PDA

View Full Version : BP's Improved Playoffs Odds Report



RedsManRick
08-20-2012, 01:48 PM
The Playoff Odds report at BaseballProspectus.com has always been a useful tool. But it just got a lot cooler with the inclusion of stock-style graphs that show the changes of the odds throughout the season.

Suffice it to say that the Reds are in pretty good shape.

http://www.baseballprospectus.com/odds/

mbgrayson
08-20-2012, 02:04 PM
I have been following this tool for several weeks now. I find the daily fluctuations troubling. How can one game won or lost with 40 games left cause a 12% swing in probability? I also don't fully understand (and really haven't tried to understand) why the Cardinals have a better chance than the Pirates given strength of remaining schedule issues, and that they are currently 2 games behind the Bucs. I guess they base this off of current position, with pythag factored in somehow.

I like seeing the Reds at 98% +, but I don't fully believe it yet. Last year, what were the Cardinals chances down to before their big run, and the Braves collapse? The Reds need to just play one game at a time, day by day.

Brutus
08-20-2012, 02:14 PM
I have been following this tool for several weeks now. I find the daily fluctuations troubling. How can one game won or lost with 40 games left cause a 12% swing in probability? I also don't fully understand (and really haven't tried to understand) why the Cardinals have a better chance than the Pirates given strength of remaining schedule issues, and that they are currently 2 games behind the Bucs. I guess they base this off of current position, with pythag factored in somehow.

I like seeing the Reds at 98% +, but I don't fully believe it yet. Last year, what were the Cardinals chances down to before their big run, and the Braves collapse? The Reds need to just play one game at a time, day by day.

Last year, the Braves' chances were about the same, but remember a 98 percent chance means that twice out of 100, the team won't make it. The Braves and Rays both just happened to be two of those 100 lol

MikeThierry
08-20-2012, 02:28 PM
I have been following this tool for several weeks now. I find the daily fluctuations troubling. How can one game won or lost with 40 games left cause a 12% swing in probability? I also don't fully understand (and really haven't tried to understand) why the Cardinals have a better chance than the Pirates given strength of remaining schedule issues, and that they are currently 2 games behind the Bucs. I guess they base this off of current position, with pythag factored in somehow.

I like seeing the Reds at 98% +, but I don't fully believe it yet. Last year, what were the Cardinals chances down to before their big run, and the Braves collapse? The Reds need to just play one game at a time, day by day.

I'm not sure how calculate playoff odds but I'm willing to guess that run differential plays a key part in it. It's probably why the Cards are ahead of the Pirates in playoff odds. I think when they went on that run last year, the Cards playoff odds were something like 1%. It goes to show you that playoff odds is a nice thing to look at but of no importance.

dougdirt
08-20-2012, 02:33 PM
I'm not sure how calculate playoff odds but I'm willing to guess that run differential plays a key part in it. It's probably why the Cards are ahead of the Pirates in playoff odds. I think when they went on that run last year, the Cards playoff odds were something like 1%. It goes to show you that playoff odds is a nice thing to look at but of no importance.

Or the 1% suggests exactly what it was, the biggest comeback in the history of the game and incredibly unlikely to be repeated.

Brutus
08-20-2012, 02:36 PM
I'm not sure how calculate playoff odds but I'm willing to guess that run differential plays a key part in it. It's probably why the Cards are ahead of the Pirates in playoff odds. I think when they went on that run last year, the Cards playoff odds were something like 1%. It goes to show you that playoff odds is a nice thing to look at but of no importance.

Mike, playoff odds are baseball's version of probability. You're basically suggesting that an entire mathematical premise and calculation has no bearing.

As Doug said, that it was the biggest comeback in the history of baseball suggests that perhaps the 1% was accurate.

One percent means that only roughly one time in 100 will a team come back from that deficit. Those odds seem to be on the money if you consider there are at least 10 teams in that situation every year and never before has a team come back from that deficit.

MikeThierry
08-20-2012, 02:49 PM
Mike, playoff odds are baseball's version of probability. You're basically suggesting that an entire mathematical premise and calculation has no bearing.

As Doug said, that it was the biggest comeback in the history of baseball suggests that perhaps the 1% was accurate.

One percent means that only roughly one time in 100 will a team come back from that deficit. Those odds seem to be on the money if you consider there are at least 10 teams in that situation every year and never before has a team come back from that deficit.

I understand what you're saying. It's just an entertainment factor for me. I don't put any real stock in it though. Sports is so fluid and unpredictable, using probability factors doesn't tell much, at least for me. If we were talking about economic data or actual tangible data that isn't fluid but set in stone, my opinion would be different. The Rays pretty much had the same percentage to make the playoffs as the Cards last year yet they did. We see teams that have a low percentage of making the playoffs make it all the time. The Rockies were probably in the same position when they went on their run a couple of years ago.

RedsManRick
08-20-2012, 02:49 PM
I have been following this tool for several weeks now. I find the daily fluctuations troubling. How can one game won or lost with 40 games left cause a 12% swing in probability? I also don't fully understand (and really haven't tried to understand) why the Cardinals have a better chance than the Pirates given strength of remaining schedule issues, and that they are currently 2 games behind the Bucs. I guess they base this off of current position, with pythag factored in somehow.

I like seeing the Reds at 98% +, but I don't fully believe it yet. Last year, what were the Cardinals chances down to before their big run, and the Braves collapse? The Reds need to just play one game at a time, day by day.

There are basically two parts:

1. Estimate how good the team really is. It does this using a higher order win % that adjusts for things like underlying stats and strength of schedule. See more here: http://www.baseballprospectus.com/standings/

2. Run a whole bunch of simulations of the rest of the season and see what happens.

Once you run enough simulations, you get to a pretty narrow estimate of the possibilities for the rest of the season.

To my knowledge, this system is essentially based on the assumption that team quality doesn't change for the rest of the season. That is, if the team plays as well as well as it has so far for the rest of the year, how are things likely to play out. So if you have a major injury or simply go on an extended winning/losing streak, things can change quickly.

In that sense, I think your intuition is right. One game shouldn't have that much influence. Still, I think it's a pretty interesting look at things. Fun to look at and think about -- but I probably wouldn't go to Vegas with it.

MikeThierry
08-20-2012, 02:52 PM
This whole conversation reminds me of why I do not and will never bet on sports. It's too unpredictable to form any sort of conclusion on what will happen or what will probably happen.

Brutus
08-20-2012, 02:57 PM
I understand what you're saying. It's just an entertainment factor for me. I don't put any real stock in it though. Sports is so fluid and unpredictable, using probability factors doesn't tell much, at least for me. If we were talking about economic data or actual tangible data that isn't fluid but set in stone, my opinion would be different. The Rays pretty much had the same percentage to make the playoffs as the Cards last year yet they did. We see teams that have a low percentage of making the playoffs make it all the time. The Rockies were probably in the same position when they went on their run a couple of years ago.

Baseball odds are based on tangible data, though. With Pythagorean win expectation and historical data, we have tangible data that has given us the ability to measure odds. Those numbers weren't pulled out of thin air.

You're correct there's some unpredictability, but when you build a 7-game lead with 40 games left, it's a lot less unpredictable. When it does happen, its' the exception and not the rule.

Ghosts of 1990
08-20-2012, 02:59 PM
"It ain't over til it's over" -- Yogi Berra

MikeThierry
08-20-2012, 03:08 PM
Baseball odds are based on tangible data, though. With Pythagorean win expectation and historical data, we have tangible data that has given us the ability to measure odds. Those numbers weren't pulled out of thin air.

You're correct there's some unpredictability, but when you build a 7-game lead with 40 games left, it's a lot less unpredictable. When it does happen, its' the exception and not the rule.

I would argue though that the tangible data in sports is still fluid. To me, strength of schedule doesn't tell the whole story, esp. in baseball. A down team can always be dangerous down the stretch. There is no Law of Baseball Epic Failure that says that a horrible team will remain horrible the rest of the season. Even if you look at historical data in sports, that's a bit deceiving because there is a great chance that what happened in the past will never happen again.

That is different than something like economics where the data given can be applied to a historical model to create a future possibility. The laws of economics very rarely change. I'm only using economics as an example because it's sort of up my alley, so to speak. The same applies for other areas where projections are made.

Again, I think these projections are fun but I put less stock in them than any projections that are made in other areas (economics, science, business, etc).

Kc61
08-20-2012, 03:09 PM
Interesting that Pirates have a larger chance of winning the division than the Cards. Yet the Cards have the larger chance of making the wild card over the Pirates.

Not sure I get that, but interesting outcome.

Brutus
08-20-2012, 03:14 PM
I would argue though that the tangible data in sports is still fluid. To me, strength of schedule doesn't tell the whole story, esp. in baseball. A down team can always be dangerous down the stretch. There is no Law of Baseball Epic Failure that says that a horrible team will remain horrible the rest of the season. Even if you look at historical data in sports, that's a bit deceiving because there is a great chance that what happened in the past will never happen again.

That is different than something like economics where the data given can be applied to a historical model to create a future possibility. The laws of economics very rarely change. I'm only using economics as an example because it's sort of up my alley, so to speak. The same applies for other areas where projections are made.

Again, I think these projections are fun but I put less stock in them than any projections that are made in other areas (economics, science, business, etc).

I honestly don't see how, in your economic example, September call-ups (for instance) would be any different than minor market fluctuations in economic analysis. It seems they both amount to the same thing. Sure, they would change the probability by a few percentage points, but not enough to discredit the model.

MikeThierry
08-20-2012, 03:21 PM
I honestly don't see how, in your economic example, September call-ups (for instance) would be any different than minor market fluctuations in economic analysis. It seems they both amount to the same thing. Sure, they would change the probability by a few percentage points, but not enough to discredit the model.

Economics is governed by known laws. Those minor market fluctuations can be predicted by other things going on with the economy (supply of commodities, tax rates, interest rates increase/decrease etc). I think it's easier to formulate a model based on given data as to whether or not the market will suffer the next couple weeks than it is to predict whether or not a september call-up will do well on the major league level.

In my previous Law of Baseball Epic Failure or LBEF, the only team I think that can apply to is Houston, lol.

George Anderson
08-20-2012, 03:23 PM
On August 12, the Dodgers had a commanding 13 game lead over the second place New York Giants. "The Giants is dead," Dressen bragged as he could sense a National League pennant. But the Giants caught fire and won 37 of their final 44 games to close the gap while the Dodgers were losing.


http://www.milb.com/news/article.jsp?ymd=20120617&content_id=33446780&vkey=news_t556&fext=.jsp&sid=t556

Sea Ray
08-20-2012, 03:23 PM
The Playoff Odds report at BaseballProspectus.com has always been a useful tool. But it just got a lot cooler with the inclusion of stock-style graphs that show the changes of the odds throughout the season.

Suffice it to say that the Reds are in pretty good shape.

http://www.baseballprospectus.com/odds/

Who cares about odds especially after last year? Anything can happen regardless of anyone formula re: odds. You've either got a chance to blow it or you've clinched it. Why waste time with someone's odds website???

Patrick Bateman
08-20-2012, 03:29 PM
Who cares about odds especially after last year? Anything can happen regardless of anyone formula re: odds. You've either got a chance to blow it or you've clinched it. Why waste time with someone's odds website???

Odds don't suggest that "anything can happen" is not true (except for mathemtical impossibilities.

The Reds have a great chance at making the playoffs. Odds are fun to look at so you have a good idea of what the future holds.

They might not make the playoffs, but everyone, including baseball prospectus already knew that. They just have a better idea of how likely that is than we could come up with in our head.

Degenerate39
08-20-2012, 03:31 PM
http://shoelace.org/pics/2010/06-15-2010/Feels_goodman.jpg

Sea Ray
08-20-2012, 03:34 PM
Odds don't suggest that "anything can happen" is not true (except for mathemtical impossibilities.

The Reds have a great chance at making the playoffs. Odds are fun to look at so you have a good idea of what the future holds.

They might not make the playoffs, but everyone, including baseball prospectus already knew that. They just have a better idea of how likely that is than we could come up with in our head.

If the odds are not 100% or 0% then they are saying anything can happen. What difference does it make to anyone if the Reds chances are 95% today or 85%?

Patrick Bateman
08-20-2012, 03:41 PM
If the odds are not 100% or 0% then they are saying anything can happen. What difference does it make to anyone if the Reds chances are 95% today or 85%?

Is this serious?

It matters because the Reds are almost certain to make it, compared to, say the Cubs, whom although are not mathematically are eliminated, do have a chance of making the playoffs if they went on an unrealistic run because, hey, anything is possible right?

I think the odds just give us a number to assign to what we see when we look at the standings every day based on mathematical and historical data.

_Sir_Charles_
08-20-2012, 03:41 PM
Interesting that Pirates have a larger chance of winning the division than the Cards. Yet the Cards have the larger chance of making the wild card over the Pirates.

Not sure I get that, but interesting outcome.

The part that isn't getting figured into that equation is the Reds. With them in the equation, the Pirates and Cards have zero chance at winning the division. :beerme:

mbgrayson
08-20-2012, 04:02 PM
There are basically two parts:

1. Estimate how good the team really is. It does this using a higher order win % that adjusts for things like underlying stats and strength of schedule. See more here: http://www.baseballprospectus.com/standings/ (http://www.baseballprospectus.com/standings/)

2. Run a whole bunch of simulations of the rest of the season and see what happens.

Once you run enough simulations, you get to a pretty narrow estimate of the possibilities for the rest of the season.

To my knowledge, this system is essentially based on the assumption that team quality doesn't change for the rest of the season. That is, if the team plays as well as well as it has so far for the rest of the year, how are things likely to play out. So if you have a major injury or simply go on an extended winning/losing streak, things can change quickly.

In that sense, I think your intuition is right. One game shouldn't have that much influence. Still, I think it's a pretty interesting look at things. Fun to look at and think about -- but I probably wouldn't go to Vegas with it.

Agreed on all counts. Thanks for the link to their standings page. It looks to me like it is mostly based on pythag percentages.

I know we have discussed this in other threads, but I really have a problem with using overall runs scored/allowed as a basis for projecting future wins.

The way baseball works, as long as a game is close, teams tend to use their 'A team' pitchers. Once a game appears to be lost, most teams will either let a struggling starter continue to eat innings (and often give up runs), or bring in a 'B team' pitcher, who does the same.

The net effect of this is that in a 5 to 1 game in the 8th inning, the winning team will often see far weaker pitching than in a 3 to 3 game. They would be much more likely to score a couple extra runs, and to run the final score up. In the world of pythag, 'runs is runs', and those extra runs are just as important as the winning run in a one run game.

Let's say the Cards catch Wrigley on a couple days when the wind is blowing out, when the Reds play a series there the wind is blowing in. Pythag would say the Cards must be strong.

Real world examples: Cards in Chicago April 23rd thru the 25th: Cards lose 3 to 2, 3 to 2 again, and win the finale 5 to 1. (Not windy, I know...) Three game series, Cubs take two of three, but the Cards outscore the Cubs 9 to 7 in the series.

Yet when Pythag looks at this series, it looks like the Cards 'won' the series since they outscored the Cubs. But look at the box scores: does the fact that the Cards scored a bunch of runs against Volstad really mean anything? I think not.

If last year there were two different teams that overcame 98% odds and made the playoffs (Cards and Rays), it could be that we saw a once in a lifetime event happen twice the same year, and we should be in awe. Or perhaps, using pythag to predict future wins is not that sound of a method. While it is amazing to watch the collapses of the Braves and the Red Sox, it might not be a 1 in 100 deal.

Homer Bailey
08-20-2012, 04:12 PM
If last year there were two different teams that overcame 98% odds and made the playoffs (Cards and Rays), it could be that we saw a once in a lifetime event happen twice the same year, and we should be in awe. Or perhaps, using pythag to predict future wins is not that sound of a method. While it is amazing to watch the collapses of the Braves and the Red Sox, it might not be a 1 in 100 deal.

If I remember correctly, the collapses of last year were the two biggest in MLB history (9 game lead on 9/1 for Boston, and 8.5 lead for Atlanta on 9/1). It absolutely was a 1/100 deal, possibly even more.

Sea Ray
08-20-2012, 04:15 PM
Is this serious?

It matters because the Reds are almost certain to make it, compared to, say the Cubs, whom although are not mathematically are eliminated, do have a chance of making the playoffs if they went on an unrealistic run because, hey, anything is possible right?

I think the odds just give us a number to assign to what we see when we look at the standings every day based on mathematical and historical data.

Yes, it's serious. Do you need a number to tell you that the Cubs are out of it and the Reds are in the catbird seat? What does their "number" tell you that you didn't know before? It's like science telling us that the sun will burn up in 3.5 billion years. Who cares? No one will be around to see if that prediction is right or wrong.

I don't see that this "number" tells us anything new. We all know that everyday we win, our chances get better

Patrick Bateman
08-20-2012, 04:27 PM
Yes, it's serious. Do you need a number to tell you that the Cubs are out of it and the Reds are in the catbird seat? What does their "number" tell you that you didn't know before? It's like science telling us that the sun will burn up in 3.5 billion years. Who cares? No one will be around to see if that prediction is right or wrong.

I don't see that this "number" tells us anything new. We all know that everyday we win, our chances get better

Exactly. I'm just saying, the projections sort of provide a more definitive answer for what the standings in the newspaper tell us everyday. It's nothing definitive, just another perspective to consider.

AtomicDumpling
08-20-2012, 04:31 PM
On August 12, the Dodgers had a commanding 13 game lead over the second place New York Giants. "The Giants is dead," Dressen bragged as he could sense a National League pennant. But the Giants caught fire and won 37 of their final 44 games to close the gap while the Dodgers were losing.


http://www.milb.com/news/article.jsp?ymd=20120617&content_id=33446780&vkey=news_t556&fext=.jsp&sid=t556

That was back when the pennant race was actually between the two best teams in the league to see which team got to go to the World Series -- a much bigger prize than today. It was a 154-162 game death match with only the single unquestioned best team left standing at the end.

Now the "playoff race" is focused on a few mediocre 2nd and 3rd place teams and the prize is much smaller -- the last spot in the playoffs rather than a World Series berth. They play 162 games to figure out who the best team is but that doesn't mean anything anymore, because they also allow a bunch of inferior teams into a crapshoot of a post season. That is why pennant races will never be as compelling and exciting nor will they capture the nation's attention like they did in the century before the Wild Card was defecated upon the world. We end up with a very long and mostly meaningless regular season, followed by a tournament that is almost always won by a team that proved during the season they weren't even close to being the best team in the league.

Baseball has chosen to maximize profits by cheapening the championship. Who cares about actually crowning the best team in the league as the champion anymore? Does a 2nd place "champion" belong in the same rank as the dominant champions of the 20th century? Does an 85 win "champion" deserve the same place in history as the 100+ win truly great World Series champions of earlier years? Winning the World Series just doen't mean as much as it used to. No, the integrity of the World Series has been pawned for cash and drama to the TV networks.

mbgrayson
08-20-2012, 04:39 PM
If I remember correctly, the collapses of last year were the two biggest in MLB history (9 game lead on 9/1 for Boston, and 8.5 lead for Atlanta on 9/1). It absolutely was a 1/100 deal, possibly even more.

Well, remember that we have only had three divisions and a wildcard slot since 1995. From 1903 to 1969, only one team in each league made it to the post-season. Then from 1969 through 1993, only two teams from each league made the playoffs. Then from 1995 to 2011, four teams made it. Now, five teams will make it.

The more slots open for playoff contention increases the odds that there will be at least one big comeback in September for at least one of the slots.

And there WERE some other big collapses... look at the 1995 Angels who had a 9.5 game lead on August 20 and blew it. What about the 1951 Dodgers-Giants race: On August 11th, the Dodgers were up by 13.5 games and blew it. The 2007 Mets were up by 7 games with 17 left to play, and blew it.

I'm not saying the Cards and Rays big runs weren't historic, they are up there too. But they were not a once in 100 years deal.

HeatherC1212
08-20-2012, 04:43 PM
The part that isn't getting figured into that equation is the Reds. With them in the equation, the Pirates and Cards have zero chance at winning the division. :beerme:

Yeah, this is the best way to look at it, LOL :lol: :p :beerme: :D

Homer Bailey
08-20-2012, 05:12 PM
Well, remember that we have only had three divisions and a wildcard slot since 1995. From 1903 to 1969, only one team in each league made it to the post-season. Then from 1969 through 1993, only two teams from each league made the playoffs. Then from 1995 to 2011, four teams made it. Now, five teams will make it.

The more slots open for playoff contention increases the odds that there will be at least one big comeback in September for at least one of the slots.

And there WERE some other big collapses... look at the 1995 Angels who had a 9.5 game lead on August 20 and blew it. What about the 1951 Dodgers-Giants race: On August 11th, the Dodgers were up by 13.5 games and blew it. The 2007 Mets were up by 7 games with 17 left to play, and blew it.

I'm not saying the Cards and Rays big runs weren't historic, they are up there too. But they were not a once in 100 years deal.

No one with an 8.5 game lead on 9/1 had ever blown that big of a lead. Isn't it fair to say that there was about a 1% chance of that happening?

I believe I read last year that the odds of both teams blowing that lead was something like 1 in 27,000,000.

Of course teams have blown huge leads before. None that big in the month of September though. In history. I think that's pretty significant.

RedsManRick
08-20-2012, 05:58 PM
If the odds are not 100% or 0% then they are saying anything can happen. What difference does it make to anyone if the Reds chances are 95% today or 85%?

It matters because, as a fan, I enjoy looking at it and thinking about it. I think that when people look at the "games behind" column in the standings, they're trying to do something like this in their heads. They don't just care about the "magic number." This brings some degree of structure to that gut process.

If you don't see any value from it, then don't look at it.

Sea Ray
08-20-2012, 06:53 PM
It matters because, as a fan, I enjoy looking at it and thinking about it. I think that when people look at the "games behind" column in the standings, they're trying to do something like this in their heads. They don't just care about the "magic number." This brings some degree of structure to that gut process.

If you don't see any value from it, then don't look at it.

I don't see any value in it above and beyond what we already know such as games behind as you mention so I'm not looking at it. I was trying to see what you folks get out of it

Homer Bailey
08-20-2012, 07:11 PM
I don't see any value in it above and beyond what we already know such as games behind as you mention so I'm not looking at it. I was trying to see what you folks get out of it

I get that, based on specific formulas used by BP that evaluate the skill level of the teams, their current standings, their remaining schedule, and other factors, the percentage chance this website believes a certain team has of making the playoffs. It's basically the same as pre-flop poker odds, IMO.

Brutus
08-20-2012, 07:25 PM
I don't see any value in it above and beyond what we already know such as games behind as you mention so I'm not looking at it. I was trying to see what you folks get out of it

I don't think value is the right word, but I think it's good for context. One can look at the standings and assume that something is likely or unlikely, but there's a big difference between 55% and 95%. It's nice to have some context to something.

mbgrayson
08-20-2012, 09:58 PM
I get that, based on specific formulas used by BP that evaluate the skill level of the teams, their current standings, their remaining schedule, and other factors, the percentage chance this website believes a certain team has of making the playoffs. It's basically the same as pre-flop poker odds, IMO.

No, the point is that the Baseball Prospectus odds formula is based heavily on the number of runs scored and allowed in prior games. If you go to the standings link that RMR posted earlier in this thread, you will see that the Cardinals and Diamondbacks are shown as being in 1st place in the NL Central and West RIGHT NOW, based on run differential (or pythag).

This same 'reliable formula' is what BP uses to guesstimate playoff odds. It is FAR less relaible than straight stats in what the next card will be in a poker game.

Homer Bailey
08-20-2012, 10:15 PM
No, the point is that the Baseball Prospectus odds formula is based heavily on the number of runs scored and allowed in prior games. If you go to the standings link that RMR posted earlier in this thread, you will see that the Cardinals and Diamondbacks are shown as being in 1st place in the NL Central and West RIGHT NOW, based on run differential (or pythag).

This same 'reliable formula' is what BP uses to guesstimate playoff odds. It is FAR less relaible than straight stats in what the next card will be in a poker game.

They use that formula to make their estimation as to how good they think a particular team is, or their "true talent." Then assume that pace to finish the season, factored into how many games behind they are in the standings and how many games are remaining.

mbgrayson
08-29-2012, 12:50 AM
Today the BP playoff odds report has the Reds odds of making the playoffs at 100%... (???) The report is HERE (http://www.baseballprospectus.com/odds/).

This breaks down to 90.8% chance of winning the division and 9.2% chance of winning a wildcard. I still find this report to have weird swings day by day, and frankly, I just don't trust it. But I will now remember that since the odds that the Reds make the playoffs are 100%, we are now guaranteed to make it.

Brutus
08-29-2012, 12:52 AM
Today the BP playoff odds report has the Reds odds of making the playoffs at 100%... (???) The report is HERE (http://www.baseballprospectus.com/odds/).

This breaks down to 90.8% chance of winning the division and 9.2% chance of winning a wildcard. I still find this report to have weird swings day by day, and frankly, I just don't trust it. But I will now remember that since the odds that the Reds make the playoffs are 100%, we are now guaranteed to make it.

ESPN has the overall odds at 98.4 currently, so either way, it's encouraging.

cincrazy
08-29-2012, 01:28 AM
We are 7 games up on St. Louis, but 9 up on the Pirates and 9.5 up on the Dodgers, so even if we somehow collapse and lose the division, it's still very likely that we get a Wild Card spot. That being said.... let's get the division :).

mbgrayson
08-29-2012, 12:01 PM
Today the BP playoff odds report has the Reds odds of making the playoffs at 100%... (???) The report is HERE (http://www.baseballprospectus.com/odds/).

This breaks down to 90.8% chance of winning the division and 9.2% chance of winning a wildcard. I still find this report to have weird swings day by day, and frankly, I just don't trust it. But I will now remember that since the odds that the Reds make the playoffs are 100%, we are now guaranteed to make it.

So how does this work: Yesterday the Reds odds were 100% for making the playoffs. Today, after the Reds win and Cards lose, they are down to 99.7%. They now have a 95.3% chance of winning the division, and a 4.4 % chance of the wildcard.

I liked seeing 100%....even though I didn't believe it.

Sea Ray
08-29-2012, 12:04 PM
We are 7 games up on St. Louis, but 9 up on the Pirates and 9.5 up on the Dodgers, so even if we somehow collapse and lose the division, it's still very likely that we get a Wild Card spot. That being said.... let's get the division :).

I want no part of a wildcard. The way things are setup this yr making the WC is like what we got in 1999. It guarantees you one more game. Big deal. What a letdown it would be if we only got a "1999" postseason out of this

Patrick Bateman
08-29-2012, 12:57 PM
Today the BP playoff odds report has the Reds odds of making the playoffs at 100%... (???) The report is HERE (http://www.baseballprospectus.com/odds/).

This breaks down to 90.8% chance of winning the division and 9.2% chance of winning a wildcard. I still find this report to have weird swings day by day, and frankly, I just don't trust it. But I will now remember that since the odds that the Reds make the playoffs are 100%, we are now guaranteed to make it.

It essentially just means that no team in baseball history has ever collapsed this late in the season with as many teams behind them that need to catch the Reds for the division and wildcard. The second wild card slot makes the difference. The Braves last year only needed one team to catch them. For the Reds, they could play bad, but there still needs to be 3 teams that catch fire big time to leapfrog the Reds.

RedsManRick
08-29-2012, 02:33 PM
So how does this work: Yesterday the Reds odds were 100% for making the playoffs. Today, after the Reds win and Cards lose, they are down to 99.7%. They now have a 95.3% chance of winning the division, and a 4.4 % chance of the wildcard.

I liked seeing 100%....even though I didn't believe it.

They run simulations for the rest of the based on their assessment of how good the team. The percents represent the actual outcomes in those simulations, so you'll see some variation from day-to-day that reflects that natural "noise" inherent in statistical prediction.

Strikes Out Looking
08-29-2012, 03:24 PM
We are 7 games up on St. Louis, but 9 up on the Pirates and 9.5 up on the Dodgers, so even if we somehow collapse and lose the division, it's still very likely that we get a Wild Card spot. That being said.... let's get the division :).

Not just the division, but home field through the playoffs.

mbgrayson
08-29-2012, 03:32 PM
They run simulations for the rest of the based on their assessment of how good the team. The percents represent the actual outcomes in those simulations, so you'll see some variation from day-to-day that reflects that natural "noise" inherent in statistical prediction.

Rick, I understand what you are saying. My point is that there is something wrong with how they are doing this. Very simple example. The Reds are up 7 games over the Cardinals. What if (heaven forbid) the Reds lose their next seven games, and the Cardinals reel off 7 wins in a row. It is possible that in about a week, the Reds and Cardinals could be tied for 1st. From there on out, all the Cardinals have to do is win 1 more than the Reds to win the division.

As to the wild card, if the Reds did lose 7 straight, it is likely that most of the other potential wildcard teams won't also win 7 in a row. However, being that the Braves, Dodgers, Cards, and Pirates are all in the WC hunt, the best two of them could still pick up enough ground to beat out the Reds for both WC slots.

I just don't see how the BP odds could EVER show the Reds chances of making the playoffs are 100% until their magic number is essentially 0 either within the division, or failing that, for a WC. And once their chances are 100%, like they supposedly were yesterday, I don't see how the same formula would show odds of less than 100% after the Reds win a game, and their closest competitor loses. Counter-intuitive.

RedsManRick
08-29-2012, 09:49 PM
Rick, I understand what you are saying. My point is that there is something wrong with how they are doing this. Very simple example. The Reds are up 7 games over the Cardinals. What if (heaven forbid) the Reds lose their next seven games, and the Cardinals reel off 7 wins in a row. It is possible that in about a week, the Reds and Cardinals could be tied for 1st. From there on out, all the Cardinals have to do is win 1 more than the Reds to win the division.

As to the wild card, if the Reds did lose 7 straight, it is likely that most of the other potential wildcard teams won't also win 7 in a row. However, being that the Braves, Dodgers, Cards, and Pirates are all in the WC hunt, the best two of them could still pick up enough ground to beat out the Reds for both WC slots.

I just don't see how the BP odds could EVER show the Reds chances of making the playoffs are 100% until their magic number is essentially 0 either within the division, or failing that, for a WC. And once their chances are 100%, like they supposedly were yesterday, I don't see how the same formula would show odds of less than 100% after the Reds win a game, and their closest competitor loses. Counter-intuitive.

Again, you and BP are doing something completely different. It may sounds like semantics (and I may have added to the confusion initially), but the "100%" number is not what you want it to be. That number is saying, based on a finite number of times in which BP ran a simulation of the season using what it knows about the quality of the teams, the Reds made the playoffs in 100% of those simulated seasons.

Your basic point, which is obviously correct, is that there's a lot of "stuff" that can happen that the simulation isn't accounting for. In none of those 10,000 simulations (or whatever the number is) did Johnny Cueto blow out his arm or did Todd Frazier fall apart at the plate.

All the sim is doing is going game by game for the rest of the season and saying:

- Team A is a .600 team, Team B is a .500 team
- Therefore, Team A has a 72% chance of winning that game
- Rinse and repeat for the rest of the season

So yeah, the basic assumption it makes is that the team's true talent level is accurate and will be that way for the rest of the season. That's not true.

oneupper
08-29-2012, 10:52 PM
I think they changed the way they are doing it from the past. Previously, they would assume that a team was as good as its current winning percentage (if it was .600, that would be the assumption going forward). Now its something else, an "expected win pct" based on "strength of team".(????).

The old way may not have been as accurate, but it had consistency and little subjectivity. I liked it better.

And I think they run 1,000,000 sims, at least before they used to say "playing the rest of the season one million times".

cincrazy
08-29-2012, 10:57 PM
I want no part of a wildcard. The way things are setup this yr making the WC is like what we got in 1999. It guarantees you one more game. Big deal. What a letdown it would be if we only got a "1999" postseason out of this

Oh I agree. I want no part of the wild card either.