PDA

View Full Version : Playoff Odds (based on Vegas point spreads)


blaze1984
12-27-2009, 09:11 PM
edited: now that pinnacle finally posted the money lines, i have much more exact calculations of our playoff chances. It started off at 13% or so when i first ran the numbers, with a math error we were down to 12%... and now that pinnacle posted their money lines, we're down to under 7%...

okay, so i've broken down the steelers playoff chances based on the math of each possible combo.... First things first, they just posted the point spreads for all the games (pinnaclesports.com is my source for info, they're the sharpest sportsbook on the net). Next, they haven't posted moneylines yet, so i took the approximate odds of victory based on moneylines of other games from previous weeks with approximately the same point spreads.

Other thoughts, the Jets and Texans being large favorites is a huge surprise to me. But i guess the Bengals / Patriots have very little to play for. Also, the Colts screwed us this week. Taking out the jets would've just about tripled our playoff odds. And if Miami didn't collapse in the first half against the Texans, we could've really improved our odds that way too. Anyways, hopes are pretty slim right now looking at it from a purely mathematical standpoint, but you never know...

Based on the early point spreads posted:

Ravens ML posted at -460 & 9.5 point favorites @ Oak (1/5.6 chance of losing or ~~18%)
Broncos ML posted at -610 & 12.5 point favorites vs. KC (1/7.1 chance of losing or ~~14%)
Jets ML posted at -485 & 8 point favorites vs. Cincinnati (1/5.85 chance of losing or ~~17%)
Texans ML posted at -385 & 7 point favorites vs. New England (1/4.85 chance of losing or ~~20.5%)

1) Steelers win @ Miami
2) One of the following
a. Jets Loss and Texans Loss -- 17% X 20.5% = 3.5%
b. Texans Win, Ravens Loss, Broncos Loss, Jets Loss -- 79.5% X 18% X 14% X 17% = 0.3%
c. Jets Win, Ravens Loss, Texans Loss -- 83% X 18% X 20.5% = 3%

6.8% chance to make the playoffs assuming we beat Miami... It's not looking good right now

Wow, i wish i hadn't actually calculated this...

DoctorCAD
12-27-2009, 09:31 PM
I still have my ticket bought in Tahoe that gives me 10 to 1 that the Steelers win the Super Bowl.

That was after the first pre-season game when everybody thought we would cake-walk in.

Steelers17
12-27-2009, 09:41 PM
okay, so i've broken down the steelers playoff chances based on the math of each possible combo.... First things first, they just posted the point spreads for all the games (pinnaclesports.com is my source for info, they're the sharpest sportsbook on the net). Next, they haven't posted moneylines yet, so i took the approximate odds of victory based on moneylines of other games from previous weeks with approximately the same point spreads.

Other thoughts, the Jets and Texans being large favorites is a huge surprise to me. But i guess the Bengals / Patriots have very little to play for. Also, the Colts screwed us this week. Taking out the jets would've just about tripled our playoff odds. And if Miami didn't collapse in the first half against the Texans, we could've really improved our odds that way too. Anyways, hopes are pretty slim right now looking at it from a purely mathematical standpoint, but you never know...

Based on the early point spreads posted:

Ravens 9.5 point favorites @ Oak (22% chance of losing)
Broncos 12.5 point favorites vs. KC (17.5% chance of losing)
Jets 8 point favorites vs. Cincinnati (25% chance of losing)
Texans 7 point favorites vs. New England (27% chance of losing)

1) Steelers win @ Miami
2) One of the following
a. Jets Loss + Texans Loss --- 25% X 27% = 6.75%
b. Ravens Loss, Broncos Loss, & Jets Loss --- 22% X 17.5% X 25% = 1%
c. Ravens Loss + Texans Loss --- 22% X 27% = 6%

13.75% chance to make the playoffs assuming we beat Miami... It's not looking good right now

Hines was spot-on with his earlier comments that each game is critical along the way! TY for the stats! :tt03:

steel9guy
12-27-2009, 09:50 PM
HOW THE HECK ARE WE BEHIND HOUSTON????????????????

steve314
12-27-2009, 10:22 PM
okay, so i've broken down the steelers playoff chances based on the math of each possible combo.... First things first, they just posted the point spreads for all the games (pinnaclesports.com is my source for info, they're the sharpest sportsbook on the net). Next, they haven't posted moneylines yet, so i took the approximate odds of victory based on moneylines of other games from previous weeks with approximately the same point spreads.

Other thoughts, the Jets and Texans being large favorites is a huge surprise to me. But i guess the Bengals / Patriots have very little to play for. Also, the Colts screwed us this week. Taking out the jets would've just about tripled our playoff odds. And if Miami didn't collapse in the first half against the Texans, we could've really improved our odds that way too. Anyways, hopes are pretty slim right now looking at it from a purely mathematical standpoint, but you never know...

Based on the early point spreads posted:

Ravens 9.5 point favorites @ Oak (22% chance of losing)
Broncos 12.5 point favorites vs. KC (17.5% chance of losing)
Jets 8 point favorites vs. Cincinnati (25% chance of losing)
Texans 7 point favorites vs. New England (27% chance of losing)

1) Steelers win @ Miami
2) One of the following
a. Jets Loss + Texans Loss --- 25% X 27% = 6.75%
b. Ravens Loss, Broncos Loss, & Jets Loss --- 22% X 17.5% X 25% = 1%
c. Ravens Loss + Texans Loss --- 22% X 27% = 6%

13.75% chance to make the playoffs assuming we beat Miami... It's not looking good right now

13.75% is probably an overestimate, as you are double-counting some scenarios.

For example
Jets Loss and Texans Loss and Ravens Loss is counted in both (a) and (c) above.

To get mutually exclusive events for which you can add probabilities, can do

a. Jets Loss and Texans Loss -- 25% * 27% = 6.75%
b. Texans Win, Ravens Loss, Broncos Loss, Jets Loss -- 73% * 22% X 17.5% X 25% = about 0.7%
c. Jets Win, Ravens Loss, Texans Loss -- 75% * 22% * 27% -- about 4.5%

This gives a total of about 12%.

Also, we're assuming the probabilities are independent. If the Patriots win, and the Bengals have the night game, they have nothing to play for, and their probability of winning could go even lower.

blaze1984
12-27-2009, 10:53 PM
13.75% is probably an overestimate, as you are double-counting some scenarios.

For example
Jets Loss and Texans Loss and Ravens Loss is counted in both (a) and (c) above.

To get mutually exclusive events for which you can add probabilities, can do

a. Jets Loss and Texans Loss -- 25% * 27% = 6.75%
b. Texans Win, Ravens Loss, Broncos Loss, Jets Loss -- 73% * 22% X 17.5% X 25% = about 0.7%
c. Jets Win, Ravens Loss, Texans Loss -- 75% * 22% * 27% -- about 4.5%

This gives a total of about 12%.

Also, we're assuming the probabilities are independent. If the Patriots win, and the Bengals have the night game, they have nothing to play for, and their probability of winning could go even lower.

They still haven't posted a line for the Steelers game, probably cause of the Ricky Williams injury. So can't really convert the odds. i didn't really care to factor that into the odds anyways... well, cause i'm just assuming we were going to win out...

Point well taken Steve, didn't factor in mutually exclusive events... i'll update first post to reflect that...

Fortunately there is no night game... all the games that matter are scheduled for 1pm on 1/3/10 (minus the Baltimore game, which is a 4 pm game). But again, Oakland and Baltimore both have just as much to play for no matter what else happens. Baltimore is in if they win, and Oakland is just playing for pride. So hopefully, they will actually be independent events. Again, this is just a rough estimate. The lines are constantly moving... The Denver line is all the way up to 13 and even 14 at some sites... so that losing % should technically drop...

Anyways, just a decent math estimate. The chances are slim. But hey, we can always hope...

One last thought... I don't see the Pats playing soft against anyone. This is a team that kept their star players in till late in the 4th quarter of 42-0 games in their perfect season

steve314
12-28-2009, 12:04 AM
They still haven't posted a line for the Steelers game, probably cause of the Ricky Williams injury. So can't really convert the odds. i didn't really care to factor that into the odds anyways... well, cause i'm just assuming we were going to win out...

Point well taken Steve, didn't factor in mutually exclusive events... i'll update first post to reflect that...

Fortunately there is no night game... all the games that matter are scheduled for 1pm on 1/3/10 (minus the Baltimore game, which is a 4 pm game). But again, Oakland and Baltimore both have just as much to play for no matter what else happens. Baltimore is in if they win, and Oakland is just playing for pride. So hopefully, they will actually be independent events. Again, this is just a rough estimate. The lines are constantly moving... The Denver line is all the way up to 13 and even 14 at some sites... so that losing % should technically drop...

Anyways, just a decent math estimate. The chances are slim. But hey, we can always hope...

One last thought... I don't see the Pats playing soft against anyone. This is a team that kept their star players in till late in the 4th quarter of 42-0 games in their perfect season

As I feared, the CIN/NYJ game is now the night game.
The NFL office didn't do the Steelers any favors there.
See http://www.nfl.com/schedules
If NE wins, CIN will probably rest most of the starters and lose.

SH-Rock
12-28-2009, 12:16 AM
Here is the playoff picture according to NFL.com if it was to start today

http://www.nfl.com/news/story?id=09000d5d8147d1c0&template=no-right-rail-with-comments&confirm=true

Baltimore and Houston is a head off the Steelers and Jets are in. At this point in time, because the NFL moved the CIN/NYJ game the Baltimore Ravens have to lose and same with Texans. This will be very difficult and the playoff picture is blurring out. C'mon you Okies, you can beat Steelers,Eagles, Broncos, Bengals you can surely beat the Ravens

markymarc
12-28-2009, 10:32 AM
HOW THE HECK ARE WE BEHIND HOUSTON????????????????

Because they have a better conference record.

Triggerfish
12-28-2009, 10:42 AM
Because they have a better conference record.

Nope....same Conference record. Houston has a better record in Common Games, from what I can gather. However, I'm still trying to make sense of that one, as the rule states there must be a minimum of 4 common games. There are four common teams that both PIT and HOU played, and HOU does have a better record against those teams. However, one of those four teams is TEN, who Pittsburgh only played once (and beat), while Houston played them twice and split. So, how can you count TEN as a "common game" when Houston got to play them twice and won one and lost one? Which result do you take? The win or the loss? Or should TEN get thrown outta the equation? Confusing, but from all reports I've seen that's how they win the tie-breaker over us.

steelerchad
12-28-2009, 10:49 AM
Nope....same Conference record. Houston has a better record in Common Games, from what I can gather. However, I'm still trying to make sense of that one, as the rule states there must be a minimum of 4 common games. There are four common teams that both PIT and HOU played, and HOU does have a better record against those teams. However, one of those four teams is TEN, who Pittsburgh only played once (and beat), while Houston played them twice and split. So, how can you count TEN as a "common game" when Houston got to play them twice and won one and lost one? Which result do you take? The win or the loss? Or should TEN get thrown outta the equation? Confusing, but from all reports I've seen that's how they win the tie-breaker over us.

Because the same holds true for Cincy. We played them twice and lost. So in the end if we beat Miami next week, we will be 2-3 against common opponents. The Texans will be 4-1 against those same 4 opponents.

Triggerfish
12-28-2009, 11:27 AM
Because the same holds true for Cincy. We played them twice and lost. So in the end if we beat Miami next week, we will be 2-3 against common opponents. The Texans will be 4-1 against those same 4 opponents.

Ok...that makes sense. I forgot to count the Bengals twice for us.

markymarc
12-28-2009, 12:31 PM
Nope....same Conference record. Houston has a better record in Common Games, from what I can gather. However, I'm still trying to make sense of that one, as the rule states there must be a minimum of 4 common games. There are four common teams that both PIT and HOU played, and HOU does have a better record against those teams. However, one of those four teams is TEN, who Pittsburgh only played once (and beat), while Houston played them twice and split. So, how can you count TEN as a "common game" when Houston got to play them twice and won one and lost one? Which result do you take? The win or the loss? Or should TEN get thrown outta the equation? Confusing, but from all reports I've seen that's how they win the tie-breaker over us.

Good deal and thanks for clarifying. Not good when Houston has the edge in common games :chuckle:

GoSlash27
12-28-2009, 12:44 PM
Question: How do you convert the point spread to a percent chance of winning?

CPanther95
12-28-2009, 03:00 PM
Nope....same Conference record. Houston has a better record in Common Games, from what I can gather. However, I'm still trying to make sense of that one, as the rule states there must be a minimum of 4 common games. There are four common teams that both PIT and HOU played, and HOU does have a better record against those teams. However, one of those four teams is TEN, who Pittsburgh only played once (and beat), while Houston played them twice and split. So, how can you count TEN as a "common game" when Houston got to play them twice and won one and lost one? Which result do you take? The win or the loss? Or should TEN get thrown outta the equation? Confusing, but from all reports I've seen that's how they win the tie-breaker over us.

It's common opponents - minimum of 4 games. Pittsburgh's record against OAK, CIN, TEN is 1-3 ........ Houston's record is 3-1.

The fact that Houston played the Titans twice - and Pittsburgh played Cincy twice doesn't matter - they go by win percentage.

It's possible that you could have a team that played two of your division opponents and you'd be counting 6 games vs 5 games. They just look at the winning percentage.

7/39/43
12-28-2009, 03:07 PM
RIGHT OFF OF THE STEELERS WEBSITE
halve to cher for the Bungals and the Cheaters do what we have to and then kick their buts when we make it in!:tt04:

Monday, December 28, 2009


PITTSBURGH STEELERS 2009 PLAYOFF SCENARIOS
(provided by the NFL)


The Pittsburgh Steelers clinch a playoff spot:

1) PIT win + HOU loss or tie + NYJ loss or tie

2) PIT win + HOU loss or tie + BAL loss or tie

3) PIT win + NYJ loss or tie + BAL loss or tie + DEN loss or tie

blaze1984
12-28-2009, 03:21 PM
Question: How do you convert the point spread to a percent chance of winning?

Technically you can't.... i compared similar point spreads from previous weeks and looked at the money lines.

Usually there's a 1-2 day delay on posting of money lines. Point spreads are usually the first thing posted. Anyways, pinnacle finally posted their money lines, so i'm going to update the original post with the moneyline calculations

CPanther95
12-28-2009, 03:23 PM
Vegas point spreads are all about the likelihood that money will be bet equally on each side. It has nothing to do with the chance of winning.

blaze1984
12-28-2009, 03:40 PM
Vegas point spreads are all about the likelihood that money will be bet equally on each side. It has nothing to do with the chance of winning.

That's not entirely true. I couldn't find the website that i originally saw the graphic on, but they have compiled data over the last 20 years that show how many % of the time a team that's a 7, 8, or any other point spread underdog will win. You can extrapolate that data to estimate the winning percentage of a team given a particular point spreads.

My original guesses based on the point spreads would've all been within 5% or so, except the Jets and Texans lines have both moved towards the Jets and Texans. Meaning all the early action has come in on those teams. So, assuming they had posted the money lines on the day i posted my calculations, they would've all been within about 2-5% of the actual data (some might've been closer, cause even the Ravens and Broncos lines have moved towards the Ravens and Broncos)

Now with the corrected odds, the Steelers playoff hopes look even more gloomy. Anyways, i'm happy to see the team playing better, and let's forget all the math and just do our part and hope for the best. Even though i'm a man who believes in numbers and logic, i still have faith...

Venom
12-28-2009, 03:45 PM
So the odds are 1,000,000 to 1 ? So we have a chance !!!!! ( taken from dumb and dumber )

steelreserve
12-28-2009, 04:03 PM
All of this is a nice amusement, but it completely ignores the fact that Vegas odds for the last week of the season are basically as accurate as the blindfold/dartboard method. You can throw a 10-point spread right out the window, because nobody knows what the hell is going to happen. The Rams could blow out the Saints by 30 points in Week 17 and I wouldn't be surprised.

JEFF4i
12-28-2009, 04:08 PM
Don't tell me the odds.

There are no odds in life.

Preacher
12-28-2009, 04:15 PM
Don't tell me the odds.

There are no odds in life.

Not true. There are always odds.

Its just that YOU don't know WHICH side YOU will be on.

So, If I play the lotto, and the odds are 1,400,235,254 to 1, I don't know if I will be the one.

The odds are still real.

If the odds are that the Steelers will lose a football game 80 percent of the time on a certain Sunday, you still play the game, because you don't know if you will end up on the 80 or the 20 percent.

Odds don't dictate life, they enumerate real events and the chances of how they will come out.

JEFF4i
12-28-2009, 04:19 PM
Not true. There are always odds.

Its just that YOU don't know WHICH side YOU will be on.

So, If I play the lotto, and the odds are 1,400,235,254 to 1, I don't know if I will be the one.

The odds are still real.

If the odds are that the Steelers will lose a football game 80 percent of the time on a certain Sunday, you still play the game, because you don't know if you will end up on the 80 or the 20 percent.

Odds don't dictate life, they enumerate real events and the chances of how they will come out.

Aha, knew someone would pick this out. You are right, of course, but you see what I'm saying?

It is what it is.

Preacher
12-28-2009, 04:30 PM
Aha, knew someone would pick this out. You are right, of course, but you see what I'm saying?

It is what it is.

If you are saying what I think your saying.. its "Don't let the odds dictate your actions in life"

To which I agree for the most part.

But there are some odds which must be paid attention to... SOME people survive a jump from a 3 story building, but I wouldn't recommend it! :chuckle:

GoSlash27
12-28-2009, 05:46 PM
Something I noticed:

I ran all the possible permutations of next week's games that affect our chances of getting into the playoffs. Assuming that we beat the Dolphins (else why bother looking), there are 32 permutations of results.

Of those 32 possible scenarios, 14 place the Steelers in the playoffs (third place behind Baltimore and NYJ, each with 16). This is far ahead of the next most likely contender.
Interesting but not important: It is impossible for the Jets to get in as #6 seed if the Steelers win.

Strong correlation:
12 of the 14 successful scenarios involve a Houston loss.

If all possible permutations were equal (which of course they're not), the Steelers would have a 44% chance of making the playoffs. With a Houston loss, this increases to 75%. With a Houston win, it drops to 13%.

Current ranking by this method:
#1 New York Jets (16 avenues, all lead to #5 seed) 50%
#2 Baltimore Ravens (16) 50%
#3 Pittsburgh Steelers (14) 44%
#4 Denver Broncos (9) 28%
#5 Houston Texans (8) 25%
#6 Jacksonville Jaguars (1) 3%

Doesn't really mean anything. Just thought it was interesting...

CPanther95
12-28-2009, 05:51 PM
Russian Roulette has great odds.

steelerchad
12-28-2009, 06:01 PM
Something I noticed:

I ran all the possible permutations of next week's games that affect our chances of getting into the playoffs. Assuming that we beat the Dolphins (else why bother looking), there are 32 permutations of results.

Of those 32 possible scenarios, 14 place the Steelers in the playoffs (third place behind Baltimore and NYJ, each with 16). This is far ahead of the next most likely contender.
Interesting but not important: It is impossible for the Jets to get in as #6 seed if the Steelers win.

Strong correlation:
12 of the 14 successful scenarios involve a Houston loss.

If all possible permutations were equal (which of course they're not), the Steelers would have a 44% chance of making the playoffs. With a Houston loss, this increases to 75%.

Current ranking by this method:
#1 New York Jets (16 avenues, all lead to #5 seed) 50%
#2 Baltimore Ravens (16) 50%
#3 Pittsburgh Steelers (14) 44%
#4 Denver Broncos (9) 28%
#5 Houston Texans (8) 25%
#6 Jacksonville Jaguars (1) 3%

Doesn't really mean anything. Just thought it was interesting...

Thanks for posting. I thought it was interesting as well. It's kind of funny how a couple of weeks ago it looked like both the Jets and Texans had impossible schedules to win out at the end of the season. But you don't think about the resting possiblilty. If the Jets win and get in, they would have possibly had basically 2 byes the last 2 weeks of the season against 2 teams who on paper were better teams than they were. When you factor in how short the 16 game NFL season is, getting 2 cheap wins against 2 of your tougher opponents is truly a gift. I think the Jets may be the worst AFC team to earn a playoff spot in recent years if they do indeed get in. Last year an 11-5 Patriot team fell short. This year 8-8 could possibly get in, but at the least 2 9-7's will be in.

When you look at wildcards in recent years they've been very tough. Colts and Ravens last year with the Ravens making it to the AFC title game. Jax at 11-5 who came to Pittsburgh and beat us the year before. The 11-5 Steelers in 05 who won the SB.

blaze1984
12-29-2009, 01:30 AM
Something I noticed:

I ran all the possible permutations of next week's games that affect our chances of getting into the playoffs. Assuming that we beat the Dolphins (else why bother looking), there are 32 permutations of results.

Of those 32 possible scenarios, 14 place the Steelers in the playoffs (third place behind Baltimore and NYJ, each with 16). This is far ahead of the next most likely contender.
Interesting but not important: It is impossible for the Jets to get in as #6 seed if the Steelers win.

Strong correlation:
12 of the 14 successful scenarios involve a Houston loss.

If all possible permutations were equal (which of course they're not), the Steelers would have a 44% chance of making the playoffs. With a Houston loss, this increases to 75%. With a Houston win, it drops to 13%.

Current ranking by this method:
#1 New York Jets (16 avenues, all lead to #5 seed) 50%
#2 Baltimore Ravens (16) 50%
#3 Pittsburgh Steelers (14) 44%
#4 Denver Broncos (9) 28%
#5 Houston Texans (8) 25%
#6 Jacksonville Jaguars (1) 3%

Doesn't really mean anything. Just thought it was interesting...

I found that interesting as well. There was a CMU math professor on the news claiming that our odds of getting in were 43.75% assuming we beat the Dolphins. Well, unfortunately the math isn't that simple. It sucks, i think every Steeler fan hates the Caldwell almost as much as Bellicheck right now

steve314
12-29-2009, 01:50 AM
Technically you can't.... i compared similar point spreads from previous weeks and looked at the money lines.

Usually there's a 1-2 day delay on posting of money lines. Point spreads are usually the first thing posted. Anyways, pinnacle finally posted their money lines, so i'm going to update the original post with the moneyline calculations

I've seen an analysis somewhere where the normal distribution was used to convert point spreads to win probabilities. Yes, the spreads are set to balance betting on each side, but the converted probabilities matched up pretty well to the actual results on the data from previous seasons. Not sure if they excluded week 17 spreads though, ha.