POLL: One on one–Steve Garvey or Bobby Abreu?
Posted by Andy on September 8, 2010
(We interrupt our weekly Hall of Fame polls for this one-on-one poll.)
On a recent post, some readers got into a debate over which player was better--Steve Garvey or Bobby Abreu.
Right off, we need to talk about what the word "better" means. Career value? Peak value? Post-season achievements?
It's hard to say, and the bottom line is that it's up to each individual reader to decide what's meant by "better". For me, it means which player I'd rather have had on my favorite team for their entire career. So that gets at career value, but not necessarily if that value was accumulated through a number of so-so years.
So, click through and let's discuss Garvey vs. Abreu.
Right off, Garvey and Abreu have fairly close plate appearance totals. Garvey had 9,466 and Abreu has 8,995 and counting. These are close enough that I feel fairly comfortable comparing career totals straight up instead of rates.
So here are arguments in support of Garvey:
- He had 272 career homers, just 1 short of Abreu's current total, despite playing in a much lower offense era.
- He has more RBI than Abreu (1308 vs 1257) despite, again, playing in a low-offense era. Incidentally, Abreu has batted primarily 3rd in his career while Garvey batted primarily 4th.
- Despite very similar career batting averages (.296 for Abreu and .294 for Garvey), Garvey had 6 different 200-hit seasons and Abreu has none. This is due to 2 factors: Garvey had a higher BA than Abreu in his peak seasons, and Abreu walked a lot, getting fewer at-bats. Despite having only about 500 more plate appearances than Abreu, Garvey has about 1,300 more at-bats.
- Garvey won an MVP award and had 2.46 career MVP shares, as compared to no wins and just 0.17 MVP shares for Abreu.
- Garvey leads in All-Star appearances 10 to 2.
- Garvey was part of 5 World Series teams and hit well in the playoffs. In the 5 NLCS, he batted .356/.383/.678 with 8 HR and 21 RBI in 22 games. (To keep some balance I should mention that although he batted .319 in 28 World Series games, he had just 6 RBI in those games and his teams lost 4 of those series.)
And in support of Abreu:
- Abreu has a career OPS+ of 130, as compared to 116 for Garvey. (Keep in mind these numbers correct for offensive level of each player's era.) This means that while Garvey was solidly above-average, Abreu has been excellent.
- The single-biggest reason for the OPS discrepancy is OBP, where Abreu leads .400 to .329, a huge gap even consider the differences in run-scoring.
- Abreu has a huge lead in WAR, at 57.5 t0 35.9.
- Abreu has stolen 367 bases at a 75% success rate while Garvey stole only 83 at a 57% success rate.
- Garvey's edge in RBI is balanced by Abreu's edge in runs scored (1346 Β to 1143) although Abreu is helped by batting higher in the lineup and playing in an era with more overall runs scored. (But getting on base a lot more is the biggest factor...)
- Both were so-so fielders but Abreu was a little better, and at a much more important defensive position (RF) than Garvey (1B).
This is just a snapshot view of the two guys, of course. Please add your own opinions and vote below.
September 9th, 2010 at 9:26 am
Again, I'd love to see the top 20-30 BB guys from the last 20 years --I strongly suspect that the divide b/w sabermetricians and others when evaluating borderliners comes from BB and defense.
I have no idea if this will work....
Rk Player BB From To
1 Barry Bonds 2274 1990 2007
2 Frank Thomas 1667 1990 2008
3 Jim Thome 1619 1991 2009
4 Gary Sheffield 1441 1990 2009
5 Jeff Bagwell 1401 1991 2005
6 Chipper Jones 1343 1993 2009
7 Manny Ramirez 1283 1993 2009
8 John Olerud 1275 1990 2005
9 Jason Giambi 1262 1995 2009
10 Edgar Martinez 1260 1990 2004
11 Ken Griffey 1259 1990 2009
12 Bobby Abreu 1254 1996 2009
13 Rafael Palmeiro 1228 1990 2005
14 Rickey Henderson 1194 1990 2003
15 Brian Giles 1183 1995 2009
16 Luis Gonzalez 1155 1990 2008
17 Todd Helton 1130 1997 2009
18 Carlos Delgado 1109 1993 2009
19 Craig Biggio 1104 1990 2007
20 Mark McGwire 1083 1990 2001
21 Bernie Williams 1069 1991 2006
22 Robin Ventura 1067 1990 2004
23 Alex Rodriguez 1060 1994 2009
24 Fred McGriff 1047 1990 2004
25 Lance Berkman 980 1999 2009
26 Tony Phillips 977 1990 1999
27 Jim Edmonds 974 1993 2008
28 Tim Salmon 970 1992 2006
29 Omar Vizquel 950 1990 2009
30 Kenny Lofton 945 1991 2007
Provided by Baseball-Reference.com: View Play Index Tool UsedGenerated 9/9/2010.
September 9th, 2010 at 9:26 am
The thing I have against the scrud is... his peak years, though being the best player on the team at the turn of the century, those teams failed miserably. It's not personal, though it sounds that way, he never made that team into a real contender
September 9th, 2010 at 9:30 am
Please, everyone, lets not pile on Abreu --he's a really good player with lots of subtle offensive positives that need a good look at. Both of these players are overrated by/for very different reasons IMO.
Perhaps BB should be factored in according to position in the order something like 1-9 as follows:
1-.30
2-.27
3-.24
4-.21
5-.23
6-.25
7-.27
8-.29
9-.31
September 9th, 2010 at 9:31 am
Nope. Anyone have any idea if we can post a table somehow? I don't know how to html.
Hopefully this is slightly better:
Rk Player BB From To
1...Barry Bonds......2274 1990 2007
2...Frank Thomas.....1667 1990 2008
3...Jim Thome........1619 1991 2009
4...Gary Sheffield...1441 1990 2009
5...Jeff Bagwell.....1401 1991 2005
6...Chipper Jones....1343 1993 2009
7...Manny Ramirez....1283 1993 2009
8...John Olerud......1275 1990 2005
9...Jason Giambi.....1262 1995 2009
10..Edgar Martinez...1260 1990 2004
11..Ken Griffey......1259 1990 2009
12..Bobby Abreu......1254 1996 2009
13..Rafael Palmeiro..1228 1990 2005
14..Rickey Henderson.1194 1990 2003
15..Brian Giles......1183 1995 2009
16..Luis Gonzalez....1155 1990 2008
17..Todd Helton......1130 1997 2009
18..Carlos Delgado...1109 1993 2009
19..Craig Biggio.....1104 1990 2007
20..Mark McGwire.....1083 1990 2001
21..Bernie Williams..1069 1991 2006
22..Robin Ventura....1067 1990 2004
23..Alex Rodriguez...1060 1994 2009
24..Fred McGriff.....1047 1990 2004
25..Lance Berkman.....980 1999 2009
26..Tony Phillips.....977 1990 1999
27..Jim Edmonds.......974 1993 2008
28..Tim Salmon........970 1992 2006
29..Omar Vizquel......950 1990 2009
30..Kenny Lofton......945 1991 2007
September 9th, 2010 at 9:32 am
I think your second table is good enough. I believe to post anything other than links in the comments, you need to be an editor of this blog (not even just an author...that's not good enough.)
September 9th, 2010 at 9:34 am
Sorry that's sort of unreadable. Anyway, you can look up the top BB guys yourself on PI.
Matt, it wouldn't be any surprise if players whose value is more concentrated in walks and defense are underrated by the general public. Defense has traditionally never been measured except in the broadest terms -- good/average/bad, with no real sense of how much it is worth in comparison to offense. And walks have historically been seen more as a pitching failure than a batting success.
September 9th, 2010 at 9:39 am
The thing I have against the scrud is... his peak years, though being the best player on the team at the turn of the century, those teams failed miserably. It's not personal, though it sounds that way, he never made that team into a real contender
Who are these individual players who can make teams into real contenders? Steve Garvey?
September 9th, 2010 at 9:39 am
Sorry, first paragraph there is quoting #102
September 9th, 2010 at 9:54 am
Well, certainly many of the most hotly debated ones are on that list Twisto (thanks) --Abreu, Sheffield, Lofton, Martinez, Edmonds and Helton --I suspect Walker's not far off this list. Additionally, there's many other players on the list that sabermetricians like to bring up from time to time like Olerud, Ventura, and Giles -- I don't think anyone in their right mind would vote any of these guys in--even the great majority of statheads --well, I've certainly heard a few Olerud arguments. I suspect if a list was done for the 1970-90 period you'd have both Evans', Randolph, Nettles, Hernandez and others on it.
September 9th, 2010 at 10:08 am
"Again, I'd love to see the top 20-30 BB guys from the last 20 years --I strongly suspect that the divide b/w sabermetricians and others when evaluating borderliners comes from BB and defense."
Even if we concede that this is true (something I'm not necessarily willing to do)... who is to say the sabermetricians are wrong...?
At this point, I think the use of the term sabermetricians has outlived its usefulness, essentially since it has become a term of disparagement in many circles. Let's just break evaluators down into two categories: statheads and seamheads. Statheads use stats, be they RBIs or WAR, to evaluate players. And seamheads use scouting, be it armchair analysis or professional expertise, to evaluate players. There are going to be excellent evaluators in both groups and horrible evaluators in both groups. Each group has its strengths and its weaknesses and both hold an important place in how we view the game, both from a professional standpoint (in terms of how teams approach player evaluation) and from an amateur standpoint (how we as fans approach it). The suggestion that one approach or the other is objectively and absolutely better is simply to ignore reality. A lot can be gleamed from stats, just as a lot can be gleamed from scouting. The better the stats or the better the scouting, the better the evaluation. When Chass went after WAR and dismissed it as "stats nonsense" or whatever while using OTHER stats to argue with it, he revealed the disingenuous attack on "stats" we see in the anti-sabermetrics argument... it isn't stats that most detractors have a problem with, just CERTAIN stats they don't like. And there definitely are certain stats that are of less use than others. They just tend to be the ones most of the establishment hold up.
September 9th, 2010 at 10:09 am
I meant to add that I have no problem with sabermetricians identifying themselves as such or using the term as they see fit. My point was that when the term is most often used outside of that group, it misrepresents what sabermetrics is and attempts to draw a line in the sand that doesn't really exist.
September 9th, 2010 at 10:13 am
Basically Walker would be nearly next on list with 913 BB.
I agree Twisto that defense and getting on base have been overlooked for far too long....but, perhaps sabermetriciains have taken it a little too far the other way as well -- Just a thought. I'm actually pretty in the middle on this stuff --more than that might come across a times. I see what I can consider ridiculous arguments from both sides. Garvey clearly isn't a HoFer, but neither is Abreu IMO. I don't see Edmonds as a slamdunk as some sabermetricians do --he's clearly borderline and his WAR of 69 overrates him some --BB and defense are at the heart of the issue.
September 9th, 2010 at 10:16 am
Excellently put BSK.
September 9th, 2010 at 10:25 am
However, seamheads and statheads come in various forms, especially the stathead group --so, this grouping is a bit too general in a sense. I do see a difference b/w sabermetricians, statheads, and seamheads, and I see merit in all forms of evaluation with no single group cornering the market on the best approach. I think the more "eclectic" stathead approach is perhaps the most moderate IMO. I certainly mean no-ill will towards sabermetricians --they have done a tremendous amount forwarding the evaluation process, and that should NOT be overlooked! WAR is a very useful tool, but it has great limitations just as the armchair smell-test does --IMO, the smell-test is worse, but as you describe, some scouts are better than others for sure.
September 9th, 2010 at 10:32 am
One other mistake people are making is the assumption that walks are the only factor responsible for the gap between them. There's more to it than that, so walks aren't valued as highly as people seem to be suggesting. Abreu also has a 10 run lead in baserunning, which if anything seems conservative, a 29 run lead in staying out of double plays, which is certainly believable given the difference, a 27 run lead in defense and a 25 run difference in positional difficulty, all balanced against Garvey's 12 run edge in reaching on error.
All of that is worth about 6 WAR, or over a quarter of the difference between them in factors that are usually ignored unless we're looking at the extreme cases for fielders or baserunners.
There are a couple more things to keep in mind as well when looking at their career WAR:
- Off-peak value added: One of the things about Abreu's walks is that it's a skill that has aged well. When Abreu's power is down, he's still a very good table setter for the hitters after him. When his batting average is down, as it is this year, he has a wide enough range of skills to prevent him from being an offensive liability. With Garvey, when his power was down, he was basically a contact hitter without an offensive identity; not enough power to be good at driving in runs, and didn't get on base enough to set the table. When his batting average dipped, he simply created too many outs to be worth a lineup spot without outstanding power or defense.
- Peak season defense: If you look at the peak seasons for Abreu, three or his four top-rated seasons by WAR are fueled by spikes in his defensive defensive numbers. His defensive rankings are all over the place during his prime, but he did have a strong arm and led the league in RF assists during two of those three years. Garvey, on the other hand, was quite stable defensively through his prime, rating mostly above average without any huge years.
When looking at Matt Y's comment:
"I gave the edge to Abreu and voted for him, but I don't see the difference being 58 to 36 in WAR. I see Garvey as a good player, perhaps borderline, and I see Abreu as a half notch better and just now edging into borderline discussion."
That really doesn't differ from WAR's assessment, once you adjust for some of these factors. Looking at hitting, Abreu rates about half a win to a win better per year during their primes. The gap extends when you factor in defense and baserunning, and it extends further when you consider that Abreu's done considerably better at remaining valuable during his decline years than Garvey did.
Abreu's always struck me as the type of player who needs extraordinary longevity to be a Hall of Famer. He's around the cusp right now for career numbers, where he has the overall numbers to be considered but far from a clear choice. He had an excellent prime, but his best seasons were a notch below what it usually takes for a peak-heavy case, and he's not yet to the point where his career totals (either traditional or WAR) make him a clear choice. He's at the point where some players get in, if they're well liked or have triple-crown-heavy skillsets, but many don't. He's better than some Hall of Famers, but not as good as other who'll never get in.
At this point, I'd compare him to Bobby Bonds; a power/speed RF who gets on base, among the better outfielders during his prime without ever being the best, not a perennial all-star, but typically one of the best outfielders who didn't make it, no signature seasons, etc. Bonds had more power, and Abreu's a better contact hitter, but I'd put them in the same class of player, and their career showings for hardware won seem to support that notion. Bonds was probably better at his peak, but Abreu's aged more gracefully and needs to continue to do so in order to get any kind of ballot support.
I figure he needs to play about 4 more years, mostly as a regular in order to have a shot. That would but him up around 2800 hits, over 300 HR and 400 SB, around 1500+ RBI, and into the range where voters start taking him seriously as a longevity candidate. If he can reach that point, I suspect he'll garner enough support to stay on the ballot until OBP becomes a more mainstream method of evaluating players, and I could see him eventually gathering support.
Edmonds is a different matter altogether.
September 9th, 2010 at 10:43 am
"Outs made is a negative stat if you're Juan Pierre or Omar Moreno.
Outs made is a positive stat if you're Cal Ripken, Derek Jeter, or Steve Garvey."
I initially thought this was a snark post, until I read further and realized that it was 100% sincere. I can't believe that I need to say this, but outs are never a good thing (well some outs are better than others yes). The Drew mini-study I linked to above I would have thought would have made some here just a bit more humble, but I guess not. They should flip OBP around by subtracting it from 1, and call it "Not making an out" percentage, because that's really what it means. Outs are the most important stat in the world, in the end analysis: if you run out of them before you beat the other team, you lose. They are the only thing in the game which remotely acts like a clock does in other sports. A wasted out is a wasted opportunity gone forever.
Look at it this way: if a guy makes the third out, the chance of scoring any more runs in that inning is zero, zilch, from that point on to the end of time and all protons in the universe decay. If he walks instead, the inning is still alive. Look at it this way: assume that it's 2nd and 3rd, two outs. By walking, he's basically loaded the bases with the inning still going, vs. the guy who went up hacking and whiffed on 3 pitches, and everybody walks (heh) back to the dugout; that has inestimable value, and the Drew study indicates that. Maybe the bozos coming up behind him don't get the job done, but the point is that they _still_ have a chance to do the damage. Garvey must have killed tons of rallies in his own inestimable fashion; sure he hit some home runs here and there, but that can't overcome all the baserunners he stranded.
I think I'll examine one of Barry's big walking seasons next, check to see what happened after he walked with 2 outs (intentional or otherwise), vs. what happened when he swung away, and compare the runs scored in the inning from that point onwards.
September 9th, 2010 at 10:44 am
But, Matt Y, how do we decide who is a sabermetrician and who isn't, outside of self-identification? SABR was originally a semi-organized group, if I understand it correctly. Many of those now dubbed sabermetricians have no formal relationship with the original SABR. Yet they carry the title anyway, deserved or not, desired or not.
Stats, and statheads, exist on a continuum, with great stats and great evaluators/developers of stats on one end and useless stats and poor evaluators/developers of stats on the other end. If we want to draw designations on that line, it would only be about quality. Some very non-SABR stats are held up quite highly in the "sabermetric" crowd. Is OBP considered sabermetric? It's been around forever. On the other hand, you have Saves which are very often used by those on the worse end of the stats continuum but are a far more "new school" stat relatively speaking (though certainly not a SABR stat). We've reached a point where these designations are pretty useless, because there is so much crossing of whatever SABR line ever existed in the first place.
Let's just look at all the stats and figure out what each one is REALLY telling us. RBIs isn't an inherently flawed stat. It tells us how many times a batter drove a run in. That has its usefulness. The problem arises when he ascribe a value to a stat that it does not or cannot possess. RBIs do NOT tell us how effective a player is at creating runs. Anyone who tries to use it in such a way is using it incorrectly. So, RBIs isn't "wrong" or "bad", it's just limited. All stats have their limits. Personally, I lean toward stats that are A) predictive and/or B) get at underlying talent/value. Some of those stats are very new and may be a direct or indirect result of SABR, some are very old, some are in between. I don't draw lines in the stat as to who created the stat or when or how. I only look at how effective a stat is at doing what it claims to do and how useful what it claims to do is.
September 9th, 2010 at 10:53 am
Many good points #115. Baserunner and avoiding DP's make a difference --not sure if I'm willing to go to quite that level of difference, but it's duly noted. With that said, I still don't think it makes up for a 22 WAR difference --these two players were still more comparable when you add in Garvey's game streak, clutch hitting and playoff performance. To some extent this exposes what i consider a pretty significant flaw in WAR --it should just add playoff stats in in some tweaked formulaic manner to account for the larger stage and importance of games. This way it would counterbalance the writers exaggeration of playoff performance (i.e Morris). Maybe Jack's playoff performance would have added another 5-6 WAR points thus bring him to a WAR of 45 --still below the line, whereas writers want to elevate him so much that it adds 15+ points of WAR. To an extent the same could be said of Garvey --maybe his playoff performance would add 5-6 WAR points thus bringing him to 43 and still below the line. Just add the dang playoff numbers to WAR and WPA. It makes sense.
September 9th, 2010 at 11:12 am
Oy vey, this is why we stat stat guys get mocked by the Joe Six Pack crowd.
John (116 regarding 92), my comment was sincere, and obviously I understand that outs are bad. My point is that while statistics are an "objective historical record" and by far the best means of measuring the value of events that have occurred, stats are not the only valuable method. Often, the human factor can be more important.
All of you remember when Derek Jeter dove face first into the stands in 2004 to make a game saving catch against the Red Sox in the heat of a pennant race. We've all seen this clip.
What is remembered by Yankee and Red Sox fans watching the game is not the story of the catch, but the story of how the camera was fixed on Nomar Garciaparra sitting like a schmuck on the bench. You see, Nomar was fighting with the Sox over his contract and upcoming free agency at the time. He refused to play in the game because of some "nagging/imaginary" injury.
Nomar at this time was the Red Sox' Derek Jeter. Jeter's bloody face and Nomar's indifference defined the difference between the two. Nomar was traded a week later in a trade that was awful on paper, but the Red Sox won the championship that year.
Make no mistake, Garvey, Ripken, and Jeter are great players because they are great players. They also play every day even against tough pitchers and play with nagging injuries. The fact that they play EVERY day even though that lessens their overall averages and leads to more outs is a positive, not a negative.
P.S. As a Yankee fan, I was thrilled when the Yanks got Abreu. I was less thrilled when he refused to bat leadoff. I was even less thrilled watching every ball hit in his direction bounce off the wall for a double because he refused to go after it. There is a reason the Phillies dumped him, the Yankees dumped him, and both teams won championships after he left. None of this will show up in stats.
September 9th, 2010 at 11:22 am
"...these two players were still more comparable when you add in Garvey's game streak, clutch hitting and playoff performance."
You need to explain why these three things SHOULD be elements of WAR.
What does the game streak matter? Is a few games off a year (which is all Abreu took off for the bulk of his career) really a difference-maker? As Andy pointed out, Abreu missed between 5 and 10 games a year, including an 8 game stretch where he missed just 22 games (about the length of Garvey's streak). For Garvey's streak to add enough value to overtake Abreu, we'd have to assume he would have single-handedly won all those games for his team. Fat chance.
How are you defining clutch hitting? If Garvey has a clear edge in this, it would show up in WPA or aLI. Please demonstrate.
You do make good points about playoff performance, but I think it should be handled separately, for a variety of reasons. First, not everyone makes it to the playoffs, often for reasons far beyond their control. Second, their is difficultly making an apples-to-apples comparison between playoff games and regular season games. The games are played quite differently, both in terms of strategy and structure; the extended schedule changes things and teams end up playing with different rules (at least recently) in the WS.
September 9th, 2010 at 11:26 am
Micker-
I just don't know that there is any evidence to support your claim. Which is not to say there isn't a psychological impact and factor at play. Just that we can't prove it. And since we can't, we should probably focus on what we CAN prove.
The problem with such narratives is that they are often set in stone. If the situation was reversed and Jeter was on the bench with a questionable injury, the narrative would have been, "Jeter must be hurting more than he's letting on, or he'd be in there. Just look at this face. He's clearly frustrated he's not on the field making a difference."
We put guys as good and bad and then only look at the pieces of evidence that support those constructs. Jeter is clutch because of a couple big hits. If he K's with the bases loaded in the bottom of the 9th of a 1-run game, people say, "He'll get 'em next time." And if every next time, he K's, they'll still keep saying it. Because of that big HR in '01 or whatever memory is burned into their retinas.
Let's take the emotion (and hyperbole) out of it. Yes, players have an impact beyond what they do with the bat and glove and on the basepaths. But I doubt that impact is large enough to overcome any major differences between what they do do with the bat and glove and on the basepaths.
September 9th, 2010 at 11:40 am
BSK, you've just defined why we need to balance sabermetrics with scouting. If I had to pick one, I'd go with the stats, but there is more going on than what we see in the numbers.
In earlier posts about Keith Hernandez, I made reference to how he was viewed while playing. Bill James wrote about the importance of doing this some 25 years ago, because revionist history can cloud our historical judgement. At the end of 1987, Darryl was comperable to Doc, and Carter was comperable to Mex. New Yorkers at the time did not consider Carter a more viable HOF candidate than Hernandez. Keith's career just ended too abruptly.
Garvey did not age well, but his career totals are much better than Hernandez and good enough for the HOF. Check out his 133 on Bill James' HOF ranker.
The fact of how he was evaluated while playing is also relevant to the discussion.
My memory of 2004 is neither emotional or hyperbole. It's confirmable, just not convertible to statistics.
September 9th, 2010 at 12:02 pm
Micker-
What happened (Jeter made the dive; Nomar sat the bench) is fact. The impact of those facts on the other players and the teams and the outcomes of games is conjecture, at best.
And, yes, I do agree that scouting and stats should be balanced (though not necessarily 50/50, depending on the goal). However, we must use GOOD scouting, and rarely does that happen from someone watching on TV, both because of a general lack of expertise and an inability to truly see what is going on on the field, due to limited angles, replays, etc.
September 9th, 2010 at 12:13 pm
Of course the streak has little to do with his WAR BSK --I'm not arguing that it should be part of his WAR, however, these other things added legitimate value to Garvey and besides there's more ways to access a players worth --which gets at one of your points you were originally making. I get the point how having Garvey in the lineup every day probably added a psychological edge to his team even if you can make the argument he was statistically hurting his team. You could make a great similar argument about Jeter right now. These two points of contetion cross (statistical value and pshychological value), and certainly if Jeter keeps this up going into next year, he'll have to be dropped in the lineup, but right now is probably not the time to do it even though he looks awful . He's statistically hurting his team and you could come up with many players in the minors that probably could add similar statistical value, however, should he really be dropped to 8th in the lineup --probably not because he adds a psychological additive affect that's hard to quantify --I think that's the point that Micker17 is getting at. None of this stuff exists in a vacuum, and there is a point where a stat becomes too distilled. WAR is reflective of a players worth, and to me it's a useful tool, but to me it's also very distilled.
I've already made a perfectly logical argument as to why playoff numbers should be added to the WAR --to me, sabermetrcians are just shooting themselves in the foot for not adding it. Until they do you'll always have writers adding too much value to a Morris or detracting too much value to a Kevin Brown b/c of playoff performance.. Just add it and it'll provide more weight to your arguments. I think it's a disservice to just say that Glavine or Pettitte's 200+ IP doesn't count much -- What if Glavine ended his career with 260 wins and people then decided he wasn't hall worthy. To me, Glavine was still a HoFer even if he ended with 260 wins. As for clutch hitting, it too adds to a players worth, and the WPA gets at some of it, but WAR really doesn't.
September 9th, 2010 at 12:28 pm
In short, Jeter's dive added immense psychological value to his team that isn't quantifiable, just as the image of Nomar on the bench subtracted immense value that isn't quantifiable. However, that psychological effect added up to --trade of Nomar and the Red sox winning a championship --do they still win with him? Very, very likely not. This cause and effect is categorically unquantifiable in many ways although I'm sure others will try. Sometimes there is a such a thing as addition through subtraction and even subtraction through addition.
September 9th, 2010 at 12:34 pm
"Garvey did not age well, but his career totals are much better than Hernandez and good enough for the HOF. Check out his 133 on Bill James' HOF ranker."
His career stats are only good enough for the hall if you are the kind of stat-ignoramus writer that James was modeling with his Hall Monitor, which was not a ranking, but a prediction of how the writers of a certain time period would view a candidate. Hall Monitor is heavy on triple crown stats and awards, and completely ignores many sources of value. The fact that Garvey failed to make it despite a 133 Hall Monitor actually suggests that the writers are somewhat better judges of good baseball players than Bill James expected when he came up with that stat.
September 9th, 2010 at 12:54 pm
Matt Y-
I think there would be value to calculating some sort of Playoff WAR, but that it should be a separate stat. So you could look at a player and say he was worth X regular season wins and Y post season wins and perhaps someone would even go so far as to tell us the relative value of X and Y in relation to one another. I just think that adding them up is adding apples and oranges. That is the same reason we don't add post-season stats to regular-season stats in general: they're not compiled in the same context.
To the psychological standpoint, I think you are overstating it. I doubt Ortiz went to the plate, poised to have a great at bat, glanced back over his shoulder and saw Nomar pouting, lost his spirits, and struck out. I think the psyche of individual players and, more importantly, an organizational unit are far too complex to say, "Pouty guys make team bad; funny guy make team good!"
September 9th, 2010 at 12:54 pm
Hmm, insults from a guy probably living in his mother's basement. We all know what the Hall Monitor is for. While the standards of voting have continued to evolve, it's a useful tool and I enjoy tracking it.
The Hall Monitor actually has Garvey at 130, which would be enough to get him voted in if his candidacy did not coincide with the glory days of the steriod era.
So tell me, are the writers "better judges", or "stat-ignoramusus"?
September 9th, 2010 at 12:56 pm
Meanwhile, through the first couple hundred votes, Abreu led about 2-to-1 over Garvey, and now with 400-some votes it's only 59-41 Abreu. Also I'm not sure why this poll has received many fewer votes than other recent polls.
September 9th, 2010 at 1:03 pm
#118
Is it a flaw in WAR that you're describing, or a lazy application in the use of WAR?
Batting average is a near perfect stat. It measures the frequency at which a player gets hits, and has a minuscule margin of error, susceptible only to the mistakes in scoring errors. If someone chooses to use it as a measure of the bulk of a hitter's offensive worth, does that mean it's a flawed stat, or that it's being used for a function beyond its capabilities?
WAR is more complex, and it's subject to a greater error margin due to the difficulty in measuring different components of a player's contributions, but it's fundamentally no different. It's an estimate of the regular season runs a player creates or saves compared to a replacement level, converted into wins based on the average run scoring of an era. Is it flawed because it doesn't include postseason numbers, or is that beyond the scope of what the stat measures?
The issue I don't think is with the stat itself, but rather with the false notion that you can rate look at WAR and rate them 58 to 36 and stop there. It's a starting point, not an end point.
From there, you can look at the shape of how players generated that WAR. Did a guy have a great peak and hang around as a fringe player, or was he a consistent performer for 15-20 years? Did he contribute due to dominance with his share of injuries, or did he build his resume due to durability? Was it offense or defense that generated the WAR? How does the player's defensive component compare with other defensive statistics out there, and does it seem reasonable? Career WAR doesn't tell you that, but you can explore these questions to get a fuller picture of a player's career shape.
Next, there are things that WAR doesn't touch, like postseason performance, or whether a player contributed to pennant races, or whether his popularity generated added team revenue. The thing about WAR is that it doesn't eliminate the need for subjectivity. It's not wrong to look at something like postseason performance or to give it a fair weight in evaluating a player's career accomplishments, but it's an element beyond simple runs created that everyone should consider separately how they treat it. Similarly, it's up to each individual to determine how he treat a player's peak as compared to his total accomplishments, or how to treat an injury prone player versus a durable one, or how to make adjustments for military service.
And I don't think any of this should be included in WAR. Consider postseason stats. Do you simply add the stats directly to a player's resume? Do you give them extra weight due to the higher leverage of the situation, or lesser weight due to the fact that opportunity plays such a key role in them? Do you adjust due to the higher level of competition faced in the postseason? It's measurable, and I think it should be measured separately, but it becomes subjective in terms of how to include that performance in the total.
A pWAR stat? (playoff WAR) Makes sense. But adding it directly to WAR would reduce WAR's usefulness by trying to be too inclusive. You'd be adding a subjective element to it, making the stat only useful if you're philosophically in agreement with the method it uses to weight the performance. It's a tool, and a good one, but no one should rely exclusively on one tool. WAR give a great starting point and captures a lot of a player's career worth, but there's always subjectivity in the refinement of our opinions.
September 9th, 2010 at 1:12 pm
John DiFool (#116),
I just got the idea to check on something, after having read your comment about "all the baserunners he (Garvey) stranded."
According to BBR, 17% of all runners who were on base scored during his at bats. That's versus an MLB average of 14% across the arc of his career.
Garvey was over 20% in four different seasons, and had a 12-year run (1973-'84) in which he was at 17% or better every year.
This stat covers all runners who scored for whatever reason when Garvey was at the plate.
This would seem to belie in part the statement that his lack of walks hurt the Dodgers and Padres during those years, although you may still argue that a walk extends some innings.
However, taking walks in some situations may have also precluded his swinging the bat and getting a hit. You can't assume that all walks negate outs.
It seems to me that the walks argument has at least one hole: it has to assume that there would have been a negative result otherwise. Even if the walk comes on a 3-1 pitch, maybe that strike -- if taken by a less-aggressive hitter -- might have resulted in a hit, if swung at.
What if the hitter takes strike three on a borderline (or, not-so-borderline) 3-2 pitch, trying to be concious of the chance of reaching base on a pass?
Working the count is fine, but a good hitter needs to be ready to swing the bat at any time. The best pitch to hit may be the first one offered. The second pitch may be a great one. You can't judge at that point whether he is even going to be in a position to accept a walk.
What about the so-called common wisdom that you'll get a great pitch to hit on 2-0 or 3-1?
How do you know if the pitcher will even throw any more pitches out of the zone during that plate appearance? It's all a guessing game, and I've seen too many hitters frozen, with the bat on their shoulders, when strike three drops right over the heart of the plate.
This is one of those great things to total-up, then point out how much the walks helped build/continue the rally. But you can't know in the moment if it'll even happen. It's no different than a Strat-O-Matic roll of the dice.
Either way, Garvey was certainly among the most efficient run producers, from the measure of percentage of total runners plated.
Comparing Garvey with Hernandez, Keith's career average was at 18%, peaking at a high of 23% in 1979. Here are a few others:
Gary Carter: 16%
Andre Dawson: 17%
Jim Rice: 17%
Dwight Evans: 15%
Kirk Gibson: 16%
Ryne Sandberg: 16%
Joe Carter: 17%
Dave Winfield: 17% (at or above MLB average every year he was in New York)
Don Mattingly: 18%
Jack Clark: 16%
Bill Buckner: 17%
Darrell Evans: 15%
Dale Murphy: 16%
September 9th, 2010 at 1:36 pm
I agree there should be a pWAR rather than just adding it to the regular --I like that concept -- I think in many ways it can be added to regular season WAR, but I like the idea of the two being kept separate to reflect different contexts. I also agree on most points #130.
I do find it a little interesting nobody bit into the :
Perhaps BB should be factored in according to position in the order something like 1-9 as follows:
1-.30
2-.27
3-.24
4-.21
5-.23
6-.25
7-.27
8-.29
9-.31
September 9th, 2010 at 1:38 pm
WilsonC for the win...
September 9th, 2010 at 1:39 pm
MattY-
I think that's because your list really didn't make any sense.
September 9th, 2010 at 1:48 pm
Matt, if you're going to adjust the value of BB by batting order position, why not homers? And why not everything else? And this eventually leads you to WPA.
September 9th, 2010 at 2:07 pm
Matt Y,
I agree with the idea that walks should be more heavily factored, based on spots in the order. A leadoff guy -- and the number-two hitter, as well -- should be looking for any means possible to get on base.
The same is true at the bottom of the order.
But the middle-of-the-lineup guys should be looking for pitches to drive. That's why they're there. At that point, especially, taking a couple of pitches to work the count may leave your top run-producers in an 0-2 hole, hurting the chance to build the inning.
One thing not mentioned in all this is the chance of even getting a walk. Even the worst control pitchers generally yield more hits than walks. The chances are often times better swinging the bat. The key is selectiveness...not going out of the hitting zone, when you do swing.
For a sub-.230 hitter, looking for a pass may be a better choice, based simply on the average. Beyond that, I'd want my hitters swinging the bat. His chances are better at getting a hit than being handed a pass.
How many hitters could draw a walk in as many as 20% of their plate appearances? Even Rickey Henderson only walked in 16% of his plate appearances.
September 9th, 2010 at 2:14 pm
"According to BBR, 17% of all runners who were on base scored during his at bats. That's versus an MLB average of 14% across the arc of his career."
That's how many people scored *during* his AB. So this stat represents how many people he drove in (or who stole home or came in on a WP or other error on his AB/BIP). It doesn't include people who come home in the inning *after* his PA, which is where we'd expect the difference to show up for people who draw a lot of walks.
September 9th, 2010 at 2:18 pm
Keith Law, in his chat wrap, just addressed this in a great way. The idea that BBs are only achieved through batters standing there, staring at pitches, hoping to get to 4 balls is generally not true (I would say entirely not true had I not watched Mark Bellhorn play for the Sox). The batters ARE looking for pitches to drive, but might not get one. Or might foul a couple off. They choose not to swing at pitches that will likely result in an out, risking taking a strike in the process. Should they get to 4 balls before a seemingly drivable pitch comes to them, they take the walk. If they get pitches to drive, they go for it. Why is that so hard to grasp?
September 9th, 2010 at 2:20 pm
Because BB is what seems to be overvalued to a degree, not homers (the list you made Twisto validates that depending on what way you want approach the evaluation process) --my list does make sense, especially if a similar concept/idea has already been added to an adjacent or cousin metric of the WAR (i.e WPA). That's good this concept is in the WPA, which was what i was trying to lead us to -- that there's more ways than one to evaluate player value. Personally I think WAR is much better than WPA, particularly for position players, but admit they tell two different stories of the same coin in a sense. If anything this just adds much more to the argument that there's many different ways to assess a players value. WAR, WPA, scouts-look, arm-chair, traditional stats etc. I think the WAR approach holds more value than the others, but it's not 90%, 80% or even 50% better than another's way of evaluation. So, not so fast on the WilsonC for the win. π
September 9th, 2010 at 2:28 pm
"WAR, WPA, scouts-look, arm-chair, traditional stats etc."
The problem is, you want to act as if these are all equally valid and effective. And they are not. And that has been proven (at least with regards to stats). Subjective analysis (aka scouting) certainly has a place, but generally a very different one than stats. Stats are better for telling us what did happen, while scouting (good scouting, at least) is better at telling us what will happen. Scouts can observe mechanics and make-up in a way that stats cannot. Put the best of both together, and you get a damn good way of evaluating players, both looking back and looking forward. The key is picking the best... and too often we are using far less than the best stats (Pitcher W-L, BA, RsBI, etc). And best doesn't mean perfect, but best means best. And if something is best, warts and all, than it's the best. Is there room for improvement? Almost assuredly. Does that mean we dismiss it until it is perfected? Absolutely not.
September 9th, 2010 at 2:34 pm
Michael E. Sullivan (#137),
Correct. It's the percentage of runners actually on base during his plate appearances who scored during those plate appearances.
Sorry if my wording left it ambiguous.
It also does not take into account where the runners were. A home run scores all runners, but a walk won't. A single with runners on second and third likely scores both, while a walk doesn't score anyone, unless subsequent hitters succeed.
With an MLB-average around 14-16%, depending on the years covered, does it make sense to pass up an RBI opportunity when it arises?
I'm assuming that the number of runners who scored "by other means" than the direct end result of the pitcher-hitter confrontation is likely to be small, across the board, and not attributable to any particular hitter's talent.
September 9th, 2010 at 2:47 pm
Where did I ever say these were equal forms of evaluation!!??. I have always said that the WAR and complex stats are a better way of evaluation, but they're not infinitely better. Look at my above posts --I even said people shouldn't bash Abreau as "scrud" and that he deserves a closer look. This attitude to push stats as mere fact, and because I have stats, then i must be right and you must be wrong is what turns people off --this is nothing more than a microcosm of our society and political arenas today. If anything I have tried to only bridge the gap b/w writer and sabermetrician. This black and white, cause and effect, direct linear relationship, if this than that, is what turns people off. Shoot, at the bottom of my post at #139 I said, " I think the WAR approach holds more value than the others, but it's not 90%, 80% or even 50% better than another's way of evaluation. Moderation is the key to me.
September 9th, 2010 at 2:54 pm
"One thing not mentioned in all this is the chance of even getting a walk. Even the worst control pitchers generally yield more hits than walks. The chances are often times better swinging the bat. The key is selectiveness...not going out of the hitting zone, when you do swing."
This would be relevant in cases where the batters who draw a lot of walks had low enough batting averages to offset the value of their walks.
But somebody like Bobby Abreu isn't sitting on lots of good pitches in order to get a few walks. If that was his deal, he'd either be striking out a lot more, or putting more weak balls in play from bad counts. A guy who draws walks in a way that increases his Rbat and OPS+ is not giving up a lot of hits to do it. All these metrics value hits much higher than walks.
Basically, Abreu *is* being selective, and it's showing up on the field, just not in his batting average. Because batting average ignores a major source of offensive value -- the free added baserunner with no outs that you get when you walk.
One thing to note is that the ratio between the run value of an average hit from an average power player, and the value of a walk defines how good a hitter should be getting intentional walks in very standard situations. I've seen tango (who designed Rbat) give .8 as the value for an average hit, and .3 for a walk. That implies that if a guy with average power bats .375, you should intentionally walk them about as often as not. Well, I think there have been three guys who batted >.375 for a year since Ted Williams and none had much power, so this jibes with our intuitive sense that the only players even worth *considering* walking close to half the time are guys like Bonds or Ruth in their best years, who have very high BA *and* amazing power.
What if we reduced the value of a walk to .25 runs and left hits alone? For reference, this would take roughly 4.25 WAR away from Abreu relative to Garvey. Well, now the BA that makes you indifferent to intentionally walking vs. pitching to an average power guy is .313. So basically, this is suggesting that pitchers should have been intentionally walking Vlad Guerrero and Todd Helton 300+ times a year in their primes, and pretty much any .300 hitter with decent power a whole lot as well. Do you think that's reasonable? Does that make sense?
This is the crux. If you think WAR and other stats are overvaluing walks by that much, then by implication you *also* must think that there should be many more intentional walks than people who agree with the WAR weights think. And if you disagree with walk value enough to make an appreciable difference in career WAR totals/ranks, then you are by implication suggesting that the people who manage baseball teams are making *major* mistakes by ever pitching to the best hitters in high leverage situations.
September 9th, 2010 at 2:56 pm
I guess in short, I don't agree with scouting is for scouts, and evaluating one's career is for sabermetricians. A scout can evaluate a player's career, as can a fervent baseball follower or writer using stats and many other things. This compartmentalizing of duties just reaffirms this lack of listening and compromise to me. Everyone at the table has merit, some have more merit than others. I think the BB thing being overvalued has some merit --do I think BB's should be discounted completely as some have even suggested --no. However, do I think some of these high BB guys are overrated --yes. Is the truth between these two groups probably more in the middle --yes. I think the list provided bares that out to a degree.
September 9th, 2010 at 2:57 pm
First time I ever saw RsBI instead of RBIs...nice.
September 9th, 2010 at 3:03 pm
Matt Y-
Scouts don't do the job that you seem to think they do. They'll watch a player or prospect a handful of times and get a sense of what his future prospects are. The best will do this in conjunction with a statistical analysis as well. This, in no way, qualifies them to evaluate an entire player's career, except insofar as the better ones have the statistical chops to ferret through the noise and figure out what is what for a guy. But to say that a scout who watched a guy in the minors only is in position to evaluate a guy's 20 year career simply misrepresents what a scout does.
You are right that there is sometimes too much black-and-white arguing going on, on BOTH sides. But, the fact is, not everything is up for debate. Some things can be empirically proven, and we shouldn't dismiss that which is because it seems to cease debate. Yes, there are lots of gray areas and lots of nuance for us to debate, but we shouldn't turn EVERYTHING into a gray area just because we don't like what the black/white facts say or because it means declaring some folks are wrong. People are going to be wrong. And that's okay.
September 9th, 2010 at 3:03 pm
TMcklev-
Not mine originally, so I can't take credit. But it is fun.
September 9th, 2010 at 3:04 pm
I don't think BB's are easy and that people get them by just standing there. It is a skill, and a valued one, and OBP is a better stat than Batting Avg --what I'm questioning is to how much better is it? To me, BB are a BIT overvalued --that didn't mean that i thought it should be thrown out. I questioned it to an extent --that's all. How much better is WAR? It's better, and it should be valued as such, but it's not infinitely better. Most of us are a lot closer on our thinking than it comes across at times --so, it's all good. Blogs aren't the best sometimes for communication. I love today's social networking atmosphere even though it's anti-social in a way. π
September 9th, 2010 at 3:06 pm
Matt Y-
A lot of your argument seems to be, "Yea, you guys are right, but I'm not comfortable with that being reality, so let's just agree that there is room to argue, even if there isn't". You want everything to be arguable. Not everything is. And I do realize that a lot of modern stats are used to end debates (sometimes rightfully so) and there is something lost from the fan experience as a result of that. I get it. That's an unfortunate externality. At the same time, modern stats have contributed a lot, and gave us new things to debate now that we can compare eras or account for park effects. And my gut tells me they've contributed more to the fan experience than they've taken away. And they're not being shoved down people's throats. Plenty of people dismiss them and so be it. But those folks probably shouldn't be on these boards. Just like I'm weary of wondering on to a lot of the boards on ESPN.com. And that's okay.
September 9th, 2010 at 3:09 pm
If we realize OBP is better than BA, why would we ever use BA? We would only do so if we were trying to get at something that BA could do that OBP can't do. Which isn't that useful. Even if OBP is only 1% better than BA, it's better and should be preferred in nearly all instances. No amount of lead in BA overcomes a lead in OBP, unless we are looking solely at a player's ability to hit the ball and get on base. Which is a skill, but a far less important one than getting on base in general.
September 9th, 2010 at 3:12 pm
I never thought a scouts job was to evaluate a player's career. Where did I say that ? Again, this points to my division of labor issue. This compartmentailizing of things. What I'm saying is, I personally would want to get a scouts take on a players career if I was having an educated discussion about a player's Hall chances. I would also want to hear what the writer sitting in the seats has to say. I would also want to get a sabermetrcians take on the player --what i wouldn't want is just the take from 10 sabermetricians. I would walk away and put the sabermetrcians perspective at the top, but I would also seriously contemplate what the scout and writer had to say as well though. I guess this is where we disagree. To think scouts don't follow players through their careers is foolish --they follow them, and I'd be willing to bet their evaluation would have legitimate points and merit that stats sabermetrics would miss.
September 9th, 2010 at 3:19 pm
wow #149 --that's just my point that's missed. There is NO absolute right or wrong even though you clearly want to make it black and white. OBP is better, WAR is better, OPS+ is better and everyone else are just fools that should go away. Even though we all know there are much greater margins of error with WAR and that not all contexts are equal. It's all good. Reality, like beauty, is in the eye of the beholder. π
September 9th, 2010 at 3:22 pm
Matt and Jeff,
I agree with the notion that walks (as well as other offensive stats) do have different contextual values, and batting order probably does affect this. I'm also going on the assumption that your weighting system is more of an example to be refined by study, rather than specific numbers that should be used.
Looking at a hitter like Abreu, I would expect the average value of his walks to be greater than average, rather than less, for two reasons: he was a good baserunner, and he typically had good hitters hitting after him.
It's important to look at a hitter based on what he does, not based on a hypothetical notion of what he should be doing. Perhaps Abreu would have been more valuable if he was sometimes more aggressive, trading a handful of walks for various types of hits to offset more outs. However, when looking at the value he did produce, what's more valuable: a walk in front of Scott Rolen, or a walk in front of Randy Wolf? Regardless of whether it's the optimal strategy, a walk in front of a good hitter is more likely to lead to runs than one in front of a bad one.
A part of it is too tight a definition of a player's "role" in the batting order. Sure you want your cleanup hitter to have power and drive in a lot of runs, but a free baserunner for the number five hitter is still typically better than a free baserunner for the number eight guy.
Let's move away from stats for a second and look at things from the opposite angle. There were a few at bats in a game the other day that made me think about this type of thing. Shawn Marcum was pitching with a runner on first, up by two, and working down and away got Vlad to chase a pitch for a weak DP. Later in the game, in a similar situation, up by three, the hitter (Cruz maybe?) laid off some change ups on close pitches to draw a walk, creating a jam with a runner on first and second and none out. The next batter chased the same close pitch down and away to ground into a double play, limiting the inning's damage to one run.
And that's not an uncommon pattern. Sometimes teams will manage strings of hits together, but when a pretty good pitcher is on the mound not making tons of mistakes, I usually feel comfortable when the ball's being put in play that the defense will do their job and limit the damage. Sometimes good hitters beat you, but it's frustrating to watch a pitcher beat himself. The most tense situations tend to be when a couple hitters work consecutive walks, especially when it's a case of not giving in on the close pitches. Many of the great tension releases take place when a good hitter chases a pitch out of the strike zone to ground into a double play. That's purely my emotional reaction when looking at things from the pitching team's perspective, but it's reflective of how I look at a walk's value. Almost everyone agrees that control problems for a pitcher are a major flaw, but for hitters it's often looked at as a passive skill and valued less.
September 9th, 2010 at 3:37 pm
"According to BBR, 17% of all runners who were on base scored during his at bats. That's versus an MLB average of 14% across the arc of his career.
Garvey was over 20% in four different seasons, and had a 12-year run (1973-'84) in which he was at 17% or better every year.
This stat covers all runners who scored for whatever reason when Garvey was at the plate.
This would seem to belie in part the statement that his lack of walks hurt the Dodgers and Padres during those years, although you may still argue that a walk extends some innings."
The thing you _really_ need to look at is the percentage of runners who scored on or AFTER the given plate appearance during the given inning. Otherwise any positive offensive occurences which don't lead directly to an RBI (run) during that specific plate appearance become non-events, but that's precisely what I'm trying to get at, so yes it appears you've missed my point. Walks DO extend innings, giving the people behind him a chance to do the damage; the batter who walks (or is HBP-same diff in the end) deserves credit for keeping the inning going if runs score later, but in the traditional stats he gets zilch credit (unless he himself scores later).
Well, I've been doing just that, for Bonds 2004 and Ortiz 2006-they provide an interesting contrast because the hitters behind Bonds sucked horribly (so far almost all of his IBB's with 2 outs paid off handsomely for the opposing team, but the ones with 1 or 0 (!) outs quite often turned into disasters). Papi had Manny behind him tho.
Don't know when I'll finish it (and to do the whole league to get a control group would take several lifetimes), but it will provide a starting point.
September 9th, 2010 at 3:46 pm
Yes, I should be doing it for Garvey vs. Abreu (for some reason I didn't think they detailed box scores for the 70's here yet). I'll choose one representative year for each (not about to do both their entire careers). I'll do Barry and Papi afterwards.
September 9th, 2010 at 4:04 pm
Michael E. Sullivan (#143),
No, I'm not. The hitter's best chance is to be aggressive, and to not expect a walk. A 25-30% chance of a hit in a player's at bats is certainly better than a 10-15% chance of drawing a walk, right?
But the pitcher should be using that 25-30% chance against the hitter as best possible, and going at him. An intentional walk, after all, is a 100% chance of the batter reaching base.
I generally feel the only reason to walk the guy intentionally is if no run will score as a direct result of the walk, and the pitcher has a better chance of getting out of the inning by facing the next guy instead.
By this reasoning, second and third with one out might merit a pass, to set up the force play/double play. Depends on the hitter/scenario.
However, the main point is that (to me, anyway) it generally makes no sense to go to the plate looking for a walk. Taking good -- "hitters'" -- pitches often hurts your team's chances of scoring runs.
As has been said several times, hitters generally go up looking for a pitch to drive, but will take a walk if given. That's fine, on the surface.
Guys who go up looking at the first few pitches as a matter of habit can also leave themselves open to situations -- such as an 0-2 count -- that present little chance of success.
Hitters who are aggressive will draw fewer walks, however that's not all bad. In the moment, who knows if the pitcher would have gone on to walk him? That's my point. Maybe he gets a hit, maybe not. A .300 hitter has a 3-in-10 chance by swinging the bat. But the chance for a walk is usually much less.
Maybe that missed opportunity to swing away will hurt. And if a hitter's chance of walking is generally so much lower than getting a hit, why take that chance?
OBA tells part of the story, but it can be misleading. Walks are essentially accidents.
WilsonC (#153),
Yes, Abreau's baserunning skills are a further benefit. However, he may also be held back by runners in front of him. The situation is everything.
Another thing to consider is, if I'm Abreau, with runners in scoring position, I'm looking for a pitch to hit. If I draw a walk, I may set up the DP opportunity. That also has to be factored into the equation.
Guys like Garvey made the choice to hack, with excellent results. Taking (or waiting out) a pass is showing a lot of faith in what is essentially a 1-in-10 chance.
Also, by swinging the bat, you open things up to the possibility of errors, and runs scoring on fly balls or fielders' choices.
September 9th, 2010 at 4:17 pm
I love WPA. I think it's the best stat to come around in a long time. However, it has one major fallacy: it doesn't consider batting order position.
Here's what I mean.
Let's imagine that your favorite team is up to bat in a scoreless game in the 6th inning. You have runners on first and second with 1 out. The next batter walks.
Is that good?
Well, if it was your #2 hitter, I would say that's damn good.You'll be licking your chops when your #3 hitter comes up with the bases loaded and 1 out.
If it was your #8 hitter and you're an NL team, it's not so good. You don't want to see your pitcher hitting in that scenario, and if the #8 batter had a pitch he could drive, you'd probably prefer that he'd swung at it rather than work out a walk. Now your pitcher is up. He's not a good hitter, plus maybe your manager is going to pull him for a pinch-hitter. Even if your pinch-hitter is a stud, he's now being used up, and if he fails in this instance, he's not available later in the game. Not a bad time to use him, necessarily, but it's limiting your options.
My issue with WPA is that you get just one number for what that walk to the hitter was worth, regardless of #2 or #8, AL or NL, when in fact there are drastically different values of true value to the team, and the true ones are hard to measure.
Thoughts on this, folks? I wonder how the WPA data would look if broken out by lineup order. Would there still be enough data to make it statistically significant?
September 9th, 2010 at 4:26 pm
John DiFool (#154),
Yes, but those runners (including Garvey) likely move farther on a hit, than a walk. A first-to-third advance on a single is better than first-to-second on a walk, right?
Again, given the 30% chance of getting a hit, versus the 10-12% chance of eventually getting a walk -- factoring in extra bases that can result from said hits -- also advances the chance that subsequent runs may result.
September 9th, 2010 at 4:59 pm
"I'm assuming that the number of runners who scored "by other means" than the direct end result of the pitcher-hitter confrontation is likely to be small, across the board, and not attributable to any particular hitter's talent."
It sounds like you're basically assuming away the value of a walk, when I thought the whole point was to try to figure out what it should be?
The value of a walk is that for the 70% of the time that swinging would have made an out (also don't forget the 2.8% of the time it makes 2 outs), you now have an extra chance for those other guys to drive in runs, *and* you have an extra runner on base that can be driven in. Keep in mind that the difference between Steve Garvey and a scrub at the plate is around .075 of batting average.
So suppose the three guys behind you in the order are scrubs and you're steve garvey at the plate with a runner at second and 1 out.
You swing vs. walk. If you swing, you get a hit 30% of the time and we'll assume they go home. The other 70 percent of the time, you make an out and get a scrub up with a 22.5% shot to drive the runner in. Between you and the next batter, you have a 45% chance of driving the runner in. But the two scrubs behind you have a 40% chance. So by taking a walk (if you have a choice), you give up a 5% chance to drive that single runner in, but you also add yourself on base. In the situations where you need more than one run, that has to be worth something also. Even with your team of scrubs, the chance that scrub 1 gets a hit is 22.5%, which will advance you into scoring position, and then 40% of the time from there, you will score in one of the next two hits. Also, scrub one could make an out, and then around 5% of the time, scrubs 2 *and* 3 will get hits in succession and you score. So you, the walked runner will score around 14% of the time in that situation even with a bunch of scrubs batting behind you. And they will drive in the runner you had only about 5% less often than if you swung. So, if you only need 1 run (tie score bottom of the ninth), you would probably go out of your way to avoid the walk. But if you can use two or more runs, the walk looks a lot better than an average swing.
This doesn't even consider that your marginal swing may not be at your batting average. It's not at all clear that the walk you could take, would be a 30% shot at a single or better if you swung somewhere in there. You are hitting in on 30% of the pitches you've chosen to take -- but that includes a bunch of fat ones along with the marginal swing/no-swing decisions where you chose to swing. The marginal ball you hit is certainly going to fall in less often than your normal batting average. In fact, it might be as low as the scrub's.
With rare exceptions (the kind where good hitters are often intentionally walked), walks are good. Deal with it.
September 9th, 2010 at 5:08 pm
JeffW, I think your analysis at 156 doesn't correctly address how hitters approach their at bats and sets up false choices.
First, I don't think any MLB hitter goes up to the plate "looking for walk." Some batters have more of a plan than others, but they are all basically trying to do the best they can with the pitches they see.
Second, they don't get a choice of the 30% chance to get a hit or the 15% chance to draw a walk. That comparison doesn't make sense because it's not how the game works. If they see a pitch down the middle, they may have a 50% chance to get a hit. If they see a pitch 3 inches off the plate, they may have a 5% chance of getting a hit. A .300 hitter doesn't have a 3-in-10 chance by swinging the bat. They get 3 hits out of every 10 AB by swinging the bat at pitches they think they can hit.
Sometimes hitters let hittable pitches go by because they weren't expecting it. Some hitters guess, and if they see a breaking ball when looking for a fastball, they can't pull the trigger, even if it ends up catching a lot of plate. Hitting is hard.
No one always takes the first few pitches. If they really did, pitchers would realize it and strike them out every time. Any actual MLB hitter is going to adjust, and if he starts seeing lots of 1st pitch strikes, he will swing at them.
Walks are absolutely not accidents. How could you possibly come to that conclusion?
Yes, some hitters sometimes let the best pitch to hit go by. Sometimes they were instructed to take, sometimes they are trying to let a guy steal, sometimes they just couldn't pull the trigger, sometimes they are being too patient against a pitcher who is pounding the zone. If Abreu, or anyone, was consistently letting good pitches go by, he wouldn't have been as successful as he's been. (Whether you like his style of play or not, he's a .300 hitter with over 800 extra-base hits. He has been successful.)
September 9th, 2010 at 5:11 pm
Hell, Andy, I'll do you one better! I think it should actually account for the specifics of the team. Walking a 2 hitter to get to Pujols batting 3rd is a lot better than walking a 2 hitter to get to Adam LaRoche. How great would it be to see the stat actually specific to the team? Probably impossible. But it sure would be something. Then, you could see something like, "The average team has a 56% chance of winning in this situation, but the Cardinals have a 57% chance because Pujols is up in the situation."
September 9th, 2010 at 5:18 pm
Matt Y-
"Reality, like beauty, is in the eye of the beholder."
Are you serious??? I don't mean to be smarmy, but there are objective, empirically proven facts in the world! A lead-off HR is better than a lead-off walk, no matter WHAT Tim McCarver thinks. SOME things are provable, black and white. Not everything, maybe not even a lot. But some are. And we shouldn't pretend that's not the case because it gives us the warm squishies to think we'll never be wrong.
JT-
Great post. Though I will quibble with this: "No one always takes the first few pitches. If they really did, pitchers would realize it and strike them out every time."
Again, I submit to you Mark Bellhorn. When he played for the Sox, he seemed to be a guaranteed 3-2 count every time. He would then proceed to take the pitch no matter what. Cockshot right down the middle and... take. It was amazing how useless his approach was once people figured it out. (I realize this is a bit colored by fan-colored glasses, but the man certainly did SEEM to be hoping for a walk most at bats. He was definitely one of those guys who got mistaken for having a good eye when he really simply got a lot of walks as a result of taking a lot of pitches.)
September 9th, 2010 at 5:51 pm
In the context you mention above you're right, but the problem is you then extrapolate it to many more complicated things that just aren't true. Yes, I am serious and live in a science world every day at work. I have debated the level that a walk is important and valued, I have not debated that a walk is of equal value to an out. And WAR is not one of those empirical things you're talking about where it's a simple matter of absolute 100% right and absolute 100% wrong. Maybe I;ve exposed something that does need a better look. Again, I've said the WAR is better, but not to the 100% exclusion of all other forms of evaluation. Most things in life, verifiable by stats or not, are not a matter of absolute right or wrong.
Yes, the sky is blue, the sun is yellow, a leadoff HR is better than a walk, a walk is better than a out, etc, etc, etc. What I debate is the level that people slap stats down as fact. Simple stats are often right or wrong, complex integrated stats are basically never about absolute right or wrong --there's lots of gray. You do get this ?
September 9th, 2010 at 6:02 pm
Matt Y-
Yes, I do get that. But I do think it is a bit disingenuous to use the chinks in WAR's armor to create larger arguments than really exist. WAR is the best stat out there right now. It's flawed, yes, but less flawed than the rest, at least in terms of evaluating the entirety of a player's contributions. So, only in the instances where we have specific reason to doubt WAR's effectiveness should we go to other, lesser stats.
September 9th, 2010 at 6:16 pm
Look, people like yourself asked to me and others to come up with specifics about the WAR b/c that's the way to improve a stat --I did that. It was not about, at least for me, exposing chinks for the sake of exposing chinks in the WAR--it's about improving a stat. I have from day one said that the WAR is the best single metric, however, it's only one metric and one way to evaluate a player. I want to genuinely improve the WAR --with that said though, I do not think it's possible to ever come up with a perfect stat --it can be improved, but not to the point of perfection. And because it can't be improved to perfection, means that others will always have merit to a degree.
Have I failed that miserably with my notion of moderation. I think there's some legitimate reasons why BB, as valued by the WAR, should be perhaps tweaked and re-evaluated.
September 9th, 2010 at 6:16 pm
Michael E. Sullivan (#159),
Specifically, the "by other means" refers to other ways a run will score during that plate appearance (WP/PB/error), not a subsequent walk. Runs that score other than as the end result of the plate appearance.
I'm not "assuming away the value of a walk." I'm arguing the in-the-moment value of taking-vs.-hacking. If you get a good pitch to hit early in the count, go for it.
The view I see here (tell me if I'm wrong) is that the hitter should always wait until late in the count, just in case he gets handed a free gift.
As I said before, that may be well and good, but it can also land a lot of otherwise good hitters in 0-2 holes.
The whole idea that Garvey's average is less valuable because of the lack of walks just doesn't work for me. You can't assume the walk. You can't assume that the count would have made it to ball four.
The average is the major, most important, component of the OBA. In many cases, it's two-thirds of the times a runner reaches base. To wait out the chance to get the walk -- at a much-lower probability -- is what doesn't make sense to me.
Sure, if the count is 3-0, I might take a pitch. But even that flies in the face of the likelihood that the pitcher doesn't want to lose the at bat, and will throw a hittable strike just to get one on the board. Then he regains his location, and throws another one (which the hitter will often take, because his mindset in now on getting the pass).
Of course, the situation is key. Johnny Twisto (#160), I agree that it's not back-and-white. Location, hittability, recognizing the pitch and being ready for it are all important.
And no (Mark Bellhorn excepted, perhaps), I don't think anyone necessarily goes up to the plate taking everything.
It's not that walks have no value, it's just what you might be giving up to get to one. If a pitcher's going to throw four straight pitches out of the strike zone, take the pass. But if he lays one in there, why not take a hack?
September 9th, 2010 at 6:32 pm
BTW Sean has agreed to put together a more detailed description of how WAR is calculated. It's on his list but I don't know exactly when it will appear (I'm guessing off-season at the earliest)
September 9th, 2010 at 6:42 pm
Bellhorn was certainly an extreme player. A type who probably couldn't have succeeded in many eras of baseball history, but was reasonably productive for a short while in his. His game had limitations, which he probably recognized and so approached his ABs accordingly, and once pitchers were able to exploit those limitations (and perhaps, his bat slowed a bit) he was quickly out of baseball.
I can understand why he'd be a hard player to watch, but he did have a very nice season for the Sox and played great in the Series. Seems like he should be remembered fondly.
In 2004, he swung at the first pitch 20% of the time. That's less than average but not extreme. 15 qualifying batters (out of 76) swung less often.
He made contact on 73% of swings. This was one of the lowest in the league.
38% of his (league-leading) strikeouts were looking. Well above-average, but not close to the highest in the league.
He swung at 38% of pitches seen, one of the lowest marks among qualifying batters. He only swung at 12% of pitches outside the strike zone, also one of the lowest.
So if he was so passive, and missed so many pitches, pitchers should have just pounded the zone. But only 58% of the pitches to him were strikes, below average and one of the lower in the league. Only 52% of pitches were in the zone. He laid off pitches out of the zone, and when he did make contact, he hit it hard (.364 BABIP, 6th highest in the league, plus a .399 BA on contact, compared to league avg of .331).
I can't find the # of full counts he saw...that info must be somewhere.
And incidentally, while it may seem like a Bellhorn-type would not be clutch -- taking walks rather than driving in runners, can be handled by better pitchers -- he had a 2.4 WPA that season, while linear weights puts him just 0.8 wins above average.
September 9th, 2010 at 6:48 pm
JeffW,
I do see some merit in what you're trying to get at in some ways, but i find your arguments in #156 mostly as way too simplistic. Yes, to a degree it's can be about a .280 chance of getting a hit vs the .120 chance of getting a walk, however, It's really not as simply as an either or. Contrary to what some others might think, I actually do value walks quite a bit, but given the similarities with players that perhaps seem legitimately overvalued, I suspected that BB were part of the issue/division. Obviously Abreu is at the heart of this issue since he's insisted at hitting 3-4 most of his career even though he really was more a 1 or 2 hitter. He a classic low risk low gain guy. He's fine with just taking the walk, sometimes to his detriment IMO. Not all walks are created equal even though the WAR treats them equally.
September 9th, 2010 at 7:02 pm
"The view I see here (tell me if I'm wrong) is that the hitter should always wait until late in the count, just in case he gets handed a free gift."
I can't speak for anybody else, but this is definitely not what I think. I also believe that hitters who think that way are much more likely to look like Mark Bellhorn (or some random never made the majors scrub) than Bobby Abreu, even if Abreu gets more walks.
Guys who get more walks but still hit well when they hit, are doing it by being somewhat more selective. There are very few guys who will fail to swing on any pitch they follow and comes where they want it. The guys who walk more are those that do less chasing of marginal or unexpected pitches. Guys who wait on everything in hopes of a freebie usually have crap for BA, and their on base is often as bad or worse than the guys who are decent garbage hitters, in spite of their walks.
The reason I look at on base percentage, and OPS is that it covers this. all those care about is how often you get on base vs. getting out, and how many total bases you take (for OPS). If you are walking *in addition to* good hitting, you will have a good OBP. If you are walking *instead* of good hitting, you will not.
September 9th, 2010 at 7:05 pm
I think a part of the problem is that there's a tendency to value a walk against the player's own skillset, rather than as an event onto itself.
I'll use an extreme example here to illustrate the point. You have Superman as your #3 hitter. Everytime he swings the bat at a reachable pitch, he hits a homerun. The only way to pitch him is to throw pitches he can't reach, and walk him every time. Occasionally a pitcher will throw a 3-o strike, and he usually just keeps his bat on his shoulder, which is a terrible strategy, since he'd guarantee a HR if he swings, but he settles for a walk instead.
He gets a day off and is replaced in the batting order by by Johnny McScrawny. McScrawny's only redeeming skill is his ability to foul off pitches. If he puts it in play, it's always an out.
Now, for Superman, the walk is always the worst possibility for a plate appearance, and for Johnny it's the best. You don't want Superman to walk, you want him to swing, yet a walk by Superman is exactly as valuable as one by McScrawny (assuming equal baserunning skills).
There may be times when you'd like your #3 hitter to be more aggressive, but he's almost always putting his team in a better his team in a better position by adding a baserunner for the next hitter.
September 9th, 2010 at 7:37 pm
JT-
As I noted, I knew I was talking as much, if not more, feelings than I was knowledge. Bellhorn did have some productive years. I remember getting quite a bit out of him in a fantasy league the first time we went to using OBP and SLUG (2002). He did have a solid 2004 for the Sox, and I vaguely remember him ringing the Pesky Poll in game 1 of the Series (I say vaguely because I was in the middle of the bi-annual Boston College pilgrimage to Notre Dame for the Blarney Bowl and everything from that weekend is a bit fuzzy). So, I definitely overstated the problem with his approach. Bet there definitely *SEEMED to be at bats that got to 3-2 and Bellhorn seemed to put the bat on his shoulder. I think he represents a guy who might acquire a lot of walks with a less-than-desirable approach, thus adding fuel to the fire of those who think walks are gained only by passive hitters.
I must ask, where did you get those numbers on Bellhorn? Are they available on this site or elsewhere? Does one need a subscription? It was kind of nice to see that my sense of Bellhorn was at least somewhat backed up by the facts.
September 9th, 2010 at 7:43 pm
How about this...
I can't think of an appreciable difference between a walk at ANY point in the game with first open and a bases-empty single. Both put a runner on first. Neither advance a runner. I suppose the single might tire a fielder out chasing the ball, but let's just pretend this is perfectly balanced by the extra pitch or two a walk is likely to elicit.
So, at the very least, a walk with 1B open is worth a bases-empty single. How much is a bases-empty single worth? That should be the starting point for the valuation of a walk. There are obviously instances where it is more valuable. But it is never LESS valuable.
Now, there are individual circumstances where a BB is less valuable than a hit, even a swinging bunt, infield single. But it is at least as valuable as a bases-empty single. Right?
September 9th, 2010 at 8:00 pm
BSK, most of those numbers are on this site, in the "Pitch Summary" section of the batting pages. A couple were from Fangraphs, and I must admit I am not sure about their accuracy, since I think they are based on PitchF/X data and I'm not sure if that was being used in every park, or how accurate it was, in 2004.
No one can know, but it's quite likely that Bellhorn's approach was the most desirable for HIM, based on his particular set of skills. It's easy to say "make more contact," but not everyone is able to do that productively.
Matt, you are focused on walks, but not all of everything is equal. A grand slam is better than a solo homer, but WAR treats them the same. (So does "HR.")
September 9th, 2010 at 8:12 pm
JT-
Sounds good. I agree that Bellhorn made the most of his abilities and used that approach to have a handful of productive seasons. But, overall, it's not the preferred approach for most ball players. And once it was exposed, and he was unable to adjust, he became a decidedly unproductive player.
September 9th, 2010 at 8:31 pm
BSK, empty-base valuation is actually the premise for some of my homebrew stats. A single and a walk are of equal value in that context. Since they are of equal value in any context unaffected by other players, I make no distinction between the two in my model.
September 10th, 2010 at 2:50 am
Michael E Sullivan (#170),
Selective hitting.
Plate discipline. Walks, in and of themselves, do have a value, but mostly if they are the result of appropriately selective hitting.
Most pitchers (at least, the good ones), however, have a significantly better ratio of strikes to balls, so the hitters better not wait around too long. As I said earlier, walks are like accidents for a pitcher, not the norm.
If a hitter sees a pitch he likes earlier in the count and goes after it, it doesn't necessarily mean he's not as good because of a perceived inability to draw walks. He sees "his" pitch, and puts a swing on it. That's what is expected of him.
Occasionally, he'll get fooled, but that happens to everyone.
If Garvey is consistently plating runners in his at bats at a significantly better rate than the league average, then what's wrong with that?
Any thought that he should have taken more pitches is pure conjecture, because you can't guarantee the end result would have been the pass you say he should be taking. You can't assume that any set percentage of the outs he made would have otherwise resulted in walks.
The question should be, was he willing to take the walk, if it came to that? Or, did he blindly hack at anything and everything that came his way, like Vlad seems to? Maybe he took the balls, but went after the strikes he felt he could handle, which is what you want.
Good plate discipline may not always mean getting more walks, but being better focused on swinging at the right pitches.
Some pitchers do nibble more than others, and will try to draw the hitter (and the home plate ump) out of the zone. That's part of pitching. But if Garvey was letting those wide ones go by, then zeroing in on the strikes he liked, that's where you'd want him to be.
Maybe pitchers knew they had to throw him strikes, since he wouldn't chase out of the zone. Any thoughts there from anyone who watched Garvey regularly? I don't remember him having a reputation as a wild swinger.
September 10th, 2010 at 7:40 am
BSK @173: "Now, there are individual circumstances where a BB is less valuable than a hit, even a swinging bunt, infield single. But it is at least as valuable as a bases-empty single. Right?"
No. Can't be. There are some times when a walk is practically indifferent, and the intentional is automatic if a weaker hitter is behind you. Think about bottom of the ninth 2 outs, tie score, runner on second. A walk barely matters. If anything it makes it easier to get an out on the next guy by putting a force up at 2nd and third. That probably balances the only negative of the walk in that situation, which is that you can't afford to walk two more guys by accident. So it's pretty close to worthless in that situation. Well in comparison to making an out it's still good, but it's basically like a null PA as far as winning chances go, while a bases empty single or walk is always positive (i.e. better than an average PA result).
JeffW@177: "If Garvey is consistently plating runners in his at bats at a significantly better rate than the league average, then what's wrong with that?"
There's nothing wrong with it. Garvey was a good player. Anybody would have been happy to have him on their team. In this thread, we're comparing him to a guy who's stats say he's got close to a HoF career. If we were comparing Garvey with some random dude, he'd come out ahead. Did you look at Abreu's percentage of runners plated? I'll bet it's similar, with his extra power making up most of the difference from walking.
I'm not saying "Garvey should have looked for walks more". I don't know what laying off more pitches would have done to his hitting. It's plausible that it would have improved his overall stats, but it's also conceivable that he didn't have the eye or the temperament or whatever je ne sais quoi to play like Abreu and get more walks without sacrificing something else just as or more important.
What I'm saying is that when you look at the sum total of what Garvey did, hits, extra bases, walks, HBP, ROE, GDP, outs and all, and compare it to the same for Bobby Abreu, Abreu added a lot more overall value from his PAs, and his extra walks (instead of outs) are a big part of that difference.
September 10th, 2010 at 8:16 am
Abreu plated 18%, compared to a league average of 15%, so a similar rate as Garvey relative to the league.
Abreu actually had a higher rate of productive outs as well, at 38% compared to a league average of 33%, with Garvey at 36% compared to 34%.
What it ultimately comes down do it that Garvey put more balls in play, but Abreu did more with the average ball he put in play, so they're close in terms of the value they created when they hit the ball. Abreu managed to add that value by using fewer outs, while giving good hitters behind him an extra 50 or so baserunners per year.
September 10th, 2010 at 9:17 am
Just as an exercise, what would Abreu's extra walks be worth if we were to use a system like Mike suggested up in 132?
Abreu has 850 more walks than Garvey, and batted 3rd most of his career, for a .24 rate.
850 * 0.24 = 204 runs, which is about 21 wins.
So, even using a system that's built to squeeze the value out of walks for middle of the order hitters, we're still looking at about 20 WAR credited just based on the difference in walks. Or, if you prefer, that's 850 free baserunners for his team's cleanup hitters.
September 10th, 2010 at 10:19 am
What's the difference in WAR using .30 and .24 ?-sorry, i don't know all the finer details of the WAR off hand.
September 10th, 2010 at 12:26 pm
Ah, I see this thread got a bit interesting.
I find it funny that people are trying to somehow evaporate the value of a walk. A walk is not an out. More than anything else not making an out is the object of your plate appearance; subsequently you will be on base. To make an out is to fail. To successfully reach base, whether by hit or walk, has value to your team. That the values of the various hits and of the walk vary is irrelevant. The best hitters make less outs. Abreu has made considerably less outs than Garvey in an appreciable number of PAs, and even if their slugging and batting averages were identical, Abreu would easily be more valuable.
I'd rather have Abreu (hitting second in my lineup) than Garvey.
To the other thing about WAR not being perfect and thus that opens the door for the use of other stats. That's true. WAR isn't perfect. I am personally skeptical about fielding data's contribution to WAR, especially for a position like 1B where if you have a very good rangy 2B on your team you may end up getting hammered by a stat like UZR. But once you start to try to make more than a superficial comparison of players, by pulling out BA and SLG and OBP and OPS and OPS+ and so on and so on you've basically just recreating in your own mind a subjective, unscientific version of complex stats like WAR. You're basically making a gut-feeling version of stats that have already been studied and tweaked, and--while imperfect--are surely better than creating one on the fly in order to satiate your gut.
Also, I'm going to completely ignore that people still love to pull out round number milestone stats in addition to hilariously unpersuasive things like "he was a star on a perennial winner!", as if, for a guy like Garvey, for instance, substituting him for Hernandez or Olerud or even some bizarro version of Abreu who is at least as competent a 1B as he is a RF takes away from the Dodgers' success during Garvey's time. Garvey was not the straw that stirred the drink.
Lastly, I like that someone above mentioned exactly what I've been telling people about the way stats are misused. All counting stats are perfect in what they do, simple stats are also perfect (I define a simple stat as a counting stat which is then used in a simple mathematical operation--ERA, for example), advanced stats are often or always perfect (a simple stat which is then combined in some fashion with another simple stat--OBP for example), and then there are complex stats, which are decidedly imperfect, but that's okay because they allow us to actually try to do what people try to do with counting and simple stats--make comparative evaluations about players.
To retread what was said before a bit, there's absolutely nothing wrong with a counting stat like RBI. It tells you what it tells you. That guy drove in that many runs. There's nothing wrong with a counting stat like wins or losses. It tells you how many games a guy won or how many guys a game lost. There's nothing wrong with ERA, or batting average. That is how many runs this guy gave up per nine innings pitched, or how often the guy got a hit per at-bat. These are all fine stats to look at. However, when you start trying to evaluate a player based upon his wins and losses, or his batting average, or his RBIs, and you use your 'gut' or 'intuition' to 'feel' that this guy's 141 RBIs batting third on a Colorado team with a potent offense is more impressive than this other guy's 106 RBI batting fifth on a Florida team with a league-average offense, you're not doing a useful evaluation. A useful evaluation requires you to embrace advanced and complex (and sometimes newer simple and counting stats) in order to determine how valuable different contributions are.
Like I said, I get that people think WAR is imperfect--I agree. But it's hilariously obtuse to think that with the way baseball is structured, with the huge reams of data that we have, that it's better to try to gut-feel one's way through counting and simple stats and derive meaning from them that lacks objectivity as a matter of course in lieu of newer advanced or complex statistics which were designed specifically to make evaluations.
There are guys out there who try to reduce the game to a bunch of numbers and that's wrong, but, if for some reason I had to choose being completely black and white on the issue, choose between being someone who sits around lauding Joe Carter because he drove in 100 every year, or being the guy who because of WAR thinks that Abreu may well end his career with a resume that seriously merits consideration for the Hall, I'd rather be that guy who is inducting guys into the Hall based upon a sterile, strictly stats-based approach rather than the guy who inducts a near-mediocre player who had a good reputation, some memorable moments, and was the face of a perennial winner.
Fortunately, I don't have to be either.
September 10th, 2010 at 12:32 pm
About 5 wins. I mentioned that in an earlier response to you Matt. That the difference between .3 and .25 over 850 walks (the difference between Abreu and Garvey) was around 4.2 WAR.
You do the same calculation Wilson just did, but with the difference of .06 instead of the .24.
That said, the thought experiment I offered in 143 demonstrates that a walk value of .24 is almost certainly too small, unless either hits are *also* overvalued, or there are nowhere near enough intentional walks.
The ratio of value of a walk to an average hit can't be too low, or you should be IBBing top guys left and right.
September 10th, 2010 at 12:52 pm
Matt Y, since you are interested in tweaking the value of BB, do you agree with my assessment that if the ratio of value of a walk to a hit does not have a lower bound at or above the BA of the best hitters, then current IBB strategy is much too aggressive (because nobody has ever been walked in half their PAs, not even 2001 Bonds)?
If we're tweaking it, and not just taking potshots at possible cracks in WAR, then let's have a discussion about the value of a BB. I'm trying to put bounds on what the value could reasonably be, and I haven't heard anything concrete from you in response.
September 10th, 2010 at 1:12 pm
Those who are unfamiliar with them should read about run expectancy and linear weights. You could start here:
http://www.tangotiger.net/RE9902.html
http://www.tangotiger.net/lwbymob.htm
There may be better or more readable articles on the subject but I don't have time to look right now.
Essentially, each offensive event is given a run value based on its average value across all base-out states in which it occurs. (Those values will vary slightly based on how much scoring there is in the league.)
You can't pick out one event and say it seems valued incorrectly. There is years of science behind these numbers. Show some evidence.
The batting measurements are the least controversial part of WAR. Figure out how many runs a player created, how many outs did he use to create those runs, and compare how many runs an average (or replacement) player create in the same number of outs. Once you start thinking about offense on a per-out rather than per-AB or -PA basis, it really changes your perspective. Because TEAMS score runs per 3 outs, 9 times per game, so the individual players making outs are costing chances to score.
September 10th, 2010 at 1:18 pm
Thanks, I missed that. This still doesn't address the issue that to the masses these guys with more subtle value appear quite overrated, and I suspect most to all will not make the Hall --basically all of the most contentious candidates are on the list JT did above, and BB appears to be pretty front and center. You can just pass that off as people just don't get it, but i don't think it's quite that simple. Personally I like the WAR a lot, and i do value it higher than all other forms of evaluation. However, as more and more players retire with WARs above 65 don't get in the Hall, the metric could become more and more irrelevant and not more relevant. Front in center at this divide are the likes of Abreu, Edmonds, Walker, Helton, Rodriguez, Lofton, Sheffield, Hernandez, Evans, Randolph, Whitaker, Nettles, etc. etc. These guys were all high BB guys that players good defense and/or hit a lot of HR. BB is a common thread. Perhaps over time this will get more traction --i hope it does, b/c I think BB should be more valued than they have been historically, I'm just not sure they're being valued a bit too much.
I did the calc. with .06, but did not have a chance to look at how much that translated in WAR. Thanks again.
September 10th, 2010 at 1:23 pm
I completely agree that the offensive stats (compared to the defensive stats) are much much less debatable and have good sound science backing them up, but that doesn't mean they're perfect.
September 10th, 2010 at 3:47 pm
#186
I think more so than walks or even defense, there are a few factors in play for why some of those guys haven't received great support:
For Abreu, Lofton, Hernandez, Evans, Randolph, Whitaker,and Nettles, there's the lack of a huge peak. They lived in the 3-6 WAR range for a long time, occasionally as all stars but never as superstars, building their value steadily over time but without any 8+ WAR seasons between them, and few seasons which scream HoF. I think in their case, the 60+ WAR is almost like a milestone in and of itself, a reflection of a long, excellent career, but for someone who favors a higher peak I can see why they'd all be on the outside looking in.
For Walker and Helton, it's more of a park factor thing. Their numbers would make them almost sure things taken at face value - they have the high batting average that voters love, along with good power numbers - but voters will deflate their numbers for playing in Colorado. How they fare on the ballots will depend on how far the voters adjust.
If Sheffield was well liked with little steroid speculation, he'd be in on the first ballot with at least 85% of the vote. A .290+ batting average with 500+ HR and 1600+ RBI will likely put him in anyways eventually, but his personality could keep him out for a while.
Edmonds is tougher one, and the only things I can think of for him are his injury history keeping him from certain milestones, and the idea that he's treated as a power hitting outfielder rather than as a centerfielder. Forget advanced metrics, from 1995 to 2004, Edmonds hit .296 and averaged 30 HR and 87 RBI even when you include his 1999 that was mostly lost to injury. Per 162 games, he averaged 37 HR and 106 RBI during that stretch, all while playing a very good CF. He didn't have any prime Griffey seasons, but even without using any advanced metrics, a decade in the middle of a career averaging 30+ HR with a batting average around .300 and a bunch of Gold Gloves at a key defensive position doesn't need any help from walks to look like a Hall of Famer. The OBP certainly helps his value, but he score well enough in the traditional numbers that I don't think it's fair to point to a BB-heavy value as hurting Edmond's case. I think it's partially a milestone thing, but more so because people are used to the term "outfielder" rather than looking at the distinction between the three different positions.
September 10th, 2010 at 5:13 pm
Good points #188 -- I had thought of many of the same ones before I wrote # 186. With that said, BB are still part of the "occasionally they were all stars but never superstars, building their value steadily over time but without any 8+ WAR seasons between them, and few seasons which scream HoF". The not having a peak is legitimate, but yeah, they racked up many solid seasons of WAR built on subtlety, which was partly (a big part) a result of them walking a lot.
With Walker and Helton it's surely a park thing, but they walked a lot like the others as well, including Edmonds. Of the entire group, Sheffield would be the only sure thing if he was more liked and didn't have the steroids issue -- Edmonds should be borderline in IMO, but he's not a slamdunk like his WAR suggests. The BB thing is a common thread, whether fair or not. And these are all guys with a WAR over 60, not 55 or 57--except Abreu, which is at 58. I've read blogs here that have said players with a WAR of 57.5+ should go in.
September 10th, 2010 at 5:30 pm
I should add that Helton's WAR is 58 as well, and that all the players on the list not only walk a lot, but were also above average defenders to stud defenders.
September 11th, 2010 at 12:36 pm
OK, took a few hours this morning to look at the numbers more closely. As you report WilsonC, what a walk is valued at seems about right --the bounds look good--I could quibble a bit, but the numbers are good --this does not surprise me given the number of years of scientific data to back it up. I did not expect to find something way off. With that said, and perhaps contradictory , I think one could still make some argument that a walk shouldn't be valued at the same fixed rate of .30 across the board b/c each person in the lineup has different roles, especially the 3-5 guys. However, this makes it purely a philosophical argument, since the science is irrefutable. Additionally, and at the very least, I still think my broader point still has merit. Perhaps the ultimate compiler is the player that racks up lots of 3-5 WAR seasons with very subtle impact and for mostly teams that never broke through....i.e. --good defense, good baserunning, and a higher than average ability to draw a walk. For such players to make the Hall, perhaps these guys need to put up a WAR of near 70 to make it, whereas, someone else that was a high HR guy, drove in a lot of runs, and his presence seemed to make more an impact on a team is more worthy even if said player has a WAR of 58 (i.e Vlad) compared to 68 (i.e Whitaker or Edmonds).... Or when you compare a Jeter with a Whitaker, which have nearly identical WARs around 69. Sure, maybe this is a little old school to some and perhaps even wise to others. In conclusion, WAR is a good metric to valued highly, but there's clearly quite of bit more to the evaluation process.
September 17th, 2010 at 2:13 am
In just a couple of years Bobby Abreu has a very good chance of having numbers better than these in his career:
Runs Hits 2B HR RBI BB SB
1500 2500 600 300 1500 1500 400
Know how many major league players have this career line? None, even if you remove the 600 doubles (and if you remove a juiced Barry Bonds from the equation ... I do). I don't know too much about WAR and the other stats, but when you are the only player to do something like this, the only one among all players that have EVER played the game, aren't you elite? didn't you have a UNIQUE (take the word literally, because he would be the ONLY player EVER to accomplish these numbers) combination of power, speed, tenure, consistency, etc etc etc that deserves the Hall?
Throw in a few other very unique achievements (only the Bonds had more 20/20 (HR/SB) seasons than Abreu, who's 2 HRs away from his 9th ... 8x100RBI seasons, 8x100R seasons..) and I think the guy should have a very good chance
I might be looking at raw numbers only, but it seems pretty hall-worthy to me