Fanhome's Dackle: World Series Odds (September 18, 2003)
Team Chance of winning World Series
Braves 24.8
Giants 21.9
Yankees 16.6
Athletics 14.0
Red Sox 7.8
Phillies 3.5
Twins 3.3
Marlins 2.9
Astros 2.6
Cubs 2.2
(Rest less than 1%)
--posted by TangoTiger at 09:58 AM EDT
Posted 10:15 a.m.,
September 18, 2003
(#1) -
Steve Buckley_s missing thoughts
As a Red Sox fan, I'm a little curious as to why the A's have almost twice as good a chance to win it all as the Sox. How exactly does one arrive at that determination? Ouija board? Magic eight ball? Please.
Posted 10:20 a.m.,
September 18, 2003
(#2) -
bunyon
For one, the As pretty much have a postseason spot sewn up. I think the Red Sox will make it, but it is less sure than for the As.
And, not that it matters froma rational point of view, but I don't think the Babe has anything against the As.
Posted 10:23 a.m.,
September 18, 2003
(#3) -
Danny
As a Red Sox fan, I'm a little curious as to why the A's have almost twice as good a chance to win it all as the Sox. How exactly does one arrive at that determination? Ouija board? Magic eight ball? Please.
Yeah, it must be wrong if it shows the Red Sox behind the A's...
Standings
A's: 92-61
Sox: 88-63
M's: 87-65
Pythag Standings
M's: 92-60
A's: 92-61
Sox: 89-62
So the A's have a better record and a better Pythag record than the Sox, which means they should be more likely to win if they reach the postseason. Meanwhile, the Sox are just 1 1/2 games ahead of Seattle for the Wild Card. The Mariners have a better Pythag than the Sox, so passing them would not be a big surprise. Additionally, the A's have a chance at the division and the Wild card, whereas the Sox only have a chance at the Wild Card. The %s seem about right to me.
Posted 10:51 a.m.,
September 18, 2003
(#4) -
Andrew Edwards
Umm, I think there's a mistake here.
It says the Red Sox and the Cubs both have non-zero chances of winning the World Series.
Shouldn't insane mystical curses be folded in here somehow?
Posted 11:20 a.m.,
September 18, 2003
(#5) -
Cubs Fans and Jim Carrey
So you're telling me there's a chance.
Posted 11:35 a.m.,
September 18, 2003
(#6) -
Joe Dimino(e-mail)
Actual Vegas Odds:
Arizona Diamondbacks 750-1
Atlanta Braves 6-1
Boston Red Sox 6-1
Chicago Cubs 12-1
Chicago White Sox 25-1
Florida Marlins 15-1
Houston Astros 18-1
Kansas City Royals 115-1
Los Angeles Dodgers 50-1
Minnesota Twins 8-1
Montreal Expos 9000-1
New York Yankees 3-1
Oakland Athletics 4-1
Philadelphia Phillies 20-1
San Francisco Giants 4-1
Seattle Mariners 50-1
St Louis Cardinals 250-1
So that would mean, the follownig are "good bets" if we assume Dackle is correct.
Braves at 6:1
Giants at 4:1
That's it. All have worse Vegas Odds (for the bettor) than what Dackle says, Vega$ gets tons of juice on these type of bets.
The worst bet is probably the Twins at 8:1, the Red Sox at 6:1 aren't great either, considering both aren't even locks for the playoffs (although they are both looking good). The Cubs at 12:1 are pretty lousy bet too.
Posted 11:36 a.m.,
September 18, 2003
(#7) -
Joe Dimino(e-mail)
The Yanks aren't a good bet either, you've always to pay extra for them. I've always wanted to track if betting against the Yankees all year would make you money, I'd imagine it would have to.
Posted 12:05 p.m.,
September 18, 2003
(#8) -
tangotiger
Good job Joe!
Here's his list in "pecentage" format. You'll note the juice adds an extra 38%. Ahh, to be a bookie.
NewYork 25.0%
Oakland 20.0%
SanFrancisco 20.0%
Atlanta 14.3%
Boston 14.3%
Minnesota 11.1%
Chicago 7.7%
Florida 6.3%
Houston 5.3%
Philadelphia 4.8%
Chicago 3.8%
LosAngeles 2.0%
Seattle 2.0%
Posted 12:14 p.m.,
September 18, 2003
(#9) -
Craig B
Man, I like the Sox better than 7.8%, but not 6-1! Now at 12-1, I think I'd bet on that.
I'm tempted to drop a C-note on Atlanta at 6-1 though. I bet if you looked hard enough, you could find them at marginally higher odds. At 13-2 or 7-1, that's a nice potential payoff.
Posted 12:22 p.m.,
September 18, 2003
(#10) -
tangotiger
We also should remember a few things that would change the odds drastically:
1 - Your top 3 starters have a higher % of innings in the playoffs than regular season
2 - Your top 2 relievers have a higher % of innings (and high-leverage innings) in the playoffs than the regular season... I think Mariano Rivera has something like 80 innings in 87 Yankee playoff games
3 - Certain types of hitters, though we haven't established which ones, may be less able to optimize their abilities against higher level of pitching
4 - The park affects every player (hitter or pitcher) differently. While in a season, the road parks balance all this out, more or less, it's not necessarily the case in a short series (say Ted Lilly at Fenway).
Because of the non-random nature of the contexts faced by the players, the "true talent" level of each team might vary greatly in the "playoff universe".
And finally, injuries.
Posted 1:35 p.m.,
September 18, 2003
(#11) -
tangotiger
And by the way, kudos to Nate Silver for putting this up:
DISCLAIMER: Because this analysis does not take into account head-to-head matchups, it may be less reliable from this point in the season onward.
It's good to see analysts establishing the boundaries of their work to the readers.
Posted 2:31 p.m.,
September 18, 2003
(#12) -
C-O
Why do the NL teams have a 60% total shot of winning it, and the AL teams only 40%? Didn't Hank Blalock give them the edge?
Posted 3:20 p.m.,
September 18, 2003
(#13) -
Replacement-level Fan
UZR? Who cares?
Go Twins and Marlins!
Posted 4:14 p.m.,
September 18, 2003
(#14) -
Curious
Why the selective editing on this thread?
Posted 4:42 p.m.,
September 18, 2003
(#15) -
mcm
Dackle has teams in the NL Central with an aggregate 100.1% chance of winning the division. I can see how a couple of .05-percentages--one digit to the right of the smallest digit he is reporting--rounds up to an error in the smallest digit he is reporting.
Dackle also has AL teams with an aggregate 100.9% chance of winning the divison. I do not see how rounding in the hundredth-of-a-percent column adds up to an error of almost a full percent. What gives?
Posted 5:08 p.m.,
September 18, 2003
(#16) -
tangotiger
Primer policy has: Comments ... may be removed...if the comment ... really does nothing to move the conversation forward.
It's (unfortunately) rarely exercised, but I thought the volume (especially by me) was just too much. I'll be happy to send anyone the exchange.
In the words of the great Leslie Nielson: "There's nothing to see here. Please move on. "
Posted 5:08 p.m.,
September 18, 2003
(#17) -
Mike Moffatt
(homepage)
If you compare these numbers to what the market thinks, there is quite a difference. I took the market values at TradeSports (and scaled them so all the probs=100) and compared it to this list. It's a lot different:
TEAM MARKET DACKLE
MLB.YANKEES 23.2 16.6
MLB.GIANTS 15 21.9
MLB.BRAVES 13.8 24.8
MLB.ATHLETICS 12.48 14
MLB.REDSOX 11 7.8
MLB.ASTROS 4.84 2.6
MLB.CUBS 5.04 2.2
MLB.TWINS 4.3 3.3
MLB.MARINERS 2.46 0.3
MLB.WHITESOX 2.1 0.1
MLB.PHILLIES 2.2 3.5
MLB.DODGERS 0.48 0.1
MLB.DIAMONDBACKS 0.24 0
MLB.CARDINALS 0.36 0
MLB.BLUEJAYS 0.08 0
MLB.FIELD 2.4 3
To me the market makes more sense, just from a "sniff test" point of view. I can't see the AL only having a 40% chance of winning when they're the home team.
Cheers,
Mike
Posted 5:20 p.m.,
September 18, 2003
(#18) -
tangotiger
The big differences are all at the top:
YANKEES 23.2 ... 16.6
GIANTS 15 .... 21.9
BRAVES 13.8 .... 24.8
ATHLETICS 12.5 .... 14
REDSOX 11 .... 7.8
I think the Yanks have a bias for whatever reason. The A's+Sox are 23.5% on the one side and 21.8% on the other. So, it comes down to the Giants+Braves being so much higher with dackle.
So, the question to ask first is:
Is the AL/NL talent even that you can do as dackle is doing? Or are some NL teams getting the benefit on beating up on worse teams than the AL does? Or are the AL teams so evenly matched that no 1 team can really stand out?
The other thing to remember is that if the Giants/Braves are really as good as suggested by their W/L record, compared to the Yanks/A's, etc, then it's no surprise that a team from the NL will be more likely to win than a team in the AL.
Posted 9:48 a.m.,
September 19, 2003
(#19) -
Lujack
Does Dackle's formula factor in injuries? If Smoltz is out, that will hurt the Brave's chances.
Posted 10:18 a.m.,
September 19, 2003
(#20) -
tangotiger
I'm sure it does not.
I would think that the best way to do these odds is to use Diamond-Mind baseball or some similar game, where the random creation of injuries, or the occurrence of injuries based on past history can be incorporated into the game.
As well, since pitcher usage changes in the playoffs, again, you can incorporate that as well.
Posted 12:20 p.m.,
September 19, 2003
(#21) -
Brent
1. It looks like HFA is not factored in.
2. I imagine that any baseball power rating comparing NL and AL teams would have only interleague play to go by. From what I remember the NL really beat up on the AL this year fwiw. A very small sample size of about 270 games btw select teams.
Posted 12:50 p.m.,
September 19, 2003
(#22) -
tangotiger
Can someone take the top 5 NL teams and top 5 AL teams, and see how they did against "same competition"?
That is, lump ATL, SF, et al into "NL leaders", and find their records against their AL opponents. Weighting only by the games against the AL opponents, how did those 5 unweighted AL leaders do?
And vice-versa. W/L, RS/RA would be nice.
Anyone, anyone? Bueller?
Posted 3:37 p.m.,
September 19, 2003
(#23) -
Danny
Lame Attempt:
SFG
1-2 vs MIN (10 RS, 14 RA)
3-3 vs OAK (30 RS, 27 RA)
2-1 vs CHW (21 RS, 13 RA)
TotaL: 6-6 (61 RS, 54 RA)
ATL
2-1 vs OAK (18 RS, 12 RA)
1-2 vs SEA (5 RS, 5 RA)
Total: 3-3 (23 RS, 17 RA)
Interleague Play
SFG: 10-8
ATL: 10-5
OAK: 9-9
NYY: 13-5
BOS: 11-7
SEA: 10-8
Yes, Tango, I realize this does not even slightly resemble the study you requested.
Posted 3:50 p.m.,
September 19, 2003
(#24) -
tangotiger
Just to add some data, NL teams are .544 when facing AL teams.
That is, an average NL team will win .544 of their games against an AL team. An average AL team will win .500 of their games against an average AL team.
RS/RA would have been better to use, but I no got.
1 standard deviation is about .030, so this may be completely due to luck.