Crashburn Alley

By Max on Sep 16, 2010

Mark Reynolds
Fangraphs: 2.4
Baseball-Reference WAR: 1.2
(http://www.baseball-reference.com/players/r/reynoma01.shtml#batting_value)
oWAR: 2.0
dWAR: -0.8
(http://www.baseball-reference.com/blog/archives/8219)

Ryan Howard
Fangraphs WAR: 1.9
Baseball-Referene: 1.8
oWAR: 2.3
dWAR: -0.5

Care for a rebuttal, Deitch?

By Dennis on Sep 16, 2010

I don’t see how Mark Reynolds is more valuable than Ryan Howard. I don’t care what the formula is. I know what I watch.

By Jon on Sep 16, 2010

So Sabermetrics is a science now?

By Mark on Sep 16, 2010

When will people realize this? When it comes to evaluating a player who has made hundreds of PAs in a year, your memory is not sufficient as the lone evidence for your opinion. It is biased and far more flawed than statistics. Additionally, Dennis, have you really watched Mark Reynolds enough this year to make a judgment like that?

By Hunter on Sep 16, 2010

“WAR is telling you that this year, Reynolds has been more valuable”

Is it really telling me this, or is it telling me what whoever created the WAR equation considers valuable?

WAR isn’t even uniform. There are two different websites I can go to that will give me two different WARs.

WAR is a judgement call, backed up with statistics, but still a judgement call.

WAR doesn’t tell me that Reynolds is more valuable than Howard or vice versa. It tells me that, given certain weight to certain things, an equation, and the creator of that equation, value Reynolds more than Howard.

This might seem like it’s splitting hairs, but to me the distinction is important. To some proponents of sabermetrics (especially the Smuggy McSmug variety) WAR is almost the be all and end all of player value. However, it seems to me that it’s simply another tool we CAN use for evalution, not the final word.

By MFP on Sep 16, 2010

But you’re arguing a tautology.

In this and the previous post, you’re replying to the statement “Something in the [WAR] equation is lethally flawed [because I do not believe that players x or y are more valuable than Ryan Howard].” But each time you do so by plugging numbers into the WAR equation, showing how it’s calculated, and saying, “WAR is telling you that this year, Reynolds has been more valuable.”

If you wanted to respond directly to the assertion (the WAR equation is flawed), you’d need to explain why you think that Reynolds is more valuable than Howard, by about the amount that WAR projects, but without simply using WAR as the means to do so.

By Hunter on Sep 16, 2010

MFP: I think WAR is flawed because it relies on defensive metrics to judge a lot of the player’s value. Defensive metrics are useless for partial seasons, somewhat useful for 1 season, and useful for 3+ seasons.

That doesn’t make WAR a useless statistic. Not in the slightest. But it doesn’t tell me that in half a season, Reynolds has actually been more valuable than Howard. The numbers are close enough that he could be. It also shows that Howard could be.

Bill mentions this in his article when he asserts the fact that given the unreliability of the defensive metric, there could be a 1 WAR shift in the value between Howard and Reynolds.

In a full season, I trust WAR enough to tell me that Josh Hamilton has much more value than Melke Cabrera. I do not trust it to tell me that Reynolds is more valuable than Howard. The numbers are too close. At most, I think we could say that Howard and Reynolds were roughly around the same value.

By MFP on Sep 16, 2010

To be a bit clearer, the post reads as:

Q: The WAR equation says that Reynolds ’10 is more valuable than Howard ’10. Is this correct?

Assertion: WAR is an equation that tells you the value of a player in a given year.

(1) Here’s Howard’s WAR and an explanation of how it was calculated.
(2) Here’s Reynolds’ WAR and an explanation of how it was calculated.
(3) Reynolds’ ’10 WAR > Howard’s ’10 WAR
(4) Reynolds ’10 is more valuable than Howard ’10.
QED.

The can’t assume the equation works to show that it works.

By Mark on Sep 16, 2010

Indeed, it is not the final word. Mostly because defensive metrics still have plenty of room for improvement. But WAR is a pretty good place to start. There is certainly no traditional stat that even attempts to do what WAR does. Some people want to scrap WAR completely or overhaul it just because it doesn’t hit the bullseye. But it comes closer than any other singular statistic. If you were to evaluate a player without using WAR, you would have to look at a few statistics and then make the judgment call in your head as to how to weight them. Just know that the people who made the judgment calls when computing WAR are really intelligent, and they put a great deal of time and effort into that decision.

By Hunter on Sep 16, 2010

“Just know that the people who made the judgment calls when computing WAR are really intelligent, and they put a great deal of time and effort into that decision.”

I think part of the problem is that some of these really intelligent people tend to get hyper defensive when people question their conclusions. When people actually have the temerity to disagree with their conclusions, they get highly offended.

Some proponents of WAR treat it almost like a number that is tied to a fact (like home runs) and therefore cannot be argued with or have a significant margin of error.

Some, obviously, like Bill are not that way, but enough exist of the former to really cloud the dialogue.

By Greg on Sep 16, 2010

MPF: http://tinyurl.com/2cy4fws

Frankly, every Johnny come lately makes the same arguments time and again. Do some research, there is alot of good stuff out there that should be read before sounding off.

By MFP on Sep 16, 2010

@Greg, I’m not sure what you’re point is. The post did a fine job of showing how it was calculated. My point was that “here’s how it’s calculated” is a far cry from “Here’s why it’s correct.”

By MFP on Sep 16, 2010

s/you’re/your/

By Hunter on Sep 16, 2010

One thing I’ve wondered about WAR…I understand there is a positional adjustment.

Via http://www.insidethebook.com/ee/index.php/site/comments/how_to_calculate_war/

These are the positional adjustments:
The positional adjustments are:
+1.0 wins C
+0.5 SS/CF
+0.0 2B/3B
-0.5 LF/RF/PH
-1.0 1B
-1.5 DH

Why pick these numbers? I mean, I understand there NEEDS to be a positional adjustment. But why these specific numbers? They seem very…round, like someone picked them out of a hat, or at the very least rounded them. Why -1.5 at DH and not, say, -1.47?

Is there an empirical reason for this or just someone best guesstimate of the positional values?

By Mark on Sep 16, 2010

Reynolds and Howard are too close in WAR to make a definitive statement about who is more valuable because of the the unreliability of one season’s worth of UZR. As Bill points out, the error involved could easily vault Howard ahead of Reynolds. I think the meaningful observation to make regarding Ryan Howard and WAR is that he cannot even sniff the likes of Votto, Pujols, Gonzalez, and Cabrera.

By DrPete on Sep 16, 2010

I’m not even sure I understand the point of comparing Reynolds to Howard. It’s like comparing the proverbial apple and orange.
What WAR is telling us, in this case, is that it’s harder to find even a decent 3rd baseman (see Feliz, Pedro) then a decent 1st baseman, so an above average 3rd baseman (Reynolds) has value than Howard, who plays 1st. This isn’t a flaw in WAR, it’s just the reality of baseball — not all positions are equally easy to fill.

What’s more amusing is that Reynolds is flat-out brutal at 1st defensively (UZR/150 -37), so he’d be pretty close to a replacement player at 1st, even if he improved his defense to Howard’s level.

By Sanj on Sep 16, 2010

How do explain why the Phils are a much better team (20 games over v. 3 games over .500) with Jimmy Rollins in the lineup even though his WAR is so low? How is it possible that a player with a poor statistical season can have such a profound winning effect for a team?

By MFP on Sep 16, 2010

FWIW, I don’t mean to sound critical of Bill here, I might not disagree at all, really.

I’m just saying that there’s still plenty of room to believe, despite what WAR says, that Howard has been more “valuable” this year than Reynolds (who, frankly, I’d only vaguely heard of before about an hour ago, and have zero opinion as to his value). Or suspect that WAR might be kind of screwy for lots of players. Or that it generally provides an accurate estimate of the number of games you’d have gained/lost by exchanging two players in the previous season. Or something in the middle, like it’s mostly useful to compare players with significant discrepancies in WAR.

By Michael on Sep 16, 2010

I’ll jump in to help Bill answer some of these questions.

First off, let’s make sure we understand what WAR is. It is a FRAMEWORK, a method of calculating value. To analyze its value, you need to break down the framework. For position players, the framework is this:

1) Add offensive and defensive value above average
2) Add a positional adjustment to adjust for scarcity
3) Add the value above a replacement level player, which gives a value for playing time.

Keep in mind that this is a framework for calculating value for a season or for multiple years if you are adding them up. It is NOT a prediction on who is a better player going forward, but rather who might have been more VALUABLE in the past.

If you think the framework (and more specifically, how it’s calculated) is sound, then you have no issues with the WAR framework, no issues with WAR.

The components that go into WAR are a different story. If you’re worried about the defensive aspect, then you have an issue with UZR/TotalZone/DRS/whatever. Take it up with that.

Also, if your issue with WAR or any of its components are with people who bring it up saying “Player X is 0.3 WAR better than Player Y, which strictly shows that that player is better,” then take it up with those people, not with WAR. The people who use WAR responsibly know that WAR isn’t really accurate to the tenths digit for various reasons. It is listed often times to the tenth digit so that leaderboards don’t look like lists with numbers between 0 and 7 in front of them. If you see someone using WAR to the tenths digit to clearly define that one player is better than another, call them out on it, because it is a misuse of the stat.

For the question on positional adjustment: the adjustment was done in various ways. Tango used UZR over a five or six year time period for players who switched positions, and used some subjective judgment to fill in the first base/catcher/DH parts. Rally (originator of B-Ref’s WAR) looked at the nth best projected player at each position offensively, where n is the rank of the first few players who would not make a major league roster (ie the 51st or so best 2B). You can check out Rally’s method here:

http://www.hardballtimes.com/main/article/replacement-level-article/

By B on Sep 16, 2010

@Bill Baer:

“What we learn from this is that Reynolds is credited a lot just for where he stands on the diamond”

It’s not just that he stands there on the diamond. The defensive calculations in WAR are done relative to the average at a players position. So Ryan Howard’s UZR is his defense compared to the 1B average, while Reynold’s UZR is his defense compared to the 3B average. This is where positional adjustments come in. 3B are better fielders than 1B, and that’s exactly what the positional adjustment is – how much better the group of fielders at each position is. So an average fielding 3B is about 15 runs better over a whole season than an average fielding 1B, and that’s where the difference in value comes from.

@Hunter – that should also answer your question where the positional adjustment numbers come from. Smart people took the time to estimate the difference in fielding ability between positions, and those are the positional adjustments used.

@MFP – yes, anyone that understands WAR knows there’s some variance in it’s measurement, which is fine. We can’t say with a significant degree of confidence (90% or higher) that Reynolds has been better than Howard. I don’t think it’s fair to call that variance “screwy”, though. Also, there’s plenty of stuff out there about why WAR works, in addition to all the stuff about how to calculate it.

@Sanj – learn the concept of random variance. It’s a useful thing to understand.

@DrPete – Reynolds has 229 career innings at 1B. That UZR number is going to have a pretty huge variance.

By Hunter on Sep 16, 2010

“@Hunter – that should also answer your question where the positional adjustment numbers come from. Smart people took the time to estimate the difference in fielding ability between positions, and those are the positional adjustments used.”

I understand the gist of what you’re saying, but firstly, it seems these are averages, and secondly, they don’t seem to contain a lot of precision. I have no problem with that, but it seems it would make the individual WAR figure a lot less valuable. It seems like WAR would be better expressed as a range.

In other words, I am questioning the usefulness of WAR for making player to player comparisons when the WARs are close…which is what most people seem to want to use it for. WAR is most precise when used to make comparisons where the WAR difference is large, but most fans don’t need WAR to tell them Josh Hamilton is a lot better than Melke Cabrera.

So WAR isn’t very good at comparing similar players and making determinations on their relative values because of the inherent volatility and lack of precision in some of the statistics it uses (like UZR). It is good at comparing wildly divergent players….but we don’t need it to do that. We can do it with our eyes.

So….why use it?

By Richard Hershberger on Sep 16, 2010

I think perhaps the unease many have with WAR is in the idea of the position adjustments. Yes, third base is a much harder defensive position than first. But you have to put someone at both positions.

Let’s stipulate that Reynolds could play first base just fine, or even excellently. (An earlier comment suggests that this is not true, but this is for the sake of argument.) Would you rather have both Reynolds and Howard on your team, or Reynolds and his identical twin? Sure, Reynolds II would do a bang-up job at first, but you would be losing Howard’s offense. Furthermore, Reynolds II would end up with a much lower WAR than Reynolds I for his effort. (Perhaps out of fairness the manager will alternate them between first and third.)

An interesting argument would be to make the thought experiment of replacing Reynolds and Howard with Howard and his identical twin. Howard II at third would obviously be a disaster, but would the extra runs from his offense offset the extra runs allowed by his defense? If the answer is no: that the defensive liability outweighs the offensive asset, then this might be an intuitive argument for Reynolds’ greater value.

But even then I’m not sure this isn’t too much comparing apples and oranges. You want your team to have both a first and a third baseman, not two of either. My question is is it harder to find a Howard or a Reynolds? My intuition is the former: Howard’s level of offense combined with his ability at first is rarer than Reynold’s offense combined with his ability at third. Show than my intuition is wrong and we will really be onto something.

By Max on Sep 16, 2010

This so-called “batting average,” (a statistic designed by pointy headed, basement-dwelling intellectuals if there ever was one) says that Derek Jeter is “batting” .262 while Omar Infante (who?) is “batting” .340. There’s clearly no way Infante is as good as His Holiness Cap’n Jetes, ergo “batting average” is a meaningless statistic.

By Hunter on Sep 16, 2010

“This so-called “batting average,” (a statistic designed by pointy headed, basement-dwelling intellectuals if there ever was one) says that Derek Jeter is “batting” .262 while Omar Infante (who?) is “batting” .340. There’s clearly no way Infante is as good as His Holiness Cap’n Jetes, ergo “batting average” is a meaningless statistic.”

Thanks for proving my point earlier and going all Smuggy McSmug on us.

By Phillies Red on Sep 16, 2010

Bill, I’m with you almost 100% on this, and find this conversation very elucidating. But I still struggle with calling WAR descriptive, and mainly because of the UZR component.

Here is what MGL wrote in his May UZR primer at fangraphs:

“UZR tries to record a player’s likely true talent and estimate his future performance based on the nuances of the batted ball and the player’s response to those nuances. It is not trying to capture exactly what happens on the field according to some arbitrary categories, like most of the offensive metrics…”

He states clearly that UZR is not descriptive, and here he goes so far as to say that UZR is actually trying to make claims about a players true talent.

So this really becomes a question about the WAR “framework” you and Tango have been talking about: why does the framework of a descriptive statistic contain a major non-descriptive component? And as it does contain this component, what is WAR actually telling us? Because from here, it looks like WAR is partially descriptive, while being grounded in some kind of true-talent estimations.

By Dennis on Sep 16, 2010

This so-called “batting average,” (a statistic designed by pointy headed, basement-dwelling intellectuals if there ever was one) says that Derek Jeter is “batting” .262 while Omar Infante (who?) is “batting” .340. There’s clearly no way Infante is as good as His Holiness Cap’n Jetes, ergo “batting average” is a meaningless statistic.

I really hope that is supposed to be sarcastic. Batting average is not at all a meaningless stat.

By Dennis on Sep 16, 2010

I think on base percentage is the most important stat in baseball. The point of the game is to get on base. But to say batting average is meaningless is ridiculous. And by the way Max, Omar Infante made the all star team and is a starter for the Braves….might want to get to know who he is being a phillies fan.

By Max on Sep 16, 2010

re: Dannis

Really? “His Holiness” didn’t trip your sarcasm detector?

By Dennis on Sep 16, 2010

no it did….just making sure….ive heard some terrible arguments before

By B on Sep 16, 2010

@Hunter – thinking of UZR as a range is not a bad way of putting it, and thus you get a range for WAR because of the range for UZR. As for your question:
“So….why use it?”

I’m not really following. Why wouldn’t you use it? If it’s the best measure of a players combined offensive and defensive value, wouldn’t we want to use it to have the best idea of a players total value as possible…?

@Richard – if you use Tango’s framework where the positional adjustments are strictly defensive, you actually expect Reynolds WAR, on average, to be the same at 1B and 3B. He loses value on the positional adjustment side but gains value on the defensive side (because he’s now being compared to the average of a worse group of fielders), and those two things are equal, on average (because that’s exactly how the positional adjustments were calculated, by looking at the defensive difference between positions). Of course, not every player will be exactly average – maybe a guy with a strong arm loses more than a guy with a weak arm making the 3B to 1B switch or something, but again, on average, it should be equal.

@Dennis – the offensive component of WAR, which is calculated with linear weights, does what OBP does – gives a player credit for getting on base, while ALSO giving them proper credit for getting on base in better ways. That is, a single is worth more than a walk, a double more than a single, a triple more than a double, a HR the most. What I’m getting at is there are better stats than OBP. Getting on base is definitely a very important concept, but we do have better ways of measuring offensive value than that these days, and really, linear weights means we don’t need BA/OBP because they don’t tell us anything linear weights doesn’t (while linear weights does tell us things they don’t)…

By Hunter on Sep 16, 2010

“I’m not really following. Why wouldn’t you use it? If it’s the best measure of a players combined offensive and defensive value, wouldn’t we want to use it to have the best idea of a players total value as possible…?”

If I was on a deserted island, and made myself a tool out of a sharp rock and a stick, that would be the best tool I had available to me, but that wouldn’t make it a good tool.

Along the same lines, I’m still trying to decide that if just because WAR is the best tool we have to use, if it has enough inherent flaws to render it meaningless enough I would rather not use it unless necessary.

By Dennis on Sep 16, 2010

@B – well then what is the difference between that and OPS??

By Michael on Sep 16, 2010

@Dennis: Re OPS vs. wOBA

wOBA properly weighs each event because it uses linear weights. OPS overvalues slugging and undervalues OBP. In addition, OPS doesn’t make any sense mathematically anyway. It’s a shortcut, it’s descriptive enough to get you most of the way there, but since there are better tools, you don’t need to use it.

By Dennis on Sep 16, 2010

sorry if i sound like I’m challenging anyone, I’m not, I just simply do not understand sabermetrics. Trying to learn about them quick though

By hk on Sep 16, 2010

A couple of points for all of those who seem up in arms by the statement that, based on WAR, Mark Reynolds has been more valuable than Ryan Howard this year:

1. There has historically been a very high correlation between team WAR and a team wins.
2. Since WAR is based on replacement levels at the various positions, it is not comparing two players to see which is better, it is comparing them to see which is more valuable relative to easily obtained replacments at their respective positions. For instance, Ryan has a .862 OPS and Reynolds has a .782 OPS so far this year – yes, I know that OPS is not really a great stat to use, but it is one that most everyone understands. Coincidentally, they are both ranked 10th in OPS among qualifiers at their respective positions. However, since the replacement level (and offensive production) at 1B is higher than at 3B, Howard’s higher OPS vs. Reynolds is mitigated. Therefore, when incorporating defense into the equation, Reynolds has produced more value relative to a replacement player than Howard so far this year.
3. WAR is not predictive; therefore, it does not imply that Reynolds will be more valuable than Howard the rest of this year, next year or at any point in the future.

By Michael K on Sep 16, 2010

I have a similar question as Richard.

To ask it a different way:
Is the distribution curve of talent the same across all positions? Are there consistently more (for example) 5+ WAR players at some positions than at others? Do some positions consistently show a higher average WAR than others?

If (hypothetically) the difference between average and replacement level is higher at one position (say 3B) than at another position (say 1B), then perhaps WAR overstates the value of the player at the first position (in this case 3B).

I say that because most decent teams will have a surplus of talent somewhere, and can use that to trade for an “average” player (rather than a replacement player) at a position of need.

By 80to08 on Sep 16, 2010

Here is the simple difference:

And why the “J-Roll conundrum” makes perfect.

Stats and Math matter in the press room.

Leadershi and the drive to win matter on the field.

Full stop.

By Steve on Sep 16, 2010

sabrmetrics is only governed by samples taken from past events. therefore, id argue that comparison between two players when looking at a relatively small sample size (i.e. ~1 season, howard w/ injury vs. reynolds) is useless.

also, thats why i feel sabrmetrics has a use in describing the long-term performance of an individual or aggregated team, but it’s nearly useless in the playoffs (when it really matters).

By hk on Sep 16, 2010

Steve,

If you can tell us what Howard and Reynolds or the Phillies for that matter are going to do in the future, please do. We can all make a lot of money on it. In the interim, past events are all we have to go on. Also, as I posted above, WAR is not predictive. Some other sabremetric stats are predictive and can be useful (obvioulsy not 100% accurate) in trying to determine which team is more likely to win a particular playoff series.

By Murgatroid on Sep 16, 2010

This might be a bit off topic, but if I’m not mistaken, Fangraphs uses FIP to compute pitcher WAR. If WAR is descriptive and not predictive, why not use ERA?

By curtwill1975 on Sep 16, 2010

It’s good to discuss stats like this because they are interesting and fascinating. But I don’t think they are the end all to be all. Someone mentioned J-Roll and Jeter and one thing both of those two bring to the table, is a feel for the game that stats can’t fully measure. They have a leadership quality that stands out above their stats. A player can bat .320, 35 HRs and 120+ RBIs and not affect the game on a given basis. They are just putting up stats.

Then you have Howard. Yeah, he strikes out a lot, and yes his BA is not really that high nor does he have a huge OBP(though it is about 35% of the time and that’s not bad). And yes, he’s not a great defensive 1st, though he has improve(his problem is throws for the most part), but people don’t fully understand the cleanup hitters’ craft. Like someone said before in another entry, he drives in runs consisently. 100 RBIs in only 129 games(remember he missed 16 games) and that’s his LOWEST in a full season.

And yes, he gets more chances to do it but again, that what’s supposed to happen with a clean-up hitter. A hitter can drive in runs and put up power numbers but not all of them are clean-up hitters. Compare all of the clean-up hitters in every line-up and Ryan stacks up with the best of them. People need to stop “clownin’” him and appreciate him more.

By AMusingFool on Sep 16, 2010

This has been an interesting discussion, and has highlighted a couple of assumptions I was making that I wasn’t aware I was making.

I guess one question for me is, why is it so hard to believe that, over the course of most of a season that Mark Reynolds (who, to be fair, I don’t think I’d heard of before reading this) has been (slightly) more valuable than Ryan Howard, who’s having a terrible year by his own standards.

It doesn’t mean that, given a choice for who to have on my team for a game today, that I wouldn’t choose Howard.

@Dennis, the difference between linear weights and OPS is that OPS assumes that OBP and SLG are equal. To be completely accurate, it would need to weight OBP * ~1.7 before adding to SLG. Also (and to my mind this decision is debatable), wOBA doesn’t include IBB, while OPS does. And almost forgot, wOBA does include reaching on error, while OPS doesn’t.

By curtwill1975 on Sep 16, 2010

And he is doing this for a team who has been the NL’s best team in the last 3 years. A team who always plays well in September. It’s no coincidence that his best months are in September and the Phils play their best baseball, so his skill set is very valuble to the team. If he was to be traded or out for the season, who would bat clean-up? Then people would understand how important Ryan is.

By Scott G on Sep 16, 2010

@Curtwill: Ugh…

By curtwill1975 on Sep 16, 2010

As for Mark Reynolds, he’s not even their clean-up hitter. Adam Laroche is. He just hits a lot of HRs. Ryan hits a lot of HRs but he is the best run producer in the NL easily.

By Sanj on Sep 16, 2010

@curtwill1975

I agree with most of these points. Sabermetrics should only be a tool for analysis (a very useful) one, but I don’t think it is the end all be all for analyzing the value a player has on a team. Specifically, players with otherworldly power like Ryan Howard psychologically impact the game differently than players like Chase Utley. People are risk averse and the threat of Howard smoking the ball over the fence forces the other team to pitch him more cautiously than other hitters. This just has to make other players on the team better as they see more quality pitches. When was the last team to win a WS without a huge power threat in the lineup?
Regarding Jimmy Rollins, he has done it for far too long for the Phillies record with him and without him just to be random variance. Look at the Phillies records over the last two years when he scores and does not score, the winning percentage is significantly higher for him than any other Philly or player in the league. There is no denying that Jimmy Rollins is the catalyst for the Phillies.

By curtwill1975 on Sep 16, 2010

You can say ugh all you want to…it’s my opinion and I can support it. When I say the best run producer…I even put him over Pujols in that way.

Yes, Pujols is the better player and hitter, but if both were in the line-up, Al would be my 3 hole hitter and Ryan would be my clean-up hitter because of the different skill-sets they have as hitters. That to ME…is what values a player, the different skill set that a player has and sometimes though stats can support a certain viewpoint, they don’t tell the whole story.

By hk on Sep 16, 2010

curtwill, your comments show a pure reading comprehension fail. No one and no statistic is saying Mark Reynolds is or has been better offensively than Ryan Howard. The point of the conversation is that, despite the fact that Howard is a much better offensive player than Reynolds (19 batting runs above replacement for Howard vs. 4 for Reynolds), this year (based on the admittedly inaccurate for one season UZR), Howard’s and Reynolds’s defense at their respective positions and the fact that Howard plays a position where the replacement level is higher make him less valuable year-to-date than Reynolds.

By curtwill1975 on Sep 16, 2010

True@ Sanj. Especially with this comment: When was the last team to win a WS without a huge power threat in the lineup?

And it is funny because back in the day…I used to be anti J.Roll but in the last couple of years, what he brings to the table with his play, goes beyond that. I just hope that he can be healthy and then stay healthy for 2011, cause I would love to see him retire as a Philly. Don’t seem right to see him in anything other than Red Pinstripes.

By hk on Sep 16, 2010

Sanj,

If his power is otherworldly, how come he’s 16th this year in HR’s per AB?

By Scott G on Sep 16, 2010

@Sanj: If people are pitching him more cautiously why is he walking less frequently than Jimmy Rollins (a leadoff man who NEVER walks). If they’re pitching him more cautiously and the pitcher’s heartrates are through the roof (since they’re scared of him), wouldn’t their control be terrible. He has lost his discipline at the plate, and this detracts from his value.

@Curtwill: If you learn anything from this site it should be that you actually CANNOT support that claim. Albert Pujols takes a backseat in any offensive category to very few current players (1 or 2 MAYBE).

If you batted Pujols 3 and Howard 4, Pujols might be intentionally walked in every PA of the season.

By curtwill1975 on Sep 16, 2010

But you’re missing MY POINT@hk

Yes, I know that Ryan is not a great defensive 1st basemen not to mention, that defensive stats at least but the WAR is more loaded positionally toward 3rd basemen than 1st basemen in this case in terms of being more valuable(like comparing a CF to a LF).

My point is because of the fact that even formulas like this is subjective, is flawed in itself. The Babe wasn’t the best defensive player but no one would argue against his value because he wasn’t a great defensive player. It’s the same with Ryan as oppose to Mark Reynolds.

By Joe on Sep 16, 2010

My only issue with Fangraphs WAR is positional adjustment and UZR. UZR is far too volitile year to year and plays to large a part in WAR for me to feel comfortable with the WAR it spits out.

By hk on Sep 16, 2010

curtwill, no one is saying WAR is perfect. As I’ve pointed out in prior posts, it is flawed. However, since team WAR has a high correlation to team Wins, it seems to be a better metric than the “I’ve watched Howard and I know he’s more valuable” one.

By curtwill1975 on Sep 16, 2010

Scott G.: I know this[Albert Pujols takes a backseat in any offensive category to very few current players (1 or 2 MAYBE)], I just said that Al is a better hitter(MUCH BETTER in fact, to be emphasized) than Ryan but one thing Ryan does and does well is driven in runs and consistently. And my point was to compare their skill sets to show their value to their respective teams.

And it is not like Mark Reynolds is a great defensive player either. Ryan is one hell of a run producer doing it, year in and year out(including this year). If nothing else he should get credit for that.

By Walter Poole on Sep 16, 2010

Speeking as a long-time Phillies fan who grew up in SNJ & worked in Philly (but now living in CNJ in the center of METS territory, living for the ESPN ticker), unfortunately, I can’t settle the question of the value of WAR. But I was IMPRESSED with the intelligence & academic nature of most of your posts.

In the final analysis, WAR values seem to come down to judgement & how the player fits in with the team. Also, obviously a Phillies fan is going to say Howard is better than Reynolds. (I’m not too sure how objective this can be, especially when one admits to only recently hearing about Reynolds).

The one thing I’m reasonably sure of is that there will be about 2M along Broad St. not too long from now !

By curtwill1975 on Sep 16, 2010

Hk, but what I am saying is all judgement is subjective. Numbers and stats have made for and against a topic, and at the end of the day they don’t fully tell the story. I just made some stats that support(IMO) why I think Ryan is more valuable than what seems to be thought of in here. Not be the best support, I can admit that.

By Scott G on Sep 16, 2010

The funny thing about producing runs is that it’s largely based on one’s teammates. Ryan Howard can’t knock in 100 runs if the people in front of him aren’t getting on base.

Howard bats fourth. That means that the most common, realistic lineup slots for him to knock in are the 1st, 2nd, and 3rd hitters. Pujols has batted 3rd about 83% of his career which means that he relies on the 9th, 1st, and 2nd slots to get on base. It’s not a fair comparison. If the man gets on base a lot more and slugs a lot more, that’s the only thing you can objectively evaluate him on.

They can’t control what happens before they come up to the plate.

I am very confident that I am correct when I say, given the same situation, Albert Pujols would be a better run producer than Ryan Howard.

By curtwill1975 on Sep 16, 2010

@Scott

I am not saying that I wouldn’t love to have Pujols in the 4th hole. Let’s say you trade Ryan for him, he’s in the 4 hole, with Chase hitting 3rd. But Al is the ULTIMATE 3 hole hitter because of his eye, his bat control and THEN he hits for power and drives in runs. Ryan isn’t on that level as a hitter. We know all of that.

But usually, the best “all-around” hitter bats 3rd and the best run producer bats 4th. Imagine Babe and Lou Gehrig, that’s the model. And Babe was a great run producer himself but he was a better ALL AROUND hitter than Lou(and Lou is great too).

You’re underrating Ryan’s ability to produce runs BIG TIME. Like I said, he missed 16 games and he is still in 4th place, 6 behind the leader. He was leading before the injury. I mean, of course, guys need to be on base to drive in runs. But the clean-up hitter still need to drive in runs and Ryan does it better than anyone in the last 4 years including even this one in the NL. That includes Albert. There is a reason why certain player hits in the 3 hole and some hit in the clean-up. If STL had Al AND RYAN, Tony LaRussa would keep Al in the 3 hole for precisely what I am saying.

By zfg on Sep 16, 2010

@AMusingFool

Nail on the head, man. Ryan was having a down year even before he got hurt. Then he missed some games, which obviously hurts your value. I don’t so what’s so incendiary about a stat that’s basically saying: “hey, Ryan Howard is have a bad year by his standards.”

By jauer on Sep 16, 2010

anyone who thinks howard is in the same offensive class as pujols should switch to football

By dwilly on Sep 16, 2010

garbage

By eh on Sep 16, 2010

you can put about 10 guys in howard’s spot in the line up and get the same or better results.

By Scott G on Sep 16, 2010

I’m actually kind of upset no one responded to curtwill while I was away.

This is not a matter of opinion. Facts are facts. Albert Pujols does EVERYTHING necessary to produce runs better than Ryan Howard. Math has proven that OBP is the stat most highly correlated with runs scored. SLG is also up there. Probably 2nd. ALL YOU CAN JUDGE A PLAYER ON IS WHAT HE DOES. PUJOLS OBP AND SLG ARE CONSIDERABLY HIGHER THAN HOWARDS. If he has a higher average, a higher obp, and a higher slg, he does everything in his power to produce runs BIG TIME.

The only other explanation as to why Ryan Howard tallies more RBIs is that the people in front of Howard are better and get on base more. This isn’t even really up for debate as Chase Utley and Shane Victorino (before this year) are undoubtedly better than people in the Cardinals lineup. Prior to the arrival of Matt Holliday in STL, Albert Pujols was essentially the only hitter on that team. Now Pujols has protection. Howard has almost always had some sort of protection (Burrell, Ibanez, and Werth).

By Van Pelt on Sep 17, 2010

I’m wondering if we’ve hit a tipping point for Crashburn Alley or if it’s just a Ryan Howard effect. What are the hits like Bill?

By hk on Sep 17, 2010

Scott G, I think we all just gave up. If you read Rob Neyer, the only new sabremetric equation you need is curtwill + sanj = theraat.

By bill on Sep 17, 2010

Pitcher WAR is based on FIP to remove defense from the “value” of the pitcher, I believe.

As far as hitter WAR, despite what the Fangraphs guys say, I still think first baseman UZR is screwed up in some way. I don’t take a lot of meaning from those UZR numbers generally, and primarily judge first baseman based on hitting, unless they have excellent (or horrible) defensive stats.

By Scott G on Sep 17, 2010

hk,

haha I don’t read Neyer (maybe I should), so that reference is over my head. I just don’t understand how people can do things like concede Pujols is better than Howard, and then insist that Howard is a better run producer; probably basing it on some intangible like “having a knack for clutchitudidity” or something like that.

I just don’t understand the progression of logic that leads people to conclusions like that. While I probably come off harsh when I make posts like that, I’m just trying to help someone out with the little that I know. However, I’m usually so perplexed at the time that it doesn’t come out that way.

By Rick on Sep 17, 2010

“If you read Rob Neyer, the only new sabremetric equation you need is curtwill + sanj = theraat.”

Boy, that was especially lame.

By Lee Panas on Sep 17, 2010

This has been a great discussion. Michael is correct that WAR is a framework rather than a measure. I think when presenting WAR to non-math people, one should never start out with the final WAR. The components of WAR need to be introduced first.

Start out by saying that Howard leads Reynolds by 14.2 offensive runs. That’s the easiest part of WAR to understand. I think every discussion of WAR needs to start with offense or readers will get lost.

Go on to explain that the gap closes when we consider that Reynolds plays a more demanding position. Then give the defensive runs number. Finally, make the playing time adjustment.

If you present the framework like that, the reader can decide for himself that he doesn’t want to use UZR. He can either throw it out and assume that all players at the position play equal defense or replace it with a fielding measure of his choice. He can do the same with any part of the framework.

If you allow people to see all the components of the WAR formula, people will be less likely to dismiss the concept and may understand it better.

By Lee Panas on Sep 17, 2010

I don’t know if anyone responded to Hunter’s question about positional adjustments. It is based on the fielding performance of players who played multiple positions. This is tricky because of the limited number of players playing an adequate number of innings at multiple positions over the years. I suspect it’s a work in progress and that it may be tweaked in the future.

By Dennis on Sep 17, 2010

I’ll agree that Pujols is the better hitter fielder and overall player, but he is not a better run producer. Pujols career high is 137, Ryan Howards career low in a full season(up until this season) is 136. Howards 162 average is 49 homers and 139 RBIs. Pujols is 43 and 128. He does everything else better, but not that

By Richard on Sep 17, 2010

what you’re still failing to appreciate is that “better run producer” is not the same as “hits more HRs or collects more RBI”

By bill on Sep 17, 2010

That is true, Albert Pujols is much worse than Ryan Howard at having other good hitters on his team.

By Dennis on Sep 17, 2010

@ Richard – what is a run producer other that driving in runs?

@ Bill – Pujols had jim edmonds, scott rolen, david ekstein. they were good hitters and he had some good 100+ win teams before…and hes had Matt Holiday behind him for the past year and a half. I dont think Howard is better then Pujols, but Pujols hasnt exactly been playing with you and me either.

By nik on Sep 17, 2010

@Scott G

“Howard bats fourth. That means that the most common, realistic lineup slots for him to knock in are the 1st, 2nd, and 3rd hitters. Pujols has batted 3rd about 83% of his career which means that he relies on the 9th, 1st, and 2nd slots to get on base.”

Since you’re Pujols’ #1 fan you should probably know that LaRussa bats his pitchers 8th, so in effect he pretty much does bat 4th.

By AMusingFool on Sep 17, 2010

@sanj: You said, “When was the last team to win a WS without a huge power threat in the lineup?”

Ok, here is the highest HR total for each WS champ since 1969.

39, 48, 35, 49, 40, 43, 32, 30, 57, 30, 29, 29, 23, 29, 27, 33, 34, 29, 25, 33, 25, 34, 27, 36, 33, 33, 19, 13, 48, 32, 27, 37, 29, 28, 32, 26, 48, 35, 26

How many of those would you call “huge power threats”? 8, maybe? Certainly 5. So is that really so necessary?

@curtwill: I’d agree with you, that given Pujols and Howard on my team at once, I’d bat Pujols first, but there’s two reasons for that, and neither is anything close to what you said.

1) 3rd batter gets more at bats
2) Pujols is more likely to get on base for Howard to drive in than vice versa.

(Note also that the second factor also means that the team gets more total plate appearances (though probably only 3-5 over a season) with Albert up first)

And about comparing the two as run producers, here’s some numbers about how well each is doing since 2006:

Pujols has 1599 PA with 2200 men on, driving in 1996 of them (OBP/SLG: .465/.615)

Howard has 1722 PA with 2406 men on, driving in 2179 of them (OBP/SLG: .404/.621)

Honestly, that’s a lot closer than I would have guessed, in terms of men driven in vs opportunities. I think that reflects the virtually identical slugging percentages, plus the added HR of Howard.

Interestingly, though, with RISP, Pujols’ OBP/SLG is .520/.679, while Howard is .418/.576 Very good numbers for Howard (All-Star starter level), but Pujols still absolutely destroys him. The OPS difference is the same as between Howard and Jerry Hairston Jr (this year). Really, it’s that big a difference.

By Scott G on Sep 17, 2010

Dennis,

If the possibility doesn’t exist for Pujols to reach those numbers it’s not a fair comparison.
I would almost guarantee that if you swap Pujols with Howard that he would produce more RBIs. He’s a better hitter. He would be in a better lineup, and naturally he would knock in more runs.

2010 PAs with RISP/RBIs:
Howard- 177/100
Pujols- 186/105

2009:
Howard- 216/141
Pujols- 189/135

2008:
Howard- 223/146
Pujols- 176/116

PA with men on base 08,09,10
Howard- 351, 354, 292
Pujols- 322, 346, 305

Career triple slash line with men on
Howard- .287/.399/.611
Pujols- .334/.447/.623

Based on their individual numbers, you cannot explain why Howard has more RBIs than Pujols. It has nothing to do with him, CLEARLY. The answer is the Cardinals surrounding lineup is worse than the Phillies lineup. This should be obvious

By Scott G on Sep 17, 2010

nik,

I’m not Pujols #1 fan. I hate the man. He won the WS and whined that he didn’t win the MVP. He’s much better than Howard though.

I do know he does that, but I wanted to see if he or someone else would call me on that. He doesn’t do it every game (about half the time). I would also say our lead-off hitter is better than their 8th (who would be batting 9th). Pulling up the Cardinals box score from Wednesday, the 8th, 1st, and 2nd hitters for the Cards:

Brendan Ryan- .226/.282/.301
Skip Schumaker- .271/.333/.350
Randy Winn- .273/.333/.417

Felipe Lopez who batted 2nd on Tuesday:
.231/.311/.341

Any more questions?

By AMusingFool on Sep 17, 2010

@Dennis – David Eckstein? Really? He of the .317 wOBA (that’s 5% below league average)? He was one that Pujols was lucky to have on his team?

And Edmonds and Rolen? Well, they were certainly very good players with the Cards, but didn’t at least one of them, and maybe both, hit behind Pujols? It’s awfully hard to knock in guys hitting after you.

By Dennis on Sep 17, 2010

@ Scott G – i would never argue Howard is better than Pujols hes just not. Pujols had some good teams and was in some good lineups before Howard was in the league. You also need to take into account that Howard always goes into slumps throughout the year and Pujols doesnt. Thats one of the many reasons Pujols is better than him. Even with those slumps Howard still usually exceeds Pujols power numbers. Howards MVP year when he hit over .300 58 HRs and 158 RBIs….pujols has never come close to that.

By Richard on Sep 17, 2010

Wow.

By Dennis on Sep 17, 2010

@ aMusingFool – they both did because he hits 3rd, but the players hitting behind you offers you protection and makes pitchers pitch to you. I have been very clear i think in this entire section that i dont follow sabermetrics too closely so i dont know what wOBA is. But I do know when he was on the cardinals he hit .294, .292, and .309. His on base percentage those years was also .363, .350, .356. Thats not too bad hitting in front of Pujols.

By heather on Sep 17, 2010

You could cut the snark with a knife in here.

By Dennis on Sep 17, 2010

@ aMusingFool – he being eckstein in that last post

By Scott G on Sep 17, 2010

Dennis,

Habla Espanol?

God. I’d love to even concede that 2006 season to you, but Howard had a .421 wOBA while Pujols’ was .436.

Howard’s BABIP was also well above his career average while Pujols was decently below his average.

By Scott G on Sep 17, 2010

I posted before you said you don’t know what wOBA is, so sorry for that. Look it up. I think it’s a very good stat that combines OBP and SLG and takes into account actual values all the hitting events that OBP and SLG miss.

By Dennis on Sep 17, 2010

Guess you really have to know these stats when you come into one of these….Ill have to look it up before next time

By Scott G on Sep 17, 2010

Dennis,

Again, sorry if I came off harsh, honestly not my intent. This is a good, informative blog don’t stop reading because I’m a jerk.

By Richard on Sep 17, 2010

you don’t need wOBA at all for this, though… the two posts (from AmusingFool & ScottG) about the disparity in number of runners on base for each hitter, and their respective slash lines w/RISP, should have settled it, but then right away we got a ref. to Howard’s great 2006.

I discerned no snark, merely exasperation.

By AMusingFool on Sep 17, 2010

@Dennis, I suppose I should have looked more closely at Eckstein’s Cardinal years. He was basically league average (within one percent, I think) over the three years. I still wouldn’t say Pujols was lucky to have him, but it wasn’t a hindrance like I was thinking.

As to the Edmonds and Rolen, having protection doesn’t really help much, if at all:
http://www.hardballtimes.com/main/article/pitching-around-batters

Getting back to the large summary of opportunities that I put up earlier, I forgot to mention PA with RISP. It favored Howard 1076 to 896, slightly more than the “Men On” PA favored him (percentage-wise, of course)

@heather, yeah, sorry about adding to that

By Jamie on Sep 17, 2010

@aMusingFool – you posted the following:

“And about comparing the two as run producers, here’s some numbers about how well each is doing since 2006:

Pujols has 1599 PA with 2200 men on, driving in 1996 of them (OBP/SLG: .465/.615)

Howard has 1722 PA with 2406 men on, driving in 2179 of them (OBP/SLG: .404/.621)”

Assuming this covers 2006 to 2010 (i.e 5 years), you’re saying that Pujols has had an aveage of 399 RBI per year. WOW! I knew he was good, just not that good…

By Dennis on Sep 17, 2010

Nah man no worries, just a difference of opinion thats all. I gotta get to kno these stats anyway cause it looks like theyre sticking around

By AMusingFool on Sep 17, 2010

@Jamie Wow, my brain went out the window for sure there (at least I was consistent, and looked at the wrong column for both, I suppose).

Actually driven in:
Pujols 392
Howard 445

Howard does come out a hair ahead; he’s driving in 18.4% of the runners he sees while Pujols only gets 17.8%. I’m curious how far ahead of average that puts each of them, but I don’t have the data. As I said before, I’m surprised Howard comes out ahead on a per-opportunity basis, even if by such a narrow margin.

Hmm… the Phillies, as a whole, are awfully good baserunners; I wonder if that factors into it (more runners scoring from first on doubles, for instance).

By YanksFan on Sep 17, 2010

doesn’t that stat favor the guy who draws less walks? shouldn’t you look at runners on base when the guy registered an AB (rather than PA) and see how often they came in?

why should pujols be penalized if pitchers don’t throw him a strike

By Danny on Sep 17, 2010

I read a lot of this stuff (I couldn’t read it all, my brain would turn into mush), and I thought of something that would improve the viewpoint of WAR:

Make it a confidence interval.

We see one number, but at the same time, we recognize the flaws it has, especially in defensive metrics. But it seems that we all seem to agree that WAR is able to come close — to some unknown degree — to a player’s value in a given year.

Bill said earlier that while Reynolds’ WAR is higher than Howard’s, it may be closer because of the inaccuracies in the data. To fix this, instead of Howard’s WAR being an arbitrary number like 1.9, why can’t it be {1.4-2.4}? Why can’t it be a range of numbers? We’d have to further develop a level of accuracy so we could say a 95% confidence interval or 90 or whatever, but even then, there is still the possibility of an outlier.

Just from reading different arguments, it seems that people are finding little flaws and ways to tweak the number. In this sense, we could encompass a range of values that are possible for a given player instead of trying to pinpoint an exact value. Scientists do this all the time in research, and that’s exactly what we are doing here.

By YanksFan on Sep 17, 2010

of course my last post speaks to the problem of simply counting up the RBI a guy has driven in, even when you account for the opportunity he has. it disregards the fact that a guy can help his team score runs even if he doesn’t get a hit or drive somebody in in that AB. from the stats provided by amusingfool, pujols makes outs less frequently in those spots, which helps his team score runs even if they don’t end up in his RBI column.

By DrPete on Sep 17, 2010

RBI have always been troublesome to me.
I understand the difficulty in comparing players — you need baserunners to get RBI (baring homeruns), which is dependent on the players ahead of you (be it on a different team or your own team).
At the same time, in the end, you need RBI to get actual wins (as opposed to WARs). If, with runners on, you make an out, or walk, maybe the guy behind you will pick you up and drive in the runs. On the other hand, if you hit a double and drive in a run or 2, then you don’t rely on the guy behind you. In the end, I guess RBI should be considered a team stat and not an individual counting stat and we should simply be glad that RBI continue to show up in bunches on the Phillies side of the box score.

By YanksFan on Sep 17, 2010

its definitely important to make the distinction between individual run production and team run production.

its possible that if we switch albert pujols with ryan howard, he would not have as many RBI batting cleanup for the phillies. i don’t know if this is the case but it very well may be. what’s important is that this isn’t really the question to be asking. the question is, which PHILLIES team scores more runs, the one with pujols batting 4th, or the one with howard batting 4th. that question i’m very confident of. and its really the only one that matters. this is where the definition of “run producer” is complicated.

By Sanj on Sep 17, 2010

Just want to clear this up, I am not anti-sabermetric, it is a great tool for determining the strengths and weaknesses of individual players and determine a proper balance to build a team. I do, however, have a huge problem with WAR. It is not even calculable for all players? WAR stands for wins above replacement. Take a player like Jimmy Rollins for instance. He is irreplaceable to the Phillies, therefore his WAR is infinite, its like dividing by 0 and busting the calculator.
I don’t like definitive statements like Jayson Werth is better than Ryan Howard, just because your fancy numbers say so. The market determines which player is better, and no one is going to give JWerth anything north of $15-17mm a year. Jayson could never be the centerpiece of a championship lineup like Utley or Howard.

By Danny on Sep 17, 2010

I don’t agree with you that the market determines which player is better. Oliver Perez makes north of $10 million, and Jimmy Rollins makes less than $10 million. By your theory, the market has determined that Oliver Perez is a better player.

Why can’t Jayson Werth be the centerpiece of a championship lineup? That’s the exact definitive statement you claimed you didn’t like. Don’t contradict yourself in the same paragraph.

Why is Jimmy Rollins irreplaceable? Again, another definitive statement with no proof to back it up. Rollins has been out for the past few games, and the team keeps winning, so apparently your theory is wrong.

By Scott G on Sep 17, 2010

Sanj,

Thanks for the cliche riddled comment. I bet Werth does get north of 17 mil.

Numbers aren’t fancy

Troy Tulowitzki
Hanley Ramirez

These men would be gladly welcomed over J-Roll at shortstop. I’m not going to nitpick, but let’s be serious: I love J-Roll as much as the next guy, but he’s pretty much only a very good defensive shortstop. His offense is almost brutal, and his age/health is causing Wilson Valdez to get wayyy too many starts. I’d take those two guys in a heartbeat.

By Jim on Sep 17, 2010

@Sanj,

Jimmy Rollins is irreplaceable to the Phillies? Ah, which 2010 Phillies team have you been watching? In the thread in Bill’s previous WAR article, someone said “while WAR is context neutral, people’s memories are not.”

Some people also like to point out Howard’s September “dominance” as proof that WAR doesn’t take into account everything. Don’t get me wrong, it’s mighty nice… And it’s worth maybe a bit extra. But I will take a player who dominates between the months of April through August over a player who “dominates” for one month a season. If Ryan Howard hit like he’s hitting right now between April through August, yeah, there wouldn’t be a “race” with Atlanta right now; no offense, I like that scenario much better. Not saying that Howard is a one-month player; I love the guy. But please, that argument is just idiotic.

By DrPete on Sep 17, 2010

Let’s not go overboard: if you check the splits, Howard was pretty damn good in June and July and then missed most of August.
In fact, checking the numbers more carefully, he was a top 5 1st baseman both of those months — and better than Pujols — not to mention the Phillies best hitter those months.
And now he’s having a very good September.

He started slow. ‘Nuff said.

By DrPete on Sep 17, 2010

I wish we could edit comments.
Using baseball-reference.com or fangraphs.com pretty much any stats are easily looked up.
Please, check the numbers before posting something completely out of left field (or first base). It’s really not hard and you may find something neat in the process.

By charless on Sep 17, 2010

bill baer, write something on chooch’s 3.6 WAR (fangraphs). what a stud, right behind mccann and mauer

By Jim on Sep 17, 2010

@DrPete,

I think you completely missed the point. I wasn’t commenting on Howard’s ACTUAL performance; I was saying using the fact that he’s “dominant” in Sept. as the argument that WAR doesn’t measure everything is stupid.

But to humor you… no one is saying Howard is doing terrible. Just not as valuable as in past years. “He started slow. ‘Nuff said.” Yeah, sorry, but April and May counts too. 2 extra wins in those two months, we would be 5 games up now as opposed to 3. That’s not important or anything.

By micah on Sep 18, 2010

The main advantage that Pujols has over Howard is that he (pujols) bats right handed. Take away the overshift, and Howard’s batting average and OBP go up .030 and his slugging goes up probably .050. Obviously he wouldn’t add any home runs, but he’d probably add 20 hits a season. you take that many hits away from Pujols and see if he is still such a stud.

By Jim on Sep 18, 2010

@micah,

Say what? So it’s actually opposing managers’ fault that Howard can only hit the ball to one side of the field? LOL.

By YanksFan on Sep 18, 2010

swings both ways. pujols gets the platoon adv in 26% of his PA, while howard gets it in 66%. if howard was facing same handed pitching 74% of the time, i bet his numbers would take a little bit of a hit.

By Scott G on Sep 18, 2010

A little bit of a hit? haha he has a .768 career OPS vs LHP. That’s a big hit. He’s essentially Ben Francisco.

Micah,

I’d love to know how you calculated the raise in his AVG, OBP, and SLG.

By micah on Sep 18, 2010

@jim

I didn’t “blame” opposiing managers for overshifting howard, just saying that it is a tactic, only useable against lefties, that costs howard probably a few dozen hits a season.

By micah on Sep 18, 2010

@ scott g

I didn’t calculate his numbers per se, just compared his numbers from 2006 (when he generally wasn’t overshifted) to his numbers since then. There’s other factors I know, but howard has certainly been hurt by the shift, and pujols hasn’t.

Also I’m not arguing that howard is better, just saying that this is a major part of why pujols is better.

By Spree75 on Sep 18, 2010

I didn’t read the last 30 or so posts so this may have been brought up… This is what I’ve picked up from both sides of the argument:

1) WAR is an overall indicator of “value” to a team which takes into account the availability of talent at a position AND the “extra” talent required to play various positions.

2) WAR is not predictive and only adds up what has been done so far.

3) Due to the variability of defensive metrics there is inherent variability in the WAR values. The best I’ve been able to determine if two players are within 0.5-0.6 WAR they can be assumed to be of equal value.

Hopefully I’ve got the above three points correct. My feeling that when people compare “value” of different players that they fall into three different categories of what they consider “value”.

1) Pure offensive “value”. This is the easiest to see from traditional stats and from watching a baseball game. “Howard hits a tonne of Home Runs, drives in runs, etc. so of course he’s more valuable than Reynolds”

2) Overall “value”. I think this category is where a lot of people that are complaining about WAR fall. Most are savvy enough to know that defense is important to the game. However when comparing two players this type of person (and I fall in here) doesn’t care how many players are available at a position or how well an average player at that position plays.

3) Overall “value” to a hypothetical replacement level team. This is what WAR is trying to describe. To me it’s useful argument for looking at contracts or for building a team or for looking at surprising/disappointing players.

By eh on Sep 18, 2010

once again, at least 10 guys, maybe 15 could do what howard does with that line up.

By Jim on Sep 19, 2010

@micah,

Hmm… I don’t see opposing managers employing that ridiculous shift against other lefties. Did I miss something?

And by the way, lefties, when they aren’t Ryan Howard, also enjoy the advantage of the first baseman playing close to the bag when there is a man on first, if you really want to talk about these “positional” advantages. Trust me, Ryan Howard has only himself to blame for that shift; it’s not a product of his left-handedness.

If Pujols hits like Howard except from the other side of the plate, trust me, managers would be moving the shortstop and second baseman way over as well. They might not be able to move the first baseman, but you can bet they will move the other three infielders. Yet they don’t… Please, stop comparing Howard to Pujols, just stop. You’re unnecessarily making a really good player in Howard look bad. Howard might be lucky if he has the numbers to make it into the HoF when he’s done playing. Pujols? He’s ALREADY there.