|
|
|
Hall of Merit— A Look at Baseball's All-Time Best
Sunday, February 11, 2007
|
Bookmarks
You must be logged in to view your Bookmarks.
Hot Topics
2014 Hall of Merit Ballot Discussion (86 - 8:02pm, May 23)Last: Ivan Grushenko of Hong KongMost Meritorious Player: 1982 Discussion (48 - 9:05pm, May 19)Last: Mr. CMost Meritorious Player: 1981 Results (11 - 3:30pm, May 16)Last: DL from MNMost Meritorious Player: 1981 Discussion (72 - 10:54am, May 13)Last: bjhankeMost Meritorious Player: 1981 Ballot (47 - 9:51am, May 06)Last: DL from MNMost Meritorious Player: 1979 Discussion (115 - 2:09pm, Apr 19)Last: DL from MNMost Meritorious Player: 1980 Results (10 - 12:23pm, Apr 15)Last: DL from MNGeorge Scales (70 - 10:52am, Apr 10)Last: Ivan Grushenko of Hong KongLarry Doby (94 - 12:28am, Apr 10)Last: KJOKMost Meritorious Player: 1980 Ballot (21 - 11:03pm, Apr 09)Last: DL from MNMost Meritorious Player: 1980 Discussion (45 - 1:04am, Apr 09)Last: lieiamMost Meritorious Player: 1979 Results (12 - 4:30pm, Mar 14)Last: TomHMost Meritorious Player: 1979 Ballot (35 - 4:06pm, Mar 12)Last: TomHNew Eligibles Year by Year (956 - 3:11pm, Mar 12)Last: Chris FluitMike Mussina (46 - 8:36am, Mar 12)Last: Rants Mulliniks (formerly Cold Prosimian)
|
|
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
Are there any stats at all that suggest that he wasn't an excellent defender? If you're going to do a realistic, as opposed to subjective, evaluation then I think it is pretty much a given. In any event a 3B contribution to team wins is about 5/6 hitting an 1/6 defense, and there is no question about his hitting.
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
Buddy Bell.......10.2, 9.8, 9.4, 9.0, 8.4...(46.8)..109.9
Tony Perez.......10.2, 10.0, 9.6, 8.7, 8.1...(46.6)..109.5
Graig Nettles....10.7, 10.2, 8.9, 8.4, 8.2...(46.4)..107.2
Bob Elliott......11.4, 10.0, 9.5, 9.2, 7.8...(47.9)...95.6
Ron Cey..........10.5, 9.6, 9.2, 9.1, 9.0...(47.4)...96.6
Sal Bando.........9.3, 8.5, 8.4, 7.9, 7.5...(41.6)...83.2
Evans is clearly the best in terms of longevity, and has the best single season by a large margin. In terms of best five seasons it is all Ken Boyer and he has the best 2nd, 3rd, 4th, and 5th best career seasons of anybody. Buddy Bell, Nettles, and Tony Perez have alost equal credentials significantly better than Cey and Elliott. Sal Bando would appear to be totally out of place in this group. Since Boyer is in it would appear that Evans has an excellent case, with Bell, Nettles, and Perez all being finge guys.
I suppose not. I don't think that there's any question that 1973 was his only great offensive year as a 3B. In which case his peak is unimpressive unless he was a truly spectacular defensive player. So we're back to defense again.
1972: Not in top 20
1973: 5th, behind Morgan, Stargell, Rose, and Perez, and tied with Bonds
1974: 6th, behind Schmidt, Morgan, Bench, Wynn, and Stargell
1975: 4th, behind Morgan, Rose, and Bench and tied with Schmidt, Simmons, and Luzinski
1978: 9th, behind Parker, Simmons, Jack Clark, Foster, Winfield, Rose, Burroughs, and Luzinski
1979: 18th
1980: 6th, behind Schmidt, Carter, Dawson, Keith Hernandez, and Dale Murphy.
1981: 18th
1982: Not in top 20
Do people really believe that Darrell Evans was as good as Mike Schmidt in 1975? Really? As has been noted many times, the list that Mulder & scully presents is based on the notion that Darrell Evans was absolutely brilliant defensively. Brooks Robinson level defense, if not better. Maybe the greatest defensive third baseman of all time.
It just doesn't pass the smell test.
Buddy Bell thread
I agree that his peak is unimpressive. That's not his calling card. He's a prime/career guy, BUT one who has just enough peak to be ahead of Nettles or Traynor or someone like that.
Very good question. I've been resting my argument on the perceptions of observers at the time. So I decided to investigate the numbers, for 1975 in particular.
First, the Braves defense as a whole was terrible. Someone had to get the chances on the team. Their defensive efficiency was awful, despite having Niekro on the team (don't knuckleballers usually have a low BABIP?). Evans was part of a bad defensive group, of who many others also had above average defensive numbers.
Second, one of the things that makes Evans look so good is his putouts - in fact, they are much higher than the rest of the league. Its also what separates him from Schmidt defensively. Personally, I am less likely to see an unusually large putout total as part of an innate ability than a large assist total.
Third, the year after he was moved to first, Jerry Royster played the majority of the games for third base for the Braves. As far as I know, Royster isn't thought to be Brooks Robinson out there. And sure enough, Royster led the league in putouts at third base (with nearly as many as Evans), and had an extremely high range factor, second only to Schmidt as far as I can tell. This strongly suggests that there was something about those Braves teams that, despite being awful teams defensively as a whole, made their thirdbasemen look good.
At the same time, his career as a third baseman was not all that long. He played almost a thousand fewer games at third than Nettles, and 400 fewer than Traynor. Evans spent about as much time at third as Bill Madlock did, or as Scott Rolen already has. He hit more than half of his career homers at someplace other than third base, for what that's worth. Three of his top four OPS+ years came after he left third base for good, at the age of 35.
It's the years at 1B/DH that make him a career candidate. I frankly have no idea what to make of that.
Evans is fun to talk about, isn't he? His career had more dipsy doodles to it than any other player of comparable worth -- which is what makes it so hard to determine who is of comparable worth.
I looked at the same thing. Looking at BPro, in 1976 Royster had a RATE of 112 at 3B. For his career, his RATE was 105 and his above-average (>100) seasons were 1976 (ATL), 1979 (ATL), 1980 (ATL), 1983 (ATL), and 1986 (SD). He mixed in a bunch of average/below-average seasons around those (some in ATL, though).
In 1977, the Braves regular third baseman was Junior Moore (first time I bet he's been mentioned in a HOM debate). His RATE was only 100, but over the rest of his career, his highest season rate was 87 (although that career consisted of 40 games for the White Sox). Then, in 1978, Bob Horner came around and put up a RATE of 113.
So Braves' 3B in general seem to do well by BPro. Given the records the Braves were putting up and their team defense in general, I do find it a little hard to believe that they had some savant-like skill at finding above-average defensive third basemen.
Phil Niekro was such a unique pitcher (with the knuckleball) and pitched SO MANY innings, I wonder if he's somehow partially responsible for this (not that I have the slightest idea how).
Alternately - just thinking out loud here - the late/70s Braves had very few starts by lefties (until 1978). I know James makes an adjustment for LHP that basically says more LHP -> more RHB -> more expected chances for 3B. This would lead one to expect fewer chances by Evans/Braves 3Bs, which would make the chances they did make that much more impressive. But maybe they either didn't face a disproportionate # of LHB or the LHB they did face were hitting the ball to 3B a lot (which could tie back to the Niekro theory).
Just some thoughts.
When Evans played in San Francisco, he did not have an unusual number of putouts. This means that if he had some weird ability to catch the ball that other thirdbasemen didn't, it didn't extend to his days in San Francisco.
My question for the group is, do win shares and WARP and RATE incorporate putouts in their calculations? Because if they do, it may be one of the key illusions.
Then again, Win Shares are one of the things that make Evans look so terrific, especially in 1975, so who knows?
I would concur that Evans' high putout totals are likely a Niekro effect, possibly combined with a tendency for players to swing for the fences at The Launching Pad. Could Niekro throwing a ton of innings raise activity at third because right-handed hitters are more likely to pull the knuckleball? It'd be interesting to compare 3B assist rates in games pitched by Niekro to games pitched by the rest of the Atlanta staff.
Given that Evans' managers in 1975 were the immortal Clyde King and Connie Ryan, I'm not going to place too much faith in their good judgment about Evans' defensive merits. I think we can conclude that, even if WS and WARP are overrating Evans' defense somewhat (and I think the jury is still out on that), he was underrated at the time. Too much attention was paid to his errors, and not enough paid to his double plays.
Here's another thought that popped into my head. James, at least (I'm much less familiar w/ BPro's formulae, but I tend to assume they're similar), uses total assists as a proxy for a pitching staff's groundball tendency. Could Evans/Braves 3B high PO total indicate a lot of infield pop-ups, which maybe don't correlate as well - maybe because of the knuckleball - to a flyball tendency of pitchers. So James/BPro are over-estimating the extent to which the Braves had a flyball staff and, hence, underestimating their expectation of Evans/Braves 3B assists, leading them to overvalue the assists that he did get.
Does that make any sense?
That's a bit of an overstatement of what either system suggests. For his career, WARP and WS agree that Darrell Evans was not one of the greatest defensive third basemen in baseball history, and no one is claiming that here. The systems are suggesting that, for three years (at a time of career where we would expect to find peak fielding value), Evans was an excellent defensive third baseman. Lots of players have a few seasons, defensively, in which they were much better than they later became.
Now, because we are interested in peak value, we have to place more weight on the seasonal findings of the comprehenseive metrics than they can bear and we should exercise caution, but we shouldn't exaggerate the apparent meaning of those seasonal findings. Here's the top 10 seasons, by FRAA, of some notable fielders at the hot corner:
B. Robinson: 32, 24, 22, 22, 21, 18, 17, 16, 12, 12 (196)
G. Nettles: 33, 21, 21, 20, 18, 14, 11, 11, 9, 6 (164)
M. Schmidt: 23, 20, 18, 17, 16, 15, 15, 14, 14, 14 (166)
K. Boyer: 21, 18, 17, 16, 15, 14, 11, 10, 4, 2 (128)
R. Santo: 25, 21, 18, 16, 15, 14, 13, 9, 1, 1 (133)
B. Bell: 20, 19, 19, 18, 18, 18, 18, 12, 11, 11 (164)
D. Evans: 23, 20, 18, 12, 8, 7, 4, 2, -6, -6 (82)
FRAA suggests that Evans had a few years that were characteristic of an excellent defensive third baseman, but he had no historically great (e.g. 25+ FRAA) seasons, and he was while still generally above average, not outstanding. He was notably weaker than all of his contemporaries and near contemporaries who had reputations as strong defensive third basemen.
We have come up with some reasons why Evans' value might be somewhat overestimated in his "fielding peak" in Atlanta, though it is also possible that his _contextual value_ is fairly accurately reflected in the metrics even though that value was not in line with his _actual skill_.
I think we can make some progress in deciding whether or not that 3-year fielding peak is accurately represented by WARP and WS by further consideration of the raw data and relevant contextual factors, but I don't believe the findings of WARP1, at least, in any way suggest that Darrell Evans was one of the greatest defensive third basemen.
Not for his whole career, no. But, as you said:
Win Shares sees Evans during these years as pretty much equivalent to Brooks Robinson, Graig Nettles, and Mike Schmidt at their best.
And the FRAA numbers show Evans's best seasons as equivalent to Schmidt's and on par to most of Nettles's and Robinson's at their best, excluding the very best season of each. If that doesn't make Evans one of the best in history at his peak, it's pretty close. And I don't think anyone ought to have so much confidence in FRAA or Win Shares to accept those numbers without a great deal of regression.
I would like to hear their answer. But EYE would be willing to bet that if you phrased the question such as "rank your opinion of FRAA and defensive win shares as it compares to TPR's defense, general public view of defense and gold glove awards", that they would put FRAA and DWS as worth MORE than others. No, it's not as good as pbp data. But it's what we have for 1973.
Personally, I'll take about 1/4th human estimate and 3/4ths metrics on Evans, et al.
For me, it's about understanding the level of uncertainty. As an example, Ralph Kiner was out of my PHoM for a number of years, the major reason being that WS and FRAA called him a horrific fielder in most of his best hitting years, which led to a significant dampening of his peak. That said, he had years of +11 FRAA sprinkled between years of double digits below average.
To test the "uncertainty", I looked at Kiner, and imagined him as always average defensively (I know, a stretch). He sailed into my personal hall. Even as a - 5 or 7 most years, he was fairly comfortably in. What I was forced to ask myself is, "Do I believe that Ralph Kiner just so happened to be a horrific defender in years W, Y and Z and an excellent one in year X?, and am I willing to be so certain of the ability of James and Davenport to accurately capture his fielding that I should not vote for him, ever?"
Dan Rosenheck, if you're reading this - as one of Evans' chief backers, can you re-run your system for Evans, but imagine him as an average fielding 3B from '71-'75? I'd love to know if he still shows up as a HoMer. It just "feels funny" if one of the main drivers of his "Merit" is his defense at third, in seasons prior to getting moved off the position due to the perception that he was not very good.
If he came up form the minors, maybe they decided they were better off with the good fielding rookie at 3B and the better hitting Evans than vice versa. Or if they acquired him in a trade, maybe the decided that it was better for themto trade for a decent fielding 3B and move Evans than to get a 1b and keep Evans at 3B. Once a 1B, Evans may have just deteriorated defensively or concentrated more on his offensive and let his defense slide.
If either of these are true, it would mean that Evans' wasn't moved off of 3B because he was poor but because Royter was pretty good (as RATE shows) and it was the best way to get both of them into the team. That doesn't mean that evans was a very good defensive 3B, but it doesn't meant that he was moved because he was awful. Other factors come into play on position changes than merely how good a certain player is at a position. Plus, if we allow that the Braves misjudged his fielding value, the scenarios above become even more likely.
And I'm not saying you shouldn't use them at all (although I myself would lean in that direction). I'm saying one shouldn't just accept them at face value.
Evans never had the reputation as a good-fielding third baseman when he played. Ever.
He wasn't even considered a primo prospect. The Braves took him as a Rule V draftee from the A's in December of 1968, at the age of 21. Evans didn't make the Braves' major league roster the following spring, and under the rules of the draft, they offered him back to the A's, but the A's said, no thanks, you can keep him.
The Braves gave Evans late-season call-up trials in both '69 and '70, but he didn't impress enough to make the team in '71, at the age of 24 -- past the point when most guys are considered A-list prospects. The Braves' top prospects at that point were Earl Williams, Ralph Garr, and Dusty Baker, not Evans.
Williams won the starting 3B job in the spring of '71, alternating with the veteran Clete Boyer. When Boyer was fired in late May in a dispute with GM Paul Richards, Evans was called up. He and Williams alternated at 3B until late June, when Williams was shifted to catcher, and Evans finally won the starting job.
He held it through 1972, but in the spring of '73, the Braves gave the regular 3B job to rookie Rod Gilbreath, and moved Evans to 1B -- hardly an indication that they considered him a strong defensive third baseman. Gilbreath only lasted a month, and Evans was moved back to 3B, although he was replaced in lated innings for defense and slid back over to 1B in several more games that year.
But Evans exploded with a great season with the bat in '73, and with the status as a power-hitting, OK-fielding third baseman, he held the starting job through 1975. But in that '75 season, he encountered a defensive slump, committing 36 errors, and it was this that prompted the Braves to move him back to first base for 1976.
Royster was guy they'd acquired in a trade with the Dodgers. He was a rookie in 1976, but hadn't been a regular third baseman in the minors; he was an infield-outfield supersub guy. He was anything but an established major league third base presence; moving Evans and replacing him with Royster speaks volumes as to how Evans' defense was perceived by his management.
One can take issue with the Braves' general decision-making process through this period, for sure. But it's important to understand that anyone who had said in 1975 that Darrell Evans had played a defensive third base from 1971 through 1975 that was excellent, as good as any other third baseman in baseball, would have absolutely been laughed out of the room. That ought to be factored into the assessment of just what Evans' defensive performance actually was.
This is a fair point, but I wonder - and this is probably a distinction without a difference - are there any current players who conventional wisdom says are <u>bad</u> fielders (or below-average, mediocre, something below "okay") that WS and BPro say are <u>excellent</u> fielders and, if so, what do more advanced metrics - UZR, Dial, Pinto - say about who's right on these guys. I guess I can see why somebody who's really bad could <u>look</u> good, but I think it's harder for a guy to really be good and still look <u>bad</u>. But I'm probably wrong - I'm just curious who a modern analog to Evans might be in terms of defensive ability/perception.
Also, Evans fielding percentage went from mid .950s to .938 with Williams his main target.
Anyone, Steve, know the reputation for Evans' accuracy in the early 70s?
But why must one "trust" either to the exclusion of the other? If you aren't totally ignoring subjective evaluations of fielding, and those evaluations are strongly at odds with FRAA/WS, then there would seem to be no choice left except to split the difference between the divergent inputs.
I watched him play only when the Giants played the Braves in that period -- this was pre-TBS! I watched him a gazillion times from 1976-83.
From what I saw, his arm was nothing out of the ordinary for a third baseman's, either in terms of strength or good/bad accuracy. When he was with the Giants, as I mentioned earlier in this thread, he had a tendency to fall into defensive slumps, when he'd commit a rash of errors in a short period. Those errors, at least when he was with the Giants, were at least as likely to be fielding errors as throwing; he could kick the ball a little.
It's a good observation that the Braves didn't exactly have John Olerud over there scooping throws at first base in that period. I'm sure that cost Evans some errors, but from what I saw in his later career, I'm very confident in saying that many of those errors he made with the Braves were fielding errors.
A point that should be made: Atlanta Fulton County Stadium had a strong reputation in those years as one of the worst fields in baseball, especially late in the season when the Falcons would be playing home games on Sunday. Without looking it up, I strongly suspect it was a fairly high-error ballpark in general.
That was often mentioned when Rafael Ramirez was having his fielding woes during the '80s.
Also pertinent to the park factor discussion: the power alleys were lengthened in 1974, making HR's somewhat less common.
ballparks.com notes various changes in dimensions at AFCS; of interest is that in 1974 the power alleys in Atlanta were increased from 375 to 385 ft, perhaps in reaction to the HR barrage of 1973.
Also a quote from the write-up:
"For many years, Atlanta-Fulton County Stadium was also known as the home of baseball’s worst playing surface. The ballpark didn’t even have a full-time groundskeeper until 1989. Before that, it had been tended by a city streets crew. "
I agree with DCW and Max above. When raw fielding stats conflict with the evaluations of comtemporaries, and the raw fielding stats are in themselves the basis for a HoM case, that seems like very shaky ground.
- Jeter hasn't been moved off of the most crucial defensive position on the field, despite his team's acquisition of a Gold Glove shortstop. Evans was repeatedly moved off and on a secondarily important defensive position, despite his team's lack of a substantial replacement (Gilbreath, Royster, O'Malley).
- Jeter's team has consistently won 100 or so games per year throughout his tenure. Evans's teams generally struggled to reach .500.
So if indeed both players' managements have been egregiously mistaken in evaluating their defensive performances, it is the case that the Yankees have won a ton of games despite having a terrible defensive shortstop, and the Braves/Giants were usually losers, despite receiving great defense from their third baseman (when they weren't too stupid to be playing him at first base or left field).
In both cases, it's my inclination to conclude that the truth probably rests somewhere in the middle.
To follow up on that point - I agree that the answer is most likely somewhere in the middle. Evans probably wasn't a B. Robinson/Rolen/Nettles quality defender for those five years, while at the same time there are contributing factors (such as the field at AFC) that cause one to question whether he was really as bad as the typical player who gets moved leftward down the spectrum (and about whom quotes like the Sporting News ones above are written...).
Which is why I'm asking Dan R. to check how Evans looks if he's average. Somewhere in the middle between All-time great, and disaster.
If, at average 3B defense, Evans still looks like a HoMer to Dan (whose player rankings I agree with for the very most part), then I'll be a lot less hesitant to vote for Evans than I am right now (he'll be around 20-25 on my '95 ballot).
Everyone agrees that contemporary opinion is of spotty quality. The Evans debate isn't about whether contemporary opinion is perfect, it's about how it should be weighed in comparison to non-PBP defensive stats. In the context of that debate, Jeter cuts both ways.
In general, I think making a HoM case entirely on non-PBP defensive stats is a bad idea. It seems like without the non-PBP stats, no one would be championing Evans. That's a lot of weight to put in them.
1. The usual 70s spiel--low replacement level for infielders, low standard deviation.
2. The combination of his hitting (using baseball-reference park factors), fielding (using BP FRAA and Fielding Win Shares), and in-season durability mean that I see his 1973-75 as an *extremely* strong peak case--with the '73 good for an MVP award in the vast majority of years, '74 played at a high All-Star level, and '75 at a good All-Star level. Of his $89M career salary, over $17M comes from 1973 alone.
If you don't consider him *the* best fielding 3B in the National League from 1971 to 1975, then he wouldn't really work as a peak candidate, and he doesn't have the 10+ All-Star seasons to make him deserving as a pure prime candidate. I think you have to believe WS and WARP about his fielding to vote him as high as I will (probably #3).
Hence, the authors of the Stats Handbook inserted a "park change break" in between 1973 and 1974, blocking their multi-year averaging of park factors, hence the wild swing between a 115 in 1973 and a 101 in 1974.
And they did that with Hank Aaron still on the team. Of course, he was sitting at 713 already, so this action wasn't going to deny him the record. But note that Aaron's own HR dropped from 40 in 1973 to 20 in 1974. (Yes, he was old.)
maybe the knuckleball induces bad contact, more pops and bloops.
Regarding Atlanta 3B
Was Terry Pendleton still a tremendous fielding 3Bman in the early 1990s?
Regarding Bill James on Evans
I wonder whether he watched Evans in his prime play 3B. Maybe not: NL, before the superstation. Maybe, then, his high opinion of Evans was derived from his early sabermetrics.
Regarding Bill James on Buddy Bell
#120 suggests that Bell was unusually consistent as a fielder, too
Bill James 1988[maybe 1985] on Evans: "a tremendous third basemen for many years"
Regarding Atlanta 3B
Was Terry Pendleton still a tremendous fielding 3Bman in the early 1990s?
Sorry, I meant something like what does modern sabermetrics show regarding his skill in relation to his range factor and other first/second generation measures
Once again 1975 is the year that seems the strangest to me. Darrell Evans is credited with far more defensive win shares than Schmidt, and I still can't figure out why. I went back and researched defensive win shares, and found the following:
First, the Phillies were credited with far more defensive win shares as a team than the Braves (about 40 to 28). This makes sense to me.
To evaluate a third baseman, win shares looks at four main pieces of information. I'm simplifying a bit but, first, what percentage of the team's infield assists did the third baseman have, and how does this compare with the expected percentage, based mostly on league average (weight of 50). Second, how does fielding percentage compared with the league average (weight of 30). Third is sacrifice bunts against the team (weight of 10). Finally, double plays against average (weight of 10). This is then adjusted to create a share of the team's defensive win shares, in part by calculating the same for all other positions.
Anyway, the Phillies thirdbasemen had a higher percentage of the team's assists than the Braves thirdbasemen (the 50 point weight), a higher fielding percentage (the 30 point weight), lower on the double plays (only a 10 point weight), and I don't know on sacrifice bunts. I didn't go all the way through all of the calculations, bu I went pretty far. When you combine the fact that the Phillies had far more defensive win shares than the Braves, with the fact that they did better on both the 50 and 30 point weights, it seems almost incomprehensible that the Braves third basemen would be awarded far more defensive win shares than the Phillies. The double plays alone wouldn't make up the difference.
Can anyone explain what happened in the calculation? It might shed some light on why win shares loves Evans so much (without using putouts in the equation. Evans appears to have had a greater percentage of his team's win shares than any other third baseman in history.
Also, Evans team outperformed ther pythag by 4 games, Schmidt's by 1. That means there were 9 more WS to hand out to Braves players. And while Evans defense seems like an odd place t put a WS, remember that WS are all whole numbers. So even half a WS coudl be the difference between say, 6 and 7 WS (I don't know the actual total).
Anyways, maybe it is because I never really got to watch him play (I did play with him in RBI though!) but I have no problem with Evans' three year defensive peak beign as good as Schmidt's. It is years 4-10 or 4-12 that makes Schmidt a great defensive 3B and Evans a 1B. I don't see why defense can't have a peak like offense can. Ther are plenty of guys who have three monster offensive seasons (I have probably voted for them) and then weren't very good afterward and guys who have 10-12 monster offensive seasons (I think we all vote for them).
So for Evans, it is a prime/career thing offensive but with a god defensive peak, which gives him enough peak to catch my attention. I am nto exactly sure where he will end up, but I know it will be in my top 10 and probably in my top 5.
I know you are right on the left/right issue (and has been said on the other thread, we know this is causing a bias toward Evans), but I am pretty sure you are wrong on the flyball/groundball issue. I followed James's example, and he didn't seem to make any additional adjustment for the flyball/groundball tendencies of the staff. The idea of taking the percentage of infield assists, rather than assists alone, already accounts for that. But maybe the answer really does lie in the left/right issue.
It should be mentioned that the Phillies had far more defensive win shares than the Braves, even after taking into account the pythag differences.
From my pathetically weak understanding of the Old Testament, it always seemed that the Lord believed that the best defence was a good offence.
Bill James described Evans as a good defensive third baseman at his best in the original Abstract, but not as good as Schmidt. I suspect that when the pbp data is in, it will show him to be in the average to good range for his peak.
You are right, the percentage of infield assists does adjust for a staff's flybal tendencies. However, I am not sure why facotring in Lefthandedness of a staff would create a bias toward Evans. Accounting for it makes sense to me, not accounting for it is leaving out some of the context in which the plays were made. If anything, its the other way around, not accounting for it would create a bias AGAINST Evans.
I wasn't clear. Its not the factoring in the L/R that's causing the bias - The Braves did face more left handed batters than the Phillies, and Evans should receive some kind of statistical benefit from that. Instead, its the degree of the benefit - James makes an assumption of how many left handed and right handed batters the Braves faced based on the number of left handed and right handed pitchers they had, and I think he overestimated the percentage of left handed hitters, at least what I understand from the other post.
In other words, I think he is correct for adjusting for L/R, but he may have over adjusted in this case.
Well, it's not fair to 'change the rules' if Evans' surface stats are way too favorable and if another candidate's are, too - but only Evans' numbers are adjusted.
But if most candidate stats are more accurate, then an adjustment for Evans as a special case is an overalll improvement.
I also think that Evans appears to be a special case.
1) Without a truly great defensive peak, he isn't much of a candidate
2) All the contemporary evidence - opinions, reports, usage patterns - conflicts with the stats
Evans is, then, a particularly extreme case. Even though I basically think the problems here are generalizable, I also understand why they've come out specifically around Darrell Evans.
I think you are right to raise the question about singling Evans out, but I think the history of the electorate suggests that voters are generally savvy about that.
Thus, thank you for mentioning it, but don't worry too much that the voters would allow themselves to be biased in this way.
The flip side of this discussion is that, if we do elect Evans this year, no one can look at his case and say we didn't put him through the wringer!
I will mention that in discussions of Tony Perez, one of the players to whom Evans has been very carefully compared, similar questions were raised about WARP's evaluation of Perez's defense at third base, and we have been fairly consistent in noting cases where there is a disconnect between reputation and systematic evaluation of fielding quality.
I am inclined to believe the following at this point:
1) Evans' PO and A numbers are raised _somewhat_ by a hidden boost in opportunities resulting from Phil Niekro throwing a zillion knuckleball innings.
2) Evans' DP numbers are still impressive, and his skill at starting a round-the-horn double play has been vouched for by Steve Treder.
3) Evans' error numbers are high, and were clearly a big part of the reason that he was moved off of third base by the Braves' management, but Evans' error rates were influenced somewhat by the bad field conditions in his home parks.
4) Evans' talents as a player were never well-recognized by the Braves until Evans' offensive performance absolutely rubbed their noses in his talent, so I am not inclined to give their juddgment much weight in assessing Evans' shortcomings as a fielder. My hypothesis is that better management would have helped get Evans out of his head about the errors (they clearly bothered him and affected his play in other ways), rather than deciding that he was inadequate at third base.
I haven't decided what, if any, formal alteration of Evans' numbers I will make as a result of these conclusions.
Overall, this has been an excellent discussion, and I'm grateful for the way the folks who are part of the HoM project go about their business.
I am not a HOM voter, but presumably I am not the only one who doesn't take FRAA and FWS at face value. I lend much greater weight to EQA, OPS+, and league rankings offensively than to non-PBP defensive stats. For most candidates, such stats have minimal bearing on which side of the in/out line they may fall. Occasionally a hitter's offense may make him borderline and a good or bad defensive record pushes him one way or the other, but especially when dealing with players outside of the middle infield positions, defense typically has only a minor impact on a player's perceived qualifications.
Darrell Evans is a highly unusual case. By your own admission, his offensive record is insufficient to qualify him as a HOMer given average defense. Here were his EQAs in his 5 best 3B seasons by WARP: 326, 290, 293, 281, 289. The first four also correspond with his top four WARP seasons overall, and only one of them is consistent with a HOM-caliber offensive peak. His peak/prime case overwhelmingly relies upon him being among the best defensive 3Bs, at a time in which more advanced fielding metrics were unavailable and his teams regarded him as below-average. From my perspective, he may be viewed as a HOMer if you are an extreme career voter over peak/prime considerations or if you have nearly complete faith in FRAA and FWS. Otherwise, he seems to come up short.
If you break yourself of the "Is he or isn't he" mindset, you'll have a better understanding of how the electorate responds to the kind of information presented on this thread.
Others (say, me) would be inclined to knock off most of his performance above average. If Evans loses 4-6 wins over that period, I have trouble seeing how he makes a top 10.
There's been no challenge brought against the metrics' evaluation of Evans D in SF, and both systems see him as a good defensive third baseman there. I see no reason to believe that Evans was not as good or better in his ages 26-28 seasons -- when we would expect to find his defensive peak, anyway -- as he was in his 31-33 seasons in SF. All the problems with the evaluation of Evans' defense that have been adduced are specific to the Braves pitching staff and the Atlanta ballpark.
I think treating Evans as only an average defensive player in these seasons is basically to disregard the statistical evidence we have, rather than to attempt to interpret it correctly.
Everyone at the time - including Evans himself! - thought it made sense to move Evans off 3B. The defensive stats say he was Graig Nettles. I'll split the difference and call him average. That's still making a significant concession to the stats - average third basemen don't get moved to first in favor of Jerry Royster.
Clearly my early-thread post of "Evans is a lot better than he's credited for, but let's not give him a pass into the HOM, either," seems to be getting granted.
I have no idea yet where I'll put him; I'll measure him closely with Elliott, Perez, Nettles, Bell, Bando, and probably Boyer and maybe Cey.
Why, baseball is full of players who defensive skills were evaluated incorrectly, even today. Human perception and memory is hopeless for this sort of thing. Reputations are made based on announcers and writers who are not the least bit objective, scientific, or even knowlegeable. Even the managers really don't have a clue. Look at the Gold Glove Awards and some of the voting over the years.
A players defense skill is based on thousands of routing and boring plays made time after time. Nobody can properly record and analyze all the balls in a players' zone. Errors may look bad but are no worse than a base hit that sails by a player or in a hole, that may have caught by 50% of his peers. A player's seeemingly spectacular plays (ofen come about because of bad positioning or a poor jump) and errors (fielding %) are horrible ways to rate defense skills. The metrics get above all that. How often does the guy make the routine play. A difference between 83% and 88% is huge and would never be noticed by a annoncer or writer who is concentrating on spectacular plays and errors. Metrics measures his jumps and positioning and reaction skills which observation doesn't.
I think that reputations and andedotal evidence is a horrible way to rate a player's defense, and the better the metrics become the more this is born out.
Why...I think both situations are equally likely.
Any player who makes a lot of errors (ie -has a low fielding %) looks bad, but oftne aren't at all. Guys with the most errors frequently have the best UZR, and that makes up for the errors several times over. When you understand that frequently the guys with the most errors make the most errors you learn that using errors as a measurint stick is silly.
It's kind of like finding our that hitters with the most strike-outs usually have the most HR and BB. Then the K's don't seem so bad. Most baseball observers simply haven't come that far yet in looking at fielding prowess.
I am not a HOM voter, but presumably I am not the only one who doesn't take FRAA and FWS at face value. I lend much greater weight to EQA, OPS+, and league rankings offensively than to non-PBP defensive stats.
Does OPS+ not rank SLG more prominantly than OBA, simply because the SLG absolute value is always higher? Has this not been shown to be incorrect? Isn't it better to use something like OBA*1.5 + SLG? I think it has been demonstrated rather conclusively that OBA is a bigger facter in team offense tha OBA. Wouldn't VORP be a better measure of a players offensive contribution, if that's all you're looking to evaluate.
That depends on your options at 1B, a lot more than Evans' percieved skills. Anyways it's been well documented that the Braves managerial and GMing decisions have been less that stellar during this time. Anyone who measures defense prowess by # of errors committed is making a major mistake, and it's likely that Evans didn't know any better himself. He was making a lot of errors playing on a terrible turf with a less than helpful 1B. Didn't all Atlanta players make more errors than other teams? They told him he was bad, so he believed them. Who knew at the time that he was making a higher % of plays than 95% of the other major league third baseman. Error...damn errors.
Let's try that again.
When you understand that frequently the guys with the most errors have the best range you learn that using errors as a measuring stick is silly. This was not the case in Evans' Atlanta days.
Does OPS+ not rank SLG more prominantly than OBA, simply because the SLG absolute value is always higher? Has this not been shown to be incorrect? Isn't it better to use something like OBA*1.5 + SLG? I think it has been demonstrated rather conclusively that OBA is a bigger facter in team offense tha OBA. Wouldn't VORP be a better measure of a players offensive contribution, if that's all you're looking to evaluate.
Yes, OPS+ gives too much weight to SLG relative to OBP, and something like the equation you suggest would be better, though I'm not sure what you mean by claiming that it has been shown to be incorrect. It's still a useful number, though it has an obvious flaw that should be considered. We know that OPS+ is a simple, "back of the envelope" calculation and we know how it is biased. The same cannot be said of non-PBP defensive stats. We cannot easily manipulate FRAA and FWS to account for their biases. They are based on insufficient data, which is a much worse problem than that of a metric like OPS+, which is overly simple and disproportionately weighted but based on the proper information. This is why I find OPS+ more useful than non-PBP defense.
I certainly could have included VORP and BRAR among the offensive stats that I consider in evaluating a player. I don't necessarily agree with all of the adjustments that BPro makes (especially the conversion from WARP1 to WARP2), but I still rely on many of their stats. Hence I listed EQA as one piece of information that I value quite highly, since it is an effective, straight-forward measure of offensive production and incorporates stolen bases. It doesn't adjust for playing time and position like some of their other stats, but I prefer it because it's less cryptic and differences in playing time and position are not difficult to see.
No doubt. But baseball is even fuller of players whose defensive skills have been evaluated correctly, so far as we are capable of knowing. The proportion of players in whom subjective evaluation/reputation lines up with objective metrics is far greater than not. To assume that when divergence occurs, the subjective evaluation is necessarily completely incorrect is unwarranted.
Didn't all Atlanta players make more errors than other teams?
No. Two separate Atlanta second basemen, Felix Millan and Marty Perez, playing on the same home infield with the same first basemen as Evans, put up extraordinarily good FAs.
There is very good reason to discipline oneself against adopting an air of certain knowledge on this particular issue.
No doubt. But baseball is even fuller of players whose defensive skills have been evaluated correctly, so far as we are capable of knowing. The proportion of players in whom subjective evaluation/reputation lines up with objective metrics is far greater than not. To assume that when divergence occurs, the subjective evaluation is necessarily completely incorrect is unwarranted.
This seems a bit like saying that sometimes W-L, ERA sometimes gives us an accurate picture of a pitcher's performance.
Sometimes it does, particularly if BABIP, strand rates, and rnu support are about normal - but K/IP, BB/IP, HR/IP, and xERA gives us a virtual iron clad method of measuring pitching performance. When there is a divergence between the W-L, ERA approach and the peripheral stats I think it is a pretty safe bet that the W-L, ERA result is the kne to be skeptical of.
Similarly when defensive metrics do not support subjective or anecdotal evidence, I suspect that it nearly always right to believe the metrics and assume that the reputations (both good and bad) are built on non-scientific data and subject to human biases.
I think it leads to incorrect conclusions, particularly in the case of players like Rice and Dawson who were much more adept at the slugging side of the game as opposed to the getting on base part to the game.
Considering the defensive metrics here, what you are saying is that because we don't have FIP et al, then we should use W-L record, rather than rely upon descriptions of a pitcher's "stuff". Considering how engrained the distrust of these particular metrics and the distrust of anecdotal evidence is, it can't be an easy decision either way.
Knock yourself out. But understand that you may very well be vividly demonstrating the very human bias you so fear.
But I think all of us have been humbled, to varying degrees, by getting insight from so many different directions since 2003.
The collective wisdom, ultimately, has been a lot more impressive than any one perspective.
Wilbur Wood was a lefty, so the comparison wouldn't hold.
I'm with TomH. I used OPS+ a lot and the real caveat is to look out for players with low playing time.
OWP ............................. OPS+
rank player.... OWP OPS+ rank
1 Babe Ruth..... .856 207 1
2 Ted Williams.. .832 190 2
3 Mickey Mantle .801 172 4
4 Lou Gehrig.... .797 179 3
5 Stan Musial.... .752 159 7
6 Charlie Keller .748 152 16
7 Jimmie Foxx.. .743 163 5
8 Mel Ott.......... .743 155 14
9 Johnny Mize... .743 158 9
10 Joe DiMaggio .741 155 13
11 Mark McGwire .737 163 6
12 Willie Mays.... .731 156 11
13 Hk Greenberg .722 158 8
14 Hank Aaron.. .719 155 12
15 Frk Robinson .717 154 15
16 Dick Allen..... .714 156 10
The top 16 are precisely the same. Biggest diff is Charlie Keller; an interesting study might be to see why so.
1975:
AL:
10th, Brett, 25
20th tied, Nettles, 21
NL:
2nd, Rose, 31
4th tied, Evans, 28
4th tied, Schmidt, 28
7th, Cey, 27
8th tied, Madlock, 26
I talked about this in the Evans thread, but this completely defies reason. By various metrics, Schmidt was either 21 (BRAA), 23 (VORP) or 26 (RCAP) runs better on offense than Evans. So, according to Win Shares, Evans was worth 20-25 runs more on defense than Mike Schmidt?
I know this was kind of responded to, but the Win Shares breakdown is as follows:
Schmidt: 22.1 BWS, 5.8 FWS
Evans: 20.5 BWS, 7.7 FWS
I don't know the conversion rate, but I'm pretty sure that a 2 WS difference is significantly less than 20+ runs. The problem may be that Win Shares is rating Schmidt's offense too low as much as it's rating Evans' defense too high.
For the record, in 1975 FRAR has Schmidt ahead 45 to 39, and in WARP1, ahead 11.3-8.3
I don't recall Evans ever being regarded as a top-notch fielder. That could have been because everyone in the NL was so far behind Schmidt.
I know for certain contemporary opinions can be pretty wrong as well, so I don't put much weight in defensive evaluations on older players at all.
Frinstance, Richie Ashburn's putout totals are just crazy. Everyone assumes he just had a million chances. Well, maybe he did and maybe he didn't - we don't know.
Cey wins, hands down, unless you put much weight on defensive metrics. Evans seems to have been an average 3B for a few years and then an averagish 1B for a few more, then had some zero defense time as a DH.
Cey was a 3Bman, through and through, and a very good one (I think he's quite a bit better than Evans).
Evans' extra career bulk (again - mostly as a part time player) probably makes up for the difference, but I think I will have Cey just above him for now.
How are you defining "regular"??
Evans played in more than 75% of the seasons' games 17 times. It's obvious that Evans was a starter for 17 years, although he missed some time in several years.
Be fair to Schmidt. 1984 was the year that it seemed he might be done at 3B --career low fielding average and DP rate (down about one-third). In 1985 he played mainly 1B.
Mike Schmidt fielding record, 3B only, 1984 and 86-87 only
Year Ag Tm Lg Pos G PO A E DP FP lgFP RFg lgRFg RF9 lgRF9
1984 34 PHI NL 3B 145 85 329 26 19 .941 .943 2.86 2.36 GG,AS
1986 36 PHI NL 3B 124 78 220 6 27 .980 .946 2.40 2.20 GG,AS
1987 37 PHI NL 3B 138 87 315 12 28 .971 .951 2.91 2.34 AS
(This forum has capability to display preformat text in normal size with bold and italic face but I messed it up last time.)
By 1988 his fielding average was down again, his DP rate still near his career average one per five games.
Everyone should expect that his fielding reputation surpassed his fielding skill by the mid 1980s. But what about the preceding decade? (actually a 1972 callup and 11 full seasons) He didn't win a Gold Glove until 1976, where Pete Palmer felt that he deserved one already in 1974.
was he platooned in some seasons?
was he platooned in some seasons?
The folks who have provided first-hand info about Evans have not mentioned anything about platooning, so I would be rather surprised if that was the case. Since they have mentioned that Evans was frequently lifted for defensive replacements in the late innings, my guess is that this sort of substitution is responsible for Evans' low ratio of PA/g in some seasons.
In 1972, he appeared in 125 games, with a 4.17 PA/g rate. His low number of games was because he was given the starting job when Clete Boyer was canned.
In 1973-75, he played in almost all games and almost all innings, with 4.55, 4.44, and 4.37 PA/game.
In 1976 and 1977, his tanking and recovery years, he appeared in 136 and 144 games ,with PA/game of 3.46 and 3.74.
In 1978-81, he appeared in 159, 160, 154, and 102 (of 108) games, with PA/game of 4.2, 4.13, 4.24, 4.11.
In 1982, his last year primary at third base, he appeared in 141 games with 3.92 PA/game.
In 1983, age 36, a full year at first and last year in SF, he appeared in 142 games with 4.30 PA/game
In 1984-87 in Detroit, he appeared in 131, 151, 151, 150 games, with PA rates of 3.69, 3.93, 3.98, 3.62.
Given that his years with fewer PA/g correspond to periods of fielding struggle and the last phase of his career when he would have needed more rest (and been lifted now and again for pinch runners), I would say that he is more likely to have been lifted in the late innings and/or benched/rested more in those seasons than he is to have been platooned. It doesn't look like a platoon pattern to me.
You must be Registered and Logged In to post comments.
<< Back to main