User Comments, Suggestions, or Complaints | Privacy Policy | Terms of Service | Advertising
|
Demarini, Easton and TPX Baseball Bats
|
|
Page rendered in 0.6481 seconds
53 querie(s) executed
|
| ||||||||||
Baseball Primer Newsblog — The Best News Links from the Baseball Newsstand Thursday, August 04, 2011Schoenfield: Time to start figuring out the MVP races
Shouldn’t send too many Primates into frothy MVP rages. jacksone (AKA It's OK...)
Posted: August 04, 2011 at 10:35 AM | 101 comment(s)
Login to Bookmark
Tags: awards |
BBTF PartnerBookmarksYou must be logged in to view your Bookmarks. Hot TopicsNewsblog: OT: NBA Monthly Thread - March 2014
(173 - 7:03pm, Mar 07) Last: The Id of SugarBear Blanks Newsblog: OTP - March 2014: Russia denies calling shots in Ukraine’s Crimea standoff (532 - 7:01pm, Mar 07) Last: Gold Star - just Gold Star Newsblog: Doug Fister injury: Nationals righty has elbow inflammation, MRI (8 - 7:01pm, Mar 07) Last: The Yankee Clapper Newsblog: Chelsea Handler Rips A-Rod, Calls Him ‘Disgusting,’ ‘Gross,’ And ‘A Buffoon’ (264 - 6:59pm, Mar 07) Last: cercopithecus aethiops Newsblog: Sports orthopaedic pioneer and Dodger physician Dr. Frank Jobe passes away (38 - 6:55pm, Mar 07) Last: The Yankee Clapper Newsblog: 1914 and 1915 seasons added (2 - 6:23pm, Mar 07) Last: Fred Lynn Nolan Ryan Sweeney Agonistes Hall of Merit: Most Meritorious Player: 1952 Discussion (16 - 6:21pm, Mar 07) Last: Moeball Newsblog: Red Sox field substandard spring lineup versus Marlins - Sun Sentinel (40 - 6:00pm, Mar 07) Last: Robinson Cano Plate Like Home Newsblog: Cardinals hof fan vote. (10 - 6:00pm, Mar 07) Last: The Yankee Clapper Newsblog: Gleeman: Brandon Phillips: “How the [expletive] am I declining?” (9 - 5:55pm, Mar 07) Last: TDF, situational idiot Newsblog: ESPN: Luke Hochevar out for Season (7 - 5:24pm, Mar 07) Last: Willie Mayspedes Newsblog: Daugherty: MVP possible for Jay Bruce (22 - 4:37pm, Mar 07) Last: Walt Davis Newsblog: Joe Poz: Top 50 Contest (34 - 4:31pm, Mar 07) Last: GregD Newsblog: Finn: Fox’s decision to hire Harold Reynolds is uninspired (7 - 4:24pm, Mar 07) Last: Lance Reddick! Lance him! Newsblog: OT: The Soccer Thread March, 2014 (56 - 3:56pm, Mar 07) Last: Mefisto |
|||||||||
|
About Baseball Think Factory | Write for Us | Copyright © 1996-2014 Baseball Think Factory
User Comments, Suggestions, or Complaints | Privacy Policy | Terms of Service | Advertising
|
| Page rendered in 0.6481 seconds | ||||||||
Reader Comments and Retorts
Go to end of page
Statements posted here are those of our readers and do not represent the BaseballThinkFactory. Names are provided by the poster and are not verified. We ask that posters follow our submission policy. Please report any inappropriate comments.
(But, Bautista or Pedroia in the AL, Kemp, McCutchen or Braun in the NL. For now.)
Fourth doesn't seem premium to me. And even elite D at 2B can't overtake the immense offensive lead Joey has. If Bautista were on a playoff team, the award would already be in his living room.
If the season ended today, in the AL, you'd have to select Bautista. His offensive production is simply too huge to ignore. However, Pedroia still has time to catch up to Bautista, because he is gaining ground on Bautista with his defense and offense. Consider this: Bautista is 1st in the league right now in WAR, with 6.8 - but 6.4 of that comes from offense. Pedroia is 2nd in WAR, at 6.2 - but only 4.9 of that comes from offense. Unless Bautista goes off on another HR binge this season, Pedroia may well catch Bautista.
If they end up at about the same place, BTW, I'd go with Pedroia, because he plays a more difficult position, and is harder to replace. Again, if they ended up with the same WAR, and you could go back to March and draft your team for the 2011 season alone, wouldn't you pick Pedroia? I mean, you would lose less value going from, say, Bautista to Ellsbury, than you would going from, say, Pedroia to about any other second baseman in baseball, right?
The gap isn't what it used to be. Bautista had a cold (for him) July.
I don't know how that effects an assessment of his value to the Jays (or if it does at all), but I think it's worth noting.
Isn't that double counting the positional adjustment in WAR?
The difference between Pedroia and Zobrist is 1.5
The difference between Bautista and Ellsbury is 1.5
Further down the OF and 2B lists you have Yunel Escobar (he could fill in at 2B) and Alex Gordon at roughly the same distance from the two leaders.
If they end up at about the same place, I'd go with Bautista, because I'd be a lot more confident that his WAR is a more accurate representation of his performance.
It's obviously too early, but I have a feeling that after a few years of getting it right, this is one the BBWAA is going to err on and go with Gonzalez.
Agreed
Why?
However, Pedroia ranks higher in B-Ref WAR since 2008 than I would've guessed among (current) AL players. Mauer (23.2), Adrian Gonzalez (21.1), Youkilis (21.0), Longoia (20.6), then Pedroia (20.1), followed closely by Teixeira (20.0) and Miguel Cabrera (18.9).
I think the assumption being we are more certain about oWAR than dWAR.
If a guy has 6.0 WAR and 4.5 of that is because he's a +20 2B
Or if a guy has 2.0 WAR, but is a -40 in LF
In either case I'd be tempted to reel those back a little bit, in a way I wouldn't if he had a ridiculous offensive component.
Defensive WAR values are far, far, far more questionable than offensive WAR values.
I'm skeptical about the overall accuracy of the defensive metrics, whereas I'm pretty confident the offensive components pretty accurately reflect the player's contributions.
Of course, my hypothetical MVP vote wouldn't be WAR-based only. Just that when it comes to that particular stat, I believe teh defensive component must be taken with more salt grains than the offensive side.
Edit: Cokes to Greg and Randy.
If he hadn't gotten hurt last year, he probably trails only Mauer.
Both players have the same amount of each, though.
Perhaps Bautista should really be at -2 dWAR.
iirc, it's a distant 3rd after shortstop. 3b and cf are similar. however, cf is sort of a special case, b/c there seems to be more space for a truly elite player to save a ton of runs.
Right, but that doesn't mean there's only .4 points worth of mistake room in Bautista's score. He could deserve a -2 dWAR, instead of a +.4. Just like Pedroia could deserve either 0 dWAR or 4 dWAR. If half of the components of the stat are questionable, the stat is questionable for everyone, not just people who score high on the questionable part.
Isn't that where regression comes in? The less certainty you have about a stat (or component), the more you would have to regress it.
And how does that invalidate the original comment by Sosh where he said if they end up about the same he would go for Bautista.
I mean most of us are about 90-95% comfortable with offensive numbers, while with defensive numbers it's probably around 30-80% comfortable. If one guy has 6.5 points with the numbers we are comfortable with and the other has 5.0 then how much value do you place with the numbers that we are uncomfortable with? Especially if all it does is make them close to each other?
Not sure about that.
I had to check.
I assume you meant June (.835), not July (1.083).
.835 OPS for June, for a sOPS+ (split OPS+) of 138.
That's crazy good for a "slump".
(Though, he's .770 OPS since the all-star break, so that's maybe what you meant.)
Well, Pedroia had a 1.010 OPS in June and a 1.188 OPS (sOPS+ 186/226). That helped the gap close a bit too.
But we all agree that there IS some correct defensive number, right? And that Pedroia's is likely better than Bautista's? So we can't really just ignore the defense entirely.
And no one suggested that. What we did say was that if their total is about the same and Bautista has significantly more offensive value then we would support him. If their offensive numbers were closer, or if Pedroia had a large overall lead, then it would be a different story.
Just that if the numbers are otherwise close, and we are more certain about the offensive numbers than the defensive numbers, the edge would go to Bautista.
IOW, assuming the numbers hold up as they are now, we can be very certain of the offensive gap between the players, but not nearly as certain about the defensive gap. So we would chose to give greater weight to the numbers we are more certain about.
Yes, it is possible that the defensive numbers are right on. It is possible that they understate Pedroia's defensive superiority. That is a function of the greater "error bars" surrounding defensive numbers. They simply cannot be trusted as much, so you give them less credence.
But you don't ignore them entirely. If the offensive gap between the players is not as great, then the assumption that defense makes up the difference becomes stronger.
But if you have 2 players with WARs that are approximately equal, and dWARs that are different, why wouldn't the error bars be the same width? They've played (just about) the same number of defensive innings.
Yeah, I should have done my homework. I really should have said "the past little while".
I'm probably the wrong guy to make this point since I know next to nothing about statistics (as will soon become evident)
But isn't this what regressing means?
The further you are from 0, or the average, the more likely there is noise in the numbers.
A guy batting .350 or .160 for a month is more likely having a fluke month than a guy batting .260. It's possible he's not...but more likely he is. A very good or very bad defensive component is more likely to be inaccurate than a relatively average one. But then again it might not be.
That's true IF the statistic itself is correctly calibrated. If it's a seriously flawed stat, then no, there's no reason to believe that.
If it's a seriously flawed stat then even more reason not to trust it.
Or maybe I'm missing your point.
Regardless of whether we have a way of measuring it, there is some number of wins that every player adds on defense. That number of wins seems like it should be between costing their team 2 or 3 games and winning 2 or 3 games for their team. Somewhere around that number.
Pedroia is almost certainly, just eyeballing it, better defensively than Bautista. I don't think a lot of people will seriously argue that point.
So while we don't trust the stat, and we assume there are large error bars on that stat, you can't just look at Bautista's offensive contributions and say that because they are higher than Pedroia's, he's almost surely got an overall better trueWAR, where trueWAR is defined as the actual, non-statistical but actual god-given-truth number of wins he really provided. You've ignored defense, there.
There are error bars on both players defensive performances, but those error bars are about the same size on both players.
It is obvious why, and I think you know why, and at this point you are just trolling.
I completely don't, and I'm not, but I'll stop the discussion.
But if the stat telling you where the error bars deviate from is flawed then how do you know?
The issue is certainty. No one's ignoring defence, just that if two guys are tied (or relatively close, I don't think WAR is meant to be taken literally to the decimal place) the uncertainty surrounding defensive numbers is a good place to think about a tie-breaker. We're talking about fine distinctions between two very near equal players. If one of the players has a lot of his value tied up in an element we're uncertain of (or is heavily penalized by that same element), it makes sense to me to err on the side of what we are more certain of.
Think of it this way. In 2008 Brad Hawpe hit .283/.381/.498 (in Coors mind you) had -1.7 WAR, mostly because he was -41 runs on defence. This year Alex Rios is -1.8 WAR, because of slightly below average defence and he's batting .207/.252/.294.
Are you pretty confident that 2008 Hawpe and 2011 Rios (to this point) are of similar value?
I guess I would say no, although "to this point" is misleading, because we're only 2/3 of the way through the season. Rios will likely end up losing another 1 WAR, and finishing near -2.7
Yeah that was a problem with the example...too lazy to find a full-season terrible year that was equivalent!
I think you also raise a good point in that the defensive component of WAR should probably be augmented by actually seeing the guy play, comparing it to other defensive metrics, comparing it to the player's previous results...all things that no one really suggests doing with oWAR.
As an aside, imagine if Dunn and Rios had met even (what I think are) reasonable expectations, maybe being 1.5-2.0 WAR players this year. They'd be right there with the Tigers.
Uhm, Pedroia had 1.1 dWAR in 75 GAMES last year.
If one of the players has a lot of his value tied up in an element we're uncertain of
But the issue is, they don't have different amouts of value tied up in their defense, not as long as they are both getting similar amounts of PA's and innings in the field. If 33% of a position players value is on the defensive end of the ball, then 33% of his value is derived from defense. It doesn't matter if you are average or 20 runs below or above it. The amount of value tied up in defense is the same.
You can obviously make an argument that positions that see more opportunities have a higher amount of value tied up in their defense, but that's not the argument anybody is making.
It's mostly a matter of perception, since offensive #s are being compared to replacement level, and defene to average. This makes it seem like his defense is a comparitively less important part of the equation.But this is like saying that a defensive wiz who is average with the bat doesn't have a lot of value tied up in his bat. But the ability to be average on offensive is of course a huge part of what allows him to be a valuable player.
But this should be done for ALL players, regardless of where they fall on the scale.
My mistake.
Different positions see different numbers of chances.
It's not a matter of perception, it's a matter of cofidence in the numbers. Most people are very confident on the reliability of offensive numbers, they are less so on defensive numbers. On top of that as mentioned, Bautista's defensive flexibility adds value that a defensive system wouldn't really be able to cover properly.But it's not about perception, but comfort in defensive numbers. When you look at other defensive systems they are going to give a lot more variance in the numbers than you will get with other offensive systems.
This is barely relevant, but some of the PA difference comes from Pedroia playing for the best offense in the league; his percentage of team PA is about 12% higher than Bautista's, not 15%.
Regardless of what you think of various defensive metrics, I think we can all agree that Pedroia is between an above-average and excellent defensive player at a position that falls on the left-center part of the defensive spectrum. The fact that he is tied with the second-highest offensive WAR in the league AND is tied for the 4th-highest defensive WAR, to me, makes him unique. In being unique, he is more valuable than a player who is, well, not unique. If Bautista and Pedroia both end up with similar amounts of total value this year, who would rather have had with your first draft pick in the AL this year? It's Pedroia, right?
I can find a number of outfielders who will give me excellent performance, but how many 2nd basemen can do what Pedroia is doing? Somebody above mentioned Zobrist, and he's a great player, but he's started 30 games in the outfield this year, and is not nearly the defensive player Pedroia is. (I will say, though, that Zobrist does bring something unique to the table - his versatility is awesome.)
At any rate, let's see if Bautista is even ahead of Pedroia at the end of the year. The gap is closing very quickly...
If it ends up tied or very close, I agree with the notion above that some credit should be given to Bautista for being able to play 3B or RF, which doesn't show up in the WAR figures.
He's streaky, and he's got his timing back. Look out MLB.
You realize that would require 30 HR in 52 games ...
I like the way you worded this.
Dude. He's Jose Bautista.
Different positions see different numbers of chances.
Well geez. If you had bothered reading to the end of the paragraph, you would have noticed I pointed that out. But again, nobody was making that argument.
It's not a matter of perception, it's a matter of cofidence in the numbers. Most people are very confident on the reliability of offensive numbers, they are less so on defensive numbers.
But the point is Bautista's defensive numbers are just as big a part of his total value, as Pedroia's defense is of his. Yes it's possible that Pedroia si 5 or 10 runs worse on defense, but the exact same is true for Bautista.
1/3 of the year is left
Sure Pedroia is on a tear NOW, that doesn't mean that he won't put up 0.5 WAR the rest of the way, nor that Zobrist puts up a 3.0 and blows by him, not likely? Of course its; unlikely- every specific outcome is - individually- unlikely
By virtue of his current lead in WAR , Bautista is merely the least unlikely to end up the year leading in WAR- given how close Pedroia is, I'd guess that Pedroia has a 30-40% chance of catching and passing him. There is also a non-zero chance that someone else from the "field" will catch and pass him as well.
Also for what it's worth, so far Fenway has been acting like it's pre-humidor Coors, 10.5r/g in Fenway, 8.7 on the road - if that keeps up Fenway's PF will be 110 or so rather than the 105 BBREF is using now and at the end of the year when BBREF updates PFs Pedroia (and all Sox hitters) may lose some oWAR (happened to Cano last year, lost nearly a whole win when PFs were updated)
I don't watch a lot of Cardinals games, but it looks like Lance Berkman is having quite a year. Is he playing as well as his numbers suggest?
Sure, it's possible. In fact, I don't really have any reason to believe these particular numbers are off.
My point, and the point of others here, is that since we have less confidence that the defensive numbers are an accurate representation of the contributions of all players, we're going to regress them some (and in my case, as with Greg above, it's most definitely a general regression rather than one that has any kind of strong statistical foundation). OTOH, since we have confidence that the oWAR portion is an accurate representation of a player's hitting contributions, we don't feel any need to regress. And once you've regressed, the guy who gets a larger part of his WAR total from his defensive contributions will suffer, even while recognizing that each player's defensive numbers are an equal part of value.
Having said that, Darren in 55 makes an excellent point about the weighted 3-year average.
Regress them to what? You don't know what the middle is! It's allegedly 0, but if you don't have confidence in the statistic, you can't say that it really truly is. So you don't know what to regress to.
No, the point is precisely that the defensive numbers are not as big a part of his total value. Because (1.3/6.2) > (0.4/6.8), right? People are saying that they'd support Bautista because more of his value comes from the part of his performance that they have more certainty about. YMMV, of course, but hopefully not because you misunderstand what they're saying.
Very, very, very, VERY wrong. Wrong in so many ways that I don't know how to describe it.
Also, pretty hard to argue that your comment was judgement-free, so I guess you're going to have to change your handle. ;)
Both players, indeed virtually all players, have the same amount of their "value" determined by their defense as their offense. If Barry Bonds had played second base as well as Dustin Pedroia, he'd have been much more valuable. So he lost value because of his defense.
This is a spurious argument. By diminishing the value of defensive numbers that are assumed to have less certainty than the offensive numbers, the equipoise that must be maintained between the two is destroyed.
For instance, let's for the sake of argument say that 70% of a position players' value is offense and 30% is defense. Using the argument quoted, one would then arbitrarily adjust that to make it 80/20 or 90/10, hence the more reliable numbers are weighted above and the defensive numbers below what they should be. That's flawed math, flawed statistical methodology. After all, it's equally plausible Pedroia's defense is being underestimated.
If a given value has less certainty than another, you don't distort their importance by re-weighting them, you just live with the uncertainty, or try to account for it in other ways. Accounting subjectively, as some have suggested here, is a much better way of handling things than doing something that is statistically wrong.
Getting back to the question at hand, BB-ref has Bautista at 6.8 and Pedroia at 6.2. Fangraphs has them at 7.0 and 6.7, respectively. That's close enough to make Pedroia a credible MVP winner. I also think, since voters give a lot of weight to team performance when selecting the MVP, you could argue Pedroia is the favorite to win right now, even if he doesn't gain any additional ground.
Finally, is it even certain Pedroia is the most valuable on his own team? Ellsbury, Beckett and Gonzalez seem to be giving him a pretty vigorous run for his money.
So it seems to me that you are double counting the positional adjustments when you start playing those games. If you'd rather compare both players to the same offensive replacement level, be my guest.
What?
I simply don't understand this.
Why must equipoise be maintained between the two? At the team level, sure (but even there, pitching is half of the defensive game). Defensive opportunities are not equal across positions. Offensive opportunities roughly are. It seems to me that all that was being argued is that since Pedroia has more defensive opportunities, and we have more uncertainty about how much value he actually produced in those opportunities, we therefore have more uncertainty about his total value. What's so wrong about that?
Getting back to the question at hand, BB-ref has Bautista at 6.8 and Pedroia at 6.2. Fangraphs has them at 7.0 and 6.7, respectively. That's close enough to make Pedroia a credible MVP winner.
Can't disagree with that part.
I admit that this changes my argument considerably (although not entirely) but also, holy ####. That's so stupid it hurts.
Didn't you just quote about positional adjustments being accounted for by oWAR? Are positional adjustments not accounted for defensively as well? I believe they are (having trouble finding the definition).
I'm fine with you having more uncertainty about defensive statistics than offensive statistics. I'm just not fine with penalizing defensive-oriented players by arbitrarily re-weighting to favor offense because of that uncertainty.
Uncertainty about a value doesn't increase or decrease its inherent weight. That might be unsatisfying for someone looking for a definitive answer but that's just the way it is.
Not so clear. Guess we'll have to ask AROM.
I'm fine with you having more uncertainty about defensive statistics than offensive statistics. I'm just not fine with penalizing defensive-oriented players by arbitrarily re-weighting to favor offense because of that uncertainty.
I'm not penalizing anybody. I was just trying to re-state what I saw the argument as being about, because I really didn't get all the fuss. Personally, I don't care about MVP awards nearly enough to go to this much trouble to decide who I'd vote for if I had a vote.
Uncertainty about a value doesn't increase or decrease its inherent weight.
I don't disagree with that either. But I also don't have all that big a problem with someone more or less saying that "given things I have very high certainty about and things I have much lower certainty about, I'll base my decision on the things I have high certainty about."
I was about to say something but it seems you may have already faced the error of your ways...
The park factors are prorated in season. As we are 2/3rds of the way through the season. I'm using 1 x 2010 + 1/3 X 2009 + 2/3 x 2011 divided by 2.
As I said in my blog post on it. The main reason is to provide a WAR number with no fielding component, so people who distrust TZR can get a number with all of the less controversial WAR values.
http://www.baseball-reference.com/blog/archives/8219
This is the post where I launched it and explained my reasoning.
See comment #10 for my reasoning behind calling it oWAR and dWAR.
If every player is an "average defender" then there's no need to adjust for position. Right? I really just don't get this.
If you look at it from the perspective of the team's offense, Pedroia being farther down the defensive spectrum adds to the Red Sox offense.
I didn't mean to ask him about the split. Just to ask him about any finer points of how Total Zone is claculated.
But the season isn't over.
Average relative to their position played. As you know there are dramatic differences between the quality of the hitters at various positions. This is what Rpos is measuring.
Doesn't that just mean that oWAR ignores the quality of defense, but not defensive position?
If you want to compare offensive value without adjusting for position, just subtract Rpos from oRAR.
I don't know about VERY misleading, but yeah, I suppose oWARP would be better.
No, it's Positional WAR.
And unlike my example, this wouldn't happen. What name would you like better?
I'm imagining so many passed balls that the Red Sox never got to bat, which I think would hurt their offense.
You must be Registered and Logged In to post comments.
<< Back to main