Look, Treder lives for this stuff. National League vs American League. The Willie Mays Giants vs anybody.

Well yeah. It's the same way Democrats want to talk about Roosevelt rather than Buchanan and Republicans Lincoln rather than Hoover. The last 50 years have given we Giants fans precious little to go on. We have to work with what we have.

Try arguing with individuals rather than some made-up club.

I was, the opinion you said that you were defending, the one you characterized as a minority opinion.

I disagree. You are claiming an injury is the same "luck" as a pebble.

So either:
1 - you think injuries are luck
or
2 - you don't think pebbles are luck

No, I'm not. And I don't have a "baseline" because one is not needed. Every outcome is influenced by players on both teams and the environment in which they play. The outcomes happen. If I'm investigating, I look to causes; if I'm analyzing, I look to find means of improvement. If I'm projecting I identify where I have to have randomize a variable because of the entropy in the model.

I don't make excuses based on output. I don't pretend that output doesn't provide useful information. I don't make myself a slave to large normal distributions, and old classical statistics. I don't just repeat something like "sample of games" just for its lyrical appeal without providing any evidence to bolster the inference. I don't selective ignore certain people when they have valid arguments.

Just because you are in one, doesn't mean I need *ANYONE ELSE* agreeing with my opinion.

Whether a closed event space has randomness is not a matter of opinion.

"What is luck?", but to ask "What is skill?".

Ask either question you want, and you end up at the same place. Every item you listed has a degree of influence from different parties and a degree of repeatability that varies among different parties.

The problem that you will frequently run into if you choose such binary criteria is that you are defining one thing as the negation of another. For instance, you will have some criteria whereby you will determine "skill"; right now that is pretty amorphous. Then does everything else become "luck". All batted balls are subject to those influences. There is a likely distribution that influences how certain batted balls will result in play. Your "pebble" can be mitigated with fly ball pitchers, dominant pitchers, or better defenders.

Right now, the only test you want to run for skill is some interval to interval variance, and you are dealing with much too large a space that has multiple different influences to extract value. That's fine for long range projections; it doesn't help to much with causal analysis; and its rather arbitrary to say something is luck and something is skill.

But in the end, more often than not, you still end up at the same place. Somebody won, somebody lost, and the fans of the losing team blame "luck" and the fans of the winning team proclaim "skill"

Those skill definitions you look for, namely repeatability, don't manifest in the micro level at all.

As I've stated many times, if what you want to say is given X number of seven game series, team A will win Y percent of the time, more power to you. I think your macro projection will be wrought with error, and any simulation will show the bias toward that error. Somebody still has a trophy, and somebody is still crying in their soup.

And if you use that to make decisions, or its precursor, causal analysis, you are going to end up making lots of errors like Little G.

This isn't consensus building, you can't decide that this is the relationship between items. You can define words however you see fit, but if those definitions don't lend to a useful categorization of elements then you aren't accomplishing anything analytically. If you determine that McCormick's injury is luck, or the A's inability to win a playoff series is luck, you miss the chance to actually improve things by not overusing young pitchers or building better teams.

If you wait for a saberman to descend from the sky to give you your precious sample size, you are going to be twenty years down the road and past your chance to do anything.

"Luck" has no definitional purpose in any of this. It has been, and continues to be employed for rhetorical purposes.

Yeah. I can turn Chris purple with rage like nothin'. Fish in a barrel.

I remember JC congratulating Chris on defending his POV in a tussle with Treder one time. I don't remember what it was about.

Okay, BL doesn't believe in luck. Feel free to divest yourself from any discussions where we talk about it.

BTW, MGL couldn't have written a better response.

The problem that you will frequently run into if you choose such binary criteria is that you are defining one thing as the negation of another. For instance, you will have some criteria whereby you will determine "skill"; right now that is pretty amorphous. Then does everything else become "luck".

I'm not defining it that way, though I can see how that wasn't clear from my post.

I assume we agree that hitting HRs is a skill. But the reason we can characterize it as one is that we see many such events performed by the same person. Repetition is the key, and repetition means large samples. In any one AB ANY hitter can get lucky enough to hit a HR -- witness Duane Kuiper. All we can say for sure is that repeatedly hitting HRs is NOT luck. That's what the phrase "sample size" means to me.

What I hear you saying is that there are always hidden variables, that attributions to chance are a way to avoid looking for those variables that would increase our knowledge. I agree that that is a danger, but I also believe that some events are inherently probablistic and that there are times when the search for hidden variables itself impedes progress. There's danger either way.

If you determine that McCormick's injury is luck

I don't think I ever claimed that. I'm not sure anyone did. Certainly it was no more or less lucky than Koufax's injury. I was only discussing the impact of the two events on the pennant race.

Certainly it was no more or less lucky than Koufax's injury.

That's the issue in a nutshell. Andy introduced the subject of the 1962 NL pennant race, and asserted that the entire reason the Dodgers didn't win it was because of the unlucky break of Koufax's injury -- in an attempt to demonstrate that complete pennant races are no less subject to the caprices of fate as short post-season series.

I pointed out that McCormick's injury was no more or less an imposition of luck than Koufax's, and did at least as much (and possibly more) damage to the Giants' effort as Koufax's injury did to the Dodgers'.

And -- this should go without saying -- no one, anywhere, at any time in the history of the world, has ever suggested that McCormick was just as good a pitcher as Koufax in 1962. The only thing more laughable than that is the notion that others will buy the lie that I said so in this thread.

Steve,

I never said that you said that McCormick was as good as Koufax. I think that if both McCormick and Koufax had been available for the entire season, and had pitched to what was their most recent pre-injury forms (McCormick's 1961 [excellent but nothing earthshattering] and Koufax's pre-All Star game 1962 [a precursor of the four years to follow; A-level HOF]), the Dodgers would have won. But your posts on the subject of 1962 were well taken nevertheless, and you presented your side of the case very well.

(out of order),

I've had an attack of terminal laziness which prevents my followup to last night's post, but I just posted my top 28 on the other thread if you're interested. It has 6 non-Series winners, including 3 in my selected top 10. If this thread continues for another day I will try to complete my response to your posts.

Andy,

Oddly enough I had an attack of terminal laziness which has caused me to post so much in these threads lately. Also, I don't think Treder was referring to you at the end of #806, but backlasher.

I've had an attack of terminal laziness which prevents my followup to last night's post,

That's a funny line.

What is it with you guys needed teams?

That's the problem with teh VBB, they miss all teh good fights.

Yeah, you don't speak in teams, right?

I'm not sure exactly whose "team" I'm supposed to be on here, since mostly it's just been me and BB-Reference trying to make sense out of the last 105 years of baseball. You never know how ignorant you can be until you immerse yourself in that truly amazing site.

And Chris, I don't recall the words "steroids" or "greenies" being used even once here, and I was under the impression that this was what we Unionists / VBB's were all about. I haven't seen all that much carryover from those threads to this one, but maybe I've been too stuck with BB-Reference to notice. But are you getting nostalgic? Do you miss those good wars of your youth? Can I help? Get you a picker-upper or something?

Also, I don't think Treder was referring to you at the end of #806, but backlasher.

Steve states that he doesn't read my posts much less respond to them.

Okay, BL doesn't believe in luck. Feel free to divest yourself from any discussions where we talk about it.

That's novel. And which conversations are you going to divest yourself from.

But your posts on the subject of 1962 were well taken nevertheless, and you presented your side of the case very well.

Yes, and it goes back to your original point. We can play "but for" in the regular season as well as the postseason. If we do, we get a whole set of "but for" causes.

I agree that that is a danger, but I also believe that some events are inherently probablistic and that there are times when the search for hidden variables itself impedes progress.

Almost all future events have a probalistic distribution. Past events are causal. Knowing what caused a fire does not necessarily make the future distribution on the likelihood of fire any different. There aren't "hidden variables" as much as there are actual causes. Something was there or not there. You have to determine that before you can know whether your future projection is valid.

I assume we agree that hitting HRs is a skill.

How about ones that are caught by Jeffrey Meyer. One's that just skim over the top of the wall. Are base hits skills? How about ones just past a diving Jeter. People want to talk about meteorites and pebbles. I haven't seen a meteorite land on the pitchers mound lately.

If you are talking about rare events, every series is full of them. Lonnie Smith scores more times than not. Mark Wohlers gets out Leyritz more times than not.

If you want to talk about injuries or unavailable players, every series is full of them. The Braves could have used Nixon; they could have used Justice.

If you want to talk about events that are largely external, every series is full of them. The Braves could do little about the Phillies sticking needles in their asses.

Many of those are "but for" causes on the series lost. Some are not, are at least subject to argument. If the results are reversed, the same is true.

As Andy mentioned, the same is also true of the regular season. And if you get to the rareness of meteorites and pebbles, then they aren't likely to be offset with the lucky acquisition of a Bam Bam.

The important thing in all of this is a seven game postseason series, or the set of postseason series are different than the regular season. They have winners and losers. If you want to use them for analytical purposes, they provide different information.

They should not be dismissed by "Luck is part of the everyday ..." The events themselves should be analyzed, and then you can determine if the future distribution is too small to care about improving.

And they should not be silenced with "When we talk about luck feel free not to join in."

If two teams tie a series 3-3, pitch duel shutouts to the ninth inning, and the pitcher gets on base with a pebble-induced base hit, and he steals the next three bases because the fielders are blinded with a meterorite, I'm not going to take away that one team sucked and the other ruled. One team still wins, and from that whole series, I'm sure there is a lot of information to be gathered from both teams.

But I'm not going to ignore every postseason series because they are just "luck" I'm not prepared to say one team is inherently "better" than another for all definitions of better just because they performed better in the first 155 games or the first 162 games. I'm not going to throw all analysis that shows injury causality out the window as "luck"

When I joined this thread, the discussion was that postseason series were not a good test because of "luck" I've since heard that injuries are luck, and that someone not performing to their median performance level is not luck. I've seen a pitcher with 2 seasons at above average supposedly establish his skill and the rest of the career as crapitude be defined as luck.

I see a pitcher that was stinking up the joint and constantly being run out to the mound as an indicator of luck. Most every team has underperforming players. What is important is the cause, and to see if you can prevent future causes.

I never said that you said that McCormick was as good as Koufax.

I know you didn't, and I sincerely apologize at implying such a thing. You always conduct yourself as a gentleman, to a degree that I can only hope to aspire.

I think that if both McCormick and Koufax had been available for the entire season, and had pitched to what was their most recent pre-injury forms (McCormick's 1961 [excellent but nothing earthshattering] and Koufax's pre-All Star game 1962 [a precursor of the four years to follow; A-level HOF]), the Dodgers would have won.

The evidence pretty clearly mounts against you, but I will certainly agree that counterfactuals are by definition a matter of uncertainty. I understand you to acknowledge that Koufax's injury is anything but a full and complete explanation of the actual 1962 finish.

But your posts on the subject of 1962 were well taken nevertheless, and you presented your side of the case very well.

As did you.

Remember, the minority opinion you are joining is saying "injury" is luck. Deviations from median OPS performance is luck.

Do you enjoy repeatedly misrepresenting people and then beating up the misrepresentations?

Nobody has said either of these things.

The people talking about luck with regards to injury are not saying that injury in a general sense is a purely random proposition. Nobody debates, for example, that Tejada is less likely to get injured than JD Drew. They MAY say that the very same broken wrist caused by a HBP has a different effect in late September than it does in mid-July, and that the timing of that injury is luck. Or that a player with known knee problems who breaks his wrist could be considered unlucky rather than an example of poor judgment on the part of the GM that signed him.

As for the idea that fluctuations in performance from year to year are solely the domain of luck, the only place I've seen such a preposterous claim in when you try to place it in the mouths of your opponents.

You are not being singled out because you have a "minority opinion" (putting aside the concept that there seems to be a fairly good mix of people on both sides of the fence here), but because you persist in pernicious, offensive, and disingenuous techniques of arguing.

But your posts on the subject of 1962 were well taken nevertheless, and you presented your side of the case very well.


Yes, and it goes back to your original point. We can play "but for" in the regular season as well as the postseason. If we do, we get a whole set of "but for" causes.

Exactly.

My point with Steve began with my post about 1962, but it ended (for me, at least) with getting him to give a bit of detailed reasoning as to why that season didn't turn on Koufax's injury. I don't mind his case for the McCormick injury being the equivalent of Koufax's; we can present facts one way or the other on that, and he may even be right. But what I objected to is the assumption that once you've defined "luck" in whatever way you want, it mostly applies just to the postseason, because of the sample size factor alone.

And I just don't buy that. If you want to argue that the 19XX regular season was less influenced by luck than the postseason of 19XX, I'll listen, but only if you show me your facts and your reasoning for that particular season. Steve did that for 1962, and whether I agree with his conclusion or not, at least I now have something to work with rather than just a theory about sample size.

Part of my thinking here may come from the fact that I'm much better grounded in the specifics of baseball history than I am in mathematics, and therefore I'm much more receptive to specific facts, and arguments based on specific facts, than I am to generalizations. As far as I'm concerned, the nuances of any given season or postseason should be able to stand on their own.

BL, I agree with all your points about the impact of chance and fluctuation in affecting the outcome of both seasons and series. Where we apparently differ is that I see the long season as a better test of greatness.

Let me analogize to indivduals. We'd like to know what mgl calls their true talent level. But we can only know that by gathering all the performance data we can. Every pitch gives new information. The more performance we can observe, the greater our confidence in our estimate of true talent.

The same is true for teams. Every game increases our confidence in its true ability level. That's why we prefer a season's worth of games to a series. It's not that the latter provides no information, it's that it provides less.

I don't recall some of the arguments you mention so I'll let those who made them respond.

But what I objected to is the assumption that once you've defined "luck" in whatever way you want, it mostly applies just to the postseason, because of the sample size factor alone.

And I just don't buy that.

Then I'm afraid you continue to fail to grasp the fundamental concept.

The sample size factor alone is precisely the reason why the vagaries of chance are likely to have a greater impact on a short series conclusion than a long series conclusion.

Let me try it this way:

Skill is skill, and always works to the advantage of the player/team who possesses the most of it. Yes?

But luck, being luck, works to no one's advantage in any predictable, systematic way. It might favor this team today, that one tomorrow, and who knows which one the next day. Yes?

Since luck is beyond the realm of the skill of teams to control it, and subject to nothing except the whims of randomness, it generally does have a way of "evening out" between two or more teams over a long enough period. Yes?

Then the whole question becomes how long is a "long enough period. Yes?

And the answer to that I would offer is: the period is never long "enough." Nothing can possibly guarantee that luck will fully even out. But it is definitely the case that the longer the period, the better chance we have of the luck coming closer to evening out. 2 games creates a better chance than 1. 3 games creates a better chance than 2. And most assuredly, 162 games presents a far better chance than 5 or 7. Yes?

And they should not be silenced with "When we talk about luck feel free not to join in."

You aren't "being silenced". You can post anything you want, but since you "don't believe in luck", or believe that "there is no luck, just "closed event states with randomness", then you aren't interested in the discussion that I am. Since you aren't going to convince me there is no luck, I'm simply letting you know that you don't have to waste your time attempting to explain "closed event states with randomness" aren't luck. I'm doing you a favor.

I know what luck is when I see it. You don't like the word - fine, is there any reason I should try to discuss it with you?

I don't see "pernicious,offensive,and disingenuous"....

I am interested in "low probability outcomes" especially as it relates to listing the best teams...in particular I just can't bring myself to support the 2001 Mariners................................. The Kuiper HR...Podsednik HR...Red Sox 0-3 comeback..injured Gibson bomb off Eckersley..Carbo and Fisk..1 strike away games that turn and go the other way..60' Pirates....K-Rod defensive indifference... all are different types of low probability events..I don't have any profound comments about them but it is amazing how each event turns into Kozinski's "Being There" in the sense that we project upon them and each fan has or can have such a varying view on what happened even when we can go inning by inning and know what happened.....I think the 95' Indians were better than the 95' Braves but they weren't.....

Steve,

I fully understand sample size, and its importance. And why in general it makes the regular season a truer test of overall, 162 game strength.

But I tend to measure team strength much more by what I perceive it to be at the end of the season (and including the postseason) than what I perceive it to be in May. In other words, I weight the end of the season. I realize you don't, but this is more a matter of choice or definition; neither is objectively right or wrong.

And this in turn makes me give special weight to the postseason: not just to the W-L (since luck can often rule), but to the overall strength as I see evidenced with my eyes. To take one example, the Cardinals may have been "better" than Boston during the 2004 regular season, but to me there was no question that the Red Sox, at year's end, were the superior team. You can call those four games a small sample size, but they were also a serious butt-kicking by a team which had showed the same ability against the Angels and the Yankees. It was no surprise to me what happened, regular season stats to the contrary. Call this 20-20 hindsight if you wish, but I doubt if I was the only one who saw it coming.

OTOH, in 1995, even though on some level I'd call the Braves a better team than the Indians, on another (and somewhat higher) level I'm pretty sure that the Indians were an overall stronger team. But that Indians team was a lot closer to the 1906 Cubs or 2001 Mariners in talent and achievement than it was to most any of the other non-winning teams I can think of.

And you won't find the 97 or 03 Marlins (or the 2000 Yankees, for that matter) on my list of 28, though the last two of those certainly didn't win their titles by any postseason "luck," Bartman aside.

IOW, I start with the assumption that the postseason results are a true test of strength, but if there's serious evidence to the contrary (which there seems to be more of lately), then I'm always willing to consider it. I don't find it surprising that of the six non-winning teams on my top 28, three of them have come in the past eleven years.

And if a team is overwhelmingly dominant during the regular season, but fails in the postseason, I try to look to see if that failure represents a somewhat flukish occurence (e.g. in the case of the 1969 Orioles, who ran into a buzzsaw of a pitching staff on the level of the 63 Dodgers), or if it represents a more serious underlying flaw in a team's makeup which was exposed during the postseason (Exhibit A being, yes, the 1954 Indians). And I take this on a case-by-case basis, which is the only way to be "fair" about it.

The one big advantage to my approach, BTW, is that it forces me to go into the particulars of each season and postseason, and learn a lot in the process. This is one of those many cases where "the journey, not the arrival" is what's most interesting. But of course I hardly have to tell you that.

But I tend to measure team strength much more by what I perceive it to be at the end of the season (and including the postseason) than what I perceive it to be in May. In other words, I weight the end of the season. I realize you don't, but this is more a matter of choice or definition; neither is objectively right or wrong.

Lightbulb goes on in Mefisto's brain.

Andy, this approach makes complete sense if we're trying to predict the outcome of the series. I'm not sure, though, that it makes as much sense in the context of this thread. The challenge here was to pick the best teams on a seasonal basis. For that specific purpose, it makes more sense, IMO, to look at the entire season -- to evaluate, in essence, the team's average strength over the course of the year. The fact that the competition will simulate an entire season (or many if they adopt BL's good suggestion) reinforces this for me.

Andy,

I understand your perspective.

But here's what I object to. You say this:

I fully understand sample size, and its importance. And why in general it makes the regular season a truer test of overall, 162 game strength.

And then you immediately say this:

But I tend to measure team strength much more by what I perceive it to be at the end of the season (and including the postseason) than what I perceive it to be in May. In other words, I weight the end of the season. I realize you don't, but this is more a matter of choice or definition; neither is objectively right or wrong.

You can't have it both ways, Andy. Either you understand the importance of sample size, or you don't. And if you do, it cannot be that "weighting the end of the season" is simply a matter of choice or definition, that neither the choice of weighting postseason results or full season results as a measure of team strength is "objectively right or wrong."

If you truly understand the importance of sample size, you simply cannot decide to focus on short-term results and draw large-scale conclusions from them. To do so is objectively wrong.

Baseball's a sport, a game, a pastime. None of it matters, and all of us are completely free to take whatever enjoyment we choose from it, in our own way. That much is indisputable.

But once we venture past the simple enjoyment of the sport into what we hope to be a rational, intelligent analysis and discussion of its results, then it's vital, if we hope to accomplish anything satisfactory, to come to a common understanding of the analytical principles of analysis. Among these are the fundamentals of the "scientific method," a basic agreement of what constitutes valid evidence, what doesn't, how hypotheses can be reasonably tested, what the differences are between hypothesis, theory, and accepted knowledge, and so forth.

In this spirit, either we accept and acknowledge that sample size is critical in the weight of evidence, or we don't. If we do, we must then live within the limitations that acceptance presents. What we cannot do is, on the one hand, say we accept and acknowledge the sample size issue, and on the other, choose to ignore it when we feel so inclined, and dismiss that choice as "simply a matter of choice."

The objective search for knowledge doesn't work that way. We either live within the constraints of analytical principles, or we don't. If we don't, we forfeit our privilege of having our conclusions and assertions being taken seriously by others who are abiding by those principled constraints.

Mark,

You'll note that in my list I included three non-winners in my top ten teams, and 6 in my list of 28. My presumptions are but the starting point of my evaluations. They are always subject to modification as the particulars may warrant.

But if you measure only the regular season, then why not just rate the teams in order of their Pythagorean ratings, or by their combined OPS+ / ERA+ numbers (throwing in fielding for a bit of seasoning, perhaps), and do away with all discussion or subjective factors? Your logic does seem to imply this, even though I doubt you'd take it to this conclusion.

Steve,

What I wrote to Mark above can be directed to you as well.

Focus on the particulars.

As to your reference to my "matter of choice," the "choice" in question was whether or not to choose to see the "team" as the team of October (or simply weighted at the end of the year, including the postseason), or to see it as the average of the team as variously constituted during the stretch from April to the end of the regular season. This has much more to do with preference than it has to do with not accepting the scientific method.

If you see, as I do, that the purpose of a team is to meld itself to a fine tune at the point of the postseason, then you will see, as I do, the regular season as more of a prelude than a finality---and even more so for the teams under consideration here, who were nearly all aiming at the postseason from early on in the year.

As you can see from my list, I'm not dogmatic about this. But it does constitute a difference of approach that (I hate to say it) we'll just have, to, yes, agree to disagree about.

And now I cede to my worthy West Coast opposition for the night. I think the sun's already down here.

Focus on the particulars.

As we all should, but never to the extent that we forget the underlying generalities, of which the particulars are mere manifestations.

As to your reference to my "matter of choice," the "choice" in question was whether or not to choose to see the "team" as the team of October (or simply weighted at the end of the year, including the postseason), or to see it as the average of the team as variously constituted during the stretch from April to the end of the regular season. This has much more to do with preference than it has to do with not accepting the scientific method.

No, to the extent you wish to present a compelling argument regarding which team is inherently superior to another, it has everything to do with accepting the scientific method.

By making the choice you do -- focusing on results over a short stretch of games (September/October) in preference of a longer stretch of games -- you unavoidably confront the sample size issue. It doesn't go away, just because you've made the choice you have. It's still there, and given the choice you've made, it's enormously there, looming over everything, rudely intruding into every tiny aspect of your observation.

Either sample size matters, or it doesn't. You say that you accept that it does. Well, if it does, then it does. If sample size matters, then it always matters.

The fundamental principles the sample size issue represents don't take every October off. They don't decide to go away, just because MLB has decided to award champions based on small samples of games, or just because TV ratings are high for post-season games. The fundamental issues surrounding sample size still apply. They always apply. The fact that this makes things difficult for us in drawing conclusions from post-season results is not a concern of sample size issues; they don't give a crap about our problems. They just stupidly, pitilessly, tirelessly, always apply.

So you can choose to focus on a small sample of games from which to draw conclusions about the quality of teams. But to the extent that you do this, you are simultaneously choosing to ignore the fundamental, eternal, omnipresent issues of sample size, and as a consequence the conclusions you draw cannot stand up to objective scrutiny to the same degree they would if you incorporated a larger sample of games into your analysis. They cannot.

It isn't simply a matter of choice, with no consequences. This is the way things are in the universe in which we're all stuck. By limiting your range of inquiry to a small sample of games, you forfeit conclusive power. And the smaller the sample of games you select, the greater the forfeiture becomes. We may wish things weren't this way, but they are this way anyway.

"You either understand the importance of sample size or you don't" implies a value judgement that obscures the fact that baseball players compete in a baseball game in a baseball park for a baseball championship..my issue with this analysis is it diminishes and distorts what happens on the field..that there is a leap from the field to the boxscore to the simulator...it is minimalist in that it segregates the competition.. animate/inanimate..at once ..with a broad stroke of self assumed superiority... and this perspective is at odds with "the residue of design" and crystallizes in this type of journey..ie..to get the best of the best..my opinion is that the 2001 Mariners team was exposed during competition..that they were somehow flawed on the field..and they lost.................and that this is different than constructing or analyzing ballclubs with the analytical tools at our disposal..UZR.OBP.K/9.WHIP.BABIP..etc........extrapolate to the absurd and when Bud Selig hands the trophy to George Steinbrenner he makes a short apology..then cites sample size...

Buzzards Bay,

I'll just say you fail to present a serious argument, and leave it at that.

Steve, I think one point from Andy's philosophy that you maybe missed is that he considers the end-of-year team a different animal than the April 5th or the July 12th team. To the extreme, if the Yanks trade A-Rod, Sheffield, Giambi, Jeter and Randy Johnson for Neifi Perez on July 13th, the performance of the pre-trade Yankees team cannot be used to correctly evaluate the October 10th Yankees. I think it's a valid proposal. And I'm pretty sure Andy agrees a 162 game sample is better than a 7 or 12 game sample. The problem is that the 162 game sample is not a sample of the end of year team.

Now, I know that usualy, teams are not totally transformed mid-season by the kind of ridiculous trade I discussed hereabove. But still, a young pitcher developing, a couple of useful players added, a key injury, normal player progression or decline, all these things can make the end of year team quite different from the "average team over 162 ganes" without a blockbuster deal. I think that's the biggest difference between your respective views, not that one likes a big sample size and the other doesn't care.

Makes sense?

Andy, I completely understand your view wrt the regular season, though it's not the way I think of it. I also agree with TYB on the implications of your view. My assumption for this thread, however, and I stated it far above, is that the sim will give players the same playing time as they actually got in the relevant season. If this is true, I think it undercuts your point for this purpose. Also, and now I'm repeating myself, this sim will treat the winner as the one which wins the most games over the course of the regular season so there won't be any advantage to teams which improve over the season. In this context, the inability to trade and the playing time restriction I'm assuming reinforce the seasonal format.

In this context, the inability to trade and the playing time restriction I'm assuming reinforce the seasonal format

It's an interesting issue. This kind of simulation is really about ideals: you take Platonic images of several superteams and then compare them. In real life, the great charm of some championship teams is that somebody unpredicted comes along and jump-starts them: Duster Mails in 1920, Barney Schultz in 1964, Marty Bystrom in 1980. Those players don't fit well into the ideal simulation (they tend to have lethal Strat cards, for instance, and have to be reined in by various restrictions) but they are a major part of real pennant races.

Yankee Blower,

Gracias. You get what I've been trying to say.

Steve,

I'm not sure you do, even now. Teams are dynamic, not static. Perhaps the best team to illustrate this from my perspective is the 1914 Braves, whom I suspect I'd rate a lot higher than you would, even if not in the top 28. When you see that team (and correct me if I'm wrong), you weigh all of their games equally. When I see that team, I see a team that went something like 68-19 over its last 87 games and then swept a very good A's team in the World Series, and I weigh that accordingly. Not to the point of "28ing" them, but certainly to the point where I question the idea that the 1914 World Series was much of an "upset."

I think maybe that one of the underlying objections you have to my pov here is that it's not "fair" to teams which dominate the regular season and then fail in the postseason. And by that I mean as much "fair to the method of proper evaluation" as "fair to the team" itself. And again, you explain your case quite eloquently.

But I still question the whole idea of "averaging" a team's performance over the regular season, rather than doing that, and then also looking at other factors in order to evaluate the whole. And among these I would include both the postseason and the latter half of the regular season, especially if it carried over into the postseason, as was the case in 1914. Call it bonus points for rising to the occasion, and for proving (to me, anyway) that the last part of the season was more representative of the team's strength than the first half.

See also my answer to Mark, below.

Mark,

I do see your point (and I think I addressed it earlier), but here I'd repeat my hypothetical question from an earlier post, to wit:

If it's only the regular season we're rating, then why all the pretense of debate and discussion? Why not just stuff the top 64 Pythags into the computer and be done with it? Or the top 64 combined OPS+ and ERA+ teams if that better suits the exercise.

In which case, what's the point of this thread? And for that matter, why not just feed the computer all the possibilities (no need to restrict it to 64 in the first stage), let it arrive at its ranking of the top 10 or 28 or 64 (whatever), and then begin our debate, with the subjective factors of luck, late season surges, and postseason performances then entered into the discussion? What was the point of having us introducing all these nominations? The computer could have done that in about five millionth of a second on its own.

And at the end of that debate, we would have two distinct lists, both interesting in their own way: Best regular season teams by computer, and best overall teams as determined by the weighing of all factors and arguments.

The problem with this thread to date is that we've been having one set of people trying to mimic the first list, and another set of people trying to come up with that second list. Two sets of people talking past each other. In other words, just another day in the life of BTF.

Well, at least nobody's brought up steroids....

Steve: You can cloak your point in condescension and "the scientific method" all you want, but it seems to me that you have an unscientific and merely asserted bias that is not preferable "scientifically" to Andy's.

When Andy says he weights the end of the season more than the beginning, that is not to abandon concerns about "sample size." Instead it is to acknowledge, I assume, the intuition that given the same goal (end of season success) all the teams have configured themselves to peak (to be their best) at the right time (the end of the season). Cruising through April and May in MLB and collapsing in August doesn't make you the best team; positioning yourself to win when everyone else agrees the winning counts most does. There are many examples of this: it happens all the time in the NBA and Larry Brown in particular is famous for it. MLB teams often will sign a guy they know can't help them until late in the season, precisely b/c that's when they know they will need the boost to make their run. It's a bit like running a marathon and reserving your strength for the last push when the pack thins out. Obviously this is a strategy embraced by Billy Beane as well.

Finally, I want to make a point about your rhetoric. You've been awfully and frustratingly condescending throughout this thread. Aside from your decision to ignore Backlasher, which I find inappropriate, your tone is dismissive and imperious. Buzzard's Bay fails to provide a serious argument? Yet another poster not worth your time? (there's no serious issue about the concern of "leaping" from the field to the spread sheet? The scientists I know acknowledge the gap between events and our observation of them.) Andy "either understands" or he doesn't. And what must he understand? YOUR particular construal of sample size issues. Andy's been extraordinarily civil throughout this, so much so that I'm sure he will deny any problems, but compare the tone in his posts to your own.

You want people to read your stuff, obviously. I assume you want them to see things as you do. And yet you respond to them by ignoring them (BL, Buzzard's Bay) or condescending.

I need to go back to grammar school.

And I'm pretty sure Andy agrees a 162 game sample is better than a 7 or 12 game sample. The problem is that the 162 game sample is not a sample of the end of year team.

I agree with that completely. The huge issue that remains, however, is that a 7 or 12 game sample isn't just slightly less preferable to a 162 game sample; it's enormously inferior in terms of the validity of the information it provides, the confidence in which we're able to place in the conclusions we draw from it.

But I still question the whole idea of "averaging" a team's performance over the regular season, rather than doing that, and then also looking at other factors in order to evaluate the whole. And among these I would include both the postseason and the latter half of the regular season, especially if it carried over into the postseason, as was the case in 1914. Call it bonus points for rising to the occasion, and for proving (to me, anyway) that the last part of the season was more representative of the team's strength than the first half.

I agree with this completely. As I've said repeatedly, the issue isn't one of ignoring post-season performance, or partial-season performance, or anything else. We should eagerly make use of every scrap of information we can get.

The issue is in properly assessing and weighting the information. And I get very uneasy when I hear you say things like your "presumption is the team that won the WS was the superior team." That to me raises a huge red flag of reading far too much significance into a series of games that is, at most, just 7 games long. I think a far more appropriate posture is to be skeptical, to begin with a presumption that neither team is superior, and to not be moved away from that position until the evidence is reasonably clear. The maximum of 7 games themselves have a very, very difficult time meeting that basic hurdle.

JC,

I appreciate your constructive criticism.

You my find my decision to ignore Backlasher inappropriate. I don't; I find it was one of the soundest decisions I've ever made, and my only regret was being as slow as I was to coming to the realization that it was the only good choice.

You my find my characterization of Buzzard's Bay as not presenting a serious argument as condescending. Perhaps it is, and I wish I had more subtle means of addressing it. But a series of posts that fail to include complete sentences, let alone complete paragraphs, and present instead some kind of a disjointed sequence of "get your head out of a spreadsheet and watch a game" cliches is, it must be said, a failure to present a serious argument.

I want people to read my stuff, as you say. However, I certainly have no desire to pretend to take seriously every poster just because I hope it might make them read my stuff. If that makes them not want to read my stuff, well then so be it. The world is a big place, I'm sure we'll do fine without each other.

Steve,

We certainly have our theoretical differences spelled out. Now why don't you list your 28 teams, so we can see how these differences play out in the specifics? Here's mine, from the other thread, just so you can compare without going back and forth:

Andy Posted: January 03, 2006 at 08:36 PM (#1805228)

I left off the 19th Century teams because I don't know squat about 19th Century baseball, and I left off the 1902 Pirates because of the AL raids that year which caused the NL to be historically weak. Possibly an injustice.

And the early deadball era may be overrepresented, but anyway....

(Non-Series winners are italicized, and my top 10 are in boldface)

1905 Giants
1906 Cubs
1909 Pirates
1910 A's (much, much better than the 1911 team---look it up.)
1912 Red Sox

1922 Giants

1927 Yanks
1931 Cardinals
1931 A's

1936 Yanks
1939 Yanks
1942 Cardinals

1948 Indians
1953 Yanks
1953 Dodgers
1955 Dodgers

1961 Yanks

1968 Tigers

1970 Orioles
1974 A's
1975 Reds

1984 Tigers
1986 Mets

1995 Braves
1995 Indians
1998 Yanks (The best of the bunch)
2001 Mariners
2001 A's
2004 Red Sox

You can see my bias against non-Series winners, but nevertheless I did stick 6 of them in there, including 3 of my top 10, and including both the Mariners (top 10)and the A's (top 28) from 2001. And I did include one wild card winner, but the 2004 Red Sox were a powerful team which had one of the great postseason runs ever against three excellent teams.

The 1910's and early 1920's are lacking, probably due to the relative balance in the game for that brief period, especially in the NL. The 1922 Giants were likely the best team in that era.

One last note: I put both the 1936 and 1939 Yankees on the list, since they were both so overwhelmingly dominant, but in a sort of reverse affirmative action, I only took the 1939 team for my top 10, even though the 1936 team almost certainly belongs there as well.

And I omitted the 1969 Orioles, since I had the 1970 team in there, and since the 1969 team blew the Series. If they'd won that year as well, I would have put them both in there, like the 1936 and 1939 Yankees.

The maximum of 7 games themselves have a very, very difficult time meeting that basic hurdle

Another obvious issue here is the match format vs. the round-robin. All kinds of strange things have happened in season series (match play, effectively) over the years. The 1927 Yankees went just 12-10 against the 1927 Indians, who featured Garland Buckeye and Homer Summa. The '86 Mets lost their season series to the Phillies (as I am very fond of remembering), and in an expanded-playoff format, the Phillies would have had a real chance of winning the pennant that year (the Astros damn near won the pennant anyway). Then again, I remember 1983, when the Dodgers absolutely owned Philadelphia in the season series (11-1) but went down meekly in a four-game LCS.

A round robin where all play all means that an outmatched or evenly-matched team can finish first, sometimes by quite a margin (the 1984 Tigers, like the 1954 Indians, played indifferently against their closest competition). But the great thing about the round robin is that everyone has the same schedule (well, used to) and the conditions, though shifting, offer a diversified test for a ballclub.

then you aren't interested in the discussion that I am.

What discussion are you interested in?

Since you aren't going to convince me there is no luck, I'm simply letting you know that you don't have to waste your time attempting to explain "closed event states with randomness" aren't luck. I'm doing you a favor.

That's very kind of you Chris. But I'm not trying to convince you of anything. I have not opined on who the best team may be. I have not opined about the probability of an event recurring based on available evidence. I've gotten a chuckle out of McCormick as an exemplar, but that is the closest thing to an opinion that I've uttered.

You are free to define whatever types of events you want as luck. If I can understand your definition, I can engage you in conversation on larger outcomes. But contrary to the rhetoric of some, I'm pretty sure that others are including events that you would not call luck, and misstating some mathematical principals in the process.

You either understand the importance of sample size or you don't

Goodness, could this be a true statement. Let's see. During the regular season how often do we see events that have these properties:

(1) Include at least one day of rest for at most every three games played. With an average amount of rest much larger.
(2) Allow a manager to set his roster so the major performers have even longer recoupment time.
(3) Facing competition that is among the best twenty percent of the league.
(4) In an environment with expanded media attention.
(5) In an environment where its unlikely you will see the lower performers on the other roster.
(6) In an environment where any mistake is critical to the ultimate goal.

I doubt you find may events that even consume that subset. The only real exception to where you may have a large number of those events present is probably a team that plays in New York City; maybe a team that has a small advantage because they can eliminate some persons who are predisposed not to be able to perform in those conditions.

Nevertheless, I find this illuminating. As I've mentioned if such a thing as "true talent" where discernable its not static. And true talent against Juan Cruz isn't quite the same as true talent against a peak Randy Johnson. You put RJ back in his 30s, give him rest and no injury, everyone isn't going to put up the same OPS they did over the magical 162 game season.

If this "sample size" is focused on samples of the mythical "true talent" the postseason's limited sample of this "true talent" doesn't hurt you in any way whatsover.

Because output is still going to be a function of player, opponent, and environment. And yes, you can very much determine some characteristics of each from the regular season event space. But you don't have that holy grail of a function that shows you how to impart these three things together when each are known independently. In fact, the only "sample" you have of this functions characteristics is the postseason event space. If RJ shuts you down, then you can't just throw that off to some mysterious event.

And we aren't talking pebbles or flintstones. Every single event in the postseason has more significance, not just a pebble. I personally don't recall many pebble events. But if pebbles are equipotential, then they are just as likely to occur in the regular season at the same frequency over any seven games as they are in the postseason. If pebbles are more likely to occur in the postseason, then there must be something about that environment that makes the event more probable. In which case it is part of the test. Again, I don't remember many pebbles. I remember Glavine shutting down the Indians. I remember Giambi and Byrnes failing to run the bases. I remember Hudson getting in a bar fight. I remember Lonnie Smith not running home. I remember Bill Buckner not fielding ground balls. I remember Justice being injured. I remember Wohlers hanging sliders.

You want to tell me that the difference between two teams is a pebble on the infield, there may be cases that I won't argue with you.

You want to tell me that during a 162 game season, the 2001 Mariners will still have the best record of all the 2001 teams. That is probably supportable.

You want to tell me that the Mariners loss to the Yankees was some flukey event. Well talk to me, but it looks like the Yanks exploited their weaknesses.

You want to specifically compare the 1991 Braves to the 1991 Pirates, I'm game. At the time, I thought the Pirates were more likely to win that series. If you go back, you might could still make a case that those Pirates were more likely to win a 7 game series against those Braves, but time doesn't change the fact that Van Slyke and Bonds wore a big choke collar. Anyway, the specific circumstances are a lot more interesting than just saying the "Braves got lucky." And just looking at Pythagorean stuff is discounting the real difference made by Alejandro Pena.

Steve again,

Just to clear up one point, when I wrote that my "presumption is the team that won the WS was the superior team," I cetainly didn't mean that to be my only (or final) thought on the matter. It's more of a way of making sure that I don't offhandedly dismiss a World's Champion with a generality before weighing the rest of the evidence, as others might. Again, you're not going to find any 1906 White Sox or 1987 Twins teams, etc., on my list, whereas you will find the 2001 A's---please do make a note of that.

Where I differ from you is our starting point, but what really matters is where we wind up, and our reasoning behind each and every choice, considered on their specific merits and accomplishments. Sorry to go all boldface on you, but that's what lies at the heart of my "method," whatever you or I might suppose that method to be.

the 1984 Tigers, like the 1954 Indians, played indifferently against their closest competition

Ahem. That may have been true for the regular season, but...

1954 Indians, World Series: 0-4
1984 Tigers, ALCS: 3-0
1984 Tigers, World Series: 4-1, knocking out the San Diego starting pitcher by the third inning (and twice in the first inning) in each of the last four games. Even the John Birch Society disowned the Padres after that debacle.

But a series of posts that fail to include complete sentences, let alone complete paragraphs

I think that's unfair to Buzzards Bay. The style might be different than with most posters, but you can just replace "...." with commas and line changes if you feel the need. I certainly have no problem understanding his posts, and, the points he puts forward in his last post are pretty well explained IMO, whether I agree or not. But I'm biased, I actually like his style, looks poetic, breaks the monotony of a long thread.

Gracias. You get what I've been trying to say.

Thank God. I was a little afraid of screwing it up and have you wake up and ask yourself "who's that idiot putting words in my mouth?"...

Now why don't you list your 28 teams, so we can see how these differences play out in the specifics?

Andy, my friend. How many times have I said this? I guess I'll say it once more: I don't know of any means of specifying a precise small number of teams, such as 10 or such as 28, that is the "all time best" with any degree of confidence. As I said before, anything beyond naming around 50 that are probably actually within the "real" top 100 is about as accurate as I think anyone could possibly get.

I don't have a list of 28 teams. If you want to put one together, have a blast.

I don't have a list of 28 teams. If you want to put one together, have a blast.

Actually I just did, and if you ever care to look, in #837 in this good old thread right here, same page as this.

As for the rest of what you wrote just now, you make a plausible point but it ain't much fun, and to argue with it is sort of like punching the tar baby. And to be honest, I'd find even your purposely hedged and fuzzy list far more interesting than what some computer simulated run might cough up. For one thing, your biases, whatever I might think of some of them, are stated up front. Will the computer's biases be equally transparent? IOW, who's programming that computer, and with what sort of input and instructions? Will it be a Bill James or a Katherine Harris behind the curtain?

You my find my decision to ignore Backlasher inappropriate. I don't; I find it was one of the soundest decisions I've ever made, and my only regret was being as slow as I was to coming to the realization that it was the only good choice.


LOL. You want to talk about a serious misrepresentation. I debuted the handle of backlasher. It was less than two weeks before Steve decided to "ignore" me. I don't even recall him making more than a couple of posts that were responsive to any post I had. One was to say, "I was pulling everybody's leg." Another was to say that he was ignoring me. Since that time, he has responded to a couple of posts saying he was "ignoring me." I don't recall him ever responding to any argument whatsoever.

And I think its pretty clear to everyone, he doesn't ignore my posts at all. Its pretty clear that he reads them, and every so often he will include a snarky remark, usually that misrepresents my position. Meanwhile others will claim I misrepresent his position, when clearly he thinks non-median performance is "luck".

I can say that Srul and Werr have convinced me it is better to ignore some posts. Those by the student crowd that deal only with the poster and not the post itself really aren't worth responding too. People that still don't understand your position after multiple posts, or that claim you say something diametrically opposite to what you have expressly stated are also not worthwhile. But meritorious points are always worth a response regardless of where they eminate.

There has been only one poster that has been condescending in this thread. As near as I can tell, there are only three people who post here that think they are smart or special enough that they don't have to respond to points. Only one of those three has contributed anything close to enough to even venture toward that conclusion.

There has only been one person in this thread who has assaulted other persons' intelligence. It hasn't been me.

Will it be a Bill James or a Katherine Harris behind the curtain?

LOL. If it is Ms. Harris pulling the strings, I'm sure the winner will be the Texas Rangers.

You want to talk about a serious misrepresentation. I debuted the handle of backlasher. It was less than two weeks before Steve decided to "ignore" me. I don't even recall him making more than a couple of posts that were responsive to any post I had. One was to say, "I was pulling everybody's leg." Another was to say that he was ignoring me. Since that time, he has responded to a couple of posts saying he was "ignoring me." I don't recall him ever responding to any argument whatsoever.

This is, of course, a stark, bald-faced lie, only the latest in a very long series of direct and knowing lies that Backlasher has posted for years regarding what I have written. He never issued a single word in response to any of the times I called him on the lies, and therefore I finally gave up on it, and gave up on him.

That, vastly more than the rude and taunting tone, is why I choose not to engage with him. He tells lies.

BL,

I'm sure you know by now that no one else is interested in the subjects you wish to discuss. So you may want to rest for a while.

I am interested however but I have difficulty understanding you. But try I do.

The topic of what is causal is fascinating. And so its frustrating that people don't understand your central point: some or all of the events that people think are caused by "luck" are actually caused by something else. In understand that you are interested in determining what causes these events.

It's a valid topic.

It's laughable that some posters here don't want to include postseason in the population of games for analysis. BL's post 39 (above) is an excellent summary of reasons that the postseason is a different type of test. Note the word "different." It is not the only test. It is another test of a team.

I don't mind that one would weight postseason differently than I. But to exclude it altogether is ludicrous.

That, vastly more than the rude and taunting tone, is why I choose not to engage with him. He tells lies.

Gentlemen, please!

If you wouldn't say it to the person's face, don't say it at all.

Does anyone know of a thread where there is a debate on Pythagorean records? I'm a skeptic about it but I want to know more before I come to a conclusion.

There has been only one poster that has been condescending in this thread.
There has only been one person in this thread who has assaulted other persons' intelligence. It hasn't been me.

Let's see, from this thread...

"I'm not sure how much weight I'd give their opinion. I'd give it more than your run of the mill fanboy, or faux baseball historian"

"I think I know now. I'm surprised I didn't see this before. Its because you are a FAN of a team that has virtually no postseason success."

"The problem is your throwing out terms you have heard others say. You like how they sound so you adopt them."

"There is malaise in the Bay so we have to blame it on something supernatural."

"I'll let you argue with Andy and Kevin about McCormick was as good as Koufax."

"You have one poster that is asserting:

(2) Injuries are random.
(3) Giving some teams credit for the best possible performance that could have occurred in the past, and sticking other teams with the most disastarous outcomes of the past"

again... "to argue that McCormick was just a good a pitcher as Koufax"

"This isn't Strat-O-Matic. If someone learned about baseball by playing that game, then I could understand those gaps in thought"

"The "luck" thing will disintegrate just as quickly as "ACE RELIEVER" "


Yeah, I could see someone deciding to ignore the posts instead of going through that bullcrap.

Andy, I think we're in agreement. Your 831 sets it out right. I also agree with BDC's posts. I've thought about trying to come up with a list, but there are some ground rules I need clarified: is there an era adjustment? how about league (an issue you noted in your list)? what about playing time?

Are we getting graded on our answers?

That last half page of the Best All Time team thread comes up to 34 8.5 X 11 pages.

Maybe I can turn that one into a graphic novel instead of PETCO: THe Graphic Novel.

That was for the Lounge. Sorry.

I'm pretty sure that others are including events that you would not call luck,

That's fine, but you are assigning me to that "minority position", when it clearly is not my position.

The cheese stands alone.

If it is Ms. Harris pulling the strings, I'm sure the winner will be the Texas Rangers

Awesome. Let me put in a vote to include the 1999 Rangers. Heck, let me put in 5,000 votes.

I prefer the 1911 A's to the 1910 A's chiefly because they led the league in runs scored and in fewest runs allowed. The 1910 team led in defense, but they were only runners-up in scoring. The 1911 team not only led in both departments, the 861 runs they put up was the biggest offensive breakout between the 1890s and Babe Ruth, and by a pretty hefty margin. Whereas the stingy ERA of the 1910 team is sort of in a bin with the Cubs teams of the decade, that 1911 batting attack towers over the two deadball decades. That to me outwieghs the 1 real win (and the 4 expected wins) the by which the 1910 team leads them.
I think how the 1910-1911 A's responded to the livelier ball says something good about the adaptive quality of that team. They won a title with a great pitching-and-defense performance. Then, they changed the ball on them, and the A's sort of said, "no problem, we'll hit the snot out of this new ball and still have the best pitching in the league." It wasn't just that runs were up all over; Philadelphia capitalized on the change better than anyone else. That's one of the reasons I think they were the best team of the various mini-dynasties of the teens, and the most interesting one to put in the simulation.

I think how the 1910-1911 A's responded to the livelier ball says something good about the adaptive quality of that team. They won a title with a great pitching-and-defense performance. Then, they changed the ball on them, and the A's sort of said, "no problem, we'll hit the snot out of this new ball and still have the best pitching in the league." It wasn't just that runs were up all over; Philadelphia capitalized on the change better than anyone else. That's one of the reasons I think they were the best team of the various mini-dynasties of the teens, and the most interesting one to put in the simulation.

That's an extremely insightful and interesting observation.

I agree that the 1911 A's did a great job of adapting to the cushioned cork baseball, as evidenced by the equal ease with which they won the pennant and World Series, again against a very good NL team.

But unless ERA+ and OPS+ are seriously flawed measures, they seem to indicate that in fact the 1910 team was quite a bit better. The numbers:

1910 OPS+ 122
1911 OPS+ 119

1910 ERA+ 133
1911 ERA+ 104

Since I wrote my time travel Series story under the assumption that the 1911 team was the best of that 1910-1914 lot, I wish that these numbers were misleading. But they are what they are. And since all the traditional ways of measuring team strength (wins, GA, World Series results) don't favor either team particularly, we're left with that humongous 29 point ERA+ gap to explain.

Perhaps that 1911 ERA+ figure was miscalculated. If so, it should be corrected. But if not, no way the 1911 team was better.

The 1910 team led in defense, but they were only runners-up in scoring. The 1911 team not only led in both departments, the 861 runs they put up was the biggest offensive breakout between the 1890s and Babe Ruth, and by a pretty hefty margin.

That's one way of looking at it. but a better way, I think, is to look at run differenials. Barely winning in both categories isn't as good as crushing the competition in one or the other.

Let's look at the two years, the A's and their closest rivals (runs scored/runs allowed):


1910 1911
A's 1.53 1.43

Rival 1.17 1.15

</pre>

Hmmm. It seems they were more dominant in 1910.

I agree that is a great team. But they might have been better adapted for deadball play than the lively ball.

Incidentally, I checked 1912 as well. The Red Sox check in at 1.47. They didn't lead the league in either but came in second both times. The Senators were next at 1.20 and the A's at 1.18. So that Red Sox team, while quite comparable to the A's, was facing stiffer competition and deserves to be stand right beside them.

The 1912 Giants checked in at 1.44, their closest rival 1.33. So, the Giants weren't all that better than the Pirates, though they smoked them by 10 games. If you go by the pythag records, they only win by 5.

I'm still flabbergasted that Neyer could pick the 1912 Giants but overlook the Red Sox, who were better in every way, including head-to-head in the World Series.

we're left with that humongous 29 point ERA+ gap to explain.

Perhaps that 1911 ERA+ figure was miscalculated.

One thing to bear in mind when looking at ERA and ERA+ figures from early eras such as this one is that there were a lot more errors then than in modern times. So ERA or ERA+ aren't necessarily as comprehensive as indicators of a team's run-prevention ability than they later became.

For instance, the 1911 A's had a 3.01 team ERA, good for third in the league, and a distant third at that, behind the Red Sox's 2.74. But the 1911 A's were first in the league in fewest runs allowed, at 3.95, far ahead of the Red Sox's total of 4.20.

ERA and ERA+, as measures of team run prevention and even of individual pitcher run prevention, don't tell nearly the story then that they would later.

Thanks for the kind remark, Steve (#859).

Andy, you may be right (#860) that the 1910 team was better, but personally I would not put that definitive "quite a bit" construction on it. For one thing, we're talking about two teams with most of the same players separated by a single year and a single real-world win. So I can't see either edition of this club as obviously superior to the other.

I'm not sure what to make of the OPS+ numbers you cite. I don't think you need to look at it through the filter of OPS when you can see how many runs they scored. That the 1910 team that finished 2nd in the league in runs scored could rate a better hitting club than a team that led the league, and the surrounding 20 period, in runs scored, by a wide margin, strikes me as dubious. Anyway, I'll think about it.

Another small reason why I'd pick the 1911 team is McGinnis istead of Davis at 1b. I like to have as many of the key members of the dynasty as possible on the representative squad. Greatness comes in different shapes, but I'm more of a dynasty guy than a single-season guy. This figures in my preference for the '95 Braves over the '98 Braves too.

If OPS+ is a flawed metric, then it's not just an opinion about the 1911 A's that may have to be changed.

Another small reason why I'd pick the 1911 team is McGinnis istead of Davis at 1b.

And now that I think about it, that's why I picked the 1911 team for my story, so that I could include the entire (inflation adjusted) "2 Million Dollar Infield."

(Hmmmm, that figure doesn't sound quite as impressive as the original "$100,000 Infield" did back in 1911. It probably wouldn't even get you the Marlins' infield these days.)

If OPS+ is a flawed metric

Every metric is flawed; it's just a question of recognizing the particular flaw(s) and how it impacts what we're trying to measure.

OPS is a very effective quick & dirty metric, because at the team level it pretty closely approximates runs scored. But it has some weaknesses, and in this particular case some of the things OPS doesn't count may be important: it doesn't count basestealing or baserunning, and it doesn't times reached on error or bases gained on error.

In the modern game, without much basestealing and very few errors, this probably doesn't matter much. But in the baseball of 1910-11, which was rampant with stealing, taking extra bases, and fielding and throwing errors, these factors might have a large impact on a team's relative ability to score. Taking extra bases (by steal or just within plays), and forcing errors, were very important skills then. It's entirely possible that because of this, OPS and thus OPS+ may over- or under-rate the 1910-11 A's to a degree it wouldn't a more modern team.

who has assaulted other persons' intelligence
Let's see

LOL. One example of a post type that I know ignore in terms of response is the above. The one where people tend to follow you around and only want to comment on the personalities involved rather than the argument. I would like to point however that the quotes cited for evidence do not assault a persons intelligence, cognitive ability, or ability to understand. And with the way that some terms have been abused, or the way that some people have dissected two sentences, I could have have gotten a lot of mileage out of those types of statements. Moreover, for the Werr's, Daly's and other counters, please note, I stayed out of the metaness of this for a long time. You have two students that have been pressing that issue.

Yeah, I could see someone deciding to ignore the posts instead of going through that bullcrap.


Then by all means ignore me. I'm not going to lose any sleep on whether or not the Yankee Blower is going to read my posts.

It's laughable that some posters here don't want to include postseason in the population of games for analysis.

Careful SS, you can't say "laughable" any more under the new civility guidelines. Thanks in part to your advocacy, I've been trying to remove that from my arsenal.

I hope you at least now see some of the things I mentioned. Its now, "assaulting intelligence" to demolish another argument. As you can see, there is no endgame to this. Give an inch, and bam, its "you don't have to post on this anymore." As you can see, it won't stop. I may as well come full force.

This is, of course, a stark, bald-faced lie

LOL. So which is it Steve are you ignoring me, or are you really just pretending to ignore me because you think it gives you some cred?

But let's examine your contention

Backlasher made his first post on April 6, 2004. Treder is right, he did respond to about 1 in 25 of my posts in that thread and let us know about the problems with my "anonymous post" I saw him respond to one other, there was no questions whatsoever. I do remember within that week, he claimed I didn't respond when there was 12 hours between post times. Goodness Gracious. I'd love to see the alleged posts I didn't respond to.

Of course, his basis for ignoring people at that time (remember, I'm not the first) was:

My experience, on this site and others, is that sensible ideas, rationally presented, will be heard and acknowledged. Rants and polemics won't, and those (like you know who) who have no interest in sincere dialogue at all will be ignored.

Now I can't find the thread where he proclaimed to the world he was going to ignore me. I'm betting it wasn't even a month, but we know that the longest it could have been was four months because we do have this:

August 10, 2004 at 01:08 PM (#787984)
Well, Backlasher, there are several interesting and well-taken points in your posts in this thread. You had me thinking that I would grant you the courtesy of a reply.

But then you had to throw in the predictable insults in post #17, reminding me why I decided awhile ago that I have many more interesting and enjoyable ways to waste my time than engaging in dialogue with you.

and here was the alleged insult, "All Steve has done, and continues do is make posts that "if it was good enough for Mike Marshall, its good enough for Aaron Sele." I don't find anything about this to be "irrefutable" "undeniable" or for that matter "convincing" and in some cases "honest". "

So which was it Steve. I doubt very many people think I was going to run away. Because almost 18 months ago you said it was for an insult, which comprises someone pointing out things are not irrefutable. But show us those posts that I ran away from.


And in conclusion, here are some words from a poster driven away from the site:

Treder doesn't respond because he is used to having his style hide the content - or lack of it. He considers it rude to address the content directly and ignore the way he has dressed it up to look pretty.

And in conclusion, here are some words from a poster driven away from the site:

Who would that be?

LOL. So which is it Steve are you ignoring me, or are you really just pretending to ignore me because you think it gives you some cred?

I think there's a chance he really was ignoring you when you started claiming that he said that McCormick and Koufax were equivalent. Other people have quoted your assertions, and that would certainly bring them to his attention even if he skipped every message that said "Backlasher" at the top of it.

I prefer the 1911 A's to the 1910 A's chiefly because they led the league in runs scored and in fewest runs allowed. The 1910 team led in defense, but they were only runners-up in scoring.

Fair enough, but that 1910 team was only 6 runs behind club with a park factor of 107, while the A's played in a very moderate pitchers park.

I don't have a calcualater handy - which by the math is more impressive - scoring 4.34 runs per game in a league that averages 3.64 in a park with a factor of 99, or scoring 5.66 in a league which averages 4.60 with a park factor of 98.

we're left with that humongous 29 point ERA+ gap to explain.

Perhaps that 1911 ERA+ figure was miscalculated.

Anyone checked their unearned runs? That's often key with deadball teams.

Kevin,<