Any ballot that doesn't have each of the following teams should be thrown out
[...]
1995 Indians
2001 Mariners
[...]


Although I have been on the postseasons-are-too-small-to-figure side of the debate in the other thread, I still instinctively feel that any club that was all that great shouldn't have lost a postseason series in six games. (Or at all, really.) In fact, a major reason I included the 1995 Braves was that they knocked off the Indians. That's also an element of my fondness for the 1954 Giants.

My Cleveland representative is 1920 over 1948, though that's idiosyncrasy, I admit. The 1920 Indians are an impressive bunch, but one always has to ask whether the White Sox were trying to win that year ...

The 1920 Indians are an impressive bunch, but one always has to ask whether the White Sox were trying to win that year ...

I voted for the 1919 Reds - who had a great regular season. I don't know that they should be punished for the lack of competition they got in the Series.

This question keeps coming up, but here's how Diamond Mind handles era adjustments

I know how they handle it. I don't think it's going to work well for this, because it doesn't really account for quality of competition (AFAICT). League averages aren't quality of competition barometers.

I would vote for the 89 or 90 A's than than the 88 one. they has " Rickeeey""

League averages aren't quality of competition barometers.

They sure aren't. Quality of competition is, of course, one of the very toughest nuts to figure. No one knows exactly how much it's changed over time, or to what degree its changes have been steady or jumpy, or to what degree there've been regressions in quality, or any of it. It's entirely theoretical, though just about everyone agrees it's real.

It's an enormous issue that kind of rains on a fun parade like this one.

A little late answering questions from page one:

I did indeed mean the 1906 Cubs.
I put the 1995 Braves on my ballot because they were the '90s Braves team on the list. I think the '97 team may have been the best, or perhaps '98 or '99.

1995 Indians
1998 Yankees
2001 Mariners

Regarding the rest, I abstain.

Quality of competition is, of course, one of the very toughest nuts to figure. No one knows exactly how much it's changed over time, or to what degree its changes have been steady or jumpy, or to what degree there've been regressions in quality, or any of it. It's entirely theoretical, though just about everyone agrees it's real.

IIRC, Palmer and Thorn tried to address this issue in The Hidden Game of Baseball. I don't remember the specifics, but I'm almost sure that they claimed that the weakest teams of today could easily beat the champions of the deadball era. Others may remember more about this book, and what the authors did or didn't say about this issue.

I don't remember the specifics, but I'm almost sure that they claimed that the weakest teams of today could easily beat the champions of the deadball era.

They may well have; it's hardly an unreasonable claim. It's impossible to prove, of course, and the harder issue for a mechanism such as Diamond Mind is to quantify it precisely. It's one thing to make the vague claim that the weakest teams today are better than the best teams of a century ago, it's quite another to get specific about it.

I strongly doubt the weakest teams of today could beat the best teams of the deadball era. I actually think it's ridiculous to be honest.

The best players of that era were every bit as good as the best players of today. The issue is with the lesser players, the bench players today are probably better, but much less so than most would think, I believe. Baseball isn't a sport like football or basketball where athleticism is everything. Hand eye coordination and timing are much more important in baseball, and I imagine that this doesn't improve at nearly the same rate as strength and running speed . . .

Anyway, here's my ballot!

1902 Pirates
1906 Cubs
1911 A’s
1912 Red Sox
1917 White Sox
1923 Giants
1927 Yanks
1931 A’s
1939 Yanks
1942 Cards
1946 Red Sox
1953 Yanks
1954 Indians
1955 Dodgers
1957 Braves
1961 Yanks
1969 Orioles - HOW COULD THEY NOT BE ON THE BALLOT? They are easily one of the 10 best teams ever, if they aren't in the tourney, the whole thing is worthless in my opinion. Something needs to be done there . . .
1975 Reds
1984 Tigers
1985 Cardinals
1986 Mets
1988 A’s
1992 Blue Jays
1994 Expos
1995 Indians
1995 Braves
1998 Yanks
2001 Mariners

1896 Orioles
1906 Cubs
1909 Pirates
1911 A’s
1912 Red Sox
1927 Yanks
1929 A's
1931 Cards
1932 Yanks
1939 Yanks
1942 Cards
1942 Dodgers
1946 Red Sox
1953 Yanks
1954 Indians
1955 Dodgers
1957 Braves
1961 Yankees
1970 Orioles
1975 Reds
1984 Tigers
1986 Mets
1995 Indians
1995 Braves
1998 Yanks

Am I a complete moron for thinking one of the 1930's Cubs teams should at least be discussed? They went to the Series in '29, '32, '35, and '38. The '35 team won 100 games (and 21 in a row). Just say'in...

I noticed Flynn voted for the 1970 O's. I think the votes of anyone who voted for the 1969 or 70 O's should be pooled and the one with the most votes be the rep (if they qualify). I imagine if they weren't on the ballot, a bunch of 1991-2005 Braves teams would split the vote, for example.

But Flynn the 1969 squad was better, no matter what a 7 game series in October says . . . :-)

My Ballot would be

1923 Yankees
1927 Yankees
1928 Yankees
1932 Yankees
1936 Yankees
1937 Yankees
1938 Yankees
1939 Yankees
1941 Yankees
1943 Yankees
1947 Yankees
1949 Yankees
1950 Yankees
1951 Yankees
1952 Yankees
1953 Yankees
1956 Yankees
1958 Yankees
1961 Yankees
1962 Yankees
1977 Yankees
1978 Yankees
1996 Yankees
1998 Yankees
1999 Yankees
2000 Yankees

and to bring it to an even 28

1904 Highlanders
1985 Yankees (Donnie Baseball gets the nod)


This is supposed to be "Count the Ringz" isn't it? As Sandy Koufax when he was growing up, it didn't seem like a real World Series without the Yankees in it.

I noticed Flynn voted for the 1970 O's. I think the votes of anyone who voted for the 1969 or 70 O's should be pooled and the one with the most votes be the rep (if they qualify).

Joe, that's my plan for when I'm tallying the votes since I assume some might think the 1907 Cubs or 1969 Orioles were better than the 1906 or 1970 version of their teams, they would still prefer to see a team from the Dynasty represented than have their vote go entirely for naught

Nice ballot, Schuey.

I actually hadn't noticed till now that the Yankees are quietly putting together their third-longest ringz drought since 1923 :)

I hadn't noticed that either. ;

I imagine that the drought has turned a lot of NYC residents into Red Sox fans.

Wow, you guys sure aren't very objective.

Someone make my list for me:

OPS+ plus ERA+ divided by 2.

The top (insert # of teams here). I don't care how many games they won, whether they finished first second or third, or when they played.

At least I think I think that. The list may tell me that's the wrong way to do it without some tweaking. Possibly just "above criteria applied to the top two teams of any season"..

But the first way first.

I don't claim any guarantee regarding accuracy.

These are the teams on the ballot plus a few I added.

1906 Cubs131
1939 Yanks129.5
1927 Yanks128.5
1948 Indians123
1969 Orioles122
1942 Cards122
1909 Pirates122
1902 Pirates121.5
1954 Indians120.5
1905 Giants120.5
1912 Giants119.5
2001 Mariners119
1995 Indians119
1931 A’s119
1896 Orioles119
1942 Yankees118.5
1912 Red Sox117.5
1953 Yankees117
1998 Yankees116.5
1897 Beaneaters116.5
1986 Mets115
1970 Orioles115
1932 Yanks115
1917 White Sox114
2004 Red Sox113.5
1978 Dodgers113.5
1977 Dodgers113.5
1955 Dodgers113.5
1921 Yankees113.5
1903 Red Sox113.5
2002 Angels113
1984 Tigers113
1925 Pirates113
1968 Tigers112.5
1961 Yanks 112.5
1975 Reds112
1940 Reds112
1947 Yanks 111.5
1946 Red Sox111.5
1934 Tigers111.5
1911 A’s111.5
1931 Cards111
2001 Dbacks110.5
1957 Braves110.5
1995 Braves110
1988 A’s109.5
2005 White Sox109
1974 A’s109
1967 Cards 108
1923 Giants103.5
</pre>

Well, I guess this list shows that our grandfathers were right. Objectivity is a wonderful thing.

But more seriously, thanks for compiling that list, strong silence. I wanted to do it myself but was of course way too slothful.

ss,
are you saying that's just the ones already on the list?

Because I hope one of these geniuses can use the Lahman db and call up hte best ones.

Eitehr way, thanks a lot. I think it goes a long way to verify the quality of these teams (despite the subsequent complaint that the difference in the quality of teh leagues that's about to crop up).

Yes, CD. Those are only the ones on the ballot plus a few; I haven't used Lahman.

Andy, sloth is my middle name

1906 Cubs
1902 Pirates
1909 Pirates
1927 Yanks
1897 Beaneaters
1998 Yanks
1931 A’s
1939 Yanks
1896 Orioles
1932 Yanks
1912 Red Sox
1905 Giants
1912 Giants
1911 A’s
1970 Orioles
1986 Mets
1975 Reds
1953 Yankees
1917 White Sox
1984 Tigers
1955 Dodgers
1948 Indians
2005 White Sox
2001 Dbacks
1974 A’s
1942 Cards
1995 Braves
1995 Indians

SS,
is that your list, or did you do what I said?

That is my list.

My list:

1986NLNYN
1939ALNYA
1927ALNYA
1970ALBAL
1998ALNYA
1906NLCHN
1902NLPIT
1990ALOAK
1974ALOAK
2001ALSEA
1969ALBAL
1998NLHOU
1975NLCIN
1985NLSLN
1971ALBAL
1937ALNYA
1948ALCLE
1984ALDET
1995ALCLE
1949NLBRO
1978NLLAN
1905ALCHA
1968ALDET
1901ALCHA
1988NLNYN
1999NLARI
1976ALNYA
1988ALOAK

This is based on creating a park-adjusted, small-strength-of-opponent-adjustment z-score for both offense and defense (pitching+fielding) for every team, then summing the z-scores to create a final z-score value. So similar to Neyer/Epstein's method, but looking at offense and defense individually first.

List is ranked in order of highest z-score to lowest, top 28 teams of all time. Disregards all, if any, post-season performance.

Let me try that again:
1986NLNYN
1939ALNYA
1927ALNYA
1970ALBAL
1998ALNYA
1906NLCHN
1902NLPIT
1990ALOAK
1974ALOAK
2001ALSEA
1969ALBAL
1998NLHOU
1975NLCIN
1985NLSLN
1971ALBAL
1937ALNYA
1948ALCLE
1984ALDET
1995ALCLE
1949NLBRO
1978NLLAN
1905ALCHA
1968ALDET
1901ALCHA
1988NLNYN
1999NLARI
1976ALNYA
1988ALOAK
</pre>

Well, crap. Sorry about that, fellas.

OK, I cleaned it up:

1986 Mets
1939 Yankees
1927 Yankees
1970 Orioles
1998 Yankees
1906 Cubs
1902 Pirates
1990 Athletics
1974 Athletics
2001 Mariners
1969 Orioles
1998 Astros
1975 Reds
1985 Cardinals
1971 Orioles
1937 Yankees
1948 Indians
1984 Tigers
1995 Indians
1949 Dodgers
1978 Dodgers
1905 White Sox
1968 Tigers
1901 White Sox
1988 Mets
1999 Diamondbacks
1976 Yankees
1988 Athletics

Note that this doesn't eliminate two of the '69-'71 Orioles squads. I think all three should be considered.

I like Tamer's list - which confirms that hte 1986 Mets were the greatest team of all time.

Note that this doesn't eliminate two of the '69-'71 Orioles squads. I think all three should be considered.

It also contains both '37 & '39 Yanks. 1988 & 1990 A's. 1976 & 1977 Yanks. And the 1986 & '88 Mets.

It also contains both '37 & '39 Yanks. 1988 & 1990 A's. 1976 & 1977 Yanks. And the 1986 & '88 Mets.

And I'm fine with that, going with the definition that "greatest team" is for one season and one season alone. Or are we suppose to, without DMB, pick one of the three Orioles squads as representative of that three-year span?

If it's frowned up to include two (or three) squads from the same franchise in a small time-frame (three or four years, say), then I'll simply update my list, picking the top z-score team from the group.

Eliminating "lesser" teams within four years of "greater" team, my list is:

01. 1986 Mets
02. 1939 Yankees
03. 1927 Yankees
04. 1970 Orioles
05. 1998 Yankees
06. 1906 Cubs
07. 1902 Pirates
08. 1990 Athletics
09. 1974 Athletics
10. 2001 Mariners
11. 1998 Astros
12. 1975 Reds
13. 1985 Cardinals
14. 1948 Indians
15. 1984 Tigers
16. 1995 Indians
17. 1949 Dodgers
18. 1978 Dodgers
19. 1905 White Sox
20. 1968 Tigers
21. 1999 Diamondbacks
22. 1976 Yankees
23. 1955 Dodgers
24. 1911 Athletics
25. 1912 Giants
26. 1998 Braves
27. 1991 Pirates
28. 1903 Pilgrims

By stretching it out to four years, I eliminated the 1901 White Sox and the 1974 Dodgers. Closest great teams of same franchise are 1955 Dodgers and 1949 Dodgers. All of these teams were at least a combined (offense and defense) 2.5 standard deviations better than average.

Tamer, thanks for using an objective measure (so I'll agree with Greg's list). That's two counts for his.

Somebody should tally these up. The deadline should be within a day. People have had enough time to vote.

Is Z-score the Neyer/Epstein standard deviation thing?

Not a big fan of that. I'm not convinced that compact leagues are by definition tougher. Didn't Bill James come up with a few examples of situations where the results were pretty ludicrous, like saying a team that scores 800 runs in a compact league that averages 750 could look better than a team that scores 850 in a spread out league that averages 700 or something?

Dial - I like your method some, but OPS+? OPS is a junk stat - I mean it's better than AVG, but how about using something that accounts for GIDPs and SB and baserunning and such? Or at least (2*O+S)+ if you must. ERA+ is nice and all, but error prone teams are going to be overrated there, no?

I'd much rather see something that just uses Runs and Runs Allowed. These aren't players, we don't need to estimate how many runs they scored or allowed, we already know.

OK, I'm lazy and I respect Dial and Tamer. I'll third their selection. I wanted to just use Neyer and Epstein's top 28 single season teams and also count winning pct as well as runs for and runs against, but never got around to it.

Somebody should tally these up. The deadline should be within a day. People have had enough time to vote.

I should've put this is the opening--my bad. The vote will be open a week, so until midday January 10th, depending on when I wake up. At that point, I'll close the thread but I figure I might as well give everyone a week to vote

I think that a big part of the difference between the teams of today and the teams of yesteryear is the training. I think it is entirely possible that over the course of a season the 1903 Red Sox wouldn't be as good as the 2003 Tigers if the REd Sox are the size, speed, etc. of the Tigers. Doesn't mean they had less baseball ability just that they were smaller, did no weight training, drank more, etc.

And even if the teams of today are that much better this whole exercise would be really really boring.

I like your method some, but OPS+? OPS is a junk stat - I mean it's better than AVG,

Joe,
are your picks in there? What did you use - team wins and losses?

Gimme a break.

First, I should mention that I've only calculate the z-scores from 1901 to 2004.

Second, regarding Joe's comments: Yes, Neyer and Epstein used z-scores in their rankings, as they compared the RS-RA difference of each team to the average RS-RA difference in the league. Instead of RS-RA, I calculated the z-score for offense (RS) and defense (RA) after adjusting for park, etc.
So the z-score would be team's adjusted RS minus the average of all the league's adjusted RS divided by the standard deviation of all the league's adjusted RS.

It seems to work fairly well. And yes, if a league average 750 RS, and almost all of the teams are close to 750 RS, yet one team has 800 RS, that team would be more standard deviations away from the average than a team that scores 850 RS in a 700 RS league average should there be a wider variance in the RS. But for that to happen, there would have to be teams scoring well below 700 RS as well for the average to be 700 RS.

It's not perfect (especially since z-scores assume a normal distribution (right?), which isn't always the case for a league), but it is better than simply comparing adjusted RS to the league average adjusted RS.

I'm still wanting to investigate this method further.

I hate not being able to edit my posts after the fact in threads I didn't create. Or perhaps I should pay more attention to Live Preview.

I'm not convinced that compact leagues are by definition tougher.

You can still be in a non-compact league and have a good z-score. In my opinion, using a z-score is better than dividing by the average or taking the difference from the average. There's probably an even better (and most likely more complicated) way to examine this, but I haven't starting looking for it yet.

Here is a ranking, based on runs and runs allowed, of the teams listed in posts 119 and 129. The Ratio column is based on the computation ((R - RA) / (R per team)) / 2. The first quotient in this computation shows how great a team's run differential is compared with the average number of runs scored by a league team. This quotient is then divided by 2, since on average half of the differential is based on runs and half is based on runs allowed.

RunsRuns Diff.LeagueNo. of R per Ratio Team
AllowedRuns TeamsTeam
7754403354476856029.902 Pirates
7053813244388854929.506 Cubs
9675564116404880125.739 Yankees
9755993766094876224.727 Yankees
7554822734784859822.842 Cardinals
102566536095201279322.797 Beaneaters
8015072945211865122.642 Yankees
6994472524541856822.209 Pirates
7805052755091863621.605 Giants
99566233395551279620.996 Orioles
77951726279601266319.769 Orioles
927627300110131478719.1 01 Mariners
965656309113651481219.098 Yankees
84058625480141266819.075 Reds
9796713086503881318.937 Yankees
8405682725841873018.648 Indians
7465042425203865018.654 Indians
7995442555495868718.612 Red Sox
8015472545512868918.453 Yankees
8616012605658870718.411 Athletics
7085042044543856818.003 Red Sox
8235712525659870717.812 Giants
7945682265161864517.547 Yankees
10027242786436880517.332 Yankees
874620254119321674617.098 Astros
74253021274721262317.071 Orioles
6564641924540856816.917 White Sox
67149217955321055316.268 Tigers
8796512285650870616.149 Dodgers
79257421881091267616.170 Orioles
840607233102251473016.095 Indians
9587082506305878815.934 Tigers
7925941985037863015.746 Red Sox
9487082406303878815.221 Yankees
78357820580961267515.286 Mets
8576502075578869714.855 Dodgers
82761221573421073414.661 Yankees
8586262326355879414.631 Athletics
8156142015537869214.531 Cardinals
908676232129661681014.399 Diamondbacks
6124511614549856914.205 White Sox
70353217175221262713.688 Mets
851644207108921477813.3 02 Angels
74757217578991265813.385 Cardinals
7075281795421867813.240 Reds
76958218785561271313.177 Dodgers
829643186100271471613.084 Tigers
8196311885874873412.801 White Sox
80062018098581470412.888 Athletics
9127151976195877412.725 Pirates
73057515577531264612.076 Yankees
72757315477421264511.978 Dodgers
7726131595426867811.757 Braves
73357016397461469611.790 Athletics
8546791755987874811.723 Giants
949768181113581481111.2 04 Red Sox
69555713862181062211.167 Cardinals
68955113879761266510.474 Athletics
81867714112186167629.3 01 Diamondbacks
6455401059329146667.995 Braves
7416459610790147716.2 05 White Sox

</pre>

Well, it sure looked great in the preview. I'll try again shortly.

Test:

Sawney, don't use Shift-(comma) & Shift-(period) for your brackets. Use the two keys to the right of the "P" key.

Just the core information: R - RA, ratio (as described in post 145), and team.

335 29.9% 1902 Pirates
324 29.5% 1906 Cubs
411 25.7% 1939 Yankees
376 24.7% 1927 Yankees
273 22.8% 1942 Cardinals
360 22.7% 1897 Beaneaters
294 22.6% 1942 Yankees
252 22.2% 1909 Pirates
275 21.6% 1905 Giants
333 20.9% 1896 Orioles
262 19.7% 1969 Orioles
300 19.1% 2001 Mariners
309 19.0% 1998 Yankees
254 19.0% 1975 Reds
308 18.9% 1937 Yankees
272 18.6% 1948 Indians
242 18.6% 1954 Indians
255 18.6% 1912 Red Sox
254 18.4% 1953 Yankees
260 18.4% 1911 Athletics
204 18.0% 1903 Red Sox
252 17.8% 1912 Giants
226 17.5% 1947 Yankees
278 17.3% 1932 Yankees
254 17.0% 1998 Astros
212 17.0% 1971 Orioles
192 16.9% 1917 White Sox
179 16.2% 1968 Tigers
228 16.1% 1949 Dodgers
218 16.1% 1970 Orioles
233 16.0% 1995 Indians
250 15.9% 1934 Tigers
198 15.7% 1946 Red Sox
240 15.2% 1921 Yankees
205 15.2% 1986 Mets
207 14.8% 1955 Dodgers
215 14.6% 1961 Yankees
232 14.6% 1931 Athletics
201 14.5% 1931 Cardinals
232 14.3% 1999 Diamondbacks
161 14.2% 1905 White Sox
171 13.6% 1988 Mets
207 13.3% 2002 Angels
175 13.3% 1985 Cardinals
179 13.2% 1940 Reds
187 13.1% 1977 Dodgers
186 13.0% 1984 Tigers
188 12.8% 1901 White Sox
180 12.8% 1988 Athletics
197 12.7% 1925 Pirates
155 12.0% 1976 Yankees
154 11.9% 1978 Dodgers
159 11.7% 1957 Braves
163 11.7% 1990 Athletics
175 11.7% 1923 Giants
181 11.2% 2004 Red Sox
138 11.1% 1967 Cardinals
138 10.4% 1974 Athletics
141 9.3% 2001 Diamondbacks
105 7.9% 1995 Braves
96 6.2% 2005 White Sox

Test

I'm pretty sure what I did wrong on that last attempt, but I've sullied this thread enough without another attempt. Thanks for the tip, though. At least I got the header right!

"Joe,
are your picks in there? What did you use - team wins and losses?

Gimme a break."

LOL Chris, I'm not a troll, Jesus Christ, give me a little credit.

I've actually thought about this for years, even sent Paul White a lengthy email about it back when it was Baseball Weekly, pitching something for the back page of Baseball Weekly, regarding a World Cup type tournament of all-time teams that had a regular season of qualifying in heats based on era (with a certain amount of teams qualifying from each era), and then having a 32 team final tournament, with 4-team groups going through round-robin play, followed by a 16-team tournament of 7-game series for the championship. Might not have been the most 'mathematically' correct way to do it, but it would have been entertaining, and teams would have had to have been the type to excel at regular season and post-season type play to win it.

Anyway, I'd studied the damn thing so hard and long figuring out what teams (I chose about 100 for the qualifying) on and off over a period of years, that I just remembered who the best teams were, I used something like Wins, Losses, R, and RA (call me crazy, for counting actual wins and runs scored as important) to sort the teams, and then just looked them over to see who might have won the pennant in a landslide (those teams usually 'pull up' at the end, like the 1984 Tigers did), etc. and get this -- I subjectively chose the ones that I thought were the best - wholly crap. I also tried to represent every team that had a good couple of year run in there with the best team they put together in any one year (they had to win at least one division title or pennant to be considered).

But I guess I should have just used OPS+ and ERA+ to decide, that would have been a much better way to do it. :-)

*****************

Greg wrote:

It's not perfect (especially since z-scores assume a normal distribution (right?), which isn't always the case for a league), but it is better than simply comparing adjusted RS to the league average adjusted RS.

I disagree. How many runs your scored in relation to the league average is much more important than how many standard deviations you are over the league.

How can you say that a team that scores 900 runs in a league that averages 700 doesn't have a better offense than a team that scored 900 in a league that averaged 750? Who cares if the SD in the first league was 80 (2.5 SD better) and the SD in the second league was 50 (3 SD better)?

Furthering my last comment to Greg . . . rant time.

In something like this, you are looking for extreme teams - obviously, in the earlier years, it was easier to construct an extreme team than it is today for any number of reasons, the reserve clause (which allowed hoarding of talent), and more teams (as a percentage) that were completely uncompetitive - note James' index of competive balance has gone up (or down, depending on your perspective) basically since the beginning of baseball - although it did tick ever so slightly back up in the 90s, the first time ever, I believe.

Just because it was easier to construct an extreme team doesn't mean it was because the overall level of play was lower (I'm not saying the play wasn't lower, just that that wouldn't result in extreme teams in and of itself). It was easier for extreme teams to pop up because of the economic structure of the game.

Using z-score is something that biases the selection in favor of compact leagues (if I understand z-scores correctly, maybe I don't). So that biases it in favor of modern teams. It gives the illusion of correcting for quality of competition, because we assume the competition is tougher in modern times.

But when you look at the z-scores of two teams in relatively close years, where expansion isn't an issue and one league was compact and one wasn't for whatever reason (heck, even look at a two leagues from the same year, where one was compact and one wasn't - I think there's a year in the early-mid-70s that fits) you can see where it is simply wrong to use.

difference in the overall quality of play (within the range of major league players) does not cause differences in the ability to assemble a super-team or the spread of teams in a league. It just appears to correlate because both have been moving in the same direction throughout baseball history, due to entirely unrelated causes.

Quality of comp - getting better due to training, etc.

Spread of teams - getting closer due to economic system changes.

I think the uptick in the 90s is an example. The competition spread out slightly. Was this because play got worse? Or was it because the economics of the game made certain teams basically unable to compete at levels not seen since the early 50s, when mom and pop ownership (Connie Mack) finally started being pushed out of the game?

"Or was it because the economics of the game made certain teams basically unable to compete at levels not seen since the early 50s"

Change 'unable' to 'unwilling' - doesn't change the point of the comment.

And by at levels since the 1950s, I don't necessarily mean in terms of a worse record in any given season (though there have been some really awful teams in the past decade, worse than any in the 1980s for example), but in terms of being uncompetitive for a very long time, like the Pirates having a losing record for 13 consecutive years (as an example).

I also think the draft has to be considered as being reason for the compacting of the leagues 1970+. It's the great equalizer, and it keeps teams from being as bad as they were before 1950s (I realize the A's of the late 50s and early 60s fit there too, but they were really the only team, it wasn't a league wide epidemic by then). It's a lot easier for a blind squirrel to find the occasional nut when he gets just as many chances as everyone else. And when the blindest keep getting to pick first.

But using z-scores to adjust for competition will tell you there's been a big jump in the quality of play since the advent of the draft, which doesn't make any since. The only thing that changes was the mechanism for spreading out the talent, not the talent itself.

". . . which doesn't make any since. The only thing that changes was the mechanism . . ."

since = sense, changes = changed . . . sorry grammar police, I need to get some sleep.

that I just remembered who the best teams were, I used something like Wins, Losses, R, and RA (call me crazy, for counting actual wins and runs scored as important) to sort the teams, ... -- I subjectively chose the ones that I thought were the best - wholly crap...

But I guess I should have just used OPS+ and ERA+ to decide, that would have been a much better way to do it.

Based on those two sentences, *YES* you should have used OPS+ and ERA+ - it would have been a much better way to do it. My teams will kick the crap out of your teams. (overlap our lists and take only the non-matches)

I know you aren't trolling Joe, we're teasing each other..but

OBJECTIVITY ROOLZ!!1!1!!

a World Cup type tournament of all-time teams

Wasn't this in fact finally done by some publication or other in the early or mid-1990s? I seem to remember the 1952 or '53 Dodgers winning the computer "tournament." Maybe I dreamed it.

1902 Pirates
1903 Red Sox
1905 Giants
1906 Cubs
1909 Pirates
1911 A’s
1912 Red Sox
1921 Yankees
1925 Pirates
1927 Yanks
1931 Cards
1939 Yanks
1942 Yanks
1942 Cards
1948 Indians
1955 Dodgers
1957 Braves
1974 A’s
1975 Reds
1984 Tigers
1986 Mets
1988 A’s
1995 Braves
1998 Yanks
2001 Mariners

It gives the illusion of correcting for quality of competition, because we assume the competition is tougher in modern times.

I don't make that assumption. In fact, it appears when using z-scores that the "great" teams ratio (# of "great" teams divided by total number of teams) from the first half of the 20th century than the second half of the century.

How can you say that a team that scores 900 runs in a league that averages 700 doesn't have a better offense than a team that scored 900 in a league that averaged 750? Who cares if the SD in the first league was 80 (2.5 SD better) and the SD in the second league was 50 (3 SD better)?

I can say it easily, as I'm more interested in how a team separates itself from the pack.

How about this extreme example:

10 teams in a league, 5 teams are clustered around 500 runs, the other 5 teams are clustered around 900 runs. The average is 700 runs. Using just a comparison to league average would claim that half the league was better than a team in another league that scores 900 runs while the rest of the league isn't as close to 900 runs, but the average is 750 runs.

But when you look at the z-scores of two teams in relatively close years, where expansion isn't an issue and one league was compact and one wasn't for whatever reason (heck, even look at a two leagues from the same year, where one was compact and one wasn't - I think there's a year in the early-mid-70s that fits) you can see where it is simply wrong to use.

I'll have to investigate this. I doubt, though, it will be simply wrong to use. I'll also have to look at both the z-score method and the straight comparison to league average, then identify which teams benefit from one method to the other. I'll then look at those teams and their leagues to examine why they were hurt by the other method.

I don't make that assumption. In fact, it appears when using z-scores that the "great" teams ratio (# of "great" teams divided by total number of teams) from the first half of the 20th century is greater than the ratio of the second half of the century.

I need to do a better job of editing my posts. I'll investigate my own claim as well.

Dial, ERA+ does not count error making. I'd rather give up fewer runs than fewer earned runs. And OPS+ ignores stolen bases, sac flies, and betting with runners in scoring position. It also weight OBP and SLG equally even thoughnthey are not equal. I know that there is some merit in using run components to factor out "luck" as you call it, but clutch hits to get runs and clutch runs to get wins do matter. The game is not determined by components. Rather a game is determined by who scores more runs and the standings are determined by who wins more games. A better objective measure would be something like actual wins + pythagorean wins + run component wins + postseason wins - 1/2 postseason losses. For pre-1962 teams, that score is multiplied by 162/154. Then, add five points if they won the World Series and subtract five if they did not win a pennent.

If you are trying to find out the best teams of alltime using Diamond Mind simulation, you need to determine which teams had the best hitting and pitching performance relative to its league after normalizing for the park. Lee Sinins team RCAA and RSAA is a pretty decent measure to use to determine this. Below are the best 50 teams of alltime according to that measure.

YEARNameRCAARSAATOTAL
1927Yankees338117455
1939Yankees28992381
1884Maroons261115376
1897Orioles29858356
1896Orioles25691347
1936Yankees28757344
1891Red Stockings27172343
1937Yankees181144325
1886Cubs171142313
1898Orioles198110308
1931Yankees29810308
2001Mariners203102305
1948Indians168135303
1932Yankees27917296
1944Cardinals176118294
1887Cardinals19993292
1976Reds27313286
1929A's115167282
1894Orioles26016276
1905Giants19878276
1995Indians19183274
1953Yankees22152273
1895Orioles122150272
1953Dodgers22050270
1969Orioles136134270
1998Yankees168102270
1998Astros153116269
1902Pirates20166267
1922Browns141125266
1886Wolverines19372265
1998Braves113149262
1928Yankees275-17258
1906Cubs115142257
1931A's115140255
1942Yankees16986255
1943Cardinals120133253
1884Cubs22228250
1889Giants16582247
1942Cardinals96147243
1911Giants137105242
1941Dodgers15189240
1970Orioles17861239
1885Giants107131238
1906Indians14197238
1930Yankees344-106238
1996Indians112126238
2004Red Sox119118237
1921Yankees15284236
1876Cubs16470234

Trying to post the list in a readable form ...


YEARNameRCAARSAATOTAL
1927Yankees338117455
1939Yankees28992381
1884Maroons261115376
1897Orioles29858356
1896Orioles25691347
1936Yankees28757344
1891Red Stockings27172343
1937Yankees181144325
1886Cubs171142313
1898Orioles198110308
1931Yankees29810308
2001Mariners203102305
1948Indians168135303
1932Yankees27917296
1944Cardinals176118294
1887Cardinals19993292
1976Reds27313286
1929A's115167282
1894Orioles26016276
1905Giants19878276
1995Indians19183274
1953Yankees22152273
1895Orioles122150272
1953Dodgers22050270
1969Orioles136134270
1998Yankees168102270
1998Astros153116269
1902Pirates20166267
1922Browns141125266
1886Wolverines19372265
1998Braves113149262
1928Yankees275-17258
1906Cubs115142257
1931A's115140255
1942Yankees16986255
1943Cardinals120133253
1884Cubs22228250
1889Giants16582247
1942Cardinals96147243
1911Giants137105242
1941Dodgers15189240
1970Orioles17861239
1885Giants107131238
1906Indians14197238
1930Yankees344-106238
1996Indians112126238
2004Red Sox119118237
1921Yankees15284236
1876Cubs16470234

</pre>

I don't make that assumption. In fact, it appears when using z-scores that the "great" teams ratio (# of "great" teams divided by total number of teams) from the first half of the 20th century than the second half of the century.

Tamer, if I read your final list correctly, here's how it works out for you:

6 teams from 1901-1912 (actually, from 1902-1912).

5 teams from 1913-67.

17 teams from 1968-now.

"a game is determined by who scores more runs and the standings are determined by who wins more games."

man you're bright. are you sure you're only in high school?

Dial, are you writing this down?

A better objective measure would be something like actual wins + pythagorean wins + run component wins + postseason wins - 1/2 postseason losses. For pre-1962 teams, that score is multiplied by 162/154. Then, add five points if they won the World Series and subtract five if they did not win a pennent.

You know, that's all well and good (and wrong), but until I started asking, no one was doing that. Derf.

What is wrong with it? That I reward WS winners and penalize non-pennent winners? That I factor in real wins? That I don't simply use the run component wins? I explained why I think that real runs and wins should count. Any, yes, I did not use any formula when making my rankings.

1896 Orioles
1897 Beaneaters
1906 Cubs
1909 Pirates
1911 Athletics
1912 Red Sox
1917 White Sox
1921 Giants
1927 Yankees
1931 Athletics
1939 Yankees
1942 Cardinals
1942 Dodgers
1953 Yankees
1954 Indians
1955 Dodgers
1957 Braves
1963 Dodgers
1970 Orioles
1974 Athletics
1975 Reds
1978 Yankees
1986 Mets
1989 Athletics
1995 Indians
1998 Yankees
1998 Braves
2001 Mariners

"Based on those two sentences, *YES* you should have used OPS+ and ERA+ - it would have been a much better way to do it. My teams will kick the crap out of your teams. (overlap our lists and take only the non-matches)"

Well, take out the 1994 Expos, since put them on there as a bone to one of my favorite teams.

And I doubt your teams will win, you are basing it on a junk stat like OPS+.

1906 Cubs
1909 Pirates
1911 A’s
1912 Giants
1927 Yanks
1931 A’s
1939 Yanks
1942 Cards
1953 Yanks
1954 Indians
1955 Dodgers
1957 Braves
1961 Yanks
1970 Orioles
1974 A’s
1975 Reds
1984 Tigers
1986 Mets
1995 Indians
1995 Braves
1998 Yanks

It would be fun though Chris, where is your list anyway?

Greg, looking forward to it!

"10 teams in a league, 5 teams are clustered around 500 runs, the other 5 teams are clustered around 900 runs. The average is 700 runs. Using just a comparison to league average would claim that half the league was better than a team in another league that scores 900 runs while the rest of the league isn't as close to 900 runs, but the average is 750 runs."

I don't see why this is wrong. The league with 5 teams at 900 just had 5 really good teams and 5 really bad ones. You can have multiple good teams in a league. I don't see how the SD is relevant.

My new list, if I'm competeing with Dial :-)

1902 Pirates
1906 Cubs
1909 Pirates
1911 A’s
1912 Red Sox
1917 White Sox
1923 Giants
1927 Yanks
1931 A’s
1931 Cardinals
1939 Yanks
1942 Cards
1946 Red Sox
1953 Yanks
1954 Indians
1955 Dodgers
1961 Yanks
1969 Orioles - HOW COULD THEY NOT BE ON THE BALLOT? They are easily one of the 10 best teams ever, if they aren't in the tourney, the whole thing is worthless in my opinion. Something needs to be done there . . .
1974 Dodgers
1975 Reds
1984 Tigers
1986 Mets
1988 A’s
1992 Blue Jays (use 1993 if they get Henderson the whole season)
1995 Indians
1998 Braves
1998 Yanks
2001 Mariners

I replaced the 1957 Braves (I wanted to see Aaron/Mathews/Spahn in) with the 1974 Dodgers, who were a better team. Took out the 1994 Expos for the 1931 Cards, same thing.

The 1909 Pirates were better than the 1985 Cards, but I figured Wagner was in for 1902 and it would be nice to see THE modern speed team.

Now I'll put list up against any other :-)

you are basing it on a junk stat like OPS+.

You are basing yours on a junkier stat of *wins*.

How can you say that a team that scores 900 runs in a league that averages 700 doesn't have a better offense than a team that scored 900 in a league that averaged 750? Who cares if the SD in the first league was 80 (2.5 SD better) and the SD in the second league was 50 (3 SD better)?

I can say it easily, as I'm more interested in how a team separates itself from the pack.

You aren't measuring how a team separates itself from the pack - you are measuring how separated the pack is.

"You are basing yours on a junkier stat of *wins*."

No I'm not. Read the words I wrote, not what you want to see. :-)

I said I used wins and runs to make the list. Then I actually looked at each team and figured out who the best ones were, based on a bunch of factors.

We need to get Sean to change B-R to (O2)PS+ since it's way to easy to get lazy and use OPS+ for more than it's worth. I do it all the time too. But to not recognize that it's still just an eyeball stat (junk was hyperbole, I was having fun) is folly.

I don't see why this is wrong. The league with 5 teams at 900 just had 5 really good teams and 5 really bad ones. You can have multiple good teams in a league. I don't see how the SD is relevant.

Let's assume the five teams that only scored 500 runs also were equally terrible at allowing runs, and thus the five "really good" teams feasted on these five poor teams. Are they "really good" or just lucky to be in a league with five relatively poor teams?

I would have greater confidnce in a team being "really good" if it significantly separates itself from all or almost all of the other teams in the league, rather than it being bunched with half of the league that is significantly separated from the other half of the league, given this extreme example.

It would be fun though Chris, where is your list anyway?

Chris seconded my list, although my list still doesn't include any pre-1901 teams. Nor 2005.

The true winner of this simulation:
1876 Cubs
Why?
135-27 Pythag record (over 162)
146/139 OPS/ERA+
Thirdbaseman prevents all colored ballplayers from taking the field.

Actually, I'm going to go with:
1924 Monarchs
1931 Grays
1933 Giants
1935 Crawfords
1938 Grays
1953/54 Yankees
1954 Indians
1955 Dodgers
1957 Braves
1961 Yanks
1964 White Sox
1968 Tigers
1969/70 Orioles
1971 Pirates
1972 or 74 A’s
1975 Reds
1977 Dodgers
1978 Yanks
1979 Pirates
1986 Mets
1988 A’s
1995 Indians
1995 Braves
1998 Yanks
2001 A's
2001 Mariners
2004 Cardinals
2005 White Sox

"Let's assume the five teams that only scored 500 runs also were equally terrible at allowing runs, and thus the five "really good" teams feasted on these five poor teams. Are they "really good" or just lucky to be in a league with five relatively poor teams? "

Well they also would have played the other 4 that scored 900, right?

"I would have greater confidnce in a team being "really good" if it significantly separates itself from all or almost all of the other teams in the league, rather than it being bunched with half of the league that is significantly separated from the other half of the league, given this extreme example."

Why? Again why is distribution of talent a proxy for quality of competition? That's what you're doing here. Why else would you care about anything but how far ahead of average they are? Why wouldn't you assume that all leagues are equal except for cases of war, expansion, etc..

I think all of this is overthinking the issue.

Tamer, use a real life example. It is not as extreame, but the 1999 NL was an extreame league. They had five teams with at least 96 wins, and all but one other were below .500, some much so. Were the Reds, who were fifth in the league and missed the playoffs, as good as a typical 96 win team? Or did those five teams just have poor competion, other than the Dodgers (86 wins). It is tough to say. In any system in which all games played by the teams are against other teams in the system, the average winning percentage is .500, regardless of how spread out they are. So, the great winning percentage of those five teams could result from an unusually weak league or an unusually high number of great teams in one season which push the others down.

You are basing yours on a junkier stat of *wins*.

So Dial, would you rather that the Mets win a hundred games and the World Series, or have OPS+ and ERA+ over 130 and miss the playoffs?

Why wouldn't you assume that all leagues are equal except for cases of war, expansion, etc..

I definitely don't assume that all leagues are equal, especially considering the set of players is separated into two leagues, not necessarily equally from season to season.

Also, the average runs scored and runs allowed varies considerably throughout the history of MLB. Is scoring 550 runs in a 500-run league average equal to scoring 880 runs in a 800-run league average?

So given using the average or using the average and the standard deviation, I'll use the latter. If I can find a better method that uses more information and seems to eliminate further assumptions, I'll go with that.

So Dial, would you rather that the Mets win a hundred games and the World Series, or have OPS+ and ERA+ over 130 and miss the playoffs?

Gagne_55, do you know what the exercise here is? This question doesn't seem to reflect that.

But I'd rather win teh WS as an 83-79 team, but I wouldn't be so stupid as to think that made them one of the greatest teams of alltime.

I think teh 130 would be a much much much much better indicator of the quality of the team *regardless of the results of the playoffs*.

Gagne_55, do you know what the exercise here is?

Yes, but I am saying that results do matter just as how those results came from.

Also, ERA+ and OPS+ are two totally different measures. OPS shows a component that correlates with scoring runs. ERA, on the other hand, is a measure of the runs scored (or more acurately the runs that were earned) themselves. OPS is determined by hits, walks, and total bases, while ERA is determined by earned runs. Hence, ERA+ includes clutch pitching, while OPS+ does not include clutch hitting. You seem to be more interested in components, as that is what matters for Diamond Mind. There for, you should look at components allowed (or OPS+ allowed if you like), rather than ERA. Plus, you are not accounting for errors made by fielders, and errors affect how good a team is.

"I definitely don't assume that all leagues are equal, especially considering the set of players is separated into two leagues, not necessarily equally from season to season."

I know all leagues aren't equal, but how can you not assume anything else for an exercise like this? Why would you assume the distribution of the talent on the teams within the leagues says anything about the overall quality of play?

"Is scoring 550 runs in a 500-run league average equal to scoring 880 runs in a 800-run league average?"

Yes. Maybe a hair worse according to PythaganPat:

550 RS, 500 RA - .5406
880 RS, 800 RA - .5453

That's why I would use if forced into an objective system, the PythaganPat record - but even that is tainted, by things like teams far ahead playing rookies down the stretch and resting vets; or by the managers propensity to sub out in blowouts (of which great teams play many).

That's why you need to look at whole picture.

"I know all leagues aren't equal, but how can you not assume anything else for an exercise like this?"

strike the not, whoops.

Getting theoretical Chris, what if a team with the 130 OPS+ is so awful baserunning that they run themselves into 2 extra outs per game. What if they are so slow that they also hit into 200 double plays. What does OPS+ say about that.

If you are looking at team from 1900, 1915 or 1930, you'd also better take errors into account. Maybe they aren't as important now, but they sure were then, since there were so many.

Why would you use anything but runs scored and runs allowed as the main basis? We care about what they did for something like this, not what they might have done had all of their components averaged out.

If you are that worried about a team that was lucky, just adjust out for clutch stats or something. Adjust for some other things too. But using only OPS+ and ERA+ can't possibly produce a better list than something like Runs and Runs Allowed.

But using only OPS+ and ERA+ can't possibly produce a better list than something like Runs and Runs Allowed.

"can't possibly"? Your hypotheticals are lovely, but *THEY DON"T EXIST IN THE HISTORY OF MLB*.

Why get theoretical about this? My idea is excellent. It's certainly better than "Wins and Loses, and teams I remember were teh best".

1896 Orioles
1902 Pirates
1906 Cubs
1927 Yankees
1939 Yankees
1969 Orioles
1992 Blue Jays
1993 Blue Jays
1994 Expos
1995 Braves
1995 Indians
1996 Braves
1996 Indians
1997 Braves
1997 Orioles
1997 Yankees
1998 Braves
1998 Astros
1998 Padres
1998 Yankees
1999 Braves
1999 Diamondbacks
1999 Mets
1999 Astros
1999 Reds
1999 Yankees
1999 Indians
1999 Texas
2000 Giants
2000 Braves
2000 Cardinals
2000 White Sox
2001 Astros
2001 Cardinals
2001 Mariners
2001 As
2002 Braves
2002 Diamondbacks
2002 Cardinals
2002 Yankess
2002 As
2003 Giants
2003 Braves
2003 Yankess
2003 As
2003 Red Sox
2004 Cardinals
2004 Yankees
2004 Red Sox
2005 Cardinals
2005 WhiteSox
2005 Indians

--------
trevise

Dial, actually, the hypotheticals do have meaning. Take the 2005 NL. The Phillies had an OPS+ of 102, while the Cardinals had one of 103. From that, you would figure that the Cardinals had a better offense. However, the Phillies had 116 SB and only 27 CS, while the Cardinals had 83 SB and 36 CS. The Phillies ouscored the Cardinal 807-805, because of their advantage in stealing bases overcame their inferior OPS+. There were also clutch hitting differences, I am sure, which you would want to filter declaring it luck. Also equal OPS+ does not mean equal offense since the environments are different. However, this does show that base stealing and caught stealing do metter in an offense.

For era, just look at the 2002 AL. The A's had the best ERA that year, but the Angels were better at preventing runs due to making fewer errors. Also, you have not justified using a run component stat on offense with a net run stat on defense.

That's terrible analysis, Gagne_55, spare me.

WRITE-IN
-----------------
1983 Orioles

And then was born the Curse of Wild Bill Hagy....

"Why get theoretical about this? My idea is excellent. It's certainly better than "Wins and Loses, and teams I remember were teh best"."

There you go only reading what you want again.

I said that I did all of the research already, way back when, and I drilled it into my memory I looked at it so much. I'm sorry I don't have my spreadsheets handy right now to give you the exact formulas I used as a guide before coming up with my final decisions.

Why the venom Chris? How about trying to add something to the discussion and point out why you think Gagne's analysis is terrible, instead of coming off like an arrogant prick?

We've given you our reasons for why we think OPS+ isn't a good way to go, you yelling louder isn't going to change anyone's mind.

Explain this from 2005, took me looking at 5 teams before I found an example:

MIL OPS+: 104, R: 726 BPF: 100
NYM OPS+: 99, R: 722 BPF: 99

Well, the Mets stole 153 bases at a 79% clip, the Brewers, 79 at a 70% clip. The Brewers hit into 137 DP, the Mets 104. But I guess OPS+ is correct in saying the Brewers had a much better offense, right?

I don't have time to lo