The Scale

[STLCards2]

I did something like this several years ago, but thought I would do it again.

Where do you fall in the observation vs. sabermetrics debate. Too me, the site has seemed to be shifting a little bit in one direction, but I won't say which. This is out of curiosity - not to cause a debate over which is better. Here is the scale (the sarcasm and satire is intended to emphasize each level and provide a few laughs- don't take it seriosly) Basicaly, it is a 1-10 scale with 1 being least saber-happy and 10 being most saber-happy:

1. Sandy Koufax is the best pitcher in the history of all history because my grandpa said so... and he knows... because he was there. I never look at stats. Mine and my ancestor's observation skills are all that matters. Gramps wouldn't lie to me! Favorite player: Sandy Koufax, of course!

2. I guess stats are alright, but they don't tell much of the story. As long as my guy hits around .300, he is having a pretty good season. I don't buy a ticket to watch guys walk - they should swing. They are called "professional batters", not "professional walkers." It doesn't take skill to take pitches anyway! Favorite player: Nolan Ryan

3. I like stats: HR, BA, and RBI. Those are the ones that matter the most, becasue they have always been the most used. Surely 120 years of experts looking at those stats couldn't have been wrong! Who needs context? The new-stuff is only for math nerds! Do they even like or watch baseball? Favorite player: Jim Rice

4. Some of the new-fangled stats are useful I guess, but I don't understand them too much, so I stick with the old faves. Hey, I just heard of this great new stat called WHIP! Wanna' hear about it?! Sounds pretty fancy! Favorite player: Andre Dawson

5. A balance needs to be reached between the two sides. Advanced data tells a lot that old data didn't, but it won't show everything. That is where observation and common sense come in. Favorite player: Jimmy Foxx

6. After many years of holding out - I am ready to concede that RBI's and Runs are (slightly) overrated and things like OB% are more important than BA. Combining a guy's OB% and Slugg% seems like a prety good idea -but then again...those big RBI totals look pretty good! Favorite Player: Lou Gehrig

7. I dabble in a few newer stats, and truly believe that they do represent reality failry accurately; but I am too lazy to research the hard stuff, so I will stick with newer, but admittedly imperfect stats such as ERA+ and OPS+. There is still some value in what you see and observe at the ballpark that stats miss on occasion. Favorite Player: Bert Blyleven

8. RSAA, VORP, WARP, and the like are my weapons of choice. I mentaly understand (sorta) and enjoy reading about linear weights, markov chaines, etc., but feel the more "basic" sabermetric stats are good enough for me to get a general understanding of what I need to know. I will, however, throw phrases like WPA, FIPx ,and baseRuns around without understanding them in order to wrack up brownie points with the #9 and #10 guys. There just so smart! Favorite Player: Joe Morgan

9. I really do understand baseRuns, Markov chaines, etc. I use them all the time when evaluating players, and have even created my own stats that (unlike the "simple people") don't always just multiply stat 1 times stat 2. Favorite Player: Jimmy Wynn

10. I used to have a shrine for Bill James- until he refused to be uber-sabermetrical enough. I have replaced my Bill James shrine with anybody who comes up with a rip-off, unoriginal, zone-based fielding metric that is only slightly different than the four created this same year. Stats tell you everything - they don't lie...ever. Favorite Player: Anybody with a great +/- or UZR score

So, in seriosness - where do you fall? I know some of these will not accurately show your beliefs perfectly, so just vote for the closest. That is why there is no "other" category. Again, these were only exaggerations- I know none of us would say these things for real. :)

[bob]

Im definitely a 7.

[BlueBlood]

Same here, 7. Maybe I'll get into those really advanced metrics at some point in the future although I have dabbled in them.

[Brad Harris]

Five years ago, perhaps, I might have checked 9, but I've not kept up with the Joneses so to speak.

[hellborn]

But, I don't like Bert Blyleven!!!

[brett]

There are two sides to the issue for me.

First off, statistical value is my baseline.

For example, when I ran my first "saber" attempt of top all time players, before ever reading anything on sabermetrics, my spreadsheet helped me produce a list of players in terms of "games worth of offense above the average for the era"

I didn't have park effects, and estimated league from a table of league ERAs over the years. Just 6-7 years ago you couldn't find data on the internet like BBRef has.

I didn't properly weight things, but I weighted them using common sense, and perhaps an inch of "difficulty" rather than raw value, but what came back was a list like this for the best offensive players of all time:

Cobb 267
Ruth 240 (much shorter offensive career)
Williams 226 (no war credit)
Aaron 206
Musial 205
Mays 205
Speaker 205
Lajoie 201
Wagner 200
Mantle 197
Hornsby 180
Morgan 172 (I overrated steals and walks)
Schmidt 145
Gehrig 145 (Gehrig's got underrated because of era issues)
Foxx 144
Yaz 135
Dimaggio 99
Brett 95

This was value above average, so anything above 0 is above average.

I was suprised to see Morgan so high, largely because offense was low in his era, but being the only player in an unexpected position, I HAD to conclude that he must at least be underrated by most. Everyone else made pretty good sense.

Anyway, to answer the question, on one side, sabermetrics is flawed because it is incomplete. We don't know how many times a player went from first to third on a single. We don't know if a player tended to produce when the production was more likely to produce a win versus a guy who hits a three run homerun with his team already up 13-2. In that sense, observation helps us to be even MORE sabermetric. We fill in the statistical gaps with other records, and as statistically conservatively as is warranted. I can say for example that I believe, based on what he did, and from first hand accounts that George Brett safely picked up 12-18 runs on Eddie Mathews with his baserunning that doesn't show up directly in the stat line. I use what Brett did (triples, steals, iphrs) and my memory of him running hard, sliding hard into third, running out 2 inside the park home runs in 2 days to help clinch the '85 division title, and Mathews' stats and perception of Mathews to guess that Brett made up at least some ground, and I use known values for great and poor baserunners to estimate where a good (Brett) versus at best an average (Mathews) baserunner would do.

Second, I also like to look at what a player did in his career in terms of what kind of talent level it required. Would his talent produce similar stats in at least some other time periods. Did a player produce stats that might not be the most saber-valuable, but which, if that player knew what was valuable they could have done.

Could Rice have drawn 100 walks a year if he was shown clearly how doing so would help his team win, even if he loses some off the raw stats? I know for example from observation that George Brett could hit the high strike. I am not sure how some hitters today could do with the high strike.

Interestigly, I have tended to lower Morgan based on stats! because of his poor hitting in certain situations, and evidence that he was an average fielder when his entire career is evaluated.

[AstrosFan]

I don't put markov chains in the same category as base runs and linear weights. Markov is much harder to understand, and frankly, I don't yet understand how to calculate it. But I do understand base runs and linear weights, and I understand why they should be used over simpler metrics such as OPS+. I at least attempt to use DIPS theory in measuring pitchers, rather than simple ERA+. However, I have never really invented a metric, rather, I simply tinker with the work of others. I'd say I'm between an 8 and a 9.

[willshad]

Im probably between a 5 and 6. I agree that sabermetrics can help see who would be best , if given an even playing field, as well as helps rate players on overll contributions...not just triple crown stats.
At the same time, I think they try to go too far in assuming how players WOULD have performed, given different circumstances, and not enough value to what they actually DID. I HATE the way it assumes the talent pool is equal during every point in major league history, even though it is obvious some eras have many better hitters. For instance, they see that Morgan had a 187 OPS+ in the 1970s, and assume he would have the same score playing in the 1990s..even though logic would tell you that wouldnt happen. He would need to hit 50 homers and bat .350 to have that kind of score during that era. They give too much credit for 'dominance' during periods of low talent or less competiition...eg Mantle, Schmidt.
It is good to know that one guy only had so many RBIs because of his lineup, or his stadium. Yet, that doesnt make all those RBIs meaningless...it is still scoring runs that wins games.
I also dont like how they change their tune so quickly, depending what they are arguing for. For instance, people say Jim Wynn was killed by his home stadiums. Yet, if one looks at the stats, you can clearly see he hit just as well at home as on the road for his career. For him , they give him credit for learing to adapt to his home stadium (wich MUST have been killing him, even though the numbers dont support it!), rather than penalize him for not playing at a hall of fame level on the road. However, for Jim Rice, he isnt given credit for hitting like a monster at home, well because EVERYONE hits well at Fenway. Youre either going to look at the home/road splits or you arent..make up your mind.
People need to learn that if you 'adjust', and play with the stats enough, you can use them to 'prove' almost everything. And the more adjusting and playing you have to do, then the less meaning the results have.

[brett]

For instance, they see that Morgan had a 187 OPS+ in the 1970s, and assume he would have the same score playing in the 1990s..even though logic would tell you that wouldnt happen. He would need to hit 50 homers and bat .350 to have that kind of score during that era.


Newer stat methods however do agree with you that it is easier to put up very high OPS+ numbers for example in eras where the average hitter is doing less-provided that the game conditions like the balls and gloves are the same. Morgan has more of a ceiling that he can tap into than say Hornsby because the baseline is set higher (although Hornsby may have benefitted because the average player in his time was not fully benefitting from the liveball advantages in terms of extra base hits and drawing walks-another issue), so a 187 in 4 run environment probably represents a talent level of
maybe 170 in a 5 run environment. The opposite also hold for pitchers, so Koufax' 131 ERA+ might be more like 140 in a 5 run setting.

On the other hand, the 187 OPS+ in Morgan's era was worth just as many wins to his team in theory as a 187 in a 5 run environment, so while the talent need may have been less, the value produced is the same. That's why I tend to rate players on a balance of value and talent, however both of these things are not flaws in sabermetrics per se, but rather flaws in the method of sabermetric analysis currently being used and a difference of opinion about whether to rate players on value or on ability/talent.

[AstrosFan]

To complement Brett's post, I will point out that translations from one era to another are not estimates of what the player would do in a different era, but simply equivalencies of value. Sabermetricians couldn't care less about what a player would do except when dealing with current players. For long retired players, prediction is useless. All they want to know is how much a player's contribution was worth to his team winning games.
Second, sabermetricians are fully aware that talent levels are not consistently even between hitting and pitching. Many are attempting to work through this assumption. However, it is much better to assume an even distribution of talent than to do what many who do not understand the situation do, which is to assume (for example) that a higher level of offense is indicative of a greater pool of hitting talent. That is based on faulty logic. Until you can prove the talent pool was greater on one side or the other, you should just work from the premise that it is about even.
In spite of this, the latest advances in sabermetrics do a good job of accounting for the difficulty of putting up dominant numbers, and are making adjustments for the pendulum shifts between offense and defense.

[willshad]

To complement Brett's post, I will point out that translations from one era to another are not estimates of what the player would do in a different era, but simply equivalencies of value. Sabermetricians couldn't care less about what a player would do except when dealing with current players. For long retired players, prediction is useless. All they want to know is how much a player's contribution was worth to his team winning games.
Second, sabermetricians are fully aware that talent levels are not consistently even between hitting and pitching. Many are attempting to work through this assumption. However, it is much better to assume an even distribution of talent than to do what many who do not understand the situation do, which is to assume (for example) that a higher level of offense is indicative of a greater pool of hitting talent. That is based on faulty logic. Until you can prove the talent pool was greater on one side or the other, you should just work from the premise that it is about even.
In spite of this, the latest advances in sabermetrics do a good job of accounting for the difficulty of putting up dominant numbers, and are making adjustments for the pendulum shifts between offense and defense.

Well, if youre just figuring out how valuable a player was/is in one particular setting, instead of tryng to figure out how he would do in another setting, arent you doing exactly what the RBI enthusiasts do? That is, rate a player on factors out of his control? For instance, Joe Morgan did what he did in the 70s, and because of the low offensive environment, he was extremely valuable. But does he really have control over how good other players in the league are? Then you look at someone like Roberto Alomar, virtually the same type of player as Morgan, but playing in a higher offensive environment. He wasnt as 'valuable' as Morgan, due to more runs being scored, but Id be willing to bet that, given the same conditions, their numbers would be very similiar. Should Morgan be given extra credit for having greater OPS+ scores, when it is obvious that they'd be about the same as Alomar's if he played during the 1990s? You can 'translate' the power numbers, and say 'well if Morgan hit 28 home runs in 1976 then he would hit about 45 in 1996' or something like that, but I think such assumptions are ridiculous. He would not be a super slugger no matter what the era. And it is also likely his steals would go down if playing in the 1990s, because steals were less of a weapon then.
So is Morgan REALLy that much better than Alomar, or is it an illusion, based on the players they are being compared to?

[AstrosFan]

No. I am rating a player based on how valuable the runs he created were. Joe Morgan was valuable because of the things he did that were under his control. Measuring offense only, and without any context adjustments, his 1976 season would be one of the top ten in baseball in 2007. Of course, ignoring context would be silly. A reasonable translation would put Morgan ahead of everyone except Alex Rodriguez. We are talking about a guy who would have a ridiculous OBP, would be the best base stealer in baseball, and would probably hit 30+ homers. Factor in defense, and Joe Morgan 1976 would be the best player in baseball in 2007. Morgan was not dependent on his environment to be a valuable player. Morgan might not have the same OPS+, but he would have created about the same amount of run value as he did in 1976. (Note: I am projecting, rather than translating here, to make a point. I know it goes against what I argued earlier.)

As for your point about RBI, I am doing the exact opposite of what RBI enthusiasts do. I am removing the illusion that Morgan was just an okay offensive player that traditional analysis creates. RBI enthusiasts add an illusion by suggesting that a stat like RBI has meaning when it really tells you nothing about the quality of a player's offense.

[BradC34]

I think I'm between a six and a seven. I voted seven though. I really don't think that the stat guys give the human aspect enough credit. That's a big reason why I sided with 7 is because I believe in some of the newer stats but I don't think you can turn humans into numbers. This isn't the movie "Pi" you know. We're not all just one big mathematical algorithim. I whole heartedly believe in superstions. Not because I'm superstious but if a player thinks he's playing well because he didn't step on the baseline on his way to the field, then by golly let him believe that.

Just my two cents.

[Los Bravos]

VII

That seems to be the consensus.

[willshad]

No. I am rating a player based on how valuable the runs he created were. Joe Morgan was valuable because of the things he did that were under his control. Measuring offense only, and without any context adjustments, his 1976 season would be one of the top ten in baseball in 2007. Of course, ignoring context would be silly. A reasonable translation would put Morgan ahead of everyone except Alex Rodriguez. We are talking about a guy who would have a ridiculous OBP, would be the best base stealer in baseball, and would probably hit 30+ homers. Factor in defense, and Joe Morgan 1976 would be the best player in baseball in 2007. Morgan was not dependent on his environment to be a valuable player. Morgan might not have the same OPS+, but he would have created about the same amount of run value as he did in 1976. (Note: I am projecting, rather than translating here, to make a point. I know it goes against what I argued earlier.)

As for your point about RBI, I am doing the exact opposite of what RBI enthusiasts do. I am removing the illusion that Morgan was just an okay offensive player that traditional analysis creates. RBI enthusiasts add an illusion by suggesting that a stat like RBI has meaning when it really tells you nothing about the quality of a player's offense.


Youre probably right, but that also has a lot to do with 2007 not being such a great year for hitters, either. I can say the same thing about Roberto Alomar. Put his numbers from his best seasons into this year, and he is the best player in baseball, all things considered. Put Morgan into 1999 or 2000 AL and he is just another guy. An All Star, and possible MVP candidate on a good team, but not really the best player in the game. I think talent pool has a LOT to do with it...moreso than a lot of people are willing to accept. It's easy to be the best player in the league , when there are no great players who are also in their prime during that time. Was Morgan a great player during 75 and 76? Yes, obviously. Was he also lucky that no other standout players were in their prime during that time? Just as obvious. He also benefitted by being in great lineups. His teams routinely led the league in scoring, so his 'environment' isnt really typical of the time and era.
I admit, Morgan was slightly better at the plate than Alomar, both in prime and career-wise, but not nearly as much as sabermetrics would indicate. When factoring in defense, I see no problem rating them about equal.

[AstrosFan]

Morgan was not just baseball's best player in 1975 and '76, but also in '73, '74, and possibly '72, though that is neck and neck with Dick Allen. In 1972, the best position players behind Morgan (in the NL, since that is who he is competing with) were Cesar Cedeno - on track for the HOF before injuries derailed him, Billy Williams, Johnny Bench, Bobby Bonds, Pete Rose, Jimmy Wynn, Al Oliver, Chris Speier (okay, you got me there), and Willie Davis (a very underrated player) by Matt's PCA system.

In 1973: Willie Stargell, Cedeno, Darrell Evans (horribly underrated), Bonds, Rusty Staub, Tony Perez, Lou Brock, Rick Monday, and Bob Watson.

1974: Mike Schmidt, Wynn, Stargell, Evans, Cedeno, Ralph Garr, Steve Garvey, Richie Zisk, and Bake McBride

1975: Dave Parker, Schmidt, Ron Cey, Davey Lopes, Garvey, Bench, Cedeno, Greg Luzinski, and Dave Cash.

1976: Schmidt, Cedeno, Wynn, Garry Maddox, Bill Madlock, Garvey, Watson, Cey, Monday.

Are all these guys studs? No, but there are a number of them, and many of these players are underrated due to the environment in which they played.

1999 AL. Since you mentioned Alomar, I'll do the top nine other than him.

Derek Jeter, Manny Ramirez, Bernie Williams, Nomar Garciaparra, Shawn Green, Jason Giambi, Rafael Palmeiro, Jermaine Dye, Randy Velarde.

It's a very good crop of players. But it's no better than the upper level competition Morgan faced, and I would dare say it's worse. If Morgan can be better than Bench and Schmidt, there's no reason he can't be better than Jeter and Manny.

[willshad]

Morgan was not just baseball's best player in 1975 and '76, but also in '73, '74, and possibly '72, though that is neck and neck with Dick Allen. In 1972, the best position players behind Morgan (in the NL, since that is who he is competing with) were Cesar Cedeno - on track for the HOF before injuries derailed him, Billy Williams, Johnny Bench, Bobby Bonds, Pete Rose, Jimmy Wynn, Al Oliver, Chris Speier (okay, you got me there), and Willie Davis (a very underrated player) by Matt's PCA system.

In 1973: Willie Stargell, Cedeno, Darrell Evans (horribly underrated), Bonds, Rusty Staub, Tony Perez, Lou Brock, Rick Monday, and Bob Watson.

1974: Mike Schmidt, Wynn, Stargell, Evans, Cedeno, Ralph Garr, Steve Garvey, Richie Zisk, and Bake McBride

1975: Dave Parker, Schmidt, Ron Cey, Davey Lopes, Garvey, Bench, Cedeno, Greg Luzinski, and Dave Cash.

1976: Schmidt, Cedeno, Wynn, Garry Maddox, Bill Madlock, Garvey, Watson, Cey, Monday.

Are all these guys studs? No, but there are a number of them, and many of these players are underrated due to the environment in which they played.

1999 AL. Since you mentioned Alomar, I'll do the top nine other than him.

Derek Jeter, Manny Ramirez, Bernie Williams, Nomar Garciaparra, Shawn Green, Jason Giambi, Rafael Palmeiro, Jermaine Dye, Randy Velarde.

It's a very good crop of players. But it's no better than the upper level competition Morgan faced, and I would dare say it's worse. If Morgan can be better than Bench and Schmidt, there's no reason he can't be better than Jeter and Manny.



You conveniently left off a bunch of guys who peaked during the same time as Alomar. Belle, Thomas, Griffey, A-rod, Vaughn, Thome, Pudge, J. Gonzalez, Sweeney, Delgado, E Martinez, all of whom played at a hall of fame level for at least several seasons....and thats not even going to the NL. Morgan was only better than Bench and Schmidt because Bench was past his prime by that point, and Schmidt was still not in his prime yet. Stargell, Rose, and Wynn were also past their prime...and Parker hadnt entered his yet. Who does that leave? Luzinski? Garvey? Cedeno?

[stejay]

I only really tend to rely on the vital stats, and I dont really take THAT much interest in Sabermatrics, although I do think they are reasonably important, I just don't care for them much.

[philipthegreat]

While I am a 6 my favorite player is Cy young not Lou Gerhig. I just haven't taken the time to understand things ERA+ etc. Those wins and innings pitched look mighty fine.

[dgarza]

5, 6, & 7 seem so similar. I'm not sure what separates the three. They all seem to acknowledge new metrics, yet hold on to or balance out with older approaches. I think I'm a combination of 5, 6, & 7 that balances out to a 6, but a different 6, not the way 6 is described in this poll. 7 seems closest.

[brett]

My real issue with this question is basically this hypothetical.

If we could take all of the "observational data" and historical accounts and actually figure out what ALL of that means in terms of bases, runs and wins, then it would be sabermetric-"value based".

I think that people are mixing 3 things together here. Its not really a one-dimensional question.

Observations and historical accounts can do 3 things:

1) They can help fill in gaps where sabermetrics is incomplete. Dickey, Sisler, and Gehringer generally look like less great fielders than historical accounts give them credit for (Dickey and Sisler basically average and Gehringer good but not all time top 5). Most around here will bump them up and assume that the stats are not complete or perfect enough to rate their defensive VALUE. They might also tell us how they influenced the performance of their teammates.

2) They can tell us about a players talent level-independent of their value based production. I would say for example that watching Joe Morgan's swing, it was nothing special. In fact most swings in the mid 70s were geared toward making level contact, and getting out of the box. I can say that Jim Rice was scary and strong and that he had the hitting talent to do just about anything, (but his particular style, and the wisdom of the time lead him to a less patient approach than would be optimal). He had the talent to put up A-Rod type hitting numbers today. (He also, by the way may be overly docked by sabermetricians for ballpark, because it put him a little closer to the ceiling than guys in other parks).

3) They can tell us about a guys personality independent of anything on the field. Greenberg, Musial and Gehrig were great Americans.

Of these, #1 is ESSENTIAL. Without it, we probably get 75% of the true value of the entirety of players who have played the game. We do better with some, and very poorly with others.

I emply #2 as a modifier. I like to look at talent level independent of how it translated into value, but it won't move a guy more than about 10% in my rankings.

#3 is not really an issue, but if two guys are basically tied in my rankings, I might give it to the better person.

[STLCards2]

5, 6, & 7 seem so similar. I'm not sure what separates the three. They all seem to acknowledge new metrics, yet hold on to or balance out with older approaches. I think I'm a combination of 5, 6, & 7 that balances out to a 6, but a different 6, not the way 6 is described in this poll. 7 seems closest.


The # 5 person would most likely use stats a lot - but almost always use "bubble-gum stats as their reference point. They will throw RBIs and Runs scored down your throat all day. They love stats - just older ones.
Whether you are #6 or # 7 depends on which stats you use. A # 6 guy is starting to OB% instead of batting average and slugging% instead of RBI's ,etc, but very little thought of park/era context. The # 7 guy, however begins to account for context, by focusing on ERA+, OPS+, etc.

I see a lot of difference between the three.

[leecemark]

--Seven was an easy choice for me. I enjoy baseball much more than mathmatics;). Seriously, I think it is imperitive to consider contextual stats to get any kind of accuracy in player evaluation. I can't go just by the numbers though. The uber stats give you a good rough estimate of where a player fits in. Looking at the shape and cicumstances of his career is what makes this fun though. Being totally committed to the numbers would make discussion pretty boring for me.

[Windy City Fan]

Seven was an easy call for me too. Some of the advanced metrics are hard to find and even harder to understand the thinking behind them. Or some don't even tell you how they are calculated at all, leaving us with no idea if the logic behind the metric is sound or not. Plus, as Mark said, I like baseball way more than I like math.

That said, I strongly believe in basic saber concepts like the value of OBP and DIPs theory (though I think that can underrate certain pitchers). That said, the unique circumstances and issues that surround a player's career need to be looked at and considered as well.

[AstrosFan]

You conveniently left off a bunch of guys who peaked during the same time as Alomar. Belle, Thomas, Griffey, A-rod, Vaughn, Thome, Pudge, J. Gonzalez, Sweeney, Delgado, E Martinez, all of whom played at a hall of fame level for at least several seasons....and thats not even going to the NL. Morgan was only better than Bench and Schmidt because Bench was past his prime by that point, and Schmidt was still not in his prime yet. Stargell, Rose, and Wynn were also past their prime...and Parker hadnt entered his yet. Who does that leave? Luzinski? Garvey? Cedeno?

It had nothing to do with convenience. I said I was simply doing a top ten. If those players were not among the top ten players in baseball that year, they're not really competing for the leaderboard spot. And I love how you write things like "Garvey? Cedeno?" when those players are clearly better than say, Shawn Green, Jermaine Dye, and Randy Velarde, and wound up with better careers than Nomar will have (though Nomar was great in 1999). And your list includes players who were not among the top 24 position players in the AL, like Thomas, Gonzalez, Sweeney, and Vaughan. Playing time and defense matter. Oh yeah, let's see, Griffey, Vaughn, and Thomas had all passed their primes.

Face it. Your argument doesn't work, because you almost entirely ignore defense and other aspects of offense (I was rating players for their overall contribution) and still seem to believe that a higher offensive environment is indicative of greater hitting talent. There were just as many good to great players in their prime in 1975 NL as in 1999 AL, and Joe Morgan would be the best player in either league. Think of the comparison to Alomar, who was a top three player (along with Jeter and Manny). Everything is the same, except Morgan has a .481 OBP (due to the odds ratio adjustment), and he's the best base stealer in the game. With Morgan's OBP, and his stealing ability, he probably steals about 60 at a better than 80% clip. I know it seems like a stretch for Morgan to top the league leader by 13, but remember, Tom Goodwin stole 39 while reaching base only 145 times. Leading off for a high powered offense (assuming he's playing for Cleveland), Morgan will compile a lot of PA, and thus he will reach base over 300 times. I'd be surprised if he doesn't steal 70+. I'm being conservative with the 60 figure. With all else equal (Morgan in 1975 was just as good defensively as Alomar in 1999), Morgan would easily vault ahead of Alomar, Jeter, and Manny as the best player in the AL in 1999.

[AstrosFan]

--Seven was an easy choice for me. I enjoy baseball much more than mathmatics;). Seriously, I think it is imperitive to consider contextual stats to get any kind of accuracy in player evaluation. I can't go just by the numbers though. The uber stats give you a good rough estimate of where a player fits in. Looking at the shape and cicumstances of his career is what makes this fun though. Being totally committed to the numbers would make discussion pretty boring for me.

I do not go full-science (10) because of the incompleteness of the stats. We don't have complete records of every play made throughout big league history, and thus we cannot make a fully accurate evaluation (or at least as close as we could hope to). How many extra bases did Ty Cobb actually take? What were the ROE rates for the top players in that category in 1894? Who benefited most from the different rules of early baseball? Gaps like those are why we can't fully trust the stats. I only voted 9 as a representation of my interest in sabermetrics, not how much stock I put into them. For that, I would be, oh, I don't know, 7 or 8. Finally, I will admit that I get frustrated with people who argue against sabermetrics that clearly have no idea what it is that sabermetricians do. We are not sitting around in front of a computer trying to figure out which players would make the best Strat-o-Matic team. We are gathering as much evidence and information as possible to take an objective look at what makes a winning ballclub in the actual game of baseball. EVERYTHING a sabermetrician does is directly related to what actually goes on on the field.

[AstrosFan]

The whole idea of a scientific endeavor is to find evidence that can improve our understanding of something. In baseball statistics, each bit of information we gather is designed to help us learn more about how teams can win more. Some add a twist of economics to it, and also apply the budget constraint of the teams to the model of winning baseball efficiently. No sabermetrician thinks, oh great, we have Win Shares and WARP-3 now! The answer to baseball's great mysteries of who is better than whom is finally here! They think, okay, let's analyze these methods and see how the evidence stacks up against it. Then they try to improve upon whatever flaws they find.

[AstrosFan]

One more thing (and I realize I could have put these all in one post): I look at baseball stats a lot because I don't really enjoy tv baseball unless I've got a bunch of friends to watch it with me. It's simply not a lot of fun watching it alone. The only way I could enjoy baseball alone is if I went to the ballpark. Of course that's going to be more fun with friends, but I can still have a great time there.

[brett]

One more thing (and I realize I could have put these all in one post): I look at baseball stats a lot because I don't really enjoy tv baseball unless I've got a bunch of friends to watch it with me. It's simply not a lot of fun watching it alone. The only way I could enjoy baseball alone is if I went to the ballpark. Of course that's going to be more fun with friends, but I can still have a great time there.

At home, I'll turn off the TV and put on the radio 9 times out of 10.

[AstrosFan]

I also have the problem of not having lived anywhere near Houston in 13 years. So as far as Astros games go, I don't catch a lot.

[dgarza]

The # 5 person would most likely use stats a lot - but almost always use "bubble-gum stats as their reference point. They will throw RBIs and Runs scored down your throat all day. They love stats - just older ones.
Whether you are #6 or # 7 depends on which stats you use. A # 6 guy is starting to OB% instead of batting average and slugging% instead of RBI's ,etc, but very little thought of park/era context. The # 7 guy, however begins to account for context, by focusing on ERA+, OPS+, etc.

I see a lot of difference between the three.It's just the way they were worded. Your explanations here, in your follow up post, are clearer that your initial explanations. I'll drop the #5 connection.

[STLCards2]

It's just the way they were worded. Your explanations here, in your follow up post, are clearer that your initial explanations. I'll drop the #5 connection.

Sorry!

A couple of observations:


It seems the 5+ group is much quicker to vote than the -5 people. I know that many of them feel the 5+ people look down upon them. I wonder if this is playing out here.

I know a few political experts who expained to me why Democratic candidates always do about 2% better in exit polls/pre-voting phone surveys (Gallop, Rassmussen, etc.) than they do in actual election - it is not as "cool" or socialy acceptable to be a Republican, so when asked about who they voted for, many out of self-consiousness, they will claim they voted for the Democratic candidate instead of the Republican. Anxiety and fear of ridicule can severley alter how/if someone votes. I really wonder if this is happening here. I don't know if this is tue or not, but it is something to think about.

[willshad]

It had nothing to do with convenience. I said I was simply doing a top ten. If those players were not among the top ten players in baseball that year, they're not really competing for the leaderboard spot. And I love how you write things like "Garvey? Cedeno?" when those players are clearly better than say, Shawn Green, Jermaine Dye, and Randy Velarde, and wound up with better careers than Nomar will have (though Nomar was great in 1999). And your list includes players who were not among the top 24 position players in the AL, like Thomas, Gonzalez, Sweeney, and Vaughan. Playing time and defense matter. Oh yeah, let's see, Griffey, Vaughn, and Thomas had all passed their primes.

Face it. Your argument doesn't work, because you almost entirely ignore defense and other aspects of offense (I was rating players for their overall contribution) and still seem to believe that a higher offensive environment is indicative of greater hitting talent. There were just as many good to great players in their prime in 1975 NL as in 1999 AL, and Joe Morgan would be the best player in either league. Think of the comparison to Alomar, who was a top three player (along with Jeter and Manny). Everything is the same, except Morgan has a .481 OBP (due to the odds ratio adjustment), and he's the best base stealer in the game. With Morgan's OBP, and his stealing ability, he probably steals about 60 at a better than 80% clip. I know it seems like a stretch for Morgan to top the league leader by 13, but remember, Tom Goodwin stole 39 while reaching base only 145 times. Leading off for a high powered offense (assuming he's playing for Cleveland), Morgan will compile a lot of PA, and thus he will reach base over 300 times. I'd be surprised if he doesn't steal 70+. I'm being conservative with the 60 figure. With all else equal (Morgan in 1975 was just as good defensively as Alomar in 1999), Morgan would easily vault ahead of Alomar, Jeter, and Manny as the best player in the AL in 1999.



Why cant people just admit theyre wrong? Ok fine, competing against Greg Luzinski, Steve Garvey and Cesar Cedeno is harder than competing against Manny Ramirez, Frank Thomas, Juan Gonzalez,Ken Griffey Jr, Albert Belle, A-rod, Pudge, Roberto Alomar, Derek Jeter, Nomar, etc. Makes perfect sense. Shall we move over to the NL? Would Morgan still have been as dominant against the likes of Bonds, Bagwell, Piazza, Sheffield, Walker, Biggio, Mcgwire, C Jones, Kent, Guerrero, Larkin, and Sosa? Keep in mind ALL of these guys had peaks around the same time. Morgan would have been lost in the shuffle. Heck even Bonds, a much better player than Morgan in his prime, didnt win the MVP from 1994-2000 (and he had to take steroids to do it the following years).
Just admit it; not all eras are equal when comparing the amount of top level talent who are in their primes all at the same time. Morgan still would have been a great player, no doubt about it; but it was impossible for any one single player to clearly be the 'best player in baseball' during that era, the way Morgan had done during his prime years. He may have been a little better than Alomar, yet he certainly would have been grouped in the same category as Alomar (probably similiar to the way jeter, A-rod and Nomar were all grouped together).