After 35 years of watching baseball, Nettles is still the first guy that comes to mind when I think about fielding excellence at third base. He was simply amazing.

An offense-only chart, using my usual system:



You've got to pack some leather to be listed here - and perhaps you can question Cey on those grounds. Harrah and Madlock were as good offensively as nearly everyone on this list and Bando and Elliott were distinctly better. This is based on RCAA, not RCAP - using RCAP would boost the older players at the expense of the more recent ones. You'll notice that every one of these players had at least one season as a below-average hitter; some of them had multiple seasons as well-below-average hitters.

Robinson, Boyer, Collins, and Groh have all been elected to the HoM. Traynor has been essentially dismissed, Cey just got inserted in to the lower backlog, and we've never entirely figured out what to do with Leach (who isn't all 3B, of course.) Although I had most of the elected ones on this list in the lower part of my ballot when they were elected, I can't say I was a big fan of most of them.

Here's a compendium to that list. First, let's take Groh out of it - I sense that his bat played a bigger role in his election than with Robinson, Collins, or Boyer. Here are the top 3 (and ties) offensive players by nth best year, taken from the above list:

1. Robinson, Collins, Boyer
2. Collins, Cey, Boyer
3. Cey, Leach, Robinson
4. Cey, Leach, Boyer
5. Boyer, Leach, Robinson
6. Boyer, Cey, Leach
7. Boyer, Leach, Cey
8. Boyer, Robinson, Cey, Leach
9. Cey, Robinson, Nettles, Leach
10. Cey, Nettles, Leach
11. Cey, Nettles, Leach
12. Cey, Nettles, Leach
13. Nettles, Cey, Robinson
14. Nettles, Cey, -
15. Nettles, Cey, -
16. Cey, -

The above is not an overwhelming argument for Nettles. If he's merely a very good defensive 3B, it's not enough. Is anyone willing to argue that was not merely very good on defense but a historical outlier?

Nettles was one of my favorites. He would make these diving catches of line drives where he would be completely horizontal in mid-air when he caught the ball. I used to practice catches like that in my front yard. This 'practice' usually involved waiting a bit so that I wouldn't catch up to the ball so that I would have to dive :-).

OCF's list above shows Nettles as similar to but a little better than Pie Traynor:


Actually that's only part true. He's distinctly similar to Pie Traynor in both of their primes and at their tail ends. Here's what's missing between them:


So reductively speaking, he's Traynor plus about 40 missing RCAA plus a 13 RCAA advantage in seasons that are similar.

Let's try it with Collins too. Here's the basic comparison:


Collins has the peak/prime advantage, Nettles, obviously, the career.



That -22 is a killer. Anyway in the common seasons, Collins leads by 8 RCAA (though -12 come on that -22 year). Nettles has an advantage of 43 "missing" RCAA above Collins.

Anyhow, just noodling. I have Nettles on the borderline for MLB-only players, but probably just on the out side after Jud Wilson is accounted for.

Comparing Nettles to Collins or Traynor is problematic, since the standards of third basemen were much different back then. We're probably better off just comparing him to the post 1950 models.

Comparing Nettles to Collins or Traynor is problematic, since the standards of third basemen were much different back then. We're probably better off just comparing him to the post 1950 models.

Or even post-1970. There are just a ton of 1970s 3B-men who compare well against a with guy like Ken Boyer. Mathews & Rosen rose the bar for superstars but Bando/Cey/Nettles certainly rose the HOVG bar.

The above is not an overwhelming argument for Nettles. If he's merely a very good defensive 3B, it's not enough. Is anyone willing to argue that was not merely very good on defense but a historical outlier?

I'm willing:

Nettles in the 70s was the best defensive 3B I have ever seen, better than Brooks (I did not see Brooks play in the 60s mind you), better than Aurelio Rodriguez (a SS playing 3B) and better than anyone since the 70s. (Rolen reminds me a lot of Buddy Bell defensively- I don't know why they really don't "look" similar, but my impression is that the overall caliber of their Dee is similar- Rolen when healthy being better - I saw both Nettles and Bell play together close in time- Bell was very good, but Nettles even aging and having lost a step was better)

Nettles lost several gold gloves he shoudl have won to Broks and he even lost one to Aurelio Rodriguez (who WAS very good, don't get me wrong).

I'm a met fan, I love David Wright, David Wright is more athletic than Nettles- but he will never ever be remotely the defensive force that Nettles was- Nettles' reaction times were unreal- that horizontal dive of his (which I used to try to imitate)- was not a belated dive in response to a ball he could have reached easier if he moved earlier, or a timed play like Winfields outfield fence leaps- Nettles' horizontal dives were 2B saving (or 1b saving in the 3b/ss hole) plays that no other human being (Brooks Robinson included) could make.

He was still a serviceable defensive 3B at 40- when he was a shadow of his earlier self.

No I don't have stats to back this up- but if I recall correctly one of Bill James' earlier work has one of Nettles mid 70s seasons (with Cleveland not the Yankees)as the highest rated defensive season of anyone (except maybe Ozzie) in the past 60 years,

The whole bunch of them suffered by comparison to a couple of roughly contemporary 3Bs. But I can't remember their names right now.

Still if any one of them rises above, it's Evans, isn't it? And given the gap after the Big 2, that pretty much makes them all HoVG until proven otherwise.

My first glove was a Graig Nettles one. My next and current glove was Jose Conseco.

Nettles was my favorite player when I first started following the game, and I agree about his defensive prowess (though I never saw Robinson, so I can't compare them).

So it pains me to say this: I had to do massive adjusting for defense to get Brooks onto my ballot, and Nettles's offense is sufficiently behind Robinson's that I can't see him getting there for me. He'd have to have been not only better defensively than Robinson, but a ton better.

Sorry, Graig. If it's any consolation, your followers as my favorite player—Mattingly, Bernie Williams—are likely to get similar treatment from me here.

WARP1 sees Robinson and Nettles at their best as being very similar in quality.

Robinson's top 5 defensive seasons by FRAA: 34, 24, 22, 22, 21
Nettles' top 5 defensive seasons by FRAA: 33, 21, 21, 20, 18

Their next 5 are also similar, though again Robinson is a little better:

Robinson: 18, 17, 16, 12, 12
Nettles: 18, 14, 11, 11, 9

It's in the remaining 8 seasons as full time starters that they really look different as defenders:

Robinson: 12, 10, 8, 7, 4, 2, 2, -3
Nettles: 6, 2, 0, -2, -8, -9, -11, -14

During his defensive prime, Nettles was so good that he was pretty much interchangeable with Robinson, though WARP finds him slightly less good each season. Robinson, however, was almost always an above average defensive player, where Nettles, according to WARP, had some genuinely bad seasons in the latter part of his career.

Comparing Nettles to Collins or Traynor is problematic

Per James and the spectrum shift, the proper comps for Nettles should be Childs and/or Doyle.
Fox is who should be compared with J.Collins and/or Traynor.

Still if any one of them rises above, it's Evans, isn't it? And given the gap after the Big 2, that pretty much makes them all HoVG until proven otherwise.

I have trouble seeing much difference between Nettles and Evans. Very similar career lengths (Evans has about 500 more PAs). Evans was a better hitter (119 OPS+ vs. 110) with a better OBP, but I think positional adjustments erase a lot of the difference--Nettles played 89.3% of his career games at third base, vs. only 53.7% for Evans, who spent most of the remainder of his career as a 1B/DH. Evans had the better offensive peak, but keep in mind that most of his best years with the bat came after he had moved to first base, so the difference in their peak value isn't as great as it first appears. And Nettles was pretty much universally considered a much better defender, correct? (Not that it means anything, but they're both each other's most similar player on Baseball-Reference.)

"Career" voters at least should consider that Nettles was platooned, apparently in Minnesota and the NL, not in Cleveland or New York; that is, 1968-69 & 1984-89, not 1970-83. It appears that he averaged only 111 games, NY 1980-83, for other reasons.

Graig Nettles at Retrosheet

Also, most of Nettles appearances not at 3B were when he was in Minnesota, and Killebrew was the 3B - he was still in all likelihood the best defensive 3B they had, but just didn't get the spot because of a superior offensive player there. With Evans, the issue was that he wasn't the best defensive 3B the Tigers had. If Evans had been capable of playing a decent 3B in his Tigers days, he would have done it, with Bergman at 1B, instead of Brookens at 3B.

As someone who watched a lot of Brooks Robinson on tv [admittedly much on very ghosty UHF] and a lot of ARod, I thought ARod was fabulous. That's the old "eyes" version of defense. How does ARod look defensively in the modern metrics compared to those leather guys above?

BTW, ARod is my just made up nickname for Aurelio of the Senators.

Fielding Win Shares per 1000 innings has Aurelio Rodriguez at 4.16 -- quite good, but behind his more famous contemporaries (Schmidt 4.51, Nettles 4.40, Evans 4.37, Bell 4.35, Robinson 4.24). I remember Rodriguez as the ideal third baseman -- though it could simply have been how damn good he looked on the field. Rodriguez had terrific poise and balance, and moved laterally with so little effort. Nettles and Schmidt both made it look harder, Nettles with those dives and Schmidt with those sweeping lunges at high chops off the turf -- but perhaps they were getting to more baseballs in the process than Rodriguez was.

A few days ago I felt old because I can remember the hype over Nettles coming up and hitting five HR in four games. Today I feel old because Alan Arkin has been nominated for an Oscar, and I remember seeing The Russians Are Coming The Russians Are Coming in the theatre in its first release. I just plain feel old.

I remember seeing The Russians Are Coming The Russians Are Coming in the theatre in its first release.

Hilarious movie!

"Vee are Norwiggian sailors on NATO training meeshun."

"Egermancy. Everyone to get from street."

"We've got to get orgainzed!"

Bob Elliott says hi.

I remember seeing The Russians Are Coming The Russians Are Coming in the theatre in its first release. I just plain feel old.

I remember that, too. That's not what makes me feel old.

I met my cousin's daughter for the first time, and learned that she is 32.

That made me feel old.

Not that turning 52 this weekend made me feel exactly young.

BTW, I have very vivid memories of Graig Nettles with the Yankees in the mid-70's (which was a mere ~30 years ago), including those horizontal stabs of line drives. He was an amazing fielder, and made some spectacular plays in the post-season for the Yankees' during the mid-70's. And he had more than a little pop in his bat. He also had more than a little rubber in it, too, at times.

He is under the border line for the HOF, and probably for you HOMers. But for those of us who saw him play for those Yankees revival years, he will always be a special player.

>I remember seeing The Russians Are Coming The Russians Are Coming in the theatre in its first release. I just plain feel old.

Because you are.

(I saw it, too. Funny movie. Almost as funny as 1942 [similar theme].)

Nettles is not so good a batter as I expected to see in the simple sabrmetrics. He was underrated as a batter thanks to "wisdom" misunderstanding the home runs and strikeouts but not far underrated.

In his three best years at bat, coinciding with a run of three Yankee pennants, he finished 16-5-6 in the MVP election.

His 68 official stolen base attempts were about four times too many: the 17 attempts in 1976 would have been just right. Are there other players with grossly anomalous SB seasons? (Mo Vaughn enjoyed a better season in 1995 but his record shows that he was generally a competent thief.) What's up with that at 32 years old?

Are there other players with grossly anomalous SB seasons?

Pudge Rodriguez in 1999 (his MVP year). Rafael Palmeiro in 1993. Sometimes managers just get the bright idea to start running a lot with someone ...

If I remember right, in Mo Vaughn's big stolen base year, at the beginning of the year, the first basemen for a lot of teams were playing off the bag a lot of the time, and he'd get a great jump.

Bill Buckner was 18 for 22 in 1985, as a 35-year old. Although from 1982-1985, 1984 was the only season he didn't steal at least 10 bases or have at least a 75% success rate. So maybe that's not anomolous, just unusual in light of how horribly slow he was at the time.

And then there are occasionally teams like the late 1970's Indians (under Torborg) that have horrible success rates because they hit-and-run a lot. This is one problem with conventional baserunning stats, that the benefits of H&Rs; are generally hidden, while the costs are right out there in the open.

"Are there other players with grossly anomalous SB seasons?"

Ryan Klesko in 2000-2001 stole more then half of his career SBs in those two years (46, 23 each year) and never stole more then six any other year.

And then there are occasionally teams like the late 1970's Indians (under Torborg) that have horrible success rates because they hit-and-run a lot.

That could be also said for the pre-Robinson generations. Success rates were much higher back then due to the hit-and-run (unless your name was Carey, of course :-)

1976 Oakland As
Bando stole 20
Don Baylor stole 52

The team stole 341 bases overall (183 the year before), Chuck Tanner was hailed as a genius- he was traded to Pittsburgh
Pitt went from 130 sb to 260 sb

gallons of ink were spilled babbling about the great rediscovery of this potent offensive weapon the SB (just as when Brock had rediscovered the SB, and when Maury Wills had, and Aparicio had...)

not knowing any better, I was fascinated by SBs, I thought, wouldn't it be great to have a 50+ sb guy in my team's lineup? (we later got one by 1982- he had no discernable effect on the team's offensive output)-
I remember an article that came out sometime between 1979-83- it was written by Earl Weaver and some BBWAA guy- Weaver & ___________ very articulately p!ssed all over stolen bases. The co-auther (wish I remembered who it was) used stats- teams that lead the league in sbs never lead the league in runs scored- the correlation between team SBs and runs was actually negative! Weaver explained why steals were useless- it's just a base- 2nd is not as important as reaching 1b, getting caught stealing is far worse (losing an out and a baserunner) than getting to second is good- batters hurt themslves (countwise) trying to "protect" the runner- for trying to let the runner steal etc... (Weaver was not a small ball guy)

I can't imagine any manager now co-authoring an article with a Sabr/stathead type guy- baseball men were much much more openminded 25+ years ago (either before statheads became insulting, baeball men felt threatened or both)

baseball men were much much more openminded 25+ years ago (either before statheads became insulting, baeball men felt threatened or both)

The fact that Earl Weaver was openminded does not mean that "baseball men" in general were more open minded. Weaver wasn't representative of anything but Weaver. In fact, I think the opposite has occurred - far more baseball men pay attention to the results of these types of studies now than did when Weaver was managing.

- far more baseball men pay attention to the results of these types of studies now than did when Weaver was managing.

Yes I think far more pay attention- but far few are willing to admit it
back then they would opnely express interest in studies- now- most baseball men will pridefully boast that they are unfamilar with or simply ignore statistical studies

Refusing to consider any type of statistical analysis was a major reason, I believe, that Frank Robinson was allowed to "retire" from the Nationals this past October.
Frank claimed he didn't need to consult stats; he went with his gut in any given situation.
I think Franks's gut was right about 40% of the time which was about the Nationals winning percentage.
Earl never wanted his guys to risk getting caught stealing second because that would cause his game-winning 3-run home runs to be 2-run HRs and he'd lose more games. Earl was a very bright guy but you couldn't convince Jim Palmer of that fact. He notoriously fought with Weaver constatntly; he never felt Weaver understood the pitching game or the people who pitched for a living.

>Don Baylor stole 52

That's only because everybody was afraid to stick their nose in there to make the tag.

kevin Posted: January 24, 2007 at 12:32 PM (#2285707)
Not that turning 52 this weekend made me feel exactly young.
Srul, I turned 52 this past weekend myself.

There must be a dozen over-50s here. Including me, but I'm a lot younger than you guys.

If it gets you down, take up duplicate bridge and play anywhere but a collegiate club. At age 50 as at age 20, I have heard (usually women at age 80) say that it's so nice to see young people playing bridge.

>Don Baylor stole 52

No anomaly. "Don Baylor steals" was a Baltimore Orioles marketing slogan, on newspaper inserts and billboards. He stole 24-32 (with declining success rate) in his four Oriole seasons before the pickup in 1976, and he completed the decade at 22-26.

Graig Nettles
11 of 17, 1976 (6.4% of career PA, age 31.8-32.1)
21 of 51! otherwise

For Baylor I count
_SB-CS
240-86, 1970s (mean full season 30-11)
_45-34, 1980s (mean 5-4)

If [being 50+] gets you down, take up duplicate bridge and play anywhere but a collegiate club.

Want to feel young? Go to your local casino any morning and play Stud Poker.
Want to feel old? Go to your local casino any evening and play No-Limit Hold'Em.

I remember an article that came out sometime between 1979-83- it was written by Earl Weaver and some BBWAA guy- Weaver & ___________ very articulately p!ssed all over stolen bases. The co-auther (wish I remembered who it was) used stats- teams that lead the league in sbs never lead the league in runs scored- the correlation between team SBs and runs was actually negative! Weaver explained why steals were useless- it's just a base- 2nd is not as important as reaching 1b, getting caught stealing is far worse (losing an out and a baserunner) than getting to second is good- batters hurt themslves (countwise) trying to "protect" the runner- for trying to let the runner steal etc... (Weaver was not a small ball guy)

I was an Orioles fan of this era (I started following the team closely in 1977), and perhaps the most remarkable baseball statistic I've ever seen is the fact that the 1973 Baltimore Orioles under Earl Weaver led the American League in stolen bases. Incidentally, they finished 3rd in the AL in runs scored, 4 behind league leader Oakland, and had the most wins in the AL that year.

Yes, the Orioles led the league in steals that year, but they were right around the league average in caught stealings, and were a close second in on base percentage. So their rate of attempts per time on first was probably about average. They had either the best or second best success rate in the AL that year (too close to Boston's to do in my head). So Weaver was really just taking advantage of the personnel he had. It's not really that he thought stolen bases were useless, it's that he thought players like Omar Moreno were useless. If you gave him a young Don Baylor who could steal bases and hit for power with a decent OBP, he'd let the guy steal.

Earl Weaver's take on stealing bases:

"Team speed? For Christ sake. You get ******* goddammed little fleas on the ******* bases, getting picked off trying to steal, getting thrown out, taking runs away from you. You get them big *********** that can hit the ******* ball out of the ball park, and you can't make any goddamned mistakes."

I saw Nettles play most of his Yankees career - good fielder, decent hitter.
Compare his offense to Elliott's just as a 3B, and try to pretend the D makes up for it. And then try again.

How does Nettles campare when defense and offense are combined:

Assuming that Schmidt, Matthews, Brett, and Boggs are no-brainers

Ron Santo........13.6, 13.5, 13.5, 11.6, 10.1 (62.2).....116.7
Paul Molitor......9.9, 9.6, 9.3, 9.3, 9.2 (47.3).....133.8
Brooks Robinson..11.4, 10.6, 10.3, 9.3, 8.7 (50.3).....123.4
Darrell Evans....12.5, 10.0, 8.5, 8.2, 8.2 (47.4).....115.6
Ken Boyer........11.2, 10.8, 10.5, 10.3, 10.0 (52.8).....102.0
Scott Rolen......11.8, 11.0, 10.8, 10.6, 9.8 (54.0)......89.3
Edgar Martinez...11.2, 10.3, 9.6, 9.2, 8.9 (49.2).....107.6
Graig Nettles....10.7, 10.2, 8.9, 8.4, 8.2 (46.4).....107.2
Chipper Jones.....9.7, 9.6, 9.0, 8.4, 8.2 (44.9)......83.9
Sal Bando.........9.3, 8.5, 8.4, 7.9, 7.5 (41.6)......82.9
Al Rosen.........12.3, 11.2, 9.1, 6.8, 6.7 (46.1)......52.4

LOL at #37 jim . . .

Allright, it's time to get some momentum behind Graig Nettles. Can anyone tell me what leaves him so far behind Brooks Robinson?

The following numbers are all standard deviation-adjusted. SFrac is the percentage of the season played (compared to a player with league average PA/G in 162 games). BWAA is batting wins above average, BRWA is baserunning wins above average, FWAA is fielding wins above average, Replc is wins above average a replacement player at the same position would have accumulated in the same playing time, and WARP is the first three minus the fourth (wins above replacement). Note that Replc drops by 0.6 wins in 1973 to account for the DH. 1972 and 1981 are adjusted to 162 games. aTTL is career totals excluding sub-replacement seasons. Sorry for the goofy formatting, but the PRE tag doesn't seem to be able to handle consecutive whitespaces anymore.


Nettles



3-year peak: 18.8
7-year prime: 38.6
Career: 62.2

Robinson



3-year peak: 20.6
7-year prime: 39.9
Career: 63.4

OK, I admit it, Brooks was better--barely. Nettles doesn't have the shiny MVP year that Brooks does. Other than that, they're virtually interchangeable. Nettles was a better hitter--25.8 BWAA after moving the DH adjustment from Replc to BWAA vs. 21.4 for Robinson. Brooks was a better fielder--or, more precisely, he maintained his defensive excellence for a longer period of time. Nettles's 1971, when he set the all-time records for assists and double plays by a third baseman in the same season, is by my measure the greatest single defensive season at *any* position since 1893 (ignoring intrinsic positional value and treating first basemen and shortstops the same). They look like virtual clones to me, especially if one factors in that the NL was much stronger than the AL during Brooks's 60's peak--just take a look at what happened to Frank Robinson when he switched leagues. Yet Brooks was elected in his second year of eligibility, and Nettles isn't close. What gives?? As with Puckett, I think some voters' attention must be distracted by Fame and Gold Gloves. Take a look at the record, and you'll see that Nettles and Brooks are one of the most closely matched pairs of players in history. They should both be in the HoM.

(cross posted from the 2006 ballot discussion thread)

I'm not seeing exactly the same thing as you in your numbers. If you eliminate the measured value of a replacement player and instead look v. average then they have the same bat, same baserunning but Brooks is 6 wins better with the glove. I agree that value is value and the lower replacement level for Nettles gives him real value. However, you're looking at an expanded league where replacement level is more than likely dropping for Nettles. League strength is a legit argument for giving Brooks extra credit.

I still agree they should both be in. Nettles is halfway between Brooks and Boyer in my rankings.

DL from MN, that lower replacement value is because of the DH! Not because of league size. My replacement calculations are done on the bottom 3/8 of regulars, so league size is never an issue.

If you want to compare Nettles and Brooks to league average on an equal playing field, add 0.6 wins per 1.00 SFrac to BWAA for league-seasons with the DH. This is how I get Nettles with 25.8 BWAA and Brooks with 21.4. Compared to average, Nettles was 4.4 wins better with the bat (25.8-21.4) and Brooks was 6.0 wins better with the glove (21.8-15.8), so a total advantage of 1.6 wins to Brooks. Since their position and career lengths (ignoring sub-replacement seasons) were identical, Robinson's 1.6-win advantage on value above average translates almost exactly to his 1.2-win advantage on WARP.

The league strength question *definitely* favors Nettles, since there was a big AL/NL league quality gap in the 1960s.

I started this project coming out of a conversation with DanR about OPS+. What I'm really interested in is WS but I'll stick with going into it via OPS+ just because that's how I started out. What I'm trying to determine is how much of an advantage selected players really provided their teams against their cohort at the same position.

For Nettles, I looked at all AL 3B 1970-1983, which is when Nettles was a regular in the AL. I did not include his NL years simply as a matter of convenience--coulda/shoulda/woulda. As a generalization, I would say that FAT at 3B was in about the 90 OPS+ range. This is based on those (few) teams that made mid-season changes for whatever reason.

1973 Cal--Bob Oliver 111 but couldn't catch the ball, Alan Gallagher 91
1974 Cal--Dave Chalk 84 Oliver 84 Paul Schaal 89
1976 MN--Dave McKay 43 Mike Cubbage 105 (Cubbage, contrary to my memory, was never a regular 3B in TX prior to this trade)
1977 Tor (expansion team)--McKay 32 Roy Howell 117 Doug Rader 104
1983 Bal--ARod (Aurelio) -9 Leo Hernandez 82 Todd Cruz 51

The mid-season replacements were 91, 84, 89, 105, 117, 104, 82, 51.

Then there were the journeymen, some of whom might have been FAT. Journeymen are defined as playing regularly for more than one team, but never more than a year at a time (no 2 years in a row for the same team).

McKay with his 43 and 32
Kevin Bell, Chi 1976 abnd 1979--104 and 75
Lenny Randle, TX 1974 and Sea 1981--103 and 68
ARod with his -9 in Baltimore but also a 72 the previous year in Chi (he was a regular 1970-1979, he qualifies here with these 2 1-year stints at the tail-end of it all)

So maybe I want to revise the FAT level downward to account for those 43, 32, 75, 68, 72 and -9.

There were also some long-career guys in Nettles' cohort--B. Robby and DeCinces took Bal through 12 years. Petrocelli and Hobson in Bos--9 years. Melton 5 years. ARod 10 years straight not counting his journeyman years. Bando 10 years in Oak and Mil. Don Money 6 years in Mil, before and after Bando. Wayne Gross (?!) 7 years in Oak after Bando. George Brett. Roy Howell. Buddy Bell.

So through those 14 years, Nettles provided the following advantage on OPS+: -3, +5, +20, -2, +6, +15, +37, +13, +2-, -1, +10, +14, 0, +18. His mean is +11.5. The closest single years to that are 1977 and 1980 when he was +10 and +13, but in 1980 he only played in 89 games. So his most typical year was 1977.

Here are his cohorts:

Nettles 124 OPS+ and 25 WS
Brett 143 and 29
Harrah 136 and 25
Soderholm 129 and 20
DeCinces 117 and 21
Howell 117 and 12
B. Bell 115 and 15
Gross 111 and 16
Hobson 100 and 13
Bando 99 and 17
Chalk 96 and 14
Cubbage 95 and 11
Stein 89 and 11
ARod 65 and 5

The non-Nettles means are 111 and 16, so Nettles provided an edge of +13 OPS+ and 9 WS (6 on offense and 3 on defense).

Due to a miscalc, I also did 1975. Nettles 115 and 21, the non-Nettles mean 100 and 14. So Nettles is +15 and plus 7 WS (5 on offense and 2 on defense).

So let's just say that in a typical year, Nettles delivered about +14 OPS+ points and 8 WS (5.5 on offense and 3.5 on defense).

This sounds pretty good to me. By comparison (so far) Larry Doyle is +28 OPS+ and +9 WS (10 on offense and -1 on defense). Rizzuto OTOH is -12 OPS+ but +1 WS on offense (above average playing time) and +4 WS on defense.

I think the critical points here are:

1. The standard deviation of defense in WS is far too small. This is positive, not normative. The standard deviation of team fielding WS is much smaller than the standard deviation of team BABIP allowed (after converting both to wins). WS is factually wrong on this front. That's why you're getting Doyle as having so many more WS above position per season than Rizzuto (that, and the fact that Doyle isn't being compared to Collins/Lajoie while Rizzuto is being compared to Joost/Boudreau etc.).

2. Rizzuto is also faring better on Batting WS than OPS+ because it's crediting him for his OBP-heaviness and baserunning (I imagine).

3. Nettles is being compared to a much stronger group (Brett, Bando, Brooks, Bell) than Doyle is.

4. I know you're a peak voter, but Nettles has ooodles of career on Doyle.

Thanks for giving him a lookover, sunnyday. Again, I finally feel like we are speaking the same language.

3. Nettles is being compared to a much stronger group (Brett, Bando, Brooks, Bell) than Doyle is.

The stuff I put at the top of this thread wasn't comparing him to Brett, Bando, or Bell. I was comparing him to Brooks and the some old-time glove-first 3B who might not be the best comparisons. But there are a number of people whose offense I would take over Nettles, notably including Bando and Elliott.

Sunnyday, oh and also I forgot to mention, are you adjusting Win Shares for the DH???

OCF, OK, but Nettles was a historically great defensive third baseman, while Bando was a poor one, making Nettles's peak meaningfully higher than Bando's. Nettles and Elliott are comparable on peak and prime, but Elliott was only half a 3B, did much of his damage during the war, and didn't stick around for nearly as long as Nettles did.

>Doyle isn't being compared to Collins/Lajoie while Rizzuto is being compared to Joost/Boudreau etc.).

True, each within his league, which is where pennants are contested.

>3. Nettles is being compared to a much stronger group (Brett, Bando, Brooks, Bell) than Doyle is.

Sure, that's who they had to beat in order to win a pennant, fair or not. (But of course we all make allowances for strong cohorts, though I doubt that most voters recognize a weak cohort that is not strictly timeline or WWII related, in which case we're not talking about a positional cohort at all. But when making allowances for strong cohorts, I'd say Rizzuto still has it over Nettles. There were a bunch of good 3B in Nettles' day but with 14 teams the mean is only the 7th/8th best player. In Rizzuto's day, the mean is the 4th/5th best.

Oh, ####. All of the numbers that I have referred to here as the "mean" are actually the "median." I agree with DanR that those are vastly more fair numbers for comparison. I always get those two confused. I meant "the other mean." Doh.

From DanR's post 44 above:

Nettles
Year SFrac BWAA RWA FWAA Rpl WARP
70-85 13.89 20.2 -0.1 15.9 -25.2 63.4

Brooks
Year SFrac BWAA RWA FWAA Rpl WARP
59-74 15.13 20.3 -0.4 21.8 -21.8 63.4

Both using their best 16 yrs, which pretty much coincide with their useful years in both cases. They DO look remarkably even.

WARP3 has Broooks ahead by THIRTEEN wins over the same 16-yr span. What's the (HUGE) diff?!?

BP DOES adjsut for the DH, giving Nettles a large EqA boost from WARP1 to WARP3.

player BRAR BRAA FRAR FRAA WARP3
Brooks 452 .... 86 .. 614 . 189 . 118.7
Nettles 493 ... 179 .. 436 .. 96 . 104.9

BP agrees that Nettles was better by rate with the bat, equiv worse with the glove. But BP seems to give many more wins-above-replacement to Brooks' durability. Robinson played 15 (lenght-adj) seasons of 150+ games. Nettles only had 9. DanR's onw SFrac shows Robinson with an entire 1.2 seasons more of play in the 16 yrs used; which, if you're a career (or durability) voter, IS a big difference; altho most of it comes in the shoulder season years.

I don't have Brooks as 13 'wins' better than Nettles; but I have him enuf ahead that Robinson would still be #1 on this ablllot, ahead of W Clark, closer to the Sandberg/Ozzie/Trammell territory (among 3B, near E Sutton and J Collins). Nettles may debut next week at #15.

TomH, that *is* a huge difference--worth analyzing step by step. The problem with WARP3 is that it's got both the DH adjustment and the inscrutable league quality adjustments, which makes it impossible to break down and analyze. So let's instead just use the much more transparent WARP1.

First, let's consider runs above average. BP has Brooks with 385 BRAA + FRAA from 1957-77, and Nettles with 276 from 1968-88, a discrepancy of 109 runs, while I have Brooks with 312 and Nettles with 303, a gap of just 9 runs. So there are exactly 100 runs unaccounted for.

I'll put in bold the biggest explanations of the difference.


1. If I strip down my run estimator to match BP's--i.e., ignoring net double plays, strikeouts, sacrifice flies, and unintentional walks, I get Brooks with 1,355 "raw" runs created from 1957 to 1977, compared to 1,347 for BP. Nettles comes out a nudge higher by comparison--I have him with 1,259 "raw" runs created from 1968 to 1988, while BP has him with 1,237. So I'm higher than BP by 8 runs on Brooks, and 22 runs on Nettles. That opens up a 14-run gap between my assessment of the pair and BP's. This is either due to using different park factors--I use baseball-reference's, BP doesn't say where they get theirs--or to different run estimators.

2. Robinson adds 2 runs for having a below-average strikeout rate, while Nettles loses 2 for having an above-average strikeout rate. This reduces the gap between my take and BP's to 10 runs.

3. Robinson adds 25 runs from sacrifice flies, Nettles 19. This shrinks the discrepancy further, to 4 runs.

4. Deducting for intentional walks drops Robinson by 11 runs and Nettles by 8 runs. This increases the gap to 7 runs.

5. Adjusting for net double plays costs Robinson 23 runs and helps Nettles by 11 runs. This is the first major difference between our systems--I count for NetDP, while BP ignores DP altogether. At this point, we have explained 41 runs' worth of the difference between my evaluation of the two players and BP's.

6. Both were below-average non-SB baserunners, but Robinson more so. He loses 13 runs on EqBR, while Nettles falls by only 6. The gap is now 48 runs.

7. My system uses an average of Fielding Win Shares and BP FRAA for defense, and regresses both to match the standard deviation of Chris Dial's Zone Rating data. This different approach to fielding is the second critical factor. Regression to the mean shrinks the absolute size of the fielding gap between them, and incorporating Fielding Win Shares is a major boost to Nettles, because FWS is much more impressed with his D (relative to Robinson's) than FRAA is. After making these adjustments, I have Brooks at +198 in the field and Nettles at +131, a 67-run edge, while BP has Brooks at +231 FRAA for 1957-1977 and Nettles at just +125, a 106-run edge. The difference between those two advantages is 39 runs. We have now accounted for 87 runs of the discrepancy between my take on the relative value of Nettles and Brooks and BP's.

8. The actual runs-per-win rate during Brooks's career was 8.8, and during Nettles's it was 8.6. BP's "standard league" has a runs-per-win rate of 9.0. So Brooks's 312 career RAA in my system get multipled by 9/8.8, for 319, to and Nettles's 303 career RAA in my system get multiplied by 9/8.6, for 317. This restores 10 runs to Nettles. So far, so good--we've accounted for 97 runs of the 100-run gap in runs above average between my analysis of Nettles v. Brooks and BP's.

Now, on to replacement level.

Brooks had 739 RAR-RAA in 2,857 games from 1957-77, so that's a replacement level of 42 BP runs/4.66 wins below average per 162 games. Nettles had 626 RAR-RAA in 2,697 games from 1968-88, a replacement level of 38 BP runs/4.18 wins below average per 162 games. So according to BP, 3B rep level was .48 wins lower per 162 games in Brooks's leagues than it was in Nettles's. I have 3B rep level in Brooks's leagues at 1.48 wins below average per 162 games, and in Nettles's leagues at 1.76 wins below average per 162 games. The DH adjustment, which I'll deal with later, accounts for .33 wins per 162 games of relative difference, so I only see 3B replacement level in Brooks's leagues as .05 wins lower than it was in Nettles's. This disagreement on the relative strength of 3B in the 1960s and 1970s is the third major factor. It accounts for .43 wins of difference per 162 games, times Nettles's 2,697 games is a beefy 7 wins--before addressing Robinson's longer career.

Moreover, BP has a lower absolute replacement level across all positions than I do, by 3 wins per 162 games. That means it rewards Robinson for the extra season he has on Nettles by 3 wins more than I do.

Finally, there's the DH. After adjusting for the 1981 strike, Nettles played in 1,040 more games in DH leagues than Brooks did, which is worth 1040*.6/162 = 4 wins.

So there you have it. To summarize: I see Nettles and Brooks as identical twins, and BP sees them as distant cousins, for the following reasons:

1. Net double plays: 4 wins.
2. Fielding quality: 4 wins.
3. Relative strength of 3B in the 1960s and 1970s (BP thinks 3B was much tougher in the 60's than in the 70's while I don't): 7 wins.
4. BP's lower replacement level: 3 wins.
5. DH adjustment: 4 wins.
6. Odds and ends (run estimation, strikeouts, intentional walks, sac flies, non-SB baserunning): 2.5 wins.

Total explained variance: 24.5 wins.
Difference between my estimation of Brooks's career advantage over Nettles and BP WARP1's: 26.7 wins.

Do let me know if this affects your assessment of Nettles at all.

excellent, excellent stuff, DanR. I'll reply when I get time, hopefully this weekend.

Not counting net double plays is bizarre. For Robinson, it's a big deal as he was among the league leaders almost every year during the 1960s.

Fielding Win Shares are viewed by some as particularly unreliable. Out of curiosity, how did Nettles' fielding win shares progress from his "peak" Cleveland years, 1970-71, through the Yankee years later in the 70s? FRAA does have the Cleveland years as signficantly better, and my intuition is that Fielding Win Shares would have them as much, much, much better.

Out of curiosity, how did Nettles' fielding win shares progress from his "peak" Cleveland years, 1970-71, through the Yankee years later in the 70s? FRAA does have the Cleveland years as signficantly better, and my intuition is that Fielding Win Shares would have them as much, much, much better.

Nettles and Fielding Win Shares:

1970 - 8.7
1971 - 9.8
1972 - 7.1
1973 - 7.4
1974 - 6.1
1975 - 6.6
1976 - 7.1
1977 - 6.3
1978 - 6.7
1979 - 3.0

They're just not included in BP's run estimator. Ask them, not me.

Nettles's '70 and '71 with Cleveland are indeed his best two seasons according to FWS--9.8 and 8.7. His '72 with the Indians (7.1) is right in line with his average for his prime Yankee years.

DanR, I wasn't suggesting that you (or anyone but BP) ought to explain it but that it is truly strange that BP wouldn't fix the exclusion of DP info in its run estimator. It is a very good reason to use RCAA (or some other metric) rather than BRAA.

Thanks, John, for the Fielding Win Share data. My intuition was entirely wrong. For what it's worth, I doubt that there are too many people who saw both Nettles and Robinson in the early 70s who thought that Nettles was the better fielder at the time. Both FRAA and Fielding Win Shares suggest that he was, and we don't have zone rating for the period.

Defensive Regression Analysis is equally fond of Nettles.

Does RCAA use actual DP opportunities data, or is it just based on actual GDP? If it's the latter, that's very unfair to hitters on good teams and far too friendly to leadoff hitters. My WARP, of course, use real net double plays for seasons they are available.

I think all RC-based calcs use actual DP.

In that case, I'll still claim superiority for my BWAA + BRWAA...absolutely everything is factored in (strikeouts, sac flies, hit by pitch, intentional walks, etc.). Double plays are net and non-SB baserunning is included. It uses BaseRuns for pre-1947 seasons, so it's flexible enough to handle extreme run environments comfortably (with estimates for things like reached on error and caught stealing). It's denominated in wins so you can compare across eras. And it's available in an easy-to-use spreadsheet on the Yahoo group. But that's just my two cents. :)

I doubt that there are too many people who saw both Nettles and Robinson in the early 70s who thought that Nettles was the better fielder at the time. Both FRAA and Fielding Win Shares suggest that he was, and we don't have zone rating for the period.

I did see both in the early 70s and I thought Nettles was better.

I think those who saw both play in the early 70s AND had not seen Brooks play in the 60s would rate Nettles higher. Those who saw Brooks first, in his prime, would not be able to overcome that initial impression- and would continue to rate Brooks higher- as long as he remained "good" in the field- I did not see Brooks play in the 60s, my memories don't go back that far, I saw both play simultaneously and for years was baffled that almost everyone thought Robinson was better.

Mike Green, I don't think your intution was entirely wrong. The standard deviation of fielding Win Shares is *extremely* small, far lower than the actual spread of defense in real baseball. After converting Nettles's FWS to runs above/below average and correcting for the standard deviation, here's how WS sees his defense:

1970 CLE +31
1971 CLE +35
1972 CLE +18
1973 NY +22
1974 NY +11
1975 NY +15
1976 NY +17
1977 NY +14
1978 NY +17

The Yankee years are eyepopping, Gold Glove type seasons--with not a single dud in the batch--but it's the first two Cleveland years that suggest Nettles may have had the highest fielding peak for a 3B ever. (His rivals for that title would be Jimmy Collins, Lave Cross, Lee Tannehill, and Tim Wallach. Brooks is a nudge down, although on *career* fielding value he laps the field at 3B).

Saw both. thought Nettles was better-- and so was Clete Boyer. Now some of that was because Brooks was so darn slow that I just didn't think he had the phenomenal range of Boyer and Nettles. Unlike Boyer who played a very good shortstop, moving Brooks to short just didn't seem possible. No knock on Brooks--Nettles and Boyer were that good. Nobody in today's game compares to any of them, with the qualification that I haven't seen Chavez enough to totally discount him. What I have seen of Chavez isn't close.

Rolen? Beltre? UZR sure loves them...

Saw both. thought Nettles was better-- and so was Clete Boyer.

Never saw Boyer.
Nettle's reaction times were unreal. You couldn't go more than few games before you saw him make a play that no one else could, period.

The funny thing was that Nettles was slow, he really couldn't run worth a damn- his first two steps were quick but then...but he could move his body from side to side faster than anyone I ever saw then or now. I have never seen anyone else catch as many balls on FULL extension while diving for a ball- not like Carney Lansford who dive for everything because he never moved his feet- you'd see a frozen rope either down the line or in the hole, and Nettles would be a horizontal blur.

For awhile there I never saw a bunt or an infield hit in his direction, he would run in at full speed (for him) barehand and throw and the throw seemed to always be right on the money- no matter what angle he threw from - you see 3Bs try that play all the time- most don't make them with any great regularity, and many you hope just don't throw the ball away- for a few years their it always seemed that Nettles put the ball in the 1b's glove on that play.

David Wright is making that play more and more...

I thought that Brooks and Nettles were equally slow. If we're talking reaction time, my money would be that prime Brooks was quicker than prime Nettles.

DanR, as you know, I have a lot of difficulty with the 70-71 FWS (and FRAA numbers) for Nettles. They would show him as having the highest defensive peak for a third baseman ever. For comparison purposes, what would the peak FWS to runs above average (and corrected for standard deviation) be for a shortstop (presumably early Ozzie), a centerfielder (presumably early Mays) and a second baseman (presumably Maz).

By the way, the inclusion of Tim Wallach in the list of all-time great peaks illustrates the problem. I saw Tim Wallach a lot during his time in Montreal. He was a good defensive third baseman once he got the hang of the position (which took a few years). But, a defensive statistic that suggests that he was better than Brooks Robinson at their respective peaks is not credible (at least over the limited 2 year time frame that we are speaking of).

*Just* using Fielding Win Shares (not averaging in FRAA), corrected for season length, and measured in wins, here are the top 3 seasons from 1893 to 1999 at each position:

1B
1. Tino Martinez 1999, 2.4
2. Charlie Grimm 1920, 1.9
3a. Fred McGriff 1995, 1.7

2B
1. Bill Mazeroski 1962, 3.5
2. Frankie Frisch 1927, 3.5
3. Hughie Critz 1933, 3.4

3B
1. Ossie Vitt 1916, 4.6
2. Lave Cross 1899, 4.5
3. Jimmy Collins 1899, 4.5

SS
1. Hughie Jennings 1896, 3.7
2. Rabbit Maranville 1914, 3.7
3. Art Fletcher 1917, 3.6

LF/RF
1. Joe Vosmik 1932, 4.6
2. Bill Collins 1910, 3.5
3. Jesse Barfield 1987, 3.3

CF
1. Devon White 1991, 3.9
2a. Marquis Grissom 1994, 3.6 (short season)
2b. Jimmy Piersall 1955, 3.3
3. Andruw Jones 1999, 3.2

Note that I set the standard deviation adjustment so that the stdev for each position *during the 1987-1999 period* matches the Dial stdev for that period. It seems clear that the standard deviation of Fielding Win Shares at 3B was much higher pre-1920 than it is in the modern game, while the modern game appears to be its high for CF and perhaps 1B.

That Vosmik season is just batty, by the way. 8 Fielding Win Shares for a corner OF? Say wha?

I looked it up. Vosmik made more plays in the same number of games (153) as the Indian centerfielder Earl Averill. Vosmik was 22, 8 years younger than Averill. It is especially odd because the Indian starting rotation was exclusively right-handed.

League Park was, of course, asymmetrical. They moved to Municipal Stadium at the end of July, 1932, where the centerfield fence was 470 feet away (to Vic Wertz' dismay). Somehow, it seem unlikely that the park played too much of a role in it.

Anyways, getting back to the point, I think the anecdotal evidence on this thread is pretty strong that the numbers aren't lying on Nettles. His case doesn't depend on whether he was +25 or +35 in 1971--just that he was a historically great fielder at the position for a decade.

After 35 years of watching baseball, Nettles is still the first guy that comes to mind when I think about fielding excellence at third base. He was simply amazing.

Nettles was one of my favorites. He would make these diving catches of line drives where he would be completely horizontal in mid-air when he caught the ball. I used to practice catches like that in my front yard. This 'practice' usually involved waiting a bit so that I wouldn't catch up to the ball so that I would have to dive :-).

Nettles in the 70s was the best defensive 3B I have ever seen, better than Brooks (I did not see Brooks play in the 60s mind you), better than Aurelio Rodriguez (a SS playing 3B) and better than anyone since the 70s. (Rolen reminds me a lot of Buddy Bell defensively- I don't know why they really don't "look" similar, but my impression is that the overall caliber of their Dee is similar- Rolen when healthy being better - I saw both Nettles and Bell play together close in time- Bell was very good, but Nettles even aging and having lost a step was better)

Nettles lost several gold gloves he shoudl have won to Broks and he even lost one to Aurelio Rodriguez (who WAS very good, don't get me wrong).

I'm a met fan, I love David Wright, David Wright is more athletic than Nettles- but he will never ever be remotely the defensive force that Nettles was- Nettles' reaction times were unreal- that horizontal dive of his (which I used to try to imitate)- was not a belated dive in response to a ball he could have reached easier if he moved earlier, or a timed play like Winfields outfield fence leaps- Nettles' horizontal dives were 2B saving (or 1b saving in the 3b/ss hole) plays that no other human being (Brooks Robinson included) could make.

He was still a serviceable defensive 3B at 40- when he was a shadow of his earlier self.

BTW, I have very vivid memories of Graig Nettles with the Yankees in the mid-70's (which was a mere ~30 years ago), including those horizontal stabs of line drives. He was an amazing fielder, and made some spectacular plays in the post-season for the Yankees' during the mid-70's.

Saw both. thought Nettles was better-- and so was Clete Boyer. Now some of that was because Brooks was so darn slow that I just didn't think he had the phenomenal range of Boyer and Nettles. Unlike Boyer who played a very good shortstop, moving Brooks to short just didn't seem possible. No knock on Brooks--Nettles and Boyer were that good. Nobody in today's game compares to any of them

Nettle's reaction times were unreal. You couldn't go more than few games before you saw him make a play that no one else could, period.

The funny thing was that Nettles was slow, he really couldn't run worth a damn- his first two steps were quick but then...but he could move his body from side to side faster than anyone I ever saw then or now. I have never seen anyone else catch as many balls on FULL extension while diving for a ball- not like Carney Lansford who dive for everything because he never moved his feet- you'd see a frozen rope either down the line or in the hole, and Nettles would be a horizontal blur.

For awhile there I never saw a bunt or an infield hit in his direction, he would run in at full speed (for him) barehand and throw and the throw seemed to always be right on the money- no matter what angle he threw from - you see 3Bs try that play all the time- most don't make them with any great regularity, and many you hope just don't throw the ball away- for a few years their it always seemed that Nettles put the ball in the 1b's glove on that play.

The funny thing was that Nettles was slow, he really couldn't run worth a damn- his first two steps were quick but then...but he could move his body from side to side faster than anyone I ever saw then or now. I have never seen anyone else catch as many balls on FULL extension while diving for a ball- not like Carney Lansford who dive for everything because he never moved his feet- you'd see a frozen rope either down the line or in the hole, and Nettles would be a horizontal blur.

sorry I'm late to the party here--Nettles was always one of my favorite players

another aspect of Nettles' fielding was that he liked to play farther off the line than other 3rd basemen (this was discussed in some detail in an article by Tom Boswell many years ago)

since (Nettles said) far fewer balls are hit to a 3rd baseman's right, it made sense to shade to the left and cut off grounders through the hole--this also allowed the SS to cheat more towards second (this could explain Nettles ungodly assist totals)

also--he ALWAYS knew when to dive and when not to; I don't remember him ever diving for a ball that he could have fielded on his feet; unbelievable reflexes

(and, by the way, Landsford didn't so much dive as fall down in place)

DanR, I agree. The real nub of the thing is that most people think of Brooks Robinson as an open-and-shut case. Once one adjusts for the net DPs, depending on your replacement level, Brooks is probably closer to the line than we commonly think.

Using Dan R.'s spreadsheet, here are the top performers at each position:

1B:
1st: 1914 Dot Miller and 1999 Darrin Erstad tied
2nd: 1920 Charlie Grimm and 1959 Vic Power tied
Tino Martinez is in a group of 6 tied for 3rd.

2B:
1st: 1908 Nap Lajoie and 1927 Frankie Frisch tied
2nd: 1933 Hughie Critz
(those are the only players above 3 FWAA2)
3rd: 1910 Eddie Collins
Maz has a 4th, tie for 5th, and is in 6th.

3B:
1st: 1971 Graig Nettles
2nd: 1899 Jimmy Collins
3rd: 1916 Ossie Vitt
Lave Cross has the 4th and 5th highest and Lee Tannehill, Terry Pendleton, and Tim Wallach are the other players over 3 FWAA2.

SS:
1st: 1906 Terry Turner
2nd: 1896 Hughie Jennings
3rd: 1917 Art Fletcher
Rabbit Maranville and Art Fletcher again are the only players with seasons over 3 FWAA2.

LF/RF
1st: 1932 Joe Vosmik
Only player above 3 FWAA2 in a year.
2nd: 1980 Willie Wilson and 1985 Jesse Barfield
3rd: 1898 Hugh Duffy, 1963 Johnny Callison, and 1977 Dave Parker

CF
1st: 1984 Kirby Puckett
2nd: 1905 Fielder Jones, 1909 Tris Speaker, and 1956 Jimmy Piersall
3rd: 1912, 1914, 1919 Tris Speaker
No player above 3 FWAA2 in any year.

Miller, Erstad, Power, and Wilson are definitely there because they played time at other positions that is being credited to them as first basemen or corner outfielders. Those are mistakes in my system that need to be fixed. The problem is that I don't have a spreadsheet that has defensive games/innings at each position for every player-season in history, so I just use the Win Shares position listings. This works fine most of the time but does cause problems with certain multi-position players--Gene Tenace and Willie Wilson are definitely overrated, and Paul Molitor is underrated. I should make clear to all voters using my system that if there is a player who split time within a season at two positions of vastly differing difficulty, they should definitely check with me so I can give them a corrected FWAA and Rep value. It's not hard to calculate them by hand for individual cases.

re: post 73, if we're looking for anecdotes, we could get about 24 bajillion of them on Brooks. He was amazing.

About positioning, Nettles likely did get more balls to the shortstop side, and possibly got to as many overall; but if anything, this ought to slide the numerical totals Robinsons's way, since he likely saved more doubles down the line (instead of singles) than any mortal. Which most metrics do not account for.

re: post 73, if we're looking for anecdotes, we could get about 24 bajillion of them on Brooks. He was amazing.

yeah, yeah, I heard them growing up, I heard him be called a human vacuum cleaner, I heard relentless references to his WS performances-

One year late in the 70s This Week In Baseball (I Think) ran a ten minute highlight reel on Brooks (this was probably 1977)
I was impressed, but no play (and they showed dozens) struck me as one that Nettles was unlikely to have made, and I distinctly recall thinking that Robinson's reflexes (or reaction time, or anticipation or whatever you want to call it) was not as good as Nettles-

I may have been biased, but I tend to think the pro Brooks camp suffers from an equal bias.
I also think that a defensive clone of Ozzie Smith would get screwed if his carer spanned 1988-2005

Brooks may have been good longer than Nettles, but I pretty much cannot be persuaded that at his best Brooks was better than Nettles at his.

re: replacement level for 3B in Nettles' and Robinson's day:

I have no empirical data for this, but it SEEMS to me that more managers in the 60s TRIED to put players at 3B who were hitters but didn't play 3B real well; hence, the offensive median/replacement level might be high then, but the typical defender to which they could be compared would be different.

Dick Allen played some thrid, not real well. Tommy Harper tried to pay some third. Killebrew, oh my. Geroge Scott. Bob Bailey. Richie Hebner. Dave Kingman! Fumble Fingers Foy. Jim Ray please-don't-hi-it-to-me Hart. Steve 'rag arm' Gravey. There might be lots more.

Just my impression.

Bob Aspromonte, Deron Johnson, Joe Torre, could include Tony Perez in this too.

>CF
1st: 1984 Kirby Puckett

Thank you.

I combined WS and BP a few years back when checking out CF defense. Best seasons I got were (in no order)

grissmom 1994 (rate only; strike year)
puckett 1984
flood 1962
an jones 1998 & 1999
mays 1954 (not even counting 'the Catch')
speaker 1919

There's no doubt Puckett's fielding was stellar his rookie year--DRA sees it the same way. The issue is that as soon as he started hitting in 1986, his defense went way south and stayed there.

continuing from Dan #54
To summarize: I see Nettles and Brooks as identical twins, and BP sees them as distant cousins, for the following reasons:
1. Net double plays: 4 wins.
2. Fielding quality: 4 wins.
3. Relative strength of 3B in the 1960s and 1970s (BP thinks 3B was much tougher in the 60's than in the 70's while I don't): 7 wins.
4. BP's lower replacement level: 3 wins.
5. DH adjustment: 4 wins.
6. Odds and ends (run estimation, strikeouts, intentional walks, sac flies, non-SB baserunning): 2.5 wins.
Total explained variance: 24.5 wins.
---
1. I agree with your use of DP; BP deos not, altho I'll note WS does handle this.
2. fielding quality: Dan noted that WS ses Nettles as even betetr. I wonder if Robinson's defensive stats are not skewed by his team; he played on one of the best range teams of all-time; Belanger, Blair, D Johnson/Grich at 2B... WS TRIES to compensate for this, but since there is Cap on overall DWS, I suspect Brooks is BETTER than his WS states. When you add in his rep, I feel that he ought to be rated higher than Bettles on defense.
3. relative league strength - I noted the 3B-poor-fielding issue before.

Overall, I've moved Brooksie a little lower in my overall estimation than I had him, and bumped Nettles a bit up; nice wrok by DanR. But I still see a clear gap between them. Brook's post-season record is good, Nettle's is not, and they each have about 1/4 of a season of ABs there; if you count those games as 4X impt as nornal, that is a full 'year' of a bonus 100 OPS for Robinson.

TomH, are you saying that putting good fielders/bad hitters at 3B lowers the overall positional replacement level (adding offense and defense together), or just changes the shape of it? If it's the former, why wouldn't that show up in the worst-regulars averages?

Brooks IS rated higher on defense--the question is just the SIZE of the gap, and how that compares to the size of the gap between their offense.

The defensive reputation and postseason issues are in the eye of the beholder, I don't really have anything to add there.

I'm saying those systems that look at avg/median/replacement 3B level from an offense-standpoint only would be missing a piece of data.

The truth is that I don't think any of us know how BP calculates FRAR-FRAA, do we?

Mike Green
My intuition was entirely wrong. For what it's worth, I doubt that there are too many people who saw both Nettles and Robinson in the early 70s who thought that Nettles was the better fielder at the time. Both FRAA and Fielding Win Shares suggest that he was, and we don't have zone rating for the period.

Not many people outside Cleveland saw a lot of Nettles. The team was weak and buried in a strong division, essentially even with Washington-Milwaukee in 5th-6th place.

Minnesota was division champion in 1969 but young Nettles played more LF (54g) than 3B (21g). Five people played at least 39 games in left and five played at least 20 at third. The team made 90 substitions and switches in left; 100 at third (total 262 player-games in 162 team-games).
1969 Minnesota Twins at baseball-reference

These two paragraphs on young Graig Nettles are written by someone who was then a young observer, who did not see Brooks until the same time (#63). Especially in what I have underlined this is what people said of Brooks in the late sixties (if not earlier) when he was also called maybe the best third basemen ever.

JWPF13 #67
The funny thing was that <u>Nettles was slow, he really couldn't run worth a damn- his first two steps were quick but then</u> ...but he could move his body from side to side faster than anyone I ever saw then or now. I have never seen anyone else catch as many balls on FULL extension while diving for a ball- not like Carney Lansford who dive for everything because he never moved his feet- you'd see a frozen rope either down the line or in the hole, and Nettles would be a horizontal blur.

For awhile there I never saw a bunt or an infield hit in his direction, <u>he would run in at full speed (for him) barehand and throw and the throw seemed to always be right on the money- no matter what angle he threw from</u> - you see 3Bs try that play all the time- most don't make them with any great regularity, and many you hope just don't throw the ball away- for a few years their it always seemed that Nettles put the ball in the 1b's glove on that play.

The second underlined quotation also fits my understanding of the Jimmy Collins sensation: swoop and pounce, control and release, all so quickly. Then for a generation or three he was called maybe the best 3Bman in history.


--
TomH
I have no empirical data for this, but it SEEMS to me that more managers in the 60s TRIED to put players at 3B who were hitters but didn't play 3B real well

Pedro Guerrero and Bobby Bonilla great batters.
How good was Howard Johnson?
Those of you who were paying attention can probably name others.


82. sunnyday2 Posted: October 11, 2007 at 06:18 PM (#2572171)
>CF
>1st: 1984 Kirby Puckett

Thank you

Check out Doctor J's rebounding in his rookie season.

One thing I see in those lists is that the greatest seasons of all-time at each position tend to be the single best seasons of numerous players. (Perhaps 8-10 players in the top ten seasons?) Except at CF where Tris Speaker shows up again and again.

but it SEEMS to me that more managers in the 60s TRIED to put players at 3B who were hitters but didn't play 3B real well

The Cardinal teams I grew up with - the '67-'68 pennant winners - had Mike Shannon at 3B. Shannon is a "tweener" case - he was a corner outfielder whose bat wasn't very impressive for a corner outfielder, but would look better at 3B.

It's not that easy to tell what his true offensive level was. For the two years prior to the move, his OPS+ was 79 and 120. In his first year at 3B, it was 93 - and the Cardinals won the pennant. In his second year, it was 113. Not long after that, his career was truncated by illness. He played 3B like an out-of-position outfielder, I think. In fact, I don't really know what his defense was like - the Cardinal experience for me was a radio experience, not TV or live. He was flanked by an amazing defensive SS in Maxvill.

To make the move happen, the Cardinals traded a 28 year old 3B with a 92 OPS+ (Charley Smith) for a famous 31 year old OF (Roger Maris) coming off a 101 OPS+ season. Maris did rebound a little for the next two years, but it's very hard to see that move as being any kind of key to the Cardinals' success. It wasn't a disaster, it wasn't great, and it's hard to see how leaving Shannon in RF and Smith at 3B would have been much different than what happened.

Now, Ray Sadecki for Orlando Cepeda ...

It looks like appreciation for Nettles, whom I strongly support, is growing, so I don't want to muddy the waters around his case, but I do want to ask, in a way that helps define Nettles' case more clearly: how does Nettles compare to Buddy Bell?

My system likes them both, but Nettles is garnering more support, though still not as much as he merits. So for those who support Nettles over Bell, what's the knock on Bell?

What makes Nettles the top modern third-base candidate among the eligibles?

Following up on 74, Nettles' innovative positioning influenced other players. By the late 70s, 3B's generally played further from the 3B line than before.

Bell didn't hit like Nettles (BWAA of 12 for Bell, 18.3 for Nettles) though the glove was comparable and held out longer. Bell was also a terrible baserunner according to Dan R (-2.3 wins career).

Cey is a better hitter than Nettles but was much closer to average with the glove.

> What makes Nettles the top modern third-base candidate among the eligibles?

The fact that we've elected Schmidt, Brett, Boggs and Darrell Evans. After them, he's the best from his era. It is a remarkably good era for third basemen and a poor era for centerfielders.

Why does WAR see such a difference in batting value between Bell and Nettles?

EQA sees them as about the same (Bell is .270/.275 W1/W2; Nettles .273/.275). Nettles is about 1 win above average ahead, but that is purely on longevity.

Bell does seem to have hit into a lot more GiDP; if that isn't just a diference in opportunities, I'd guess there's a few lost wins there in his batting record.

DL from MN, remember that if you want to make apples-to-apples comparisons using BWAA, you have to adjust for the DH (since the DH adjustment is applied to the replacement level in my system). Failing to do so makes Nettles' offnsive advantage look bigger than it is (this is why Bell's Replc column is higher than Nettles's despite less playing time and a stray season in the outfield).

Charts:

The following numbers are all standard deviation-adjusted. SFrac is the percentage of the season played (compared to a player with league average PA/G in 162 games). BWAA is batting wins above average, BRWA is baserunning wins above average, FWAA is fielding wins above average, Replc is wins above average a replacement player at the same position would have accumulated in the same playing time, and WARP is the first three minus the fourth (wins above replacement). Note that Replc drops by 0.6 wins in 1973 to account for the DH. 1972 and 1981 are adjusted to 162 games. aTTL is career totals excluding sub-replacement seasons. Sorry for the goofy formatting, but the PRE tag doesn't seem to be able to handle consecutive whitespaces anymore.


Nettles



3-year peak: 18.8
7-year prime: 38.6
Career: 62.2

Bell



3-year peak: 18.9
7-year prime: 36.7
Career: 57.2

Chris, you're right that Nettles and Bell (and, by extension, Brooks Robinson) are strikingly similar players. Long careers, terrific gloves, above-average bats, same position, same league, same era. In terms of their profiles, Nettles has two subtle offensive advantages on Bell: double plays and baserunning. First, Bell hit into 32 more double plays than a league average player would have given his opportunities over his career, while Nettles hit into 21 fewer. That's a three-win gap. Second, while Nettles never should have been allowed to steal (32/36 career SB/CS), Bell practically ran his teams into the ground (55/79), representing a further two wins' difference. Those two factors, plus two points of OPS+, show Nettles as a meaningfully superior offensive player. On defense, Bell didn't have a historically superlative peak like Nettles did, but he remained a Gold Glove-level fielder well into the 1980's, enabling him to actually accumulate more career fielding value above average than Nettles had.

In terms of their relative ranking, Nettles simply had more good years in him than Bell did. Their top five seasons (1971 and 1975-8 for Nettles, 1979-82 and 84 for Bell) are virtually identical, but after that, Nettles opens up a lead. Nettles played at a consistent All-Star level in the first half of the 1970s, while Bell's pre-peak years in the second half of the decade were above average but not as strong as Nettles had been five years earlier. Nettles's prime was a little longer and the back half of it was a little higher, and that's enough to distinguish them clearly for me. If you regress the 1981 strike season, the gap opens up a little more, since it was Bell's best season and Nettles's 10th-best. Nonetheless, Bell is a serious candidate--he's just on the wrong side of the borderline.

Thanks for the more detailed breakdown. It's the great difficulty of ranking near (and below) the in-out line: finding the difference between players like Nettles and Bell in GiDP and baserunning!

That stuff can make a big difference. The gap between Willie Davis and Joe Torre in combined DP avoidance and baserunning is equivalent to 15 points of career OPS+.

If you are mooting about modern third basemen on the borderline, don't forget about Robin Ventura (who will be coming up soon). He's in the same pack with Nettles, Bando, Bell, Cey...the only advantage we have with him is that we have good zone rating data. Dial has him as the top-rated AL third baseman of the last 20 years.

Personally, I prefer the starting pitchers though...

Ventura is indeed a similar profile to Bell and Nettles, but didn't last as long as either of them. That's a big difference when you're talking about players without much of a peak.

"I looked it up. Vosmik made more plays in the same number of games (153) as the Indian centerfielder Earl Averill. Vosmik was 22, 8 years younger than Averill. It is especially odd because the Indian starting rotation was exclusively right-handed."

Most fly balls go the other way. We just remember the HR, which are pulled.