POLL: Hall of Famers on playoff teams
Posted by Andy on October 5, 2010
I'm switching up our Hall of Fame poll this week to take a look at the 8 playoff teams. I picked one position player and one pitcher from each team as candidates. I avoided guys who are quite obviously 'yes' already based on their past accomplishments. Some of these guys are pretty close already while others are off to great starts but still very young.
One other difference: you can vote for as many players as you like in this poll. Click through to vote and discuss!
October 6th, 2010 at 10:47 am
MIke Felber says
"Awards are almost universally acknowledged to be often just wrong"
HILARIOUS! The sports writers and others who watch baseball for a living are "universally wrong" and the SABRE-heads who might never watch a single game are right. That is precious.
But another example of "fuzzy math". Please mathematically justify the phrase "Awards are almost universally acknowledged to be often just wrong"
Let'sface it. We're all guilty of the same thing. We pick a player we like, and THEN we promote them via our favorite prism- me, I talk about awards and visible impact of players I've seen; and the rest of you go to whatever stat justifies the player you like. But I think nearly all of us begin with the player.
October 6th, 2010 at 11:34 am
Again Barkfart, by whatever means you analyze, Thome is worthy of the Hall. You don't need the WAR to come to that conclusion. The only types that think he's not are ones that are still DH-prejudice, which is partially silly since it's been around for nearly 40 years now...and, all this about a guy that's only been a DH the past 4-5 years. I was one the fence about Thome (partly b/c of his one-dimensional aspects and DHing) before this year, but what he did at 39 to get his WAR over 70, his HR's within shouting distance of 600 (589), and doing this in a pennant race as a bat replacement for Morneau puts him over the top!!!!! He simply had a great year --yes, he only had 340 AB, but he raked this year. 25 HR, 59RBIs, .283 avg, and .412 OBP, .627 slugging, 1.039 OPS and 178 OPS+ is doing some serious damage with his limited chances. Also, many of his hits this year were really clutch!!
2010 39 MIN AL 108 340 276 48 78 16 2 25 59 0 0 60 82 .283 .412 .627 1.039 178
October 6th, 2010 at 11:48 am
Bf, you just (#101) took the phrases "almost universally acknowledged" and "often just wrong," left out the "almost" and "often" qualifiers, and mixed the phrases to come up with "universally wrong" as a straw man.
Mr. Felber did not contend that awards voters are "universally wrong."
Yes, we (probably) all start with preferences and buttress our arguments with whatever evidence is most favorable to those preferences. But what you've done here seems an example of "fuzzy English."
October 6th, 2010 at 11:55 am
#102 Matt
even you admit to being recently on the fence about the guy
#103 Toledo
(I used to live in Rossford- loved it)
Let's go to the guy's original phrase.Can something really be universally recognized as often wrong?
October 6th, 2010 at 12:01 pm
#83
Regarding McLain and Gibson:
One thing that's important to keep in mind is that WAR is based off of RUNS allowed, rather than EARNED runs. ERA makes its defensive adjustment by factoring in errors into the formula, whereas WAR uses the total runs allowed, and credits each pitcher for a portion of the team's defense.
That's one of the limitations in how WAR is calculated, and it's important to keep in mind when evaluating WAR for pitchers. Using McLain as an example, his team rates highly on defense in 1968, but a greater than average number of his runs allowed were unearned. The nature of the WAR calculation assumes that a team's defense plays equally well behind each pitcher in a given year, which isn't always the case. In 1968, McLain's low BABIP was consistent enough with the team's .217 BABIP, but the number of unearned runs makes it quite possible that the defensive credit in his WAR is inflated. In 1969, the reverse is true and he allowed 4 unearned runs, so it's possible the defense played particularly well against him.
In addition to the defensive quirks of the system, there's also the league's context. 1968 was just a strange year, with about 3.4 runs per game, compared to over 4 one year later in 1969. A part of that contextual adjustment is due to expansion, which can artificially cause a blip in the run scoring levels and create a little oddness in adjusting for context.
Both pitchers do illustrate why we shouldn't be slaves to the WAR number, especially for pitchers. For a single season, it makes a great starting point, but you need to dig deeper to see if there's anything that could be causing a player to be over or under rated. In the case of both McLain and Gibson, it's fair so say that the number of unearned runs, especially in a context where every run is worth so much, probably cause WAR to underrate their 1968, and similarly, the opposite likely causes it to overrate McLain's 1969.
October 6th, 2010 at 12:04 pm
"HILARIOUS! The sports writers and others who watch baseball for a living are "universally wrong" and the SABRE-heads who might never watch a single game are right. That is precious."
Can we stop this BS narrative? Have you heard of Joe Morgan, who is the crown jewel of the anti-SABRE crowd who regularly admits that he does NOT watch most games or teams, yet still feels qualified to comment on them? Have you heard of Keith Law, a former professional scout who works as a scouting analyst for ESPN, attends dozens (if not more) of Major and Minor league games a year, and employs a healthy dose of advanced statistics in his analysis?
Yes, with some stats, we've reached a point where you can calculate them without watching a game. That has always been true. You could figure out how many HRs Babe Ruth hit without ever seeing a game... just look at the boxscores. You can also calculate WAR or WPA without watching a game if you have the right input. What you CANNOT do is come up with any of those stats without a deep understanding of the game. People didn't start counting HRs randomly from box scores. Folks watched the games and said, "Hey, if he hits it over the fence, they automatically score! That's pretty good!" And people didn't pull WAR out of a hat. They watched games and analyzed results and said, "These things seem to correlate highly with winning and players who do more of those are more valuable, etc, etc, etc."
So, yea, I suppose it's possible that statisticians (which include folks who just count HRs or Ws or BA, since those are in fact stats) can do their analysis without watching the games. But the stats are only as good as their relation to the gameplay itself. And to derive the more sophisticated stats requires a more sophisticated understanding of the game. It's easy to count HRs. You could stand on the other side of the fence, see none of the game, and just count the balls that go over. You could never come up with a stat like WAR without taking the time to learn the intricacies of the game.
October 6th, 2010 at 12:06 pm
"Let's go to the guy's original phrase.Can something really be universally recognized as often wrong?"
Sure it can.
As an example, on any challenging test in school, students often get some answer wrong. This can be universally recognized without saying they always get it wrong.
October 6th, 2010 at 12:11 pm
I have a different take on the "east coast bias" thing. I agree (with Dupp @48) that "the Yankees, Mets, Red Sox, etc. don't benefit from Hall of Fame voting." However, the original statement was about "once someone wears Yankee pinstripes for six or seven years"; I think there is a very real benefit for someone's career in playing for one of the consistently-good teams (like the recent Yankees or Red Sox). A very good pitcher, backed for 6-7 years by the Yankees offense, can be a spectacular pitcher in the HOF voters' eyes.
Consider Blyleven's fate: had he played for the A's over 1971-74 (during which years he went 70-66 while posting a 136 ERA+ and completing nearly half his starts). Instead of pitching for a .500, middle of the pack team, he could have pitched for a team that won 3 WS in four years. 3 rings and certainly another 13 wins (to get him the magic 300) and he's a lock for the HOF voters.
This is a slightly different argument than a player getting unfairly voted in due to playing for good teams (or, even more incredulously, well-publicized teams). I have to believe that rarely happens. I do believe that there are certainly benefits to playing for excellent teams, and that those benefits can potentially push a bubble player into immortality (Tony Perez?).
October 6th, 2010 at 12:15 pm
So I had hoped to break this down by number of ballots so we could see who got at least 75%, but it turns out my free polldaddy account doesn't let me see that and I'm not going to ask Sean to pay for a membership 🙂
As of the moment, I would guess there are 900 ballots cast, meaning that a guy would have to have 675 votes to be on 75% of them. Mauer has 701, followed by Sabathia at 504. So only Mauer is close to 75%.
October 6th, 2010 at 2:14 pm
"Without the DH, he has, max, 450 homers, maybe 1200 runs, 1200 RBI, less than 400 doubles."
If you cut off his career at the end of 2005 (after which he has played 65 WAR.
The guy wasn't a bad fielder, he just got old, and extended his career by DHing. In reality, it's pretty clear he could have played 1B for much of this time if he'd been in the NL. He had positive TZ fielding numbers in 2003-2005, it's not like he was terrible. He just got traded to teams who already had well established 1Bs who couldn't be easily moved, and nobody great in the DH slot, so they chose to use him as a full time DH. Had there been no DH, it is *very* unlikely that he would have had to hang it up after 2005. Instead he would have gone to a team that could use a 1B.
He may have gotten fewer PAs due to injuries or platooning/subbing to keep him healthy, but the idea that his career would have ended or he would have become little better than a replacement player without the DH rule is utterly ridiculous. And you have to think something close to that to suggest that DHing his last 5(+?) years should keep him out of the hall.
October 6th, 2010 at 2:24 pm
Barkfart,
Yeah, on the fence before this year. He's done enough, get with it. This year puts him clearly over the top.
October 6th, 2010 at 2:27 pm
The only guys on your list that I'd give even money to make the hall are Joe Mauer (Mauer I'd probably give 2 or 3-1) and CC, because they are already about 2/3 of the way there.
The rest of these guys are either still very young, or not really on pace to make it with standard career/age arcs (Howard has done very well but started late, and he's no ichiro). It's very likely that at least a couple of them do ultimately make it, but I don't think any are individually a smart even bet. I voted for Longoria and Lincecum anyway, although comments have made me think Longoria, like a few other 3Bs of note, has a decent chance to make *my* hall but not the cooperstown one.
October 6th, 2010 at 2:45 pm
I don't know what happened to the first part of my comment about thome.
I can't recover it completely but I was trying to say that even if you cut his career off at the end of 2005, he's at least a borderliner.
Look at the list of guys who are close or over his OPS+ and have between 6000 and 8000 PAs. Here's thome's stat line if you cut him off there:
.281/.408/.562 .970 OPS, 149 OPS+ 1151 R 1193 RBI, 430 HR 7281 PAs.
Here's who'd be ahead of him in OPS+ for 6-8k PAs: Joe Dimaggio, Hank Greenberg, Johnny Mize (all HoF). Albert Pujols (lock future), Mark McGwire (not lock without PEDs but probably in, and horrible defensively, lower WAR), and Dick Allen (not in, but most saber people say he should be).
Even with him in OPS+ is Ralph Kiner with fewer PAs and worse defense, and he's in the hall.
Behind him are Mike Piazza (should be in, but a catcher), and some guys who don't belong, but were very good, and are behind him in nearly every category (Berkman 145 for 6600 PAs, Belle 143 for 6300 PAs).
So that's where he'd stand if he'd hung it up in october of 2005. Everybody ahead of him offensively is either in, or probably should be, one guy that is close behind him is in, and a couple guys a little further behind him are not in. So he'd be borderline.
And that assumes that his last 5 years DHing are completely valueless, which is ridiculous. Even if you penalize him an extra 10 runs per 700 PAs for DHing (so his positional adjustment becomes as bad as the Rpos + Rfield for the worst 1Bs in any given year), he still would wind up with >65 WAR which should be clearly in.
October 6th, 2010 at 3:05 pm
BSK @ #106.
"What you CANNOT do is come up with any of those stats without a deep understanding of the game."
That's actually not at all true, if anything, it's the opposite.
People come up with these "advanced" stats as a way to understand something they don't know. Bill James has said as much, as have quite a few of those who have followed him.
If one understands baseball, there is no need for stats, which are nothing more than documentation of what has transpired.
The only reason to look deeper is failing to understand why.
October 6th, 2010 at 3:10 pm
Another way to look at Jim Thome is with Runs Created. He's 25th on the career list with 1918, and everybody ahead of him is either in, going in, or will be out largely for off-field reasons (palmeiro, rose, will they really keep Bonds out?). Less than half of these guys did it at a similar or better RC/PA rate. Jim Thome is quite simply one of the best offensive baseball players of his time, and behind only the inner circle greats all-time. He deserves a spot in cooperstown.
October 6th, 2010 at 3:10 pm
Thank you for the analysis Wilson, that is incisive information. That seems a basic problem: why cannot WAR easily account for/excise unearned runs? It is one thing if they do not use defensive metrics like range factor, but to not even distinguish which runs are unearned seems negligent. And thanks for the support on my comment Wilson, BSK & David in Toledo.
BF, their comments anticipated my own. You were really fuzzy actually, even getting wrong the clear meaning of my words. Such as missing the "almost" & "often" qualifiers, even after you copied my quote! I assume without intending to be dishonest. It is not possible to "mathematically" prove my statement, since it is one of opinion: but the opinion of ALMOST all who follow baseball, including the traditionalists, is that awards are often wrong. Given out due to emotional biases, team/context dependent reasons, & misperceptions of real total value. If you like I & others can provide innumerable examples of this, but it is so obvious I doubt anyone really needs this done. Certainly we do not WAR to show how so many players were poor selections, & so many clearly best ones not selected.
While All Star games & gold gloves, to some extent, are often about reputation/popularity, ironically the best, most prominent players who deserved more MVPs do not get it, do to a yearning for novelty/a good new story. Bill James said Mantle deserved all MVPs '54-'62 save '59, & guys like Musial & Mays were close. Williams famously cheated a few times, & more recently, so many guys getting awards due to gaudy context dependent #s like RBIs, (era, team/opportinity, & park factors) & a poor focus on total production. We could have a proverbial Field Day with the Cy Young award, starting w/the utter folly of making wins & W-L a litmus test for excellence.
Anyway, I hope you finally realize not only what I actually said, but that nobody but you claimed such a bifurcation between saber guys (who do often watch games), & other observers-it would be rare to find ANYONE who thought awards are not wrong at least fairly often. I will bet nobody else on this site will argue otherwise. Now this proceeds sometimes from fan bias-but also very often from just disinterested observation.
Specifically, I do NOT believe their is evidence that "nearly all of us" start with a favorite player & then confirm our bias. This is not uncommon, & their arguments might still be completely correct. But many look at properly adjusted #s, consider all aspects of the game, & THEN pick a player as best (sometimes even on an "enemy" team). That might be when they begin to like a player more-after they better appreciate their true value.
October 6th, 2010 at 3:29 pm
I think you are correct that many use stats to understand something they have limited knowledge of. Though it is not merely documentation, there is figuring adjusted/real value, hence interpretation of the raw #s. But it is not a black & white, all these guys need stats 'cause they know little, & it is extreme to think that any one can just observe games & no stats will add any understanding or new insights. Even if you could see more than a small fraction of all games, all use stats to "count" what is produced. And while many traditionalists have some intuitive idea re: the nuances & value of various aspects of the game, virtually all can learn things from looking closely at context. No stats are perfect, yet you cannot expect that good observers know all precisely via just looking.
To think any "side" monopolizes all truth is at best silly. But guys spent forever worshiping BA, RBIs,W-L.
It was the rare man who did not "need" a bracing corrective to perceive real value.
October 6th, 2010 at 4:21 pm
I expect that WAR uses all runs, because they believe looking at their own fielding metrics behind the pitcher adjusting the total runs gives a better answer than looking at earned runs (which gives the pitcher carte blanche on things judged an error, but does not distinguish between excellent and average fielding behind.
I'm not sure I agree, given that we are working with fairly poor fielding numbers. But honestly, I doubt it's any worse than just looking at ER and ignoring unearned runs. The judgements on what is earned versus not is pretty arbitrary and surely introduces pretty large errors. The whole point of earned versus unearned runs is to look at what is due to the pitcher versus what is due to the fielders. But we know that looking solely at errors is a pretty poor way to judge fielding. And we also know that once a runner is on base, no matter how, what the pitcher does going forward has a great deal to do with whether they score or not.
If we ever get a really good fielding stat, then surely the B-R way will be a big improvement over just using earned runs. Right now, it might not be much of an improvement, but I doubt it's really worse, just different.
October 6th, 2010 at 4:35 pm
Yeah I really love hearing about how stats people know nothing because we don't watch the games from a guy who's only stated reason for doubting the value of ERA+ and WAR is ...
wait for it ...
a different stat.
If a pitching coach tells me that they watched all or most of Denny McClain and Luis Tiant's games in the 1968 season, and Denny McClain was clearly better with some coherent non-stat explanation of why he thinks that and which doesn't consist mostly of "OMG 31 WINS!!!" -- I'll give that serious consideration.
Some blowhard who just repeats "31 wins" over and over is not worth paying attention to.
October 6th, 2010 at 5:08 pm
"Some blowhard who just repeats "31 wins" over and over is not worth paying attention to."
Neither is anyone who doesn't believe Felix Hernandez should win the Cy Young.
Such as Keith Law.
October 6th, 2010 at 6:18 pm
Chuck-
You make a really interesting point, though I don't think we're really talking about the same things. Yes, advanced stats help us try to quantify the apparently unquantifiable, to make order from chaos, sense from nonsense, music from noise. But one cannot do that successfully without some understanding of the inner workings of the game. If you had no idea what baseball was or how it was played, you would never be able to come up with a sophisticated metric that had any credibility or reliability. The notion that the stats were borne of pure ignorance of baseball, by people who never watched the game, is simply wrongheaded.
To your latter post, unless I misunderstood your double negative, KLaw certainly does advocate on behalf of Felix for the Cy Young: http://insider.espn.go.com/mlb/blog?name=law_keith&id=5647358
October 6th, 2010 at 6:46 pm
I don't know if I saw it in this thread, but Andy you were close, Halliday with a no hitter -- not quite the perfect game. I think i saw that somewhere.
October 6th, 2010 at 7:13 pm
BSK @ # 121
"you would never be able to come up with a sophisticated metric that had any credibility or reliability."
Most of them don't, which is my point.
They're always changing. There are different versions of WAR. There have been three or four amendments to UZR. I remember reading something recently that said Baseball Prospectus is overhauling its PECOTA formula for next year's publications.
Why are saber stats constantly changing?
Because the more the "inventors" learn about their subject, the more they realize the flaws in what they initially have said.
It's like learning a foreign language.
You start with nothing, and as your learning improves, so does your methods.
Eventually, you become self sufficient and no longer require things like WAR.
October 6th, 2010 at 7:34 pm
Do you have the same problem with physics, Chuck?
Because it changes, it can't be trusted? Obviously these physicists don't really know what they are talking about. Perhaps we should just stick with what the greeks thought? Or aristotle?
How about medicine? Should we use leeches to cure nearly everything? You can't really trust doctors, can you? What they say now, is a lot different from what they were saying 30 years ago, let alone 100 or 200 years ago.
The fact that things are changing is a good reason not to take any given pronouncement as gospel truth -- if it's changed before now, it's likely to change again. But the chance that what a given field believed 30 years ago is actually more accurate than what they believe now, is pretty small.
Traditional stats don't change because they aren't intended to measure something that is difficult to measure. They are intended to measure exactly what they do measure, and what they measure was chosen based on how easy it is to measure.
WAR on the other hand, is an attempt to put the entire value of a player on a single number scale. That's pretty hard. If it was easy, they wouldn't have to pay people lots of money to evaluate baseball players and build teams. Any random college grad could be trained to be a competent GM, and they'd get paid like bookkeepers -- just show them the stats and away they go.
Given that it's hard, any given attempt to do it is probably not going to be perfect. If you start with something that isn't perfect, then changing it has the potential to make it better. In any endeavor that is interesting enough for smart people to spend any time on it, prevailing understandings are likely to change with time, and most of the time, they'll be changing to something more accurate than what they started with.
October 6th, 2010 at 8:09 pm
So let's see... evolving a stat when new information is gained is WORSE then refusing to evolve?
There is room to quibble about the validity and reliability of WAR and other sophisticated stats. But it is simply ignorant to insist that anyone who creates and/or uses such stats never has or will watch a baseball game. It's insulting and stupid and the most useless form of arguing ever.
October 6th, 2010 at 8:25 pm
That's kind of the problem I have, though, Mr. Sullivan, because while medicine and physics, etc, have changed in the past 80 years, baseball hasn't.
Ty Cobb could play today just as easily as Albert Pujols could play in 1920 because the game is the same.
Travel is better, obviously.
Nutrition is better, money is better, conditions are better.
The game is the same.
You reincarnate Grantland Rice and stick him in the booth in Philadelphia today and he'd write just as good a story as Jayson Stark because the game is the same.
Sabermetrics allows people to look at baseball differently,but no understanding comes from it.
If some UFO landed in your backyard and you invited the aliens inside to watch the Yankee game with you, would you whip out your copy of BP and log onto BR and use those tools to explain the game to them?
It would be impossible.
October 6th, 2010 at 8:32 pm
"But it is simply ignorant to insist that anyone who creates and/or uses such stats never has or will watch a baseball game. It's insulting and stupid and the most useless form of arguing ever."
I agree.
I also don't see anywhere on this thread where somebody has said, or even suggested, what you are implying.
Although it must be said my eight year old will sit and watch with me and she has no clue to what she is "watching."
I know plenty of adults who fall into the same category.
October 6th, 2010 at 8:43 pm
Chuck Says:
October 6th, 2010 at 8:25 pm
That's kind of the problem I have, though, Mr. Sullivan, because while medicine and physics, etc, have changed in the past 80 years, baseball hasn't.
Baseball hasn't changed in the last 80 years.... except for the DH, divisional play, interleague play, wild card, steroid era, pitcher dominant era, lowering of the mound, artificial turf (come and gone), free agency....
October 6th, 2010 at 8:54 pm
Chuck, I'll go back to 126- you said Ty Cobb could play today, I know you meant that more as metaphorical than carved in stone, but on what basis do you assert that?
When I say people think they can judge players without watching the game, here's what I mean....
It's one thing to use sabremetrics to compare Pujols to, say, Frank Thomas. Most of us have seen both play and can mix the statistical with the experiential. But many in the Sabre crowd think that they can accurately players well before their time whom they have no concept o, except for the stat sheet.
Conjecture, yes, but others will speak as though WAR has a sense of certainty.
Probably a half dozen times, people have compared Thome to Killebrew, etc and cited nothing (or next to nothing) than WAR. I'm old enough to have seen both players, can appreviate the stats, and compare the different eras. That's a lot of information.
October 6th, 2010 at 9:05 pm
"Chuck, I'll go back to 126- you said Ty Cobb could play today, I know you meant that more as metaphorical than carved in stone, but on what basis do you assert that?"
Wait..you're not seriously suggesting he COULDN'T, are you?
October 6th, 2010 at 9:21 pm
"I'm old enough to have seen both players."
Me too, Bark, me too.
October 6th, 2010 at 9:55 pm
Chuck, your argument is very ignorant. As stated above, it's ridiculous to think that a formula offers little insight because it has been tweaked or will continue to be tweaked. Re-analyzing a formula and the data and tweaking it is part of the process of making it better. Also, the idea that sabrmetrics has offered no insights into the game is such a ridiculous statement to make it borders on lunacy. All the teams use some form of analysis owed to the sabrmetric movement in their player evaluations--some more than others, but all do it. The idea that a .285 hitter who has a .425 OBP is likely to be more useful than a .335 hitter with a .360 OBP is not something that is derived from traditional player evaluation, but rooted in the earliest attempts at sabrmetric-type analysis.
October 6th, 2010 at 11:54 pm
#123,
If two scouts disagree in their evaluation of a given player, does that invalidate the entire usefulness of scouting? If advances in bio-mechanical analysis help a scout understand with greater accuracy the injury risk associated with a given pitcher, should he ignore that information because he thinks he knows everything there is to know about mechanics already? If a fan who has no idea how to evaluate a player's tools properly tries to give an online scouting report, does that negate the value of real scouts?
We have different variations to many of the stats because the people who make them are, in fact, human. As humans, sabr folk do have opinions. Just because two people study the same data doesn't mean they'll come to the exact same conclusions or use the exact same methods, any more than two scouts will always agree with each other. The open-source nature of most sabermetric tools lends itself to a broad community of people making refinements for more accuracy. One of the major factors in the continuing change in some of the formulas is the amount of raw data people now have access to.
One of the unfortunate things about the open-source nature, however, is that you do have peculiarities like the different versions of WAR. This is particularly glaring for pitchers, since we don't yet have a way to accurately separate pitching from defense in a given year. Different versions have their own flaws, and the creators of the stats will typically readily admit that they do have flaws. Flaws in a stat are only a problem if you don't recognize them and know how to adjust for them.
The WAR cited on the B-R site runs into the problem of assuming an equal defense behind all pitchers on a team. Over a career, that's something that's not necessarily a major flaw, because it does adjust for team defensive support, and would be rare for a pitcher to always have his defense perform better or worse than average in a given year.
If you're not comfortable with the assumption of equal defense, but ARE comfortable with absolving the pitcher of responsibility for, say, a home run after an error on the would-be third out of an inning, it would be possible to adjust WAR to a format based on earned runs. For example, taking McLain's 1968, if you strip his defensive support out of the replacement value and use earned runs instead of runs, he'd be 74 RAR, or about 1.54 times as valuable, and then you'd have to adjust to convert it back to the RA scale (since wins are based on runs, not earned runs) - running the numbers quickly, I get a WAR a little over 8 for that season, which is quite a difference from his 5.9. The flaw with that method, aside from a philosophical difference of how much pitchers should be accountable for unearned runs, is that it makes no adjustment for range.
Fangraphs uses a completely different method of calculating WAR, in attributing the variances in BABIP to defense and chance, rather than the pitcher. It's based off a linear weights type of system, where it takes the defense-neutral elements of a pitcher's performance (K's, BB, HR, HBP, etc.) to estimate what a given pitcher would be expected to do with neutral defense. It runs into a few obvious flaws, in that many balls in play are the result of how a pitch is hit, rather than how the fielders perform, and perhaps more glaring is that it strips sequencing out of the equation.
It can be confusing to have different versions of a stat, but it can also give you different perspectives - as long as you understand each one and the built-in limitations. B-R's WAR can only be as good as the best defensive metrics it has, and even the best ones rarely have pitcher-specific splits. Fangraphs uses an abstract method which will paint a decent picture in the general case, but it misses out on almost all the little things a pitcher can do to help himself out. Neither WAR is anywhere near good enough to be a definitive answer, but it can provide a good starting point to get a general idea of value.
October 7th, 2010 at 12:51 am
Many people with Big Smarts here! Maybe using all runs yields about the same results Mr. Sullivan, though given the problems with advanced fielding metrics, maybe they should stick to earned runs & range factor for now. ER must have some 1/2 way decent correlation with "the truth".
If McLain's example is typical, where one WAR version is over 1/3 higher than the other, then to even get a ballpark accurate #, perhaps we should look at various versions of WAR & triangulate to what degree each is valuable in these cases. Even if the rules had not changed at all, that does not show that the analysis of players & value should not change.
October 7th, 2010 at 6:30 am
Chuck-
Barkfart said "HILARIOUS! The sports writers and others who watch baseball for a living are "universally wrong" and the SABRE-heads who might never watch a single game are right. That is precious."
To your point that no understanding comes from statistical analysis, I think that is simply wrong. Statistical analysis helps us determine whether stealing bases is conducive to winning or not. Obviously, an observer can pick up on times that a SB helped in a specific situation or a CS hurt. But without a more thorough analysis, how can we judge the approach in general and determine when, if ever, to employ it? How do we determine if certain players were good or productive or valuable if we have no way of measuring "goodness" or "producitivity" or "value"? Whether we use HRs and BA or WAR and WPA, we are still using statistics to measure performance. The more effective a statistic, the better understanding it yields. A guy's SB total is a statistic that is pretty easy to determine. But how do I know if his 150 SBs and 70 CSs helped or hurt his team? I probably can't determine this from those raw numbers alone, and even watching every game will still only yield so much data. By knowing the 'break even point', where a guy's base stealing goes from positive to negative, is necessary. Going further and using a stat like WPA (which is far from perfect) gives me more specific understanding of the context of those steals and what this individual player did.
SBs is a perfect stat in the sense that it tells me exactly how many bases a guy stole. Great. Grand. What is sucks at is telling me what type of baserunner that guy is, or how fast he is, or how much he contributed to his team. If you stick to the former use of it, so be it. When you try to use it to argue the latter, the stat is piss poor. More advanced metrics, despite all their warts, are far better at getting to those deep understandings, even if they aren't perfect.
October 7th, 2010 at 7:57 am
or example, taking McLain's 1968, if you strip his defensive support out of the replacement value and use earned runs instead of runs, he'd be 74 RAR, or about 1.54 times as valuable, and then you'd have to adjust to convert it back to the RA scale (since wins are based on runs, not earned runs) - running the numbers quickly, I get a WAR a little over 8 for that season, which is quite a difference from his 5.9.
It seems like you are, but I want to make sure you're adjusting replacement level to consider earned runs instead of total runs.
It looks like if you do a similar calculation for luis tiant, his numbers don't change much, suggesting that McClain may indeed have had a slightly better season if the defense numbers aren't helpful.
A quick check on that for me was to realize today that there is a perfectly good way to look at how to balance ERA+ and IP in determining seasonal or career value (something I was doing by the seat of my pants in a previous thread on 1968 pitching). Simply multiply the two.
It's not normalized to anything we care about like runs or wins or distance from replacement level, but looking purely at relative numbers between two above average pitchers it's a pretty good proxy for what you're looking for (total value based on using purely earned runs, with no accounting for good/bad fielding other than whether runs are scored earned/unearned). If you do that, McClain does come out slightly better than Tiant for the year at ~51k vs. ~48k. This still tells you that despite the win difference, the seasons are close.
And we don't really know how bad/good the fielding numbers were for that year. A quick look at the important positions indicates that a couple of guys had bigger numbers than their avg for Rfield that year, but no really huge outliers. So there may be some fielding error making McClain look worse than he was in the WAR calc, but there's no really big bit in there. The variances from average among the starters, account for maybe 1 WAR. The team seemed to have pretty good fielders that year, so the truth of his season probably lies somewhere in between the two numbers 5.9 and 8.
October 7th, 2010 at 9:57 am
# 135 BSK
"To your point that no understanding comes from statistical analysis"
I hope you're not referring to me? I've never said anything remotely like that. Unless you mean that my saying that "WAR is way overused" as an example.
October 7th, 2010 at 10:01 am
To: Chuck re: Ty Cobb
More accurately, what I'm saying is this. Not having seen him play, I cannot point to his WAR and say with certainty that he could. In fact, where old time players are concerned I can't say anything with certainty. I can conjecture, I can dream, I can shoot my mouth off with the best of them.
October 7th, 2010 at 10:09 am
#119 Sullivan- regarding McClain
I don't have a pitching coach to quote, but we have Al Kaline and all the other veterans who are still so in touch with the Tiger fans. Now, before you shout that their all homer-fans, remember- most people on the Tigers didn't like Denny McClain. He brought a lot of shame to this team and community- and continues to do so.
To a man, they all say that McClain's 68 was the best season they ever saw from a pitcher. And people like Kaline battled against the best that the 60s had to offer. Ask them if they thought his season didn't rank in the top 500 ever and they'd laugh in your face.
Why don't you go the next step and find a real person who played in 68 and quote them as saying; "eh, he was overrated".
THERE'S MY REAL OBJECTION TO WAR. Not one of McClain's critics on this page bothers to research the thoughts of an actual baseball player!
October 7th, 2010 at 10:21 am
Barkfart-
I was resopnding to Chuck. I quoted you earlier to show where the original assertion came from.
October 7th, 2010 at 1:46 pm
"That's kind of the problem I have, though, Mr. Sullivan, because while medicine and physics, etc, have changed in the past 80 years, baseball hasn't."
The laws of nature haven't changed in the past 80 years. The human body hasn't changed, at least, not any more than the game of baseball has.
What has changed is our understanding of how those things work, and our mental and mathematical models. And the reason they have changed is because of what we have learned about how to make them closer to the underlying phenomena they are trying to explain.
The fact that things are and have been changing, is an indicator that current medical practice or physics understanding is very unlikely to be TheTRUTH(tm), and the same is true for advanced baseball stats. But I'll still take current advice over what was current 30 years ago.
October 7th, 2010 at 4:49 pm
BSK,
Referring to your comment from post #121
"Yes, advanced stats help us try to quantify the apparently unquantifiable, to make order from chaos, sense from nonsense, music from noise. If you had no idea what baseball was or how it was played, you would never be able to come up with a sophisticated metric that had any credibility or reliability."
That's the thing, though, most things that happen during a baseball game ARE unquantifiable, the fact that people attempt to quantify them with sophisticated metrics shows they have little understanding of how the game is played.
An example;
From #135
"But how do I know if his 150 SBs and 70 CSs helped or hurt his team?"
Not all stolen base attempts are equal.
Although advanced metrics consider them to be.
Shane Victorino attempting a steal in the first inning is different than in the eighth inning, even if the score is the same.
Could be a different batter. Or pitcher. Later in the game means less margin for error because, theoretically, his team has fewer at bats remaining to try and score. Conditions could be different.
You also have to look at baseball rules.
If, during an attempted pickoff play, the runner is put out at second base, it is scored a caught stealing and an attempt, even if there was none. Or, if he beats the throw to the next base, he is given credit for an SB, despite being in a rundown for half an hour.
A runner advancing on the backend of a double steal is given a successful attempt even if he walked into an unoccupied base because the play was on the lead runner.
"But without a more thorough analysis, how can we judge the approach in general and determine when, if ever, to employ it?"
I listen to sports radio on the way to work everyday, in the morning, it's Colin Cowherd. He had Keith Law on today, and Law was bragging about "criticizing Gardenhire for having Hudson bunt in the first inning" during Game One.
While I personally would have preferred Span to steal second THEN have Hudson bunt, I agree with Gardenhire's decision.
CC Sabathia hadn't pitched in eight days. Baseball players, especially pitchers, and especially starting pitchers, are creatures of habit. Giving a pitcher an extra three days off is significant in his mindset and preparation. Gardenhire felt Sabathia would have rythym problems in the first inning and being aggressive both on the bases and at bat may lead to an early run.
That's an unquantifiable.
Advanced metrics can't say that was an optimum time to be aggressive, either with a stolen base or a bunt, although it clearly was.
Law believes no team should ever bunt in the first inning.
He is wrong.
"How do we determine if certain players were good or productive or valuable if we have no way of measuring "goodness" or "producitivity" or "value"?"
When the WAR believers put together their results from 2009 and came up with the numbers they did for Ben Zobrist, that should have been an indication to napalm the formula and start over.
At the beginning of the season, Zobrist was a utility player forced into full time play due to injuries to starting players. He FAR outperformed his own talent level and FAR outperformed his salary, so, to those extents, he was valuable (unquantifiables), but you can't throw him into a bucket with Pujols and ARod and Howard, etc, and expect anyone to take the result seriously.
"By knowing the 'break even point', where a guy's base stealing goes from positive to negative, is necessary."
I remember a conversation twenty some odd years ago with Bobby Bragan, who was a teammate and manager of Hank Aaron with the Braves. Bragan said, when talking about Aaron's baserunning ability, that, "every one of his stolen bases led to a run."
Hank Aaron stole 240 bases in his career, or roughly ten per season. His success rate was 77%.
Rickey Henderson stole 1406 bases in his career, or roughly 55 per season. His success rate was 81%.
Henderson stole five times as many bases as Aaron, yet there is only a four percent success rate difference.
Basically, their effectiveness was the same, although they were completely different type base stealers. Aaron ran to benefit his team, Rickey ran to benefit Rickey.
Aaron had the lower break even point despite 1400 fewer attempts, because, individually, his attempts meant more to his team. Aaron was running in 2-1 games, Rickey was running in the eighth inning with his team down 8-2.
"What is sucks at is telling me what type of baserunner that guy is, or how fast he is, or how much he contributed to his team."
No stat can tell you how fast a runner is, unless the data is obtained using a stopwatch.
WAR and these other stats treat everything that happens during a game as being equal, which they most certainly are not.
If his numbers were exactly the same but he played for the Yankees instead of the Pirates, would Andrew McCutchen's WAR be different?
And, if not, why?
October 7th, 2010 at 7:43 pm
Chuck-
You've left me nothing to respond with because you responded with anecdotes and beliefs. How can I argue that?
You argue that pitchers are creatures of habit. How do you know? How do you know the mindset of a guy who had an extra 3 days off? You don't. And to pretend you do is bullcrap, honestly. Zobrist doesn't fit your preconceived mold of what an elite ball player is. No one is throwing him in a bucket with those other guys in terms of talent level. But, for one season, Zobrist produced at an elite level. Just because you don't like it doesn't mean it's not true.
I realize there is evidence outside the statistical. But you have not provided any of it. Please get back to me when you can.
Also, as you reference KLaw, he is far from a hardcore sabremetrician or statistician or whatever windmill you are tilting at. He is a former professional scout who heads up ESPN's baseball scouting department. He bases his analysis on far more than what the numbers say. So, if a guy with both the statistical and scout background suggests a strategy is wrong, I'm likely to believe him. And while I'm sure you will respond with something about Gardenhire's experience and expertise and such, let's not pretend that being an MLB manager necessarily qualifies one as the most knowledgeable about baseball strategy. Managers are hired for many reasons, only some of which have to do with actual on field results.
October 7th, 2010 at 8:40 pm
"He is a former professional scout who heads up ESPN's baseball scouting department."
Keith Law was a book nerd who wrote for BP and got hired as a "special assignment consultant" through a meeting with JP Ricciardi at the Winter Meetings.
He advanced to an assistant to the GM role in a NON BASEBALL related capacity and did some special assignment scouting, specifically at the lower minor league level.
He is NOT the head of ESPN's Scouting Department, matter of fact, no such department exists.
Law writes for ESPN blogs from both a stat, and business perspective, which is where his expertise lies.
His scouting experience is limited, at best.
Let's not give credit where none is warranted.
October 7th, 2010 at 9:55 pm
Keith is the Senior Writer for ESPN's Scouts Inc. I guess calling it their "scouting department" was too confusing for someone who seems only capable of copy and pasting from Wikipedia. Read his columns, his chats, and his blogs (both personal and on ESPN); listen to his radio pieces; watch his television interviews. The man is sent all around America, scouting major and minor league talent. Unless you think that is all a lie, I think it is fair to say the man works in the capacity of a professional scout. I guess since it didn't appear in Wikipedia, that means it never happened, write?
Regardless, you've offered nothing to counter my assertion, which is that you are either unwilling or unable to offer any REAL evidence. First, you start with mostly made-up stuff. Next, you source out of Wikipedia without even taking the time to conduct a simple Google search to get a wee bit more data. Fail.
October 8th, 2010 at 9:09 am
Interesting debate here that has also been going on in other threads. I will not go nearly as far as some here, since I think the WAR is the single best metric to evaluate a player. It is perhaps worthy of 50% of the evaluation process, however, I will not support the abuse by some in the ways that they apply the WAR as if it is fact and 90% of the evaluation ---as if the evaluation process starts and stops with the WAR --it does not. There will always be nuanced areas of baseball that can be critically assessed fairy accurately --just b/c you can't put a number to it doesn't mean you should just ignore that part of the game.
October 8th, 2010 at 9:58 am
Not all stolen base attempts are equal.
Although advanced metrics consider them to be.
Untrue. Some do, some don't. I'm quite certain all raw stats consider them equal, and have for 150 years. Why not direct your complaints at the raw stats, which obviously don't fulfill your needs, instead of the advanced stats which are trying to improve on them?
you can't throw [Zobrist] into a bucket with Pujols and ARod and Howard, etc, and expect anyone to take the result seriously.
When the HR believers put together their results from 2010 and came up with the numbers they did for Jose Bautista, that should have been an indication to napalm the formula and start over.
At the beginning of the season, Bautista was a utility player forced into full time play due to injuries to starting players. He FAR outperformed his own talent level and FAR outperformed his salary, so, to those extents, he was valuable (unquantifiables), but you can't throw him into a bucket with Pujols and ARod and Howard, etc, and expect anyone to take the result seriously.
Bragan said, when talking about Aaron's baserunning ability, that, "every one of his stolen bases led to a run."
Aaron had the lower break even point despite 1400 fewer attempts, because, individually, his attempts meant more to his team. Aaron was running in 2-1 games, Rickey was running in the eighth inning with his team down 8-2.
What a load of isht. All the evidence you need to prove or disprove this is right on the site. Why not do a little research and support your statement? You're really gonna throw something like that out as proof of anything? Commentary like that is the reason sabermetrics exists. Unfortunately, even if I do the work and should find your claim is completely wrong, we know you will still twist the evidence and say it's missing something about how the heart and spirit in Aaron's slides inspired his team, even as the umpire called him out.
October 8th, 2010 at 10:01 am
Dammit, the second-to-last paragraph should be italicized as well. Honestly, I don't think I'm screwing up my tags every time, I think this site has some problem with paragraph breaks.
Bragan said, when talking about Aaron's baserunning ability, that, "every one of his stolen bases led to a run."
Aaron had the lower break even point despite 1400 fewer attempts, because, individually, his attempts meant more to his team. Aaron was running in 2-1 games, Rickey was running in the eighth inning with his team down 8-2.
What a load of isht. All the evidence you need to prove or disprove this is right on the site. Why not do a little research and support your statement? You're really gonna throw something like that out as proof of anything? Commentary like that is the reason sabermetrics exists. Unfortunately, even if I do the work and should find your claim is completely wrong, we know you will still twist the evidence and say it's missing something about how the heart and spirit in Aaron's slides inspired his team, even as the umpire called him out.
October 8th, 2010 at 10:30 am
#142:
"Not all stolen base attempts are equal.
Although advanced metrics consider them to be.
Shane Victorino attempting a steal in the first inning is different than in the eighth inning, even if the score is the same.
Could be a different batter. Or pitcher. Later in the game means less margin for error because, theoretically, his team has fewer at bats remaining to try and score. Conditions could be different."
Agreed on all accounts except the second. There's Win Probability stats that are designed largely to examine these types of differences in context, and they weigh stats according to how each action contributes to the likelihood of a win in a given situation.
* * *
"CC Sabathia hadn't pitched in eight days. Baseball players, especially pitchers, and especially starting pitchers, are creatures of habit. Giving a pitcher an extra three days off is significant in his mindset and preparation. Gardenhire felt Sabathia would have rythym problems in the first inning and being aggressive both on the bases and at bat may lead to an early run.
That's an unquantifiable.
Advanced metrics can't say that was an optimum time to be aggressive, either with a stolen base or a bunt, although it clearly was.
Law believes no team should ever bunt in the first inning.
He is wrong."
Considering the Twins failed to score even one run in that inning, the results certainly don't back up this assertion. Even if we assume that the assessment is correct and C.C.'s rhythm was off, though, why would a bunt make sense in that situation? If the idea is to get to him early, why use up an out and limit your potential for a big inning at exactly the time when you feel he's most vulnerable?
* * *
"When the WAR believers put together their results from 2009 and came up with the numbers they did for Ben Zobrist, that should have been an indication to napalm the formula and start over.
At the beginning of the season, Zobrist was a utility player forced into full time play due to injuries to starting players. He FAR outperformed his own talent level and FAR outperformed his salary, so, to those extents, he was valuable (unquantifiables), but you can't throw him into a bucket with Pujols and ARod and Howard, etc, and expect anyone to take the result seriously."
I suppose, then, that Willie McGee's 1985 means that we should throw away batting average, because clearly the stat must be wrong if it says Willie McGee is in the same class as Tony Gwynn. And home runs must be wrong too, because Davey Johnson hit 43 HR in 1973, and he's certainly no Hank Aaron (who hit only 40 that year).
Look, a player can have a career year in WAR just like in any other stat. I have never seen a stathead suggest that Ben Zobrist is as good a player as Albert Pujols or any other true elite player. Even for just 2009, the people who actually USE the tool will generally tell you that you should use caution when looking at a single season's worth of defensive data, because defensive metric take more time to stabilize than offensive ones. But, like with offense, it IS possible and not uncommon for a player to over-perform on defense in a given year.
What some people don't seem to realize is that stats are tools. When you argue that all statheads are wrong because WAR isn't perfect, without considering that the people who use the tool do have opinions and do differ in how they adjust when looking at WAR, you're essentially arguing that a hammer is a useless tool because it won't build a house for you. Statheads are people, and like in anything, different people have different levels of skill when using the tools of their trade.
October 8th, 2010 at 11:43 am
We can just agree to disagree. I love the complex numbers, but I also think you can critically assess things with reasonable accuracy without the aid of numbers. I know people that are all about the "knowing" are uncomfortable with that, but stats are just a tool --they are not definitive. I've seen way to much statistical abuse in my day to know that stats aren't the end all be all. I don't think stats are the only way to 100% evaluate a player or come to reasonable conclusions. Overall, I think you and i come to the same conclusions on most players. Maybe 5% is where there'd be a difference. In those cases you'll look at the stats and really not go much further, but i want to go further --maybe that's flawed, maybe it's not. I think clutch does exist, I think there are different levels of pressure and certainly pitching in a 1-0 in May when you're 10 games out does not match pitching in a 4-2 game when you're team is down 1 game to the wild card leader in October. I think it's reasonable to say that the pitcher pitching in the 1-0 game in May 10 games out certainly has personal pressure --they should if they're professional. However, the pitcher pitching in a pennant race in a 4-2 game with a chance to tie the wild card leader on October 3 has a whole other level of team pressure added. I don't think that's a stretch.
October 8th, 2010 at 10:17 pm
Matt Y-
It's not agreeing to disagree. You are being disingenuous and intellectually dishonest.
Your post is full of things you "think". And you are certainly entitled to your opinion. But you don't know. You have no idea. None of us do, unless there are former professional athletes among us. There are a lot of things we DO know. And we ought to use those as much as possible. As for the things we DON'T know and CAN'T know... there is certainly room to discuss them. No one is saying those are out of the conversation. But if your point is that we should put more emphasis on conjecture than on fact, you might as well believe in fairy tales. And if you further assert that facts are not facts because they disagree with your conjecture, you have officially jumped into the uselessness that is cyclical logic. There is simply no prove to what you "think". That doesn't mean it is necessarily wrong. But there is also no proof that unicorns don't exist.
October 9th, 2010 at 10:30 am
Nothing i said is disingenuous. In fact, it seems you have some hellbent desire to push this stuff with no deviation --that seems disingenuous and intellectually dishonest IMO.
As you wrote: "But if your point is that we should put more emphasis on conjecture than on fact"
You continue to twist, contort and flat out misquote. Where did I say that "conjecture" should hold more weight. You argument here should be held for the far more WAR-weary people that just dismiss it. And yes, b/c you can't put a number to it means that it should categorically be put in the unknown conjecture category. I disagree. I'm not going to be a slave to stats as you are. They are a tool, a good tool, a tool to be used as a primary guide, but they are still nonetheless just a tool.
October 9th, 2010 at 11:13 am
As you implied --" Where did I say that "conjecture" should hold more weight than complex stats".
This statement is what continues to baffle since you act as if because you have stats you therefore have the facts and only facts and therefore everyone else is just talking or guessing-- As if you can't make good sound arguments in other ways. I guess to you philosophy, sociology, psychology etc. are fields that aren't based in much fact or even any fact. The difference here isn't that I think conjecture should hold more weight than complex stats (I don't and have said so many times), but that I think "conjecture" (obviously __some___ arguments, even ones without stats, are based in more fact than other arguments) holds more weight than what YOU think it should. This gets back to the lets just agree to disagree.
My approach is to give some legitimacy (it would be less legitimacy than people that use stats) to other ways of evaluation, yours is to push your stats only driven mentality as fact so that everyone else's is nothing more than guesswork. That is rather disingenuous and intellectually dishonest IMO.
October 9th, 2010 at 10:05 pm
Matt Y-
You equate WAR with stats-minded people, ignoring:
A) that there are so many stats besides WAR, some of which are favored by folks who favor WAR and some of which are favored by folks who hate WAR
B) that your "WAR = 50%" argument is based in nothing other then your desire to retain the right to offer nonsensical arguments into the discussion
C) that WAR has never been argued as the end-all, be-all of stats
You have constructed a strawman, which is, in and of itself, an intellectually dishonest approach to discourse. Engage me fairly or not at all.
October 11th, 2010 at 10:05 am
Matt Y,
I actually agree that there's a lot to be learned outside the numbers, especially when it comes to projection. I suspect that a lot of why some prospects develop and others stagnate has as much to do with psychological factors as it does physical ones.
The difficulty I think is threefold:
1. The research that's been done into the non-statistical elements of player profiles is limited.
2. The public access to the type of information that would help us do this research is limited.
3. Much of the information that we go by in these area is based on faulty sources.
Points 1 and 2 really tie in together. It's certainly not easy to get the kind of direct access to professional athletes that it would require for a real psychological study, so we're left with a lot of guesswork. We know that people do respond differently to pressure, which suggests a likely variance in clutch performance, but what we don't know is whether that difference manifests itself at the top level. How does the pressure of performing to actually earn a job in the first place compare with the pressure of playing in a pennant race? Is it even possible for a player to have a MLB career if he's adversely affected by pressure? In addition, we hear a lot about clutch performance, but we shouldn't overlook the ability to perform outside of pressure situations. In a lot of industries, it's easy to find people who get motivated around deadline time and rise to the occasion when the pressure's on, but people who are able to stay motivated outside of pressure situation and consistently do their best work tend to be more valuable than those who need pressure to perform at their best. Again, I question whether there's a significant variance at the MLB level here, because we're dealing with the best.
If we were to study the effects of pressure, the above is one of the confounding problems with looking at statistics. If a player does perform better in clutch situations, is that because he's risen his level under the pressure, or is it because he tends to party a little harder during the lower pressure situations of the season-long grind and performs at a lower level without the added motivation? To some degree, we probably can identify clutch players by those whose performance doesn't change in higher pressure situations, but it's important to be careful not to let the popular narrative dictate the assessment of a player. It's also important not to be too quick with applying a characteristic to the player, rather than the play. Sometimes a single defining play or game or series series creates a narrative for a player that's not necessarily indicative of the player himself.
As an example, compare the popular perceptions of A-Rod and Jeter. Now we haven't heard the A-Rod "unclutch" criticisms so much since his big performance in last year's win, but the traditional perception of the two is that Jeter's a postseason superstar, and A-Rod becomes a bum in October (or at least did until last year). Both players have had excellent careers, and have been consistent. Both seem to be hard workers, and neither missed many games outside of real injuries. Both have an overall postseason line that looks like a typical season for them. Both got their first big contract after the 2000 season, and both were extremely handsomely rewarded with a premium on their star-power.
The big differences between the two are that Jeter has a likable public personality and A-Rod doesn't, that Jeter played for a huge market team that was willing and able to give him a massive contract, whereas A-Rod had to go elsewhere for his, and that Jeter was surrounded by a championship-caliber team throughout his entire career. From a narrative standpoint, Jeter's a hero and A-Rod's a villain, and much of what we notice about their performance is fueled by that narrative. It would be worth looking at whether there are real differences between the two beyond the numbers that translates to additional value on the baseball field, but it's not easy to effectively gauge that.
How do you look past the narrative and the player's media charisma in order to remain objective and apply only the known information that is relevant to the evaluation of a player?
October 12th, 2010 at 2:00 pm
Please BSK, yes, there's lots of other stats, but the WAR is the one that's really, really being pushed at times at the expense of nearly all context,,,almost as if to re-write things and completely ignore the past. Sabermetricians want to act like most Hall votes have been done by bums that just get it wrong --really, most people in the HALL, except for the really dubious Vet picks of 50-60's are pretty good overall. People act like electing Hall players is an easy endeavor that takes place in some vacuum. I'm not convinced that it would get better with these more complex stats --i think there's plenty of evidence it would be pretty comparable at the end of the day. That gets to the next point, I agree that A-rod wasn't nearly the playoff clutch -bum some wanted to make him out to be, but he clearly also was not the clutch playoff player that Jeter was --really, Jeter's more ridiculously consistent than anything.
I get the whole overblowing a players perspective, why do you think I proposed a pWAR to counterbalance traditional voters overblowing a Morris Game 7 performance. However, one of the big, very big problems with the complex stats is it boils the crap out of things that nearly no context or history remains. Again, i love the WAR, but it shouldn't be nearly all of the evaluation process --again, I weigh it the most, but it might make up 50% of the evaluation process regardless of what some want to push.
October 12th, 2010 at 2:31 pm
"How do you look past the narrative and the player's media charisma in order to remain objective and apply only the known information that is relevant to the evaluation of a player"?
As for this comment above, you do it b/c clutch, pressure performance, playing in a pennant race, winning championships and how you deal with the media does indeed count or matter in some way. It's all part of the narrative (as are the complex stats) and I would argue that it's not wise to completely look past the "full" narrative. That's the problem here --people want to just look at the part of the narrative they are most comfortable with. Look at the WAR first, look at the complex stats as well (these are part of the narrative), count this stuff higher if you want, but also look at the rest of the narrative, b/c whether you want to acknowledge it or not, the rest of the narrative matters too, and in most cases for a legitimate reasons!! Yes, we can overblow the small sample size, but certain things transcend the stats alone regardless of some thinking the complex stats are lightning in a bottle. --like getting the big hit or winning championshps and being a major player in that, or even being good with the media so that you take the pressure off of other players that might be pressing. Now, I'm not advocating that this should make up the other 50% of the evaluation process, but numbers alone can only say so much. I have both a statistical background and some years on this planet so that i can see both sides perhaps a bit more than some 22 year old can (no knock on age here, just age does matter to a degree, sorry).
I have always lobbied for a more moderate approach, I get equally annoyed when traditional voters overblow things, but I also get annoyed when these complex stats are pushed by some so much that they want to re-write everything including history.