Well Al I will fix that now that I was treated at the Wafer Wire and I am under Contract to the NY Yankees for nine years and 3.6 Million Smackers. That is Million Al read it and stand Agog. I expect I will toss about a Ten WAR season next year and then join the rotation and get about 20 WAR for the next 8 Years. Wins Above Replacement is a Concept Al it is like If there Was no pitcher and the other guy could just toss the ball up and whack Fungoes, how many games would you win. I am easily worth more than a Fungo Bat Al. If I can make it here I can make it unywhere. Its up to you Noo York New York.

Edit sorry about that bad Spieling Al I just got off the Train at Pencilvania Station and I may have had a nip or two at the Pub by the Subway stop. I looked up after I orderd and hey theres a game at Yankee Stadum if you know what Subway to catch to get there send me a Tex Massage.

It's possible that the Mets could have the leading NL leaders in WAR for both pitchers and position players and end up under .500 for the season. That's weird.

It's possible that the Mets could have the leading NL leaders in WAR for both pitchers and position players and end up under .500 for the season. That's weird.

What's odder is that it almost happened in 2011 with Kemp and Kershaw!

(EDIT: really not meant as snark, I'm assuming it is a rare thing to have two great seasons on a mediocre team. And even rarer to have them back to back)

It's possible that the Mets could have the leading NL leaders in WAR for both pitchers and position players and end up under .500 for the season. That's weird.

Cueto has a huge lead by bWAR. I'm not sure why fWAR is so down on him.

Cueto's advantage over Dickey appears to come primarily from his HR rate, .4/9 against .8/9. That and park effects seems to be the gap in WAR.

Cueto has a huge lead by bWAR. I'm not sure why fWAR is so down on him.

Because fWar is utter crap.

Because fWar is utter crap.

May I direct you to this post by Tango?

I asked MGL to run UZR at the pitcher-level. And he came through, and you can see the data in post 33 of the same thread. And, wouldn’t you know it, we got results consistent with our expectation.

The fielders behind Verlander had a UZR of +8 runs. They played great for him. Verlander has about 500 balls in play, and 8 runs translates to about 11 extra hits-into-outs, or about 22 points in BABIP. Therefore, Verlander’s .270 BABIP, after you adjust for how well his fielders did (as per UZR) would have had a .292 BABIP.

Porcello’s fielders performed at -9 runs, so they played terribly for him. Porcello also has a similar number of balls in play, so this affected his BABIP by about 24 points. His .350 BABIP therefore would plummet down to .325 if he had fielders playing as average fielders.

Scherzer. Poor Scherzer. His fielders performed at -11 runs, and he had only about 400 balls in play. His BABIP is affected by about 37 points because of bad fielding. His .350 BABIP would otherwise plummet down to .313.

So, if you are trying to use a team’s overall fielding as an INDICATOR as to how much fielding support an individual pitcher received: don’t! It’s very well possible that if you see a pitcher with a low BABIP on a bad-fielding team that he could have still gotten GOOD fielding behind him. Just as you wouldn’t presume a good hitting team provides good run support to all its pitchers, or a good bullpen helps out all of its pitchers, then neither should you presume that having a good set of fielders behind you means that ended up receiving good fielding support.

The truth, as always, is somewhere in the middle.

Now, perhaps you hate UZR as well. If that's the case, then I doubt that there's any numbers-based argument that will convince you. But this is a genuinely important piece of evidence, albeit with the usual caveats, that the assumption of fWAR (fielders perform much worse on some days than on others, making pitchers prone to having high BABIPs that aren't the fault of bad pitching OR bad luck) is more accurate than you think.

Love UZR.... But when evaluating a pitchers performance backwards looking, only thing to look at is results... not some theoretical crap.

Measure the hits allowed, strikeouts, walks, clutch performance all you want. The second you start saying "well with a league average defense he would have allowed 5 more hits..or whatever" then you are creating a fictional stat. I do not know why people want to cite a fictional stat. fWar is utter crap for pitchers.

But this is a genuinely important piece of evidence, albeit with the usual caveats, that the assumption of fWAR (fielders perform much worse on some days than on others, making pitchers prone to having high BABIPs that aren't the fault of bad pitching OR bad luck) is more accurate than you think.

How? This shows that fielders achieved better results behind Verlander than behind Porcello or Scherzer. fWAR assumes that difference is "luck" (from the pitcher's perspective) and discounts it entirely, but couldn't that difference be real - couldn't it be that Verlander's pitching was responsible for (at least some of) that difference?

The truth, as always, is somewhere in the middle

Yep, so why give the pitcher credit for all of the "bad" fielding. UZR is nowhere near accurate enough to do with it what Tango does there. Did Porcello's fielders play poorly or did the balls he gave up happen to be in the unfieldable (or rarely fielded) part of their zone or be hit harder than the ones given up by Verlander? UZR doesn't have a clue (nor does DRS or whatever).

Sometimes drilling down deeper and deeper into detail hurts -- forest, trees and all that.

You can think of it in this way:

ERA -- gives the pitcher full blame/credit for the result of BABIP
bWAR -- gives the pitcher a sizeable chunk of blame/credit
fWAR -- gives the pitcher minimal blame/credit (but not quite zero I don't think)

If the truth is "somwewhere in the middle", bWAR is the one closest to the middle.

I understand the fWAR concept and I'm OK with it. If Verlander and Porcello both give up a 6-hop ball up the middle and, for some reason, Verlander's gets fielded but Porcello's doesn't -- well, the pitchers did equally good jobs and there's no reason for the pitcher to take the blame for the random variation in defensive performance. But I don't think UZR or any defensive stat measures the "fieldability" of a BIP with sufficient accuracty to parcel out what's random variation and what's a harder-to-field ball. bWAR may go too far in regressing it all to the team mean but at least you have confidence that the team mean is pretty accurately measured.

I'd be a bit more confident if these pitcher-to-pitcher differences could be tied to consistently different defensive lineups behind them. I mean if you're mid-80s Sid Fernandez and you've got Howard Johnson at SS because you don't give up a lot of GB to SS, I'm not going to blame you if those GB to SS you do give up turn into hits a lot.

Is fWAR still FIP or xFIP based or is it doing some something with UZR similar to what bb-ref does with BIS runs save/total zone?

Is fWAR still FIP or xFIP based or is it doing some something with UZR similar to what bb-ref does with BIS runs save/total zone?

My understanding is that it's based on FIP. The argument, as I understand it, is that FIP is pitcher, UZR is fielder, and any differences between reality and FIP+UZR are "luck" and aren't credited to anybody.

Back to the premise of the article ... 8.2 WAR seasons (position players) by decade 1962-2011

62-71: 18 (21.2 teams per season)
72-81: 18 (25 teams per season)
82-91: 17 (26 teams per season)
92-01: 19 (28.6 teams per season)
02-11: 18 (30 teams per season)

Pretty even spread by decade but that means we've gone from about .9 per team per decade to .6 per team per decade. Still, given they've been around for the whole period, that would mean you'd expect about 6-7 of these seasons to take place in NY. The slackers here are of course the Mets who have generally sucked during this period.

The Red Sox it turns out are outliers on the position player side with 9 such seasons compared to an expectation of about 3.4 -- 3 Yaz, 4 Boggs, 1 Petro, 1 Lynn.

Then there's the Big Red Machine:

72: Morgan and Bench
73: Morgan and Rose
74: Morgan
75: Morgan
76: Morgan (he was pretty good)
77: Foster

You can also add FRob in 62 to the Reds list. Reds pitchers might not have ever won the CYA but they do really well in MVP.

But maybe not so surprisingly, the raw total champs are the Giants with 12:

Mays 5 (remember, only 62 onwards)
W Clark 1
Bonds 6

It seems it helps to have possibly the two greatest players of the last 60 years. The Cards have 5 from Pujols and 1 from Rolen.

But the rate champ is presumably the Mariners 8 in their 35 seasons ... OK, the Giants still win. Anyway, 3 AROD, 3 Griffey, Boone and Ichiro.

Lots of teams without one:

White Sox
Tigers (!!)
Indians (kinda surprising)
Blue Jays
DBacks
Mets

AROD's the only one to do it for 3 teams. Rickey is the only player traded in the middle of such a season.

Given that 29 of the 90 seasons come from just 5 players (Bonds, AROD, Pujols, Morgan, Mays) it's not shocking that there are a number of teams without one or with just one. All it really tells us about the Yanks is that they haven't developed super-duper star position players in the last 50 years (and that Jeter's defense sucks) and they aren't buying them until after they've hit their prime. If the Mets had signed Bonds, they'd have had a lot of these seasons.

In short, the draft and (to a lesser extent) requiring 6 years to become an FA are pretty effective in making sure that superstar talent gets evenly distributed around the league.

That's kinda interesting now that I think of it. I don't know how many we would expect but there aren't many foreign-born seasons on this list:

Pujols 5; Sosa 1; Beltre 1; Ichiro 1; Clemente 1; and Canada's greatest hitter Larry Walker. And Pujols and Walker came through the draft and Carew would have if he was born a couple of years later.

Love UZR.... But when evaluating a pitchers performance backwards looking, only thing to look at is results... not some theoretical crap.

Measure the hits allowed, strikeouts, walks, clutch performance all you want. The second you start saying "well with a league average defense he would have allowed 5 more hits..or whatever" then you are creating a fictional stat. I do not know why people want to cite a fictional stat. fWar is utter crap for pitchers.

While I disagree, I respect this point of view. But I sure hope that in that case, you hate the concept of errors and ERA just as much.

How? This shows that fielders achieved better results behind Verlander than behind Porcello or Scherzer. fWAR assumes that difference is "luck" (from the pitcher's perspective) and discounts it entirely, but couldn't that difference be real - couldn't it be that Verlander's pitching was responsible for (at least some of) that difference?

Yes, it's certainly possible, and perhaps likely, that Verlander's balls in play have been slightly harder to field even beyond what UZR accounts for. But UZR does account for whether the ball was recorded as being hit hard/medium/soft and approximately where it was hit, and yet it still says that Verlander's fielders did a better job with what they were given. I would very much like to see larger samples of data, and obviously on more than just these few pitchers, because then UZR's "coarseness" will be overwhelmed by the sample size and let us be pretty confident in saying whether the fielders did a better or worse job behind different pitchers. Of course, there will always be inherent limitations and biases in the UZR data that will prevent us from making definitive, sweeping statements on the matter, such as the scorer bias in assigning hard/medium/soft. But you would have to be pretty stubborn if you wanted to try to claim that Scherzer and Porcello consistently gave up balls hit to the very most unfieldable edges of zones.

But I sure hope that in that case, you hate the concept of errors and ERA just as much.

I have no problem with errors, have a problem with era, but it's a good enough tool that I live with it. I find it to be utterly silly that a pitcher isn't charged with any runs after the third supposed out failed to happen. I'm sorry but it's his job to get out of that jam, yes there are runs that they can't avoid allowing, but I've seen a pitcher allow 4 unearned runs on a grand slam....if you give up a homerun, you should be charged for the homerun at the minimum and realistically all the runners who you put on through non-errors. (heck even the runners that are put on from an error should be subjected to a basic "would he have scored with normal performance" by normal performance I mean one base advanced per out remaining. Guy reaches second base on an error no outs, if he eventually scores, that is a justified unearned run in my book. Guy reaches first on an error and eventually scores, it should be an earned run---same thing with inherited runners, inherited runners should be charged to the person's era based upon expected one base advanced per out rule. If you come in to relieve me with two outs in the inning and a man on second, promptly allow a hit, then get the third out, that man on second should be charged to your era)


I'm fine with the theoretical that is basically one level removed from the real events. Component era type of stuff I'm fine with. But fip/xfip goes two steps beyond the real events. It doesn't even look at the real events, instead it just assumes what should have happened, if what should have happened didn't happen it assigns the difference to luck and defense. By fip a no hitter with only one strikeout is an impossibility.

Measure the hits allowed, strikeouts, walks, clutch performance all you want. The second you start saying "well with a league average defense he would have allowed 5 more hits..or whatever" then you are creating a fictional stat. I do not know why people want to cite a fictional stat. fWar is utter crap for pitchers.

Right. The only stat that matters is W-L. Everything else is fiction. You start saying that a pitcher who allows 2 runs should win 70% of the time and therefore get a fraction of a win -- that's utter crap.

(Of course, we've been having this argument on this site for ten years, and many people have just stuck with whatever made sense to them then, and dismissed everything else as crap without ever making a convincing argument.)

Right. The only stat that matters is W-L. Everything else is fiction.

You're responding with snark, but CFB's position is actually a totally reasonable one when considered in good faith. We don't really have a very good grip on precisely how much credit or blame to give a pitcher for his components, and things like xFIP are useful as gauges of likely future performance, they also obviously miss big chunks of the pitching population that limit runs in ways that have little to do with their K rates. Basing a value stat like WAR on a moderately useful predictive one like xFIP is not much more useful than including something asinine like WPA, in my estimation.

UZR is also based on subjective judgements, so even though it's a number it's not at all immutable truth.