a smaller sample will decrease the likelihood that the sample accurately represents the whole.

A sample by definition is not the whole. A larger sample is more of the whole than a smaller sample. But none of that is really relevant.

You have a set of outcomes; you are making a hypothesis about some thing. You determine certain characteristics of that set.

If you really want to find something out, you compare hypothesis against each other to see which is more strongly weighted, and what is your reliability in the best hypothesis.

If you don't know how to do that, you choose out of the blue some hypothesis to adopt, and then conduct a new hypothesis and run a test on it. If the result of that test is below a threshhold (which is based on someone pulling a number out of their ### about 100 years ago), you adopt the new one. You have no basis to know the weight of that hypothesis or the validity of that hypothesis based on those techniques.

So, will a smaller sample size decrease the likelihood that the sample accurately represents the whole?

strongsilence- I don't see what people say or even feel is relevant. This isn't a contest to find out which teams had the happiest players.

Ben – I’m quoting the people that play and manage the games. Wouldn’t they know a lot about the game? Wouldn’t that additional information help us come to agreement on the “random/luck” issue?

A comparison, using the criteria (some much more useful than others) in post 273, of the 1995 Indians and the 1998 Yankees. The first two categories, for Cleveland, are prorated to a 162-game schedule and rounded.

Wins: CLE 112 (100), NYY 114
Games ahead: CLE 34 (30), NYY 22
Postseason: CLE 9-6, NYY 11-2
OPS+: CLE 117, NYY 116
ERA+: CLE 121, NYY 117
Runs: CLE 1st, NYY 1st
ERA: CLE 1st, NYY 1st
Fewest hits allowed: CLE 2nd, NYY 1st
Fewest runs allowed: CLE 1st, NYY 1st
Offensive walks: CLE 9th, NYY 1st
Fewest offensive strikeouts: CLE 1st, NYY 5th
Fewest walks allowed: CLE 1st, NYY 2nd
Strikeouts: CLE 3rd, NYY 4th
Shutouts: CLE 1st, NYY 1st
Offensive shutouts: CLE 1st, NYY 1st
Complete games: CLE 7th, NYY 1st
Stolen bases: CLE 1st, NYY 2nd
Home runs: CLE 1st, NYY 4th
One-run games: CLE 28-14, NYY 21-10
Blowouts: CLE 29-11, NYY 42-13
Defensive efficiency: CLE 3rd (tie), NYY 1st
Fewest errors: CLE 8th (tie), NYY 3rd
Fewest home runs allowed: CLE 2nd, NYY 1st
Payroll: CLE 5th, NYY 2nd

So, will a smaller sample size decrease the likelihood that the sample accurately represents the whole?

I'm not terribly fond of repeating myself, the statement was responded to with relevant explanatory information. If you have trouble understanding the explanation, I will tell you that the statement is false, but can be valid, and is meaningless as an assertion. The why has already been explained.

List of teams that I can't even imagine wouldn't get in regardless of how it works out

Here's the trouble with "indisputable" lists: I disagree with three of your teams, 53 Yankees, 55 Dodgers, and 84 Tigers. It's not that these teams might not get in, it's that I don't consider them indisputable.

As for season v. series, IMO performance is distributed around a mean. Even the most consistent players don't perform exactly the same way in identical situations. A season tests their average ability. A series may test something else, but it doesn't test that.

I'm puzzled why so many are adamant about the importance of the postseason when we're discussing a sim which will run a full season.

Hey BL,

How ya doin'?

What would be your initial hypothesis for testing "greatness"? If the result is a victory of 4 games to 3 for some team, would you adopt a new hypothesis?

What is your definition of a great baseball team? Mine is 285 wins over a 3-year period with a minimum of 1 WS victory and a minimum of 2 WS appearances in that 3-year period.

Mefisto,

We are arguing: 1) whether postseason results should be a criterion; and, 2) what weight to give postseason results for admitting a team to the simulation.

Regarding the sim: I understand the movement toward including a single, semi-arbitrary team for cases such as the 1990s Braves, but I don't like that at the expense of any genuinely better single-season team. The sim, after all, can only simulate a team for an individual year; mixing all-time-dominant, single-season teams with "representative" teams from a franchise's dominant era seems to make the sim try to do two different things, neither of them optimally.

The 1995 Indians and the 2001 Mariners belong in the sim. It surprises me there is any debate at all. These teams were juggernauts, on a level rarely seen. Cleveland's statistical dominance speaks for itself. Seattle played 18 teams during the regular season and had a winning record against all of them.

strong silence Posted: December 29, 2005 at 12:10 AM (#1798035)
Hey BL,

How ya doin'?

What would be your initial hypothesis for testing "greatness"? If the result is a victory of 4 games to 3 for some team, would you adopt a new hypothesis?

What is your definition of a great baseball team? Mine is 285 wins over a 3-year period with a minimum of 1 WS victory and a minimum of 2 WS appearances in that 3-year period.

before the wild card, I would have agreed with you, after the wild card it has become more difficult to be that dominant, and I just don't see how you can fault a team that, says win 300 games over a 3 year period but doesn't win a world series.

I have to agree with Sawney about the Mariners(haven't looked at the Indians though, but intuitively they seem like a genuinely great team)

ME: Players who say "wrong time to have a bad week" are either using luck as an excuse or not giving credit to a better team.

CFB: I don't get it, so now that we have you admitting that other players besides beane think that way, you are now claiming they are using an excuse.

CFB, if you really think players think the playoffs are a crap shoot, switch your allegiance to Beane's team. How much motivation would a player with this view have? Very little. If a manager thought his player had that view, how much confidence would the manager have in that player?

We look back on a year or a season, pick one: 1975, 1963, 1912, whatever year, and do we not think first of the World Series Champion? Why is that? We know, with some exceptions, that the Champion is the better team, not merely some random winner.

Who remembers which teams had the best regular season record for a given year? I'll admit I know a few: 1927, 1954, 1975, 1998, 2001.

(Personally I think the WS should be 9 or 11 games but many times 7 games are enough to determine the better team.)

As for a particular series:

a. I know with absolute certainty that the winner of a 4-0 sweep is the better team.
b. I know with a little-less than absolute certainty that the winner of a 4-1 series is better.
c. I know with no certainty and some doubt, that the winner of a series 4-2 may not be the better team. The winner of a 4-2 series may have had some luck.
d. I know with no certainty and much, much doubt that the winner of a 4-3 series is better. The winner of a 4-3 series was fortunate or lucky.

Tournaments are the only objective way of determing the better team. When you can normalize unbalanced schedules and strength of competition, with absolute certainty, then you can use regular season results.

The argument for 28 teams (28 lends itself perfectly to home-and-home series with each of the other teams, for a total of 162) from post 453 is pretty good.

I'd like to see the format resemble an actual MLB season (pre-realignment) as much as possible. So two leagues, randomly seeded, 162 (or 154) games, no interleague play, and a Super Duper World Series matching up the top two teams.

The 1995 Indians and the 2001 Mariners belong in the sim. It surprises me there is any debate at all. These teams were juggernauts, on a level rarely seen. Cleveland's statistical dominance speaks for itself. Seattle played 18 teams during the regular season and had a winning record against all of them.

I'm not arguing that the 2001 Mariners don't belong in the sim. I'm arguing that they are not great. I'll be interested in the results of the sim.

I'm not arguing that the 1995 Indians don't belong in the sim. I'm arguing that they are not great. I'll be interested in the results of the sim.

The 1997/1978 Dodgers belong in the sim. I'm not arguing that they are great. I'll be interested in the results of the sim.

I doubt any of these teams will win the sim.

Tournaments are the only objective way of determing the better team. When you can normalize unbalanced schedules and strength of competition, with absolute certainty, then you can use regular season results.

You seem to be holding a 162-game span to a higher burden of proof ("absolute certainty") than even a 5-game series ("little less than absolute certainty"), let alone a 6- or 7-game series.

Another team that should be mentioned is the 1886 or 1887 St. Louis Browns, part of four straight Browns pennents. However, if you consider 1893 as the cutoff, then do not include them. I think they should be on the ballot.

About close plays being determined by luck: given what the players do, such as hit the ball with X force with a Y projectory into a wind with Z velocity, and the defender takes the exact rute that they do, then there is only one possible outcome. And there are factors not under players controll, such as whether and umpiring, but they are not controlled by luck either. I will, however, give you that in a short series the lesser team can win, and that somebody's true talent is clearer with a greater sample. Contributing events to luck, though, should not be done.

I'm not arguing that the 2001 Mariners don't belong in the sim. I'm arguing that they are not great.

The straight, simple odds that an average team would win 18 of 18 season series is about 0.0004%. The Mariners were really that much of a fluke?

I'm not arguing that the 1995 Indians don't belong in the sim. I'm arguing that they are not great.

Besides the Indians' postseason record in post 503, how do the Indians not compare favorably to the 1998 Yankees? By your own admission (by implication) in post 510, the Indians might have been better than the Braves despite the 4-2 series loss.

Tournaments are the only objective way of determing the better team

Tell that to Division I-A Football.</BCSrip>

Bill James said that he could not consider the 1906 Cubs or 1954 Indians for greatest team ever, because they did not win the Worlsd Series, for whatever that is worth. I feel the same way, thoguh I think these teams should be in the sim.

As an aside: when will the ballot thread be posted?

Tournaments match two teams, and only those teams, against each other. Inherently better.

Paraphrasing Andy, or someone else earlier, we may be measuring different things when we evaluate regular season that when we evaluate playoffs. Regular season can measure greatness - but there is a lot of noise when we compare records of different teams. Playoffs measure greatness - and depending on the result - may or may not have noise.

I think my previous standard of absolute certainty is too high. Yes I agree with you. We should have reasonable certainty that the extraneous factor can be measured.

Tournaments match two teams, and only those teams, against each other. Inherently better.

Only if the weaknesses each team exploits in the other are equivalent in scope and significance to what each team did against the rest of the league. It's quite plausible that a team may simply show an Achilles' heel against one team, but the team that wins the short series actually has more weaknesses overall. I don't believe one can equate a head-to-head competition with a definitive determination of overall excellence. There are a lot of teams out there, each with a unique combination of strengths and weaknesses; I would argue that the ability to beat such a variety of teams is more significant than the ability to beat a single team, even an excellent one.

Playoffs measure greatness - and depending on the result - may or may not have noise.
The Mariners and Yankees played a three-game series in April 2001. Seattle scored seven runs each game and won all three. The Yankees won the LCS, 4-1. Between the regular season and playoffs, the teams split 14 games. I can't conclude a thing from any of that. But I can conclude that the Mariners were undoubtedly the better team, having won 21 more regular-season games. You can remove any "noise" and, assuming the half chance it goes in the Yankees' favor, there's still plenty of margin left in favor of the Mariners.

I think my previous standard of absolute certainty is too high. Yes I agree with you.

I believe that's the first time in three years of occasional posting that I inspired a concession from anyone. Thanks.

I would argue that the ability to beat such a variety of teams is more significant than the ability to beat a single team, even an excellent one.

Really? The ability to beat one of the top 4 teams, such as in a World Series, is very significant. The only exception I can see is when the strength of a league or division is much, much higher than the other. The champ who emerges from the stronger league could be a better team, regardless of what happens in the WS.

But I can conclude that the Mariners were undoubtedly the better team, having won 21 more regular-season games.

We should agree to disagree - you prefer the season, I prefer the WS - blondes, brunettes, etc. This is okay. We won't get anywhere arguing further. Let me ask you this: When the regular season ends - what do you do? Do you determine, by some method of analysis, which team is the best? Do you not care who is the best? Do you celebrate if its your team? Do you think, aw shucks, if only we had beat Tampa Bay in May we would have had a better record and shown that we were better? I'm curious.

I believe that's the first time in three years of occasional posting that I inspired a concession from anyone. Thanks.

You are welcome. Either you are getting smarter or I am getting dumber. Seriously though. I make mistakes and I'm glad you pointed it out.

Are you a Mariners fan?

When the regular season ends - what do you do?

I have always cared much more about the regular season than the playoffs. Honestly. Perhaps 15 years ago, I told people I would rather my team have the best regular season record and not win the World Series than win the World Series but not have been the best team over 162 games. My wish came true, twice.

Are you a Mariners fan?

No, I just believe that it may be a long time before anyone sees such a quick succession of three teams for the ages (1995 Indians, 1998 Yankees, 2001 Mariners), and two of these teams are not like the other in terms of public perception. So this topic inspired me to pop into the BTF fray for the first time since January.

Without justifying much reasoning here, I consider these three teams to be, as likely as any, the best MLB teams ever. I'm not sure in which order. There are way too many unknown variables that prevent any legitimate claim for any team of any era, but these three teams steamrolled their league in an era when the talent pool has been as great as it's ever been.

I was an Indians fan until 2000. I've rooted for Tampa Bay since 2003.

These are the worst 10-consecutive game postseason records for selected teams:

1906-1908 Chicago Cubs: 7-2-1
(Lost 1906 WS 4-2 then won 1907 WS 4-0-1)

1921-4 New York Giants: 3-7
(Lost last 3 games of the 1923 World Series and lost 4 of 7 of 1924 World Series)

1926 – 1928 Yankees: 7-3
(Lost 1926 WS 4 games to 3 then won first 3 games of 1927 WS)

1936 – 1939 Yankees: 7-3
(Won 1936 WS 4 games to 2 then won 3 of first 4 games of 1937 WS)

1939 – 1940 Cincinnati Reds: 3-7
(Lost 1939 World Series 4-0, then split first 6 games of 1940 World Series)

1949 – 1953 Yankees: 6-4
(Several 10-games stretches over 1951 – 1953)

1957 – 1958 Milwaukee Braves: 5-5

1969 – 1971 Baltimore Orioles: 6-4

1972 – 1974 Oakland A’s: 5-5

1975 – 1976 Cincinnati Reds: 7-3

1988 – 1990 Oakland A: 5-5
(Lost 1988 WS 4-1. Won 1989 ALCS 4-1)

The Mariners record of 4-6 in the postseason games of 2001 looks good, but only superficially. These games were played in the ALDS and ALCS, whereas all teams up to 1969 played stronger competition in the World Series.

Those 1988 - 1990 Oakland A's were a "mediocre" dynasty. They got skunked by two NL teams in 1988 and 1990 - 1 win in 9 World Series games against the Dodgers and Reds. (Oh it was fun to be an NL fan back then!)

The A's shut out the 1989 Giants 4-0 but I don't think they were that good. The earthquake gave them a 12-day layoff which allowed them to give two starts each to their two dominating pitchers - Stewart and Moore. Thus, they avoided having to pitch their 3 and 4 starters.

To some degree, it takes different abilities to be the best regular-season team and the best postseason team. In the regular season, it's probably better to have 5 above-average starters than 2 aces and three questionable starters. In the playoffs, you probably prefer the latter.

I think there's definitely some value in the postseason because it shows ability to succeed to in two separate environments (you can't be a terrible regular season team and make the playoffs). I tend to weight it a bit less heavily (like the 2001 Mariners have to be a better team than the 2000 Yankees) because a short series has a lot of luck involved. I also think the wild-card has to dilute the value of a WS victory relative to the regular season success because of the extra series, the fact that the extra series is even shorter, and also because making the playoffs doesn't necessarily demonstrate the degree of regular season dominance that it used to.

Really? The ability to beat one of the top 4 teams, such as in a World Series, is very significant. The only exception I can see is when the strength of a league or division is much, much higher than the other. The champ who emerges from the stronger league could be a better team, regardless of what happens in the WS.

One can imagine a situation where teams A, B, and C, play 100 games against each other, with Team A going 80-20 vs. Team B, Team B going 80-20 vs. Team C, and Team C going 80-20 vs. Team A.

Now imagine that Team A is in one league and Teams B and C are in the other. Teams A and B go to the WS and Team A wins, but is inferior to Team C.

BL, I think we're talking past each other. I don't understand how any of your posts are responsive to mine.

I'd explore that in more detail, but I hurt my left arm and have to type with one hand. I'll just say that regardless of RL situations, the sim here will be affected by randomness because that's what sims do. IIRC, you yourself proposed running multiple seasons in order to reduce this problem.

We are arguing: 1) whether postseason results should be a criterion; and, 2) what weight to give postseason results for admitting a team to the simulation.

Agreed. My question is, why are postseason results relevant for a sim that will run regular season only?

CFB, if you really think players think the playoffs are a crap shoot, switch your allegiance to Beane's team. How much motivation would a player with this view have? Very little. If a manager thought his player had that view, how much confidence would the manager have in that player?

I guess nobody wants to play for the Braves, then. Stan Kasten has expressly stated that the playoffs are a crapshoot.

As someone pointed out above, LOTS of inside baseball people say this.

The playoffs are a crapshoot. not a 50/50 crapshoot of course, but they are a crapshoot. I don't think the players care one way or the other about it, as they are out to win year in year out. winning the world series does not mean you were the best team that season, it means you were the best team that week.

The world series is nice, and as mentioned before, they are the teams that are remembered in history, so of course everyone wants to be the champions, but I really don't think anyone outside of the team that wins the world series, really believes that the best team is the world series champions, some years like 2005 and even 2004 it's probable that the world champs were the best team in baseball, other years say marlins 2003, it's pretty likely that the best team didn't win the world series.

I just don't see how a short series is more conclusive than a long season.

I think every single player who has ever played baseball thinks their season is a failure if they don't win the World Series (unless they knew going in that their team sucks).

I doubt that. I don't think the 1984 Cubs considered their season a failure. I don't think the 1950 Phillies conisdered their season a failure. I don't think the 1991 Braves considered their season a failure.

They knew going in that their teams sucked.

What would be your initial hypothesis for testing "greatness"? If the result is a victory of 4 games to 3 for some team, would you adopt a new hypothesis?

What is your definition of a great baseball team?

Good to see you too ss. I think your questions aren't necessarily in the proper order.

Great has good connotational meaning. It has little analytical meaning because "great" is subject to many different interpretations particularly at the boundaries.

Before you get to your hypothesis, you have to decide what it is you want to measure. From this thread, there is no consensus. It sounds as if the closest test is:

"(1) Given a team's roster in its roster construction, (2) presuming that differences in era do not impact the outcome, (3) placing each an every team in an environment where their individual skills are not impacted by the environment, (4) Presuming they perform with the same distribution of output (5) which roster throughout baseball history would produce the most wins over a 162 game schedule."

Now, you guys are arguing on the sample to subject to that test. (And its not intractable to run that on the entire universe rather than just a sample.)

Second, that is a game, and not analytical; condition 2 is clearly false. Condition 3 is impossible. Condition 4 is not real plausible.

Third, their is no consensus on meaningful criteria, even inside the DMB simulator. For instance, how do you handle utilization, does it have to match exactly?; does it have to be within a range?; or will you use a poisson distributed function to change the distribution in order to model injury.* If you do that, do you try to model the utilization factors more, do you use DMB's generics (Prone, Normal, Stud), etc. How do you handle other intrinsic skills (Normal, Clutch, Terror)?

If that is your test, then the postseason will have relatively little weight. It should be factored in for intrinsic skills (like Clutch), and it should be factored for its extra outcome space. But if you are testing over an arbitrary length of games, your selection function for the sample need not overweigh the result of a series. It won't matter too much in the test. (See Acey, the saberists do some cluster analysis.)

I think that addresses part of questions 1, 2, and 3. But you did ask how I would do things. To be honest, I don't know that much of it would interest me beyond the fan stage. I'd root, root, root for my Bravos. I'm not sure how the test produces value beyond entertainment. However, I've never had that obsessive/compulsive list making thing that is so vogue in this generation. I'm a throwback guy. I just look at tiers of greatness. It doesn't matter much to me if Jimmie Foxx was a better hitter than Ted Williams. I know they are both great and distinguishable from others in a meaningful way.

Nevertheless, if I adopted the test, I presume that is being used above, I would not think that a single 4-3 series victory provides a lot of information on THAT TEST.

If I were trying to be analytical, I'd probably be searching for things that make it more likely to win the objective that matters, those high intensity short series. In which case, my selection function would be entirely different, as would my test. Because then "wins" don't matter as much as "series wins" DMB is probably not the best simulator for that test. And in that case 1, 4-3 series lost still may not provide a lot of information, but it provides far more than in the former case.

*-Maybe this will help a little Mark. The simulator does randomize things that it has not expressly modelled. So the output of this simulation will depend upon the values of that random variable. That does not mean its "random" when it happens in the real world.

Before you get to this, a straight model may also have entropy, and that entropy increases with time t. You handle entropy in a simulation sometimes by randomizing a variable. That does not mean its random when it happens.

If you want to project JD Drews performance next August, there are a lot of unknown events between then and now. If you now he gets hit on the elbow on July 30, you have a different projection than if you know that he is perfectly healthy on July 30. Well today, you don't know this and their are too many interdependent events. So your model will have entropy based on health; if you simulate it, you may randomize whether or not he has an injury. On July 31, you know these things, and you have neither entropy nor is their a need to randomize the August prediction. What has happen is not random, and its not pursuant to some mystical force.

"sample" and "whole" do not have any meaning outside of context. It is likely that given two random samples and a random hypothesis, the larger of the two random samples will give you more certainty in your hypothesis. But that's not nearly true in this context. You aren't even trying to get a random sample, you are fighting about a selection algorithm for an undefined test.

If what you want to figure out is if given an infinite number of 7 game series, with x,y,z,a,b,c, who do you project to win between teams S and T. Prior results of S and T aren't samples of a whole. They are the whole. A previous coin flip isn't a sample of the space of coin flip, its a completed trial.

The only way the language makes sense, is if you take it close to Tango's level. If you are searching for this metaphysical "True Talent" (which would not be static) previous trials provide you some evidence on the space of ALL POSSIBLE OUTCOMES. When you forecast, you may use some randomizer. And it want be that random either. If you want it to be close, go talk with Daly about his semiprime stuff. Most likely, it will have a discernable distribution.

A 7 game series has little value in forecasting who will win the next 2,000,000 games between the teams (although it does have some value). But all of that is fictious too.

You may agree with Treder, but what he is saying is prima facially wrong.

The ballot thread will be posted January 3rd, to give everyone a chance to return from their vacation/sober up

'76 Reds...the end.

Thanks BL. Now I understand your point. I disagree in part, but my typing is so cumbersome right now that I'll take a rain check on the discussion. You might want to save your last post against that day.

For what it is worth, my opinion:

Sim a season. Every team needs to play every team multiple times. There is more of interest here than which team is number 1. It is also interesting to find out which team is number 2, 3 etc. If you do a tournament format, the second best team could drop out in the first round.


If I was running a sim like this, for my own amusement, I would try to gather the best teams I could, however many, but I would include one of the Jays WS teams, the Expos strike season, and the '99 Spiders.

Simply because I would want to see how the first two stack up and I'd like to see if the third would win any games.

I think the 1901-03 Pirates need representation; if pressed I would choose th '01 squad. 1909 Pirates are also a logical choice, but for some reason I prefer the turn-of-the-century version. The 01 Pirates pitching staff was insane, and they had the best SS-3B combo of the deadball era, as well as a great outfield in Beaumont and Clarke

Interesting that the very first post in this thread mentions the '60s Dodgers, but they are not one of the 64 teams listed in post 442. Were they overrated because of Dodger Stadium and the Koufax mystique, or do they belong on the list of 64? They didn't have great hitting, but they did win 3 pennants in 4 years.

Interesting that the very first post in this thread mentions the '60s Dodgers, but they are not one of the 64 teams listed in post 442. Were they overrated because of Dodger Stadium and the Koufax mystique, or do they belong on the list of 64? They didn't have great hitting, but they did win 3 pennants in 4 years.

Not to mention that the early 60's NL was an overwhelmingly superior league. Or that the Dodgers missed a fourth pennant by one bad inning of relief pitching in the 1962 playoff. Or that they completely shut down a 104-58 Yankee team in a four game sweep in 1963.

In a best of seven Series, these LA teams would have been a threat to beat just about every team on that 64 team list. This doesn't make them one of the 10 best teams of all time, but it should put the best of those teams (likely 1963) on the big list. You can safely remove both of those 1954 AL teams for starters; neither of them would have had a prayer against those Koufax-Drysdale teams, regardless of what any computer might say.

And I still would like someone to explain what makes winning a 154- or 162-game race all that special, in terms of rating a team's strength. If the point is to eliminate "luck" or "randomness" as a factor in this "contest," why not just run the top 64 Pythagorean teams through the computer, and not pay any attention at all to pennants or division championships?

In truth, running only the top 64 Pythags makes perfect sense, following the logic of many of the arguments I've read here about the meaninglessness of titles, etc. And I think that the main reason it isn't going to be done that way is that many of you don't want to face the logical conclusion of your own belief system, which holds that concepts such as "strength" and "best" can be determined more by statistics than by concrete team accomplishments, such as championships.

And I still would like someone to explain what makes winning a 154- or 162-game race all that special, in terms of rating a team's strength.

It isn't the be-all and end-all of metrics of a team's strength. But it's a useful one, as are many others, including Pythag record, post-season performance, and quality of competition. On its own, I would give it more weight than winning a best-of-5 or best-of-7-game series, in terms of rating a team's strength.

If the point is to eliminate "luck" or "randomness" as a factor in this "contest,"

I'm not certain that was the point. I think the more central point in the discussion of regular season vs. postseason records is to recognize the reality of "luck" or "randomness" as factors in real-world baseball.

why not just run the top 64 Pythagorean teams through the computer, and not pay any attention at all to pennants or division championships?

That might be fun. Of couse, Pythag has its own set of problems.

many of the arguments I've read here about the meaninglessness of titles

I don't recall reading many arguments about the meaninglessness of titles. Seriously. Maybe I've missed them, it's been an extremely long thread.

I recall quite a few arguments (I'm often making them myself) about the importance of understanding the inherent limitations of titles as indicators of team strength. I think they absolutely are important and should be incorporated. I don't think that in and of themselves they tell the whole story.

I think that the main reason it isn't going to be done that way is that many of you don't want to face the logical conclusion of your own belief system, which holds that concepts such as "strength" and "best" can be determined more by statistics than by concrete team accomplishments, such as championships.

Seriously, I don't know exactly who it is you're addresssing this too. If I'm getting in someone's way, please pardon me. For myself I'd say that concepts such as ultimate "strength" and "best" among MLB teams are elusive, indeed likely impossible to confidently determine by any single means, and something close to impossible to confidently determine, period. I strongly believe that both statistics and "concrete team accomplishments, such as championships" are necessary but not sufficient, on their own, in that ambitious endeavor.

strong silence- The real goal of baseball teams is to make money, as I've noted. Must we consider profitability now?

Actually, I'm sorry, I thought you weren't trolling. Most people who aren't children can grasp idea that the better team doesn't always win.

In truth, running only the top 64 Pythags makes perfect sense, following the logic of many of the arguments I've read here about the meaninglessness of titles, etc.

And then declaring the Pythag winners as champs, rather than following the standings.

And then declaring the Pythag winners as champs, rather than following the standings.

RDF!

Andy- That's the old sportswriter conceit about favoring "stats" over "what really happened". Statistics are just what happened. Liking holding your opponent to 21% fewer runs than league average isn't a concrete accomplishment?

Most people who aren't children can grasp idea that the better team doesn't always win.

Ben - I never said the best team always wins. Where did you get that impression? In Post #10 above I sum up my position quite clearly. Did you miss it?

That's the old sportswriter conceit about favoring "stats" over "what really happened". Statistics are just what happened. Liking holding your opponent to 21% fewer runs than league average isn't a concrete accomplishment?

I think Andy might say it's a means to an end, not an end in itself, which is winning the championship.

Which is true, of course, but the point remains that both pieces of information are relevant and useful to the assessment of the strength of teams.

We know, with some exceptions, that the Champion is the better team, not merely some random winner.

From post #10 above.

Ben, I wrote Champion. I mean World Series Champion specifically.

Steve,

In truth I believe several things:

---Championships are the only true team goals: World's Championship, pennant, and division championship, in that order. A team which loses the Series isn't necessarily a "failure," but I would say that often a losing Series team falls short of achieving what its talent might have brought it. By this standard, I'd consider the 1987 Cardinals much more of a "failure" than the 1998 Padres, which by beating the Braves in the LCS had already accomplished a goal few would have thought possible.

---True team strength is measured, as you say, by a combination of factors, including both the postseason and the regular season

---I give extra weight to postseason results because I believe that the "best" team means the best team at the end of the year. This doesn't necessarily mean that I think that the best team by definition always wins the World Series (1906, 1985, 1987 prove otherwise), though it probably does the great majority of the time. But I do weight the postseason because I believe that it's the best measure of team strength that we have, again, at the end of the season. Regular season championships are often won by teams which build up a lead early on, often with significantly different rosters, and then hang on, only to go out quickly in the postseason. I see no reason to believe that such teams are "better" (or "stronger") than (for instance) a strong wildcard World Series winner (the 02 Angels or 04 Red Sox being prime examples), since much of their 162-game achievement can fairly be credited to a "team" which no longer exists at the end of the year.

---There are a few truly exceptional regular season teams which fail in the postseason (06 Cubs, 01 Mariners) which certainly deserve to be placed on a short list of "greatest teams," and I can see a case being made that they really were the "best" team in baseball that year. But I can't see a reason to put teams like that at the top of such a list, over teams which did win the World Series.

And I also can't see more than a small handful of non-Series winning teams being put on a list of the top 15 or 20 teams. The two I just mentioned, in fact, are the only ones I'd consider that high up. Most of these non-championship teams on that list (in post 442) like the 1912 Giants, the 1925 Senators, the 1929 Cubs, the 1946 Red Sox, the 1954 Indians (and 1954 Yankees, for Chrissakes!), etc., are interesting to plug into a computer, but teams like these are, relatively speaking, fairly common, compared to great teams which dominate both the regular season and the postseason.

For myself I'd say that concepts such as ultimate "strength" and "best" among MLB teams are elusive, indeed likely impossible to confidently determine by any single means, and something close to impossible to confidently determine, period. I strongly believe that both statistics and "concrete team accomplishments, such as championships" are necessary but not sufficient, on their own, in that ambitious endeavor.

On that subject, did you get that e-mail I sent you a couple of days ago? I got an auto-reply from your office, and I'm not sure that means I was supposed to re-send it. It concerned that very point about the elusiveness of "best," at least about the impossibility of proving time traveling "best" by statistics, without considering playing conditions. I'd be interested to get your reaction.

Liking holding your opponent to 21% fewer runs than league average isn't a concrete accomplishment?

I think Andy might say it's a means to an end, not an end in itself, which is winning the championship.

Indeed I would.

but the point remains that both pieces of information are relevant and useful to the assessment of the strength of teams.

Indeed they are, but to me it takes an overwhelming set of dominant regular season statistics (such as the 06 Cubs or 01 Mariners had) to get it into any serious discussion of "10 or 15 best teams ever."

strong silence- We know nothing of the sort. That's the point. Winning the World Series tells us that a team won a 7 game exhibition against another good team. No more, no less.

Andy- "Acheiving goals" is irrelevant, beyond that it requires mindreading. Winning every game is the player's goal, general managers regularly trade current strength for future strength, managers bring in a bad pitcher to eat innings in blowouts.... Everyone has different goals.

Having more expensive tickets doesn't mean that playoff games represent a more "real" test of baseball ability.

Steve,

Just got your e-mail. Glad you liked it.

Winning the World Series tells us that a team won a 7 game exhibition against another good team. No more, no less.

And in the Land of Self-Esteem, we're all hot chocolate drinking winners.

Having more expensive tickets doesn't mean that playoff games represent a more "real" test of baseball ability.

No, in fact there are no "real" tests of baseball ability other than sticking your stats into a computer.

---"Live from Silicon Valley, the Pythagorean World Series is on the air!"

I assume that Diamond Mind has "era normalization"; I ran a "best 32 teams of all time" with the teams imported from Lahman and ran a replay league on OOTP6 and had some glitches:

6 players hit 100+ HRs in a 162 game season:

Babe Ruth of the '27 Yankees hit .506/175/300 (yes, that's AVG/HR/RBI)
Roger Maris, '61 Yankees: .450/160/280
Paul Sorrento (?!?!?!): .315/120/222
Albert Belle, '95 Indians: .358/119/201
Lou Gehrig, '27 New York: .428/110/219
Jimmie Foxx, '29 A's: .429/109/204


All in all, about 60 players hit 60+ dingers. And David Cone of the '92 Jays struck out 412 guys. Apparently, I forgot to fix something. :D

On its own, I would give it more weight than winning a best-of-5 or best-of-7-game series, in terms of rating a team's strength.

What is a "team's strength"? If you are deconstructing to figure out the probability of winning some interval of 300 games, its also pretty useless. At that point, you are better off looking at the other components, and reaggregating. If you are trying to figure out if the team itself has a premium above its parts, its best to look how it performs under stressful circumstances or intervals. Is it prone to losing streaks? How does it handle quality competition. It doesn't provide you any information whatsoever, unless you are too lazy to look at the components. And if your selection algorithm is this precise, you can't afford to be lazy.

I mean it may sound good to say, "its one more piece of information", but its not. Its an aggregation of important information that loses precision. Those "7 game exhibitions" are going to carry more information, unless you totally ignore the players, and totally ignore everything that Andy has been talking about.

Apparently, I forgot to fix something. :D

You have to turn off the "players take steroids" feature.

Interesting that the very first post in this thread mentions the '60s Dodgers, but they are not one of the 64 teams listed in post 442. Were they overrated because of Dodger Stadium and the Koufax mystique, or do they belong on the list of 64? They didn't have great hitting, but they did win 3 pennants in 4 years.

Not to mention that the early 60's NL was an overwhelmingly superior league. Or that the Dodgers missed a fourth pennant by one bad inning of relief pitching in the 1962 playoff. Or that they completely shut down a 104-58 Yankee team in a four game sweep in 1963.

I thought about mentioning them, but they just didn't quite seem to measure up. Comparing them to say, the 1967 Cards (another bubble team that's only been mentioned by me as near as I can tell in this thread) and they come off looking inferior every way possible. The Cards blew away the league when winning the pennant, while the Dodgers usually snuck in by a few games. I know Dal[e]y's mentioned he like the teams that survive tough pennant races, but I'm more impressed with the teams that show themselves to be ahead of everyone else all year long. Those Cards won 100 games. The only time that the Dodgers won 100, they didn't even go to the post-season. Both teams won the Series twice, and lost a third. Granted the Cards never dominated a Series the way the Dodgers did in '62, but then again the Cards never got anally raped by another team the way the Dodgers were by the '66 Orioles pitching staff. The Cards had no Koufax, but they did have this fellow named Gibson.

Part of me wants to say that the only difference between the '60s Dodgers and the 1953-65 White Sox is that the former didn't get stuck in the same league with the Evil Empire right when they were at their most dominant. I'd like to believe that but the NL did integrate faster & thus was the stronger league, and the Go Go Sox got their tails handed them by a '59 Dodgers squad that has not been mentioned in this thread for good reason.

Babe Ruth of the '27 Yankees hit .506/175/300 (yes, that's AVG/HR/RBI)

Well, in 1927, Ruth hit more homeruns than any other team did, so 175 makes sense for a linear timeline adjustment. Paul Sorrento was with the '95 Indians, so I guess that they had two hundred homerun sluggers. Seriously, Craig, you must have set something up incorrectly. Either that or Diamondmind sucks.

Some more thoghts for the simulation. There should be an equal number of AL and NL teams, and they should have the modern playoff format. Post-1973 AL teams should have the DH at home, while NL and pre-1973 AL teams should not. Homefield in the World Series should be determined bu the result of an all-star game pitting the all-time NL ledgends against the all-time AL ledgends. If the 1887 St. Louis Browns make the cut, they should be placed in the AL.

Andy- I know that Primer, for some reason, frowns on being mean to civil idiots. And you do remain civil. But I never said anything that could be paraphrased as:
"No, in fact there are no "real" tests of baseball ability other than sticking your stats into a computer."

I, in fact, don't think sticking "my" stats(as most of you know, I did invent the concept of "a two base hit", later called a "double")into a computer is the only real test of baseball ability. "My stats"... Are we channeling Plaschke today?

RE: 442 -- a minor question:

How about the 1942 Dodgers? 107 Pythag wins (pro-rated to 162), 308 pro-rated wins from 1940-42. No Postseason, but they did win pennants in '41 and '47 with much the same core. WWII also had an impact on the dynasty.

I also think there's an argument for the 2001 A's

Andy- I know that Primer, for some reason, frowns on being mean to civil idiots. And you do remain civil. But I never said anything that could be paraphrased as:
"No, in fact there are no "real" tests of baseball ability other than sticking your stats into a computer."

That perhaps was poetic license, but since you did write this....

Winning the World Series tells us that a team won a 7 game exhibition against another good team. No more, no less.

Andy- "Acheiving goals" is irrelevant, beyond that it requires mindreading. Winning every game is the player's goal, general managers regularly trade current strength for future strength, managers bring in a bad pitcher to eat innings in blowouts.... Everyone has different goals.

Having more expensive tickets doesn't mean that playoff games represent a more "real" test of baseball ability.

....I'm not sure that what I wrote was all that much off the mark. Perhaps you can tell us what does define greatness beyond what the computer tells us, since tangible on-the-field markers such as championships don't seem to mean much to you beyond revenue enhancement.

I've stated in other posts that there are other factors in determining greatness beyond championships, but that they are one essential way of separating two otherwise closely matched teams. So it's not just a "count the rings" strawman that I'm asking you to address, but what weight you assign to the computer vs what weight you assign to the various levels of championships---regular season and postseason---and your reasoning for alloting them their respective weights. This is what I was trying to do in #547 above, and I'd like to see an answer to it beyond references to "civil idiots." I don't mind the personal note, which was certainly mild enough, but you didn't follow it up by really addressing the substantive issue.

I give extra weight to postseason results because I believe that the "best" team means the best team at the end of the year.

That's a valid definition. But perhaps this is revealing of our difference in perspective: I consider that the "best" team means the best team over the course of the complete baseball season, most likely to win a game at any point in the year, early, middle, or late.

This doesn't necessarily mean that I think that the best team by definition always wins the World Series (1906, 1985, 1987 prove otherwise), though it probably does the great majority of the time.

Based on my defninition of "best team," obviously I'm not nearly as convinced as you that this is so. Particularly in the modern era of three tiers of postseason tournament, I see no reason to expect that the best team wins the WS a great majority of the time. Before 1969, they more likely did.

Regular season championships are often won by teams which build up a lead early on, often with significantly different rosters, and then hang on, only to go out quickly in the postseason. I see no reason to believe that such teams are "better" (or "stronger") than (for instance) a strong wildcard World Series winner (the 02 Angels or 04 Red Sox being prime examples), since much of their 162-game achievement can fairly be credited to a "team" which no longer exists at the end of the year.

Rosters don't typically turn over nearly as much as this implies. The performance of individual players does fluctuate quite a bit, of course. And obviously by my definition of "best team," wins and losses in spring and summer count equally with wins and losses in autumn, as indicators of team quality.

strong silence- We know nothing of the sort. That's the point. Winning the World Series tells us that a team won a 7 game exhibition against another good team. No more, no less.

Ben, do you really think it's an exhibition. Using the word exhibition puts the WS in the same class as the ASG. And there is no comparison between the two because the players in the WS are giving 110%.

I'm not sure what you mean by exhibition.

If what you want to figure out is if given an infinite number of 7 game series, with x,y,z,a,b,c, who do you project to win between teams S and T. Prior results of S and T aren't samples of a whole. They are the whole. A previous coin flip isn't a sample of the space of coin flip, its a completed trial.

What do x,y,z,a,b,c represent, BL? the games played?

What is the best model for doing this exercise, DMB or the players themselves? Say Ruth, Wagner, Foxx, Munson, et al were alive today and agreed to an ordinal list of the top 10 teams, I think I would weight their opinions higher. Sounds like a Monty Python sketch:

Wagner: I hear this Derek Jeter dove into the stands did he? To catch a foul ball.

Arod: Yep, you should have seen ....
[Wagner interrupts indignantly]

Wagner: Hold on, young fella. I'll get to you in a minute. Back in my day, to catch a foul, ball we had to risk the broken shards of glass from the bottles of the damn Irish fans! That is a feat.

Arod: But..

Wagner: !. The fans today actually get out of the way of the player and make it so easy to catch a lazy foul pop. They want to protect their Chardonnay. Why you couldn't even seen the pop ups because of all the damn smoke from the Homestead and Edgar Thompson steel mills! Some games were cancelled due to darkeness because the smoke didn't let in any sunlight.

Arod: [Holding tail between legs] Ok, you're right. I apologize for interrupting.

That's a valid definition. But perhaps this is revealing of our difference in perspective: I consider that the "best" team means the best team over the course of the complete baseball season, most likely to win a game at any point in the year, early, middle, or late.

Steve,

How do you account for unbalanced schedules, strength of division, injuries throughout the season, changes in rosters due to player movement after the trading deadline, among others?

These are the things that make it very, very difficult (if not impossible) to find a best regular season team?

The 1995 Indians could have tied the 116 mark had the season been complete, but they played in a very weak division. The 01 Mariners played in a very strong division. It's problematic, as you know, but I think you've given a lot of thought to equalizing these factors.

I give extra weight to postseason results because I believe that the "best" team means the best team at the end of the year. This doesn't necessarily mean that I think that the best team by definition always wins the World Series (1906, 1985, 1987 prove otherwise), though it probably does the great majority of the time. But I do weight the postseason because I believe that it's the best measure of team strength that we have, again, at the end of the season. Regular season championships are often won by teams which build up a lead early on, often with significantly different rosters, and then hang on, only to go out quickly in the postseason.

This is very well said and I agree. One note: the 1984 Tigers started the season 35-5. They played .500 the rest of the season. Critics point this out as some reason to discount their performance but they don't realize that usually there is a stretch in a season when very good teams play .500. Why is a season of 35-5 and 61-61 worse than 16-12, 15-11, 16-11, 15-12, 16-11, 16-11?

The end of the season, the World Series, is much more important because players look forward to it and want to be healthy for it, teams set up their rosters for it, and managers try to be ready for it.

I give extra weight to postseason results because I believe that the "best" team means the best team at the end of the year.

That's a valid definition. But perhaps this is revealing of our difference in perspective: I consider that the "best" team means the best team over the course of the complete baseball season, most likely to win a game at any point in the year, early, middle, or late.

Point granted, difference noted. I have to admit that in October I have little interest in what the best team had been in April or June, other than to the extent that April or June helped teams qualify for October. We've been through this before and have agreed to disagree, but I see the regular season as more of a prelude to October than a special entity in itself. This isn't to say that I don't enjoy the regular season, but I enjoy it a hell of a lot more when the Yanks are in contention. I doubt if I could survive too long as an Orioles or Royals fan. My only defense is that I was born across the street from Morningside Park, and when I was a kid most of my friends were Italians and the Yanks' four best players were Dimaggio, Berra, Raschi and Rizzuto.

Having said that, I have to admit that as a Yankee fan I may be biased, but OTOH I've never tried to make any claim for any of the 2001-2005 Yankee teams, even though they won(*) the AL East every year. And in 2004 there's no question in my mind that the Red Sox were the best team, even though the Yanks won the division and the LCS was only decided by one game. Within the confines of my own standards, I'm pretty objective.

How do you account for unbalanced schedules, strength of division, injuries throughout the season, changes in rosters due to player movement after the trading deadline, among others?

With difficulty, as everyone else does. What I don't do is throw up my hands and decide the complete regular season isn't important, because of its complexities.

I would also say that teams' differing capacities to deal with all these sorts of things is one of the important ways in which the best rise to the top. The long, long season, its marathon aspect, is one of the essential, and most compelling, elements of the sport.

These are the things that make it very, very difficult (if not impossible) to find a best regular season team?

It's generally fairly simple within a given league and a given season. Comparison across leagues and seasons is, agreed, where it gets extremely difficult.

The end of the season, the World Series, is much more important because players look forward to it and want to be healthy for it, teams set up their rosters for it, and managers try to be ready for it.

That's so, but "importance" doesn't automatically provide a superior indicator of innate team quality. The Achilles heel of the postseason's capacity to reveal team quality is, always and crucially, the very small sample of games played.

One note: the 1984 Tigers started the season 35-5. They played .500 the rest of the season. Critics point this out as some reason to discount their performance but they don't realize that usually there is a stretch in a season when very good teams play .500. Why is a season of 35-5 and 61-61 worse than 16-12, 15-11, 16-11, 15-12, 16-11, 16-11?

Good point. In the case of the 1984 Tigers, though, they certainly re-demonstrated in the postseason that their early start wasn't any illusion. And actually they went 69-53, not .500, after that 35-5 start.

If they had folded in the postseason, I'd certainly discount their regular season consderably. By going 7-1 in October (and by their convincing manner of winning), they made a good case that their late season letdown was more a case of resting than weakening. But this wouldn't have been the case if they'd lost to KC or San Diego.

We've been through this before and have agreed to disagree, but I see the regular season as more of a prelude to October than a special entity in itself.

Then this distills the differences in our perspectives: I see the postseason as a fun tournament, but something of a fluffy dessert, a sweet trifle capping off the main course. The baseball season begins in April, and its length is one of its major charms. As any long-distance runner will attest, each step in the marathon counts just as much as every other; the race can be won or lost every bit as easily in the fifth mile as in the twenty-fifth.

This isn't to say that I don't enjoy the regular season, but I enjoy it a hell of a lot more when the Yanks are in contention.

And I certainly enjoy it more when the Giants are in contention. But perhaps this is the difference between being a Yankees' fan and a Giants' fan: I've learned (been forced to learn?) to pay close attention and take delight in every month and every team.

But perhaps this is the difference between being a Yankees' fan and a Giants' fan: I've learned (been forced to learn?) to pay close attention and take delight in every month and every team.

Insightful.

Up in the Northwest we take delight in a trade that improves the Mariners. Also, we've been forced to pay close attention to the trading deadline - since Gillick always stood pat when we needed one more piece. Also, the Hydro races. Fans love the Hydro races.

Steve,
Who in your opinion was the best team in 2005?

!OFF TOPIC!

Happy Birthday Sandy Koufax!

"Hall of Fame left-handed pitcher with the L.A. Dodgers." The Historynet.com - Where history lives on the web.

I would also say that teams' differing capacities to deal with all these sorts of things is one of the important ways in which the best rise to the top. The long, long season, its marathon aspect, is one of the essential, and most compelling, elements of the sport.

But consider: Two teams are tied with seven games to go. One team wins seven, the other six, falling one game short.

I'm not sure what makes this more of a marker of team strength than if the same two teams meet in the LCS and the second place team sweeps it.

In fact, since they are now meeting head-to-head, and not (often) meeting teams of differing strengths, I'd be even more reiterative of the question.

This goes to the heart of our differences about the relative weighting of the earlier parts of the season, and to my view that "luck" doesn't necessarily even itself out over 162 games any more than it does over a short series. Mathematicians might say this, but baseball history is replete with counterexamples, as I'm sure you'll acknowlege.

It seems in your view that "luck" in the playoffs is a significant factor, whereas a season-ending injury to an ace pitcher in July which costs a team a pennant is just "part of the game." I'm at a loss to see the practical difference. But correct me if I'm misreading your position.

Steve,
Who in your opinion was the best team in 2005?

Probably the Cardinals, but as I've said, it's extremely difficult to know with confidence.

Good point. In the case of the 1984 Tigers, though, they certainly re-demonstrated in the postseason that their early start wasn't any illusion. And actually they went 69-53, not .500, after that 35-5 start.

Bill James, in the 1985 Abstract (IIRC), writes that right after the 35-5 start he should have known that the Tigers were the best team simply because it's rare for any team to have such an outstanding 40-game stretch.

I wish I had the old Abstracts available right now.

But consider: Two teams are tied with seven games to go. One team wins seven, the other six, falling one game short.

I'm not sure what makes this more of a marker of team strength than if the same two teams meet in the LCS and the second place team sweeps it.

Neither am I. But far more revealing than any performance in any given week is both teams' performance in the 155 previous games. From what's presented here, there isn't a dime's worth of difference between them.

to my view that "luck" doesn't necessarily even itself out over 162 games any more than it does over a short series. Mathematicians might say this, but baseball history is replete with counterexamples, as I'm sure you'll acknowlege.

I'll acknowledge the counterexamples, certainly. But anecdotal counterexamples to every general rule abound. They don't invalidate the rule at all.

It seems in your view that "luck" in the playoffs is a significant factor, whereas a season-ending injury to an ace pitcher in July which costs a team a pennant is just "part of the game." I'm at a loss to see the practical difference. But correct me if I'm misreading your position.

Luck is just part of the game, period. So are injuries. But the key difference between an injury to a key player occurring in July and one occurring in October is that a team has a realistic opportunity to adjust and rebound from the July injury; their capacity to do so is a significant indicator of team quality. It's vastly more difficult to make a serious adjustment to an injury occurring during a short series -- meaning that the bad luck of the injury has far more impact on the short series.

The likelihood of equally critical injuries to competing teams is also far greater, the longer the span of time. One need not be a mathmetician to perceive this.

This goes to the heart of our differences about the relative weighting of the earlier parts of the season, and to my view that "luck" doesn't necessarily even itself out over 162 games any more than it does over a short series.

And now you know why the plan is to have each team play 162,000 games...

Steve,
I assume you would put the Yankees next, then Boston, then the White Sox.

The Pale Hose went 11-1 in the postseason. The last time that happened was in 1999.

They swept the Indians in the last season series. It was must win. I remember a few people saying the Indians were the best teams last year. Even with two weeks left, I think Dave Cameron called them the best team in baseball. It turns out they were 4th or 5th best in the AL.

Steve,
Who in your opinion was the best team in 2005?

Probably the Cardinals, but as I've said, it's extremely difficult to know with confidence.

C'mon Steve. Take a stand. Make a case for the Cards. Or are you admitting that your system can't determine the best team?

Luck is just part of the game, period. So are injuries. But the key difference between an injury to a key player occurring in July and one occurring in October is that a team has a realistic opportunity to adjust and rebound from the July injury; their capacity to do so is a significant indicator of team quality. It's vastly more difficult to make a serious adjustment to an injury occurring during a short series -- meaning that the bad luck of the injury has far more impact on the short series.

I guess I'm not quite sure how a team "adjusts" to the loss of (say) a Koufax in the middle of the 1962 season, or how the Dodgers' failure to "adjust" (i.e. replace an irreplaceable asset) proves much of anything about their strength that year. I'd call the Giants' ability to tie the Dodgers that year every bit as "lucky" as the placement of that Willie McCovey line drive.

The likelihood of equally critical injuries to competing teams is also far greater, the longer the span of time. One need not be a mathmetician to perceive this.

Or an economist (Keynes) who noted that in the long run we're all dead. But I'd be willing to bet that you could find just as many pennant races where "luck" (in the form of disproportionate injuries or other flukish circumstances at a key moment---a broken bat bloop triple by Gerry Coleman) played a role every bit as critical as you could find in the entire history of postseasons.

I assume you would put the Yankees next, then Boston, then the White Sox.

No. Why do you assume that? I'd put the White Sox next, and beyond that, I really don't know.

The Pale Hose went 11-1 in the postseason. The last time that happened was in 1999.

Yeah, they were red hot. It was fun to watch. I especially loved the dominating starting pitching performances, one on top of the other. It was a great demonstration of how tough to beat a team can be when it's got an outstanding starting rotation hitting on all cylinders.

And now you know why the plan is to have each team play 162,000 games...

With Julio Franco having played in all of them.

I guess I'm not quite sure how a team "adjusts" to the loss of (say) a Koufax in the middle of the 1962 season, or how the Dodgers' failure to "adjust" (i.e. replace an irreplaceable asset) proves much of anything about their strength that year.

Obviously, losing a key star is always a bad thing. But just as obviously, different teams have different capacities to adjust to it. A team with greater depth, on the current roster and/or in triple-A, can make the adjustment better than the team with lesser depth. Depth is a critical element of team strength.

I'd call the Giants' ability to tie the Dodgers that year every bit as "lucky" as the placement of that Willie McCovey line drive.

I wouldn't. The Giants' performance over 162 games was a vastly more revealing test of the quality of their team than one at-bat could ever possibly be. The fact that luck was an element in both doesn't mean luck was an equal element in both.

But I'd be willing to bet that you could find just as many pennant races where "luck" (in the form of disproportionate injuries or other flukish circumstances at a key moment---a broken bat bloop triple by Gerry Coleman) played a role every bit as critical as you could find in the entire history of postseasons.

Well, have at it. But bear in mind that the existence of luck isn't the same thing as its equal impact.

I assume you would put the Yankees next, then Boston, then the White Sox.

No. Why do you assume that? I'd put the White Sox next, and beyond that, I really don't know

So you are simply using regular season wins as a basis for your opinion.

If you combine the regular season with the playoffs which is justifiable.

White Sox 110 - 64
Cardinals 105 - 66

They swept the Indians in the last season series. It was must win.

They'd clinched the division before that series even began.

There are a few truly exceptional regular season teams which fail in the postseason (06 Cubs, 01 Mariners) which certainly deserve to be placed on a short list of "greatest teams," and I can see a case being made that they really were the "best" team in baseball that year. But I can't see a reason to put teams like that at the top of such a list, over teams which did win the World Series.

The problems is that such reasoning would lead you to prefer the 1908 Cubs to the 1906 Cubs, and I don't really think the 1908 Cubs were better. In fact, it's quite possible that the 1908 Cubs weren't even best in the NL in 1908, since both the Pirates (who would win in 1909) and the Giants (who lost the pennant on Merkle's Boner) finished only one game back.

So you are simply using regular season wins as a basis for your opinion.

No, I'm not, although certainly regular season wins is pretty much the starting point. I'd also consider strength of schedule, Pythag record, postseason record, and anything else that seems relevant, such as the degree to which the team's key players appeared to be playing significantly above or below their likely "true" talent.

If you combine the regular season with th