Wednesday, September 15, 2010

Straightening Out the AL Cy Young Race

Recently, there's been growing debate surrounding this year's American League Cy Young race. The three most popular candidates are David Price, C.C. Sabathia and Felix Hernandez, all of whom are having outstanding years. One could argue that a few other starters deserve to be part of the mix, including our own Francisco Liriano, but for today we'll stick with the trio mentioned above. Let's compare some key numbers for those three hurlers:

Pitcher A
: 217 IP, 3.03 ERA, 1.18 WHIP, 7.4 K/9IP, 2.8 BB/9IP, .653 opp OPS
Pitcher B: 186.2 IP, 2.75 ERA, 1.21 WHIP, 8.1 K/9IP, 3.5 BB/9IP, .640 opp OPS
Pitcher C: 225.2 IP, 2.39 ERA, 1.10 WHIP, 8.5 K/9IP, 2.5 BB/9IP, .599 opp OPS

Undeniably, all three of these pitchers are having terrific seasons, but it should be clear from those numbers who's been best. It's Pitcher C, and it's not even close. He's markedly better than both of his opponents in every single category.

One factor that is not reflected above, however, is the situation surrounding all three players. Pitcher A (Sabathia) pitches for a playoff-bound team with an elite offense, and as a result, his win-loss record sits at 19-6. Pitcher B (Price) works under similar circumstances, and sports a 17-6 record.

Meanwhile, Pitcher C -- which is obviously Hernandez -- plays for a last-place team with a historically terrible offense. This has led him to an 11-11 record despite his spectacular performance on the mound. On 12 occasions this year, Hernandez has pitched seven or more innings, allowed two or fewer earned runs, and come away with a no-decision or a loss. This has happened to Sabathia and Price a combined total of six times.

The Cy Young Award is meant to go to the best pitcher in the league, and there's zero doubt that Hernandez has pitched far better than Sabathia or Price this year. Yet, because of the flawed logic that goes into the voting process, he could well finish third in the voting. That's because the writers who vote on this award have historically weighed W/L as the most important measure of success, despite it being the statistic over which a pitcher has least control.

Sadly, some of the leading voices in the baseball community have been trying to lead fans astray on this matter. Yesterday, Jon Paul Morosi of FoxSports.com posted an article in which he suggests that Hernandez doesn't merit consideration for Cy Young honors due to the fact that he isn't pitching for a playoff-bound team. An excerpt from Morosi's column:
There’s an award for a pitcher such as Hernandez. It’s called the ERA title. Not the Cy Young Award, as voted on by the Baseball Writers’ Association of America.

To be the best, one must do what Sabathia and Price have all season — compete against the best lineups, in postseason-type atmospheres, before crazed crowds at hitter-friendly ballparks.

Of course, as I pointed out above, Hernandez leads his opponents in not only ERA, but essentially every measurement of pitching aptitude other than win/loss record. I'd certainly disagree with Morosi's second assertion; I'd argue that "to be the best" one simply must be better than everyone else. Hernandez has been. And, ironically, Morosi last year supported the case of rightful Cy Young winner Zack Greinke, who of course pitched a ton of playoff-type games in Kauffman Stadium for the last-place Royals.

In a chat on ESPN.com yesterday, Joe Morgan stated that it's "a joke" that there is even debate about this year's Cy Young race, proclaiming that "the name of the game is to win and [Sabathia has] won," all while ignoring the fact that Sabathia pitches for the top scoring offense in the league while Hernandez pitches for a team whose OPS this year is lower than Nick Punto's career mark. Then he made some point about Cliff Lee not winning games since being traded to Texas while ignoring the fact that Lee has been battling a back injury since switching clubs.

The Baseball Writers Association of America, which votes on the Cy Young Award, hit an all-time low for me back in 2005 when they selected Bartolo Colon over Johan Santana in a situation that was eerily similar to this year's Sabathia/Hernandez juxtaposition. Santana rated significantly better than Colon in essentially every metric other than W/L record, and yet Colon coasted to an easy victory based solely on his 21 wins and his team's success.

The BBWAA seemingly showed that they'd finally moved away from their fixation on win/loss record last year when they awarded Greinke with the Cy Young despite his 16-8 record (not to mention Tim Lincecum in the NL, who finished with just 15 wins). This year, they will undo all that progress if they hand it to the undeserving Sabathia while punishing Hernandez for the crappy offense Seattle's front office put together.

I want to care about the Cy Young Award. I really do. It's an important part of the way the game's history will be written and one day it may be the deciding factor in a Hall of Fame case. (Does anyone doubt that Bert Blyleven would have been inducted long ago if he had a few Cy Youngs sitting on his shelf?)

Unfortunately, it's getting harder and harder to care when the voters show so little consideration for the numbers that actually indicate how well someone has pitched. And to see respected national figures like Morosi and Morgan present the kind of woefully misguided arguments linked above is disheartening, to say the least.

45 comments:

Eliot said...

I found it funny that on sabermatric focused sites Sabathia and Price aren't even in the top 3 discussion (Lee and Liriano taking those slots). It just seems insane that wins are still used by so many voters in deciding the Cy Young in this day and age where so many superior stats are readily available. I guess it will take much longer than I thought for "mainstream" baseball to adapt to a more sabermatric line of thinking.

Nick N. said...

To me, this isn't even a matter of sabermetrics. It's common sense.

Anonymous said...

There is a fairly large number of sports writers that still put stock in a barbaric statistic like w-l to judge pitcher skill. But I will argue that giving hernandez the cy young because of era is only marginally better because era, while better than w-l record, is still a bad way to measure pitcher performance. Its heavily influence by things like ballpark, defense, opponent ability, luck, ect ect, and these things make it impossible to standardize which makes it a terrible choice to compare pitchers on different teams even if you believed it was a good measure of pitcher skill, which it isnt. I believe francisco liriano has pitched better than david price this year because he strikes out more, walks fewer and gets more groundballs. I dont care at all what their respective eras are because they represent far more than pitcher skill. I could make the same compelling argument for liriano over hernandez. My point is that you are right that people should look past brutal w-l record stats, but stopping at a bad statistic like era isnt going to lead to significantly more meaningful analysis.

And the awards are already meaningless. Every night dick bremer tell me about this weeks bad defensive first baseman, with 120 rbi and 35 hr, who should be the mvp if his team can make the playoffs. The highest award in baseball is given to a player based on some of the worst stats you could think off ( BA/rbi/hr), without regard to defense, routinely. Dont even start on gold golves.

Nick N. said...

I guess I feel like a pitcher should be rewarded for how he actually performed -- in terms of allowing runs and base runners -- rather than how he should have performed or would have performed under neutral circumstances. Sure, the fact that Francisco Liriano had a .350 BABIP at one point indicated that better results should be expected going forward, but it doesn't change the fact that he allowed those hits and allowed runs to score because of them.

(Incidentally, Liriano's case is looking better and better all the time as BABIP continues to come back down to Earth.)

Anonymous said...

Personally I'd prefer this award goes to a Yankees player every season as long as there is a decent pitcher on their roster. They have the largest fanbase and therefore can extract the most joy from this honor.

For the hoard!

Anonymous said...

I guess my point is that k rate, bb rate and gb rate are how the pitcher "actually pitched". ERA, like w-l record, is very much a team stat that can be disguised as a pitching performance stat. Obviously ERA is much more focused on pitching, but i think its inaccurate to say that hits and runs scored are a great reflection of how a pitcher pitched. Do you really think that if liriano pitched with the rays defense behind him that he wouldnt have given up fewer hits and in turn fewer runs? That has nothing to do with how liriano pitched. Without even taking into account batted ball luck, ball park and stuff like double plays turned, ERA is still not a very good pitching stat.

Jack S. said...

I laughed when I read the quote from Jon Paul "To be the best, one must do what Sabathia and Price have all season — in postseason-type atmospheres, before crazed crowds" after watching the game Monday at Tropicana Field that had less than 27,000 in attendance. If that is the biggest game Price has pitched in, he certainly doesn't deserve the Cy Young award according to Jon Paul.

Anonymous said...

Devil's advocate...

The numbers you cite describe how a pitcher performed. W/L describes the actual results - whether the performance of the pitcher was enough to help earn his team a win, which is ultimately the only thing that matters. The voters decide what the award means.

Steve L. said...

While I agree somewhat that W-L shouldn't matter as much when talking about a Cy Young Award pitcher, I do think that Hernandez's Win total hurts him. He's won just as many as he's lost, and that doesn't sit right with me if we're going to give him an award.

And I agree with the idea of rewarding actual results over would-have/could-have/should-have prognostications.

But Since we're including Liriano in the Cy Young discussions, I think Carl Pavano's name has to be in there too. You could argue that the Twins have 2 of the top 5 contenders in the AL for the award are on our team. Pavano's not as flashy, but has been just as successful.

Anonymous said...

brian duensing should win.

wwcd said...

Funny isn't it that just a few years ago all the Yankee fans were so upset at Morneau winning the MVP because Jeter's stats were better.

Anonymous said...

ERA is heavily influenced by team defense and hitting environment, which are different for every team. These are concrete issues with era that make it a poor statistic to evaluate how a pitcher pitched, and thats on top of the influence things like batted ball luck, dp rate, hr rate ect have on it. I think people are confused about how well era illustrates what actually happened. It seems like people believe era is a good indicator of what happened while not necessarily being a good indicator of the future. It's neither. Use ERA to pick a cy young winner is a better effort than W-L record but is a bad choice for all the same reasons W-L record was, the statistic has a sizable dependence on things besides pitcher skill. Rewarding CC sabathia because he plays for a great offense is bad but rewarding Price or hernandez becaues the they play for teams with great defense is alright? And its not just defense. Hernandez and liriano's era are helped by pitching in extreme pitcher friendly environments.

Nick N. said...

W/L describes the actual results - whether the performance of the pitcher was enough to help earn his team a win

But this essentially eliminates pitchers playing for bad teams from being able to contend for the award. Hernandez's low win total has nothing to do with the way he's pitched, why should he be punished for having teammates who can't hit?

He's won just as many as he's lost, and that doesn't sit right with me if we're going to give him an award.

Why? The .500 record is a reflection of his team's performance, not his. Cy Young is an individual award.

Scott Weber said...

All Felix can do is limit baserunners, limit runs. The result of a win or a loss flat out isn't up to him. It's up to his offense or his bullpen to hold. That's why you can't just look at his record and make a conclusion - we should be judging pitchers on factors they can control, not things they cannot. Great article, Nick!

Nick N. said...

Rewarding CC sabathia because he plays for a great offense is bad but rewarding Price or hernandez becaues the they play for teams with great defense is alright?

You'd have a point with these lengthy ERA diatribes if anyone was actually trying to award Hernandez solely for his ERA, but I don't see anyone doing that. I pointed out in the post that he leads Price and Sabathia in just about every category other than W/L, and those include neutralizing metrics like FIP, xFIP and ERA+.

Anonymous said...

If felix hernandez had a 3.5 era instead of a 2.4 era youd overlook him for sure just like you did liriano in the opening of your post. Hernandez tops sabathia and price in the advanced metrics category but gets crushed by lee and liriano.

Nick N. said...

I respect your thoroughly sabermetric approach to this discussion, but I'm just not going to agree. If we're talking about predicting future performance or judging a pitcher's true skill level, then I will defer more to the advanced metrics.

To me, though, the Cy Young is about awarding the pitcher who had the best season. I don't think W/L record is a remotely good indicator, but the other ones I listed in the initial comparison are.

Agamemnon said...

As ludicrous as this probably sounds, perhaps we need two titles, one for the "win champion" and one for the "best pitcher" taking other (more important) things into consideration. It'd be a bit more like the offensive side of things were we get a batting champion based on stats, and an MVP based more on value.

I'm sure this idea will probably get shot down immediately however by the many experts on this blog who probably know a lot more than me.

Anonymous said...

Take for instance the kanerko misplayed triple on the weakly hit pop up to right center. Most teams in the league record an out there, most of the rest play that ball into a single. Lirianos result was a triple. He could make an identical pitch with identical ball in play and get a much better result if he were pitching for a better team. That would lower his whip, baa, obpa, slga, and in long run era. Everyone of the stats you listed as valid indicators are influenced strongly by defense and hitting environment. If liriano has that ball turned into a triple and david price has that exact same ball turned into an out you cant believe that david price pitched better.

Pa Torg said...

Re: Anonymous who keeps talking about the ERA title being the King's consolation prize, the opposite could be suggested: the Win title is Sabathia's prize then. Let's get back to actually talking about pitcher performance.

The thing that gets under my skin is the discussion about level of competition because it's thrown in there off-hand with a lot of presumption without really looking at the actual games. If you look at their actual starts you get a much different picture, also keeping in mind Hernandez is pitching FOR a terrible team, but he has actually happened to match up against a lot of the best teams in the league multiple times. Sabathia has obviously pitched against good teams too, but he NEVER pitches against one of the very best teams in their very scary ballpark. Joe Posnanski discussed this thoroughly and at length last week, but to cherry pick, King Felix and Sabathia have matched up three times this year. That would be King Felix pitching against .268/.350/.438, Sabathia against .234/.298/.338. In chronological order:

Game 1 in NY:
KF: 9 IP, 0 ER, 2 H, 11 SO, 3 BB
S: 8 IP, 2R (1 ER), 5 H, 4 SO, 2 BB

Game 2 in SEATTLE:
KF: 9 IP, 1 ER, 10 H, 9 SO, 2 BB
S: 7 IP, 1 ER, 6 H, 1 SO, 1 BB

Game 3 in NY:
KF: 8 IP, 0 ER, 4 H, 11 SO, 3 BB
S: 6 IP, 0 ER, 3 H, 8 SO, 0 BB

Totals:
KF: 25 IP, 1 ER, 16 H, 31 SO, 8 BB
S: 21 IP, 2 ER, 14 H, 13 SO, 3 BB

That would translate to: a fantasy first round playoff series between the Yankees and the Mariners, the teams get to start their very best pitchers every game ... and the Yankees get swept by King Felix , 1 inning from the bullpen, and 9 Nick Puntos who have to come up with a total of five runs. Sabathia is an awesome pitcher, don't get me wrong. But if you want to talk about stepping up in "playoff-type atmospheres" against "the highest level of competition" ... Yeah ... for this small sample size (but from what I've seen, it would look pretty much the same if you lined them up vs every above-average offensive team they've each faced), it's a no-brainer.

Ed Bast said...

The outfield defense argument doesn't hold water. If Liriano's numbers are hurt by bad outfield defense, than surely Felix Hernandez's are helped by superb outfield defense (their OF is pretty good, right?). Should we have to de-rate his numbers?

Or how about this: since CC had such good run support, and played with a lead a lot, maybe he "let up a bit", knowing he could afford to make a mistake or 2, and that if he'd played for a worse team with lower margin for error his numbers would be better?

You could go on and on and on with all the hypotheticals, but in the end, you need to go with what actually happened in the actual games played by actual teams.

All this attempting to "even out the variables" is futile. While I agree that wins don't tell the whole story, you shouldn't need a background in calculus to determine the Cy Young either.

Anonymous said...

The example about the triple was just to illustrate defense effecting different result oriented metrics. The example itself is not important.

"but in the end, you need to go with what actually happened in the actual games played by actual teams." W-L record does exactly this. But moving from w-l record to era, your just going from one stat that doesnt represent how the pitcher pitched to another. You dont need an advanced math degree or crazy sabermetrics to understand how worthless era is. Its much more of a team statistic than any one seems to give it credit for. Id guess that team era would probably correlate pretty closely with advanced team defense metrics.

Nick N. said...

Take for instance the kanerko misplayed triple on the weakly hit pop up to right center. Most teams in the league record an out there, most of the rest play that ball into a single. Lirianos result was a triple. He could make an identical pitch with identical ball in play and get a much better result if he were pitching for a better team.

Yes, but the fact remains that Liriano gave up a deep line drive and it would have been a tough play. Span and Kubel did a poor job of getting to that ball, but it happens. It happens to every pitcher. It's part of the game. Liriano (and every other pitcher) also has gotten plenty of help from defenders making good plays on balls that reasonably could have turned into hits.

I agree that ERA has it's flaws, for the reasons you mentioned, but over the course of a full season things tend to even out and I think for the most part it accurately reflects a pitcher's performance.

Kevin said...

http://www.fangraphs.com/blogs/index.php/liriano-is-lost-in-the-cy-young-world/

This criticizes Morosi's garbage article, using his own criteria Liriano ranks ahead of both Price and Sabathia. Here are their numbers since the all star break, you know, the stretch that matters most and closely resembles a postseason atmosphere. Note that three of Liriano's starts came against the White Sox, their closest rival and playoff contender.

Liriano 10 GS, 7-0 W-L, 2.22 ERA, 2.95 SO/BB
Sabathia 12 GS, 7-3 W-L, 2.93 ERA, 2.78 SO/BB
Price 11 GS, 5-2 W-L, 3.28 ERA, 2.23 SO/BB

Anonymous said...

"but over the course of a full season things tend to even out and I think for the most part it accurately reflects a pitcher's performance." This is just wrong. The pitcher with the better defense is obviously going to have a larger percentage of plays made for him, and thats going to impact his era. And if pitchers with great defenses have an era edge thats not related to their pitching ability it makes comparing pitchers based on era much less meaningful.

Nick N. said...

And if pitchers with great defenses have an era edge thats not related to their pitching ability it makes comparing pitchers based on era much less meaningful.

You're treading in murky waters, my friend. Defensive metrics are just not reliable enough for us to evaluate fielding as definitely as we can hitting and pitching, so it can only factor so much into consideration. I'm not prepared to say Liriano's inferior defense makes up for Hernandez's significant edge in other categories (most notably innings pitched; Felix has 50 more innings and five complete games to Liriano's zero).

Anonymous said...

I would never try to uncouple the pitching skill and other things portions of ERA because im certain it cant be done accurately. But the fact that i know era has a fairly significant dependence on things other than pitcher skill is enough for me to know its not a very good pitcher evaluation tool.

Im very confident that liriano has been the better pitcher this year because his pitcher control peripherals are much better than hernandez'. And even with hernandez significant innings lead, liriano still has a higher WAR. Im not sure why people seem to have put a moratorium on using the best pitcher evaluation statistics (k, bb, gb)? Why do strikeouts, walks and groundballs not count as how a pitcher "actually pitched"? Instead the consensus seems to be that arbitrarily choosing ERA as THE indicator of how a pitcher pitched is just fine. Why not choose ERA+, or wRC against? Both of those would be better than ERA, but still not the best you could do.

Anonymous said...

Ed Bast said it best, you need to go with "what actually happened in actual games against actual teams".

This is after all, "The Cy Young Award", you know..., in honor of the guy with 511 WINS. Example? Steve Carlton managed to go 27-10, getting about 45% of the total number of wins for a horrible '72 Phillies team. He found a way to actually win games against actual teams, probably with one of the worst fielding/hitting set of team metrics in MLB history. I know he had a low whip and ERA and a high amount of innings pitched (346) and strikeouts (310), all "barbaric stats", but pretty incredible, nevertheless. Zach Greinke, to some extent in this new era, had a Carlton-type season last year, and was recognized for it.

Many of the stats used for this article are valuable in a black box sort of way to assess pitching effectiveness. I suggest following Agamemnon's idea and create another additional award, rather than graft new criteria onto a very old-fashioned award with it's own very deep-seated particularities which surround it.

Anonymous said...

I meant to support ed bast's comments further, by pointing out that what is expected, and what CC Sabbathia must deliver, statistically for the Yankees, is far different than what he had to deliver for the Brewers and Indians. The situational reality of, for whom, and where, one pitches can really muddle, and ultimately diminish, the value of these black box statistical analyses.

Ed Bast said...

To be clear I'm not advocating W-L as the ultimate determinant of the award. I'm just wary of the claims that if X played with this defense his numbers would be better, or if Y pitched in a different part his numbers would be worse, etc. Stats like xFIP, WAR and what have you are hypothetical. It's fun to speculate how one pitcher would do if all variables were equal, but it's just speculation. We can never know what Y's ERA would be playing for a different team, so we shouldn't pretend to know for the sake of handing out an award.

Anonymous said...

Take cliff lee as an example example. Playing with a very good outfield defense in a very pitcher friendly environment he posted a 2.3 era with the mariners. Without his peripherals changing much his era has ballooned to 4.3. Some of that increase is of course sample size and a shift in batted ball luck, but an increase in his era was almost a certainty after the trade. Its a lot easier to hit homeruns in texas and their fielders get to fewer balls. How is cliff lee for the mariners the unquestioned best pitcher in baseball and cliff lee for the rangers a failure? A pitchers ability to generate an ERA is very dependent on the team he plays for.

And just to be clear, there is very little actually hypothetical about xFIP. Its a moderately regressed statistic that correlates strikeout rate, walk rate and flyball rate. Not one of those things hypothetical. xFIP is a better predictor of future performance than era because it does a better job of assessing how a pitcher pitched and uses that as a guide to future performance.

Ed Bast said...

How a "predictor of future performance" is a) anything but the very definition of hypothetical, and b) relevant to an award given based on past performance is beyond me.

Nick N. said...

Take cliff lee as an example example. Playing with a very good outfield defense in a very pitcher friendly environment he posted a 2.3 era with the mariners. Without his peripherals changing much his era has ballooned to 4.3. Some of that increase is of course sample size and a shift in batted ball luck, but an increase in his era was almost a certainty after the trade. Its a lot easier to hit homeruns in texas and their fielders get to fewer balls. How is cliff lee for the mariners the unquestioned best pitcher in baseball and cliff lee for the rangers a failure? A pitchers ability to generate an ERA is very dependent on the team he plays for.

OK, now you're starting to lose credibility with me. Lee has been HURT since that trade. His back acted up and he's had starts skipped because of it. This is fact, not some hypothetical conjecture that you're putting forth. Making judgments based on his sample with the Rangers is beyond irresponsible.

C.J. Wilson and Colby Lewis have managed to put together very strong seasons while playing in that park and with that defense all season. I don't think anyone could sanely argue that either is a better pitcher than (a healthy) Lee.

Ed Bast said...

Anon is making the assumption that Cliff Lee is a robot, that the only thing that affects his pitching outcomes are external factors such as ballpark, defense, weather, etc. This is one of the fallacies of hard-core sabermetrics: expecting players to behave according to the laws of statistics rather than as human beings. It's possible Cliff Lee simply isn't pitching as well as he did last year.

Anonymous said...

Ed, a stat like xFIP which is a good indicator of future performance is not hypothetical because indicators and predictors are not hypotheses, which is the definition of being hypothetical. Most stats can be and routinely are used to make hypotheses about the future, the numbers themselves are not the hypothesis, they are of course derived from real data. For instance, say i make the prediction that joe mauer is going to hit above .300 next season because he hit about.300 this season and the season before. My guess at next years batting average is hypothetical, the batting averages from the past 2 years i used to make the prediction are not. xFIP is a composite that correlates k, bb and fb. It by itself represents a measure of things that have actually happened. It doesnt become hypothetical until you use it to start predicting future performance or it you were to say compare it to era in order to make a qualitative assessment of whether a pitcher was better or worse than his era. There is nothing intrinsically hypothetical about it. As for the second part of your question, i think an "advanced statistic" that corelates bb, k, gb, adj hr rate and a few others in some meaningful way is certainly a better method of determining how a pitcher 'pitched' than era.

As for my lee example, i admittedly did almost 0 research on that. I didnt even look at IP. I personally dont believe any number of innings validates era but if there was some number it would be way over 1 seasons worth so the example would be meaningless even if he had been healthy. Still shotty work on my part. I was just trying to give an easily understandable example of pitcher era not translating from team to team, to illustrate the point i didnt think people were understanding. I see that was a total failure as your take home seems to be that i dont understand sample size, i dont think anyone can post a low era in texas and that i think cj wilson is better than cliff lee, none of which were not intended. If ive lost credibility with you nick, i suppose that alright, im not really sure why i had credibility in the first place. I tried my best to present a my point of view. I think for the most part my claims have been disregarded as overzealous saber nonsense (despite advocating the use of simpler statistics than era), or have been refuted be people that dont really understand the nature of advanced pitching metrics or era and refuse to think critically about either. I do appear to be in the small minority so its possible im way off base and dont understand what im talking about.

Anonymous said...

"This is one of the fallacies of hard-core sabermetrics: expecting players to behave according to the laws of statistics rather than as human beings." Im pretty sure this is largely untrue. What are hardcore sabermetrics anyhow? Sabermetrics with penetration? I am not a sabermatrician as i have no affiliation to the organization, and ive never really read specifically what they are about. But most intelligent people are capable of understanding that no statistic is going to be even close to 100% predictive. What i try to do is break down whatever facet of the game im interested in so that i can try to better understand why something happened and be able to make more educated guesses of future results. This typically involves a little more research that ERA, BA, W-l, clubhouse leardership, but really not much. Instead of being satisfied with the twins made the playoffs in 09 because of orlando cabreras club house leadership, I try to determine the actual factors that lead to their playoff run.

Anonymous said...

Here's the Bill James model which weighs heavily on IP, K, W minus L, Shutouts, Saves, playing on a championship team.

http://espn.go.com/mlb/features/cyyoung

Here's the AL list. Cahill, Lester and Bucholz surprised me, but after looking further at it, I don't feel that bad about Pavano and Liriano at 8 and 9. Like seeing how you can put a couple relievers in the mix quantitatively. No Hernandez there, doubt there ever will be of this type of year, he has to breakthrough on some of these other statistics, no matter how impressive his saber stats are.


1 CC Sabathia NYY
2 David Price TB
3 Rafael Soriano TB
4 Jon Lester BOS
5 Clay Buchholz BOS
6 Trevor Cahill OAK
7 Neftali Feliz TEX
8 Carl Pavano MIN
9 Francisco Liriano MIN
10 C.J. Wilson

Ed Bast said...

Okay. xFIP normalizes a pitchers home run rate to a league average home run rate somebody determined. So if hypothetically a pitcher were to give up x fly balls, he'd give up y homers instead of the acutal homers he actually gave up.

It's the definition of a hypothetical stat. Again, use it as a predictor of future ERA, fine. It should have nothing to do with evaluating a pitcher's Cy Young candidacy because it is fundamentally based on things that did not happen.

Anonymous said...

You could use fip or even better SIERRA and avoid that nasty hypothetical adjusted flyball metric (sierra is probably adjusted too but i dont feel like checking). Without a doubt era+ would be a better tool to evaluate a pitcher for the cy young than era, and it too is adjusted to league and ball park. The essence of era+ and xfip is still real data. I suppose youve maintained that as little bias as possible should be removed from a statistic to keep it as worthless as possible.

Ed Bast said...

You seem to be a little thickheaded, but let me try to explain one more time. Advanced metrics are very useful in theoretically comparing talent, value, or skill of certain players. More straightforward statistics are useful in measuring performance. The Cy Young is a performance-based award. I don't think this is very complicated.

Think of it in terms of video games. Pitcher A is a 99 overall. This rating is analagous to sabermetrics - all things equal, he rates as the best pitcher in the league. You play a season with Pitcher A, and he does okay, but Pitcher B has a better ERA, IP, Ks, WHIP, etc. across the board, even though he's only a 92 overall.

According to your convoluted logic, you're going to give Pitcher A the Cy Young because has the better overall rating.

Now do you understand how silly that sounds?

Anonymous said...

Ive always understood what you are saying. What i have been saying the entire time, and what was my original complaint, is that i believe the statement "More straightforward statistics are useful in measuring performance" to be almost entirely false, and im yet to get a compelling argument that im wrong. ERA isnt an effective way of measure how a pitcher performed, it measures how a pitcher performed, and how the defense played, and what type of ball park he plays, etc. Its a team statistic, just like w-l record. Its better than w-l because its focus is the subject, but its dependent enough on none pitcher factors that it in my opinion does not to a good job representing what a 'pitcher' actually did.

I also never intended to lobby for "advanced metrics" to decide the cy young. Id like to see people use stats like k rate, walk rate, gb ld fb rates what the evaluate pitchers because i believe those are the factors that pitchers have the most independent control over. Instead of looking at the season as a whole with w-l record and taking a finer look with a stat like era, evaluate the pitcher based on the results of the pitch not the results of the play.

Kinda funny that you can think you understand my argument just fine that im a moron that doesnt get what youre saying and I can think the same thing about you.

Ed Bast said...

"i believe the statement "More straightforward statistics are useful in measuring performance" to be almost entirely false"

You don't think earned run average measures the average amount of earned runs a pitcher gives up per 9 innings? You don't think strikeouts measures strikeouts? What does WHIP measure? How about Innings Pitched? I don't know how to help you there.

I'll rephrase: more straighforward stats are useful in measuring WHAT ACTUALLY HAPPENED.

If ERA is a team statistic, every pitcher on a given team would have a similar ERA, right? Or is that, once you look at every roster in the league, actually "entirely false"?

Anonymous said...

Haha, as i was writing my last comment i guessed you take issue with me saying that basic statics didnt represent what actually happened. ERA does of course measure average runs allowed per 9 innings. Its is most definitely not standardized independent measure of what a pitcher has done. So in the case of awarding the cy young to the pitcher who actually pitches the best, i think era does not represent how a pitcher actually pitched very well.

As for your nonsense complaint about by my terrible logic suggesting every pitcher on a team should have the same era, this obviously not what i said. And indeed entirely false. ERA is still largely a function of pitcher skill, which as i said is why its better than w-l, but there is still a significant dependence on fielding, park, luck etc. Its impossible to determine exactly how large the non pitching element is but it exists and makes comparing pitchers with era very inexact.

Ed Bast said...

"Its a team statistic, just like w-l record."

You're contradicting yourself nicely here.

Have a nice weekend, go Twins!

Anonymous said...

Youre a moron Ed. There is nothing contradictory there. I call era a team statistic because it has a significant dependence on team defense, even though the focus is pitching. You could be an idiot but i suspect youre just trying to make me upset. Ive laid out my argument a 1000 times, ive logically question your claim a ton. Youve never supported your argument with anything but baseless opinion. All youve got is wild paraphrasing and quoting my stuff, not to refute some piece of logic in my argument, but to argue meaningless semantics, like just how hypothetical a statistic is, as if that really matters, or whether ive contradicted myself, which i havent. You speak in idiotic absolute terms: sabermatricians believe this, anon fails to realize that. You dont know what youre talking about. There is no point debating stupid comment board trolls, you just end up debating your own terribly paraphrased argument.