I'm doing. Let's break this down by possible exact series scores:

4 - 0 = p(game 1,2,3,4 win) = .03646
4 - 1 = p(game 1,2,3 win and game 4 loss) + p(game 1,2,4 win and game 3 loss) + and so on. Multiple the resulting sum of this by p(game 5 win) + p(game 5 loss and game 6 win) + p(game 5,6 loss and game 7 win).
... and so on for 4-2 and 4-3.

Add up all of the percentages.

I got ~35%, which is about +183. Could others double-check?

I vageuly see what you're doing, but in reality, I can't follow. So you came up with 35% Bulls win series? That surprises me, but I can't contradict your math because it's over my head.

[Blax0r]

I'm really just listing every single possible sequence of events that will lead to bulls victory. And I made an error in arithmetic; it should be 45% my bad.

[d2bets]

I'm really just listing every single possible sequence of events that will lead to bulls victory. And I made an error in arithmetic; it should be 45% my bad.

That number intuitively makes more sense to me.

And I do think those odds should be roughly the likely hypotehtical odds. Books should be putting up Miami -135/+115. But they won't. It's going to be hyperinflated up to -180 to -200. Bulls series will have great value.

[Blax0r]

Getting those initial MLs for each game takes far greater skill and analysis than what I just whipped up. If you're confident on those, I'll throw a few on Bulls ML if it's > +122. We'll see though.

[MonkeyF0cker]

I posted a no-vig/series calculator a couple of years ago that does exactly what you're asking.

http://forum.sbrforum.com/handicapper-think-tank/190052-no-vig-series-calculator.html

[LT Profits]

Assuming you knew what the ML would be for each game in a best of 7 series, is there a formula for calculating the correct series price? Hypothetically speaking, let's say the "no-vig" line for a Bulls/Heat series per game is:

Game 1: Bulls -150
Game 2: Bulls -140
Game 3: Heat -220
Game 4: Heat -200
Game 5: Bulls -150
Game 6: Heat -220
Game 7: Bulls -150

What would the correct no-vig series price be, and how do you arrive at that?

Are those No Vig odds? If yes, then I get Bulls +123

[LT Profits]

I posted a no-vig/series calculator a couple of years ago that does exactly what you're asking.

http://forum.sbrforum.com/handicappe...alculator.html

And that is what I used to get the +123. I wish I would have trusted it when I came up with Miami -200 over Boston,

GREAT tool. Thanks.

[d2bets]

I posted a no-vig/series calculator a couple of years ago that does exactly what you're asking.

http://forum.sbrforum.com/handicappe...alculator.html

Thanks! I'll check that out.

[mebaran]

You didn't read my question. That was just a hypothetical. I'm not saying those game numbers are accurate. That's not important. My question is let's assume here are your numbers to work with, whatever they are. Based on those numbers, then how do you mathematically calculate a correct series price. This is a math question, not a capping question.

I am saying that it is hard to cap games that are a week into the future (game 5, 6, 7)...I do understand your question, and yeah I'd attack it like Blaxor did. It's, in essence, a probability tree. You list all possible outcomes (4-0, 4-1, 4-2, 4-3, 3-4, etc) and find probabilities of each, then add up all the ones where Bulls win to find Bulls series win%, then calculate to moneyline price.

[mebaran]

I actually just looked at Monkey's Series calculator, and he does the same thing. He only finds out each team's expected Home Court win %, and then does what I said in my last post. Actually a very nice little program.

[cyberbabble]

NBA series calculator for Excel

sharpsportsbetting.com from Stanford Wong has a downloadable Excel macro for this calculation also. Look for prop tools on the web site.

You can search Stanford Wong and King Yao here or in Google if you haven't heard of them. Both have written good books on sports betting. King Yao has a section in his book on NBA playoff series analysis. The web site is not very active, but the spread sheet still seems to work OK.

[smoke a bowl]

Chi 44.9% to win series or +123 is the correct series price according to those inputs. However the Heat will be in the -300 range in their home games I believe.

[Blax0r]

I posted a no-vig/series calculator a couple of years ago that does exactly what you're asking.

http://forum.sbrforum.com/handicappe...alculator.html

Ah I didn't know about this; I guess I'm still relatively new to SBR. Great app!

I'll likely write a tennis version of it pretty soon in either c# or matlab; probably leaning towards matlab since UI isn't particularly important to me.

[TomG]

To what extent can someone start with the probability of winning a single game on a neutral court for a specific match-up, and then use historical trends to back into home/away and zig-zag theory specifics?

[d2bets]

Chi 44.9% to win series or +123 is the correct series price according to those inputs. However the Heat will be in the -300 range in their home games I believe.

I assume you mean -300 no vig (let's say -320/+280). That would point to a 7/7.5 point line. Really? Bulls as 7 point dogs? Miami was -5 when the Bulls went in there (and won) in March, and both teams were healthy. Miami was -5 first games against Boston. Of course, they won't be in Miami until Game 3 and a lot can happen before that. But all else equal I don't think it's terribly likely that Miami will be laying 7/7.5 points. If the Bulls win a couple close games and it's a must win type deal then maybe it's 6 points, tops. Just my opinion. I'm fairly confident that the no-vig lines for Bulls home games would range from -130 to -150 and Miami would range from -200 to -240. I'm going to play around with that calculator with various inputs later. I have this suspicion that the series line is not going to be well coupled with the reality of the game lines in this series, presenting excellent value.

[smoke a bowl]

I assume you mean -300 no vig (let's say -320/+280). That would point to a 7/7.5 point line. Really? Bulls as 7 point dogs? Miami was -5 when the Bulls went in there (and won) in March, and both teams were healthy. Miami was -5 first games against Boston. Of course, they won't be in Miami until Game 3 and a lot can happen before that. But all else equal I don't think it's terribly likely that Miami will be laying 7/7.5 points. If the Bulls win a couple close games and it's a must win type deal then maybe it's 6 points, tops. Just my opinion. I'm fairly confident that the no-vig lines for Bulls home games would range from -130 to -150 and Miami would range from -200 to -240. I'm going to play around with that calculator with various inputs later. I have this suspicion that the series line is not going to be well coupled with the reality of the game lines in this series, presenting excellent value.

I think the Heat will be -7ish at home and +2ish on the road but I've been wrong before. My database actual makes the no vig ml at -7 flat about 72.8% or -268 however the no vig ml in the playoffs for a game lined on 7 has been about -300 so far this year. I'm not sure if playoff games are played a little tighter to the vest thus making the points worth a nut hair more or if it's just a flaw in the market but I'm guessing the market conversions are pretty close to correct being how big these markets actually are.

[LT Profits]

More like Heat -5 at home, Bulls -2 at home

[LT Profits]

Remember, Heat were only -5 and -4.5 at home vs. Celtcs

[d2bets]

More like Heat -5 at home, Bulls -2 at home

That was basically what I thought, but I cannot reconcile that with the apparent predicted series price from Pinnacle. Just doesn't mesh.

[suicidekings]

That was basically what I thought, but I cannot reconcile that with the apparent predicted series price from Pinnacle. Just doesn't mesh.

It's easy to lower the bar if need be, but setting the asking price high from the start makes sense. From the public perspective, the Heat just smoked the Celtics and the Bulls are struggling against the Hawks.

[smoke a bowl]

More like Heat -5 at home, Bulls -2 at home

I can only pray that you are right. I'll go dead off on the Heat -5 at home vs the Bulls.

[LT Profits]

I can only pray that you are right. I'll go dead off on the Heat -5 at home vs the Bulls.

They can't make Heat a bigger home favorite over the Bulls than they were over the Celtics, Bulls are rated much higher than Boston.

[LT Profits]

That was basically what I thought, but I cannot reconcile that with the apparent predicted series price from Pinnacle. Just doesn't mesh.

Yeah, I am thinking that if the Bulls wins Game 1, then Miami will actually be a small favorite in Chicago in Game 2. If Miami wins Game 1, then it reconciles pretty closely if you leave the rest of the series as you priced it.

[d2bets]

Yeah, I am thinking that if the Bulls wins Game 1, then Miami will actually be a small favorite in Chicago in Game 2. If Miami wins Game 1, then it reconciles pretty closely if you leave the rest of the series as you priced it.

Could be. Although it does depend. Let's say the Bulls wipe the floor wtih Miami by 15 points. Miami is favorites in Game 2? That's ballsy. Now if it's real close but the Bulls pull it out, then yeah, I can see that.

[LT Profits]

I don't think the winning margin will matter, only WHO wins matters. The Celtics were -3 vs. Miami in Game 3, blew them out and then Miami was -2 in Game 4. The Celtics blew out the Lakers in Game 1 of the NBA Finals at home a few years ago, and then the Lakers were -1.5 in Boston in Game 2.

[sneak-a-peak]

The one thing I know is its not often you see the favorite in game 1 be the underdog for the series

[LT Profits]

The one thing I know is its not often you see the favorite in game 1 be the underdog for the series

Eh, it happens more often than you think when the underdog has home court advantage, and it is almost always in this small spread range.

[smoke a bowl]

They were 5, 6 then 7.5 vs Celtics at home I thought. Obviously Rondo helped with the 7.5. I'll stick with Heat 7 maybe as low as 6.5 at home unless something drastic happens in games 1 or 2. Anything under 6 and I will be shocked. I'm down at -176 on Heat series for a nice bet for this reason. Go Heat.

[smoke a bowl]

The one thing I know is its not often you see the favorite in game 1 be the underdog for the series

Only 100% of the time when the fav is at home laying only 2 in game 1.

[LT Profits]

They were 5, 6 then 7.5 vs Celtics at home I thought. Obviously Rondo helped with the 7.5. I'll stick with Heat 7 maybe as low as 6.5 at home unless something drastic happens in games 1 or 2. Anything under 6 and I will be shocked. I'm down at -176 on Heat series for a nice bet for this reason. Go Heat.

They were 5 and 4.5 the first two games. Game 5 would have been close to same if series was 2-2 instead of 3-1

[JustinBieber]

I understand that Bulls can be underdogs in this series but still don't get how they can be +160. Do most people get the correct price to be around +120 when using conservative estimates?