Wednesday, October 27, 2010

Ace in the Hole

I've touched briefly on the subject of a contract extension for Francisco Liriano a couple times, both in a guest post on Rob Neyer's SweetSpot blog at ESPN.com back in early August and in my offseason blueprint which appears in the GM Handbook.

Today, I'd like to delve a little deeper into the matter, as I feel that it is arguably the team's top priority this winter.

One of the main talking points among fans frustrated by the Twins' early playoff exit is the need for a legitimate ace to head the rotation. Fans watched the Twins' starters helplessly bow to the Yankees' potent hitters during the ALDS while Cliff Lee has shredded every lineup he's faced in helping carry the Rangers to a World Series berth, so that reaction is understandable.

However, many people seem to have a vague and mystical conception of what an "ace" is. It's not Liriano, they say, because while he was one of the league's most dominant pitchers this year (fifth in the American League in strikeouts, sixth in ground ball rate, best home run rate in baseball), he wasn't consistent enough, didn't pitch deep enough into games, and -- most importantly -- didn't come through in the postseason.

The problem is that you can count on one hand the number of pitchers who would satisfy that definition of an ace. There's only one Cliff Lee, folks. And while you can add Roy Halladay, Tim Lincecum and perhaps a few others to the list of consistently dominant starting pitchers who have outstanding postseason track records, these guys don't grow on trees. They are extremely tough to come by.

Take a look at a couple top Cy Young contenders in the AL and how they performed in the playoffs this year. CC Sabathia was not at all sharp, posting a 5.63 ERA and 1.81 WHIP over three starts. David Price went 0-2 with a 4.97 ERA for the Rays. Liriano's first five innings against the Yankees in Game 1 of the ALDS were probably better than any stretch of pitching by either of those two this October.

Things came unraveled for Liriano in the sixth inning against a good offense, but that happens often to pitchers in the playoffs, even the ones that can legitimately be labeled aces. And while Liriano's inconsistency and inefficiency on the mound are marks against him, let's not forget that this was his first full, strong season since Tommy John surgery and he's still only 27 years old (younger than any pitcher I've mentioned in this article, save for Price and Lincecum). It's hardly fair to assume we've seen the best of him.

Aces are exceedingly difficult to come across and Liriano gives the Twins their best shot at having one over the next several years, so it's imperative that they keep him under team control. He's entering his second year of arbitration this offseason, so the Twins have the option of either continuing to go year-to-year with him or offering a multi-year deal.

In the Handbook, I suggest that the Twins sign Liriano to a three-year, $21 million deal. They pay him $4 million in 2011 (a bit less than he'd probably get through arbitration), $7 million in 2012 and buy out his first season of free agency in 2013 for $10 million. For a pitcher capable of the Liriano's dominance, that's a bargain, and considering his history of arm troubles I have to think that the southpaw would jump at the financial security. The downside, for him, is that if he starts to flourish his big payday comes one year later (when he'll still only be 31 -- younger than Lee is now) and the upside is that he's guaranteed $21 million no matter what happens to his arm.

It seems clear that Liriano wore down at the end of the year. Halfway through September, he was 14-7 with a 3.28 ERA, positioning himself as a sneaky Cy Young contender, but over his final four starts (including the playoffs) he went 0-3 with a 7.58 ERA. Given that Liriano racked up such a large workload between winter ball, spring training, the regular season and the playoffs, that shouldn't come as a huge surprise. But it doesn't really affect his long-term outlook, and in fact, the experience of logging all those innings may ultimately be beneficial for his arm strength. Now that he can finally spend an offseason resting rather than rehabbing or pitching in winter ball, I suspect he'll come out better than ever next season.

If that happens, he'll be a lot more expensive a year from now. Bill Smith would be wise to show foresight and buy low on the left-hander. The reward far outweighs the risk.

81 comments:

Anonymous said...

can we sign curt schilling?

Anonymous said...

Id try for more than 3 years. 4 yrs 40 mil perhaps. Because he ended sluggishly, didnt have an elite era (which doesnt really matter), and has had serious injuries in the past, i think the twins could buy out a couple of lirianos FA years for well under market value if they are willing to take on some risk.

Ben said...

Dont you think he is going to be hesitant due to his contracts with both Span and ... Blackburn last year.

Nick N. said...

Id try for more than 3 years. 4 yrs 40 mil perhaps.

I wouldn't necessarily be opposed (they gave Santana a very similar contract at the same point) but given Liriano's fragility that might be a little more risk than I'd be comfortable taking on.

Dont you think he is going to be hesitant due to his contracts with both Span and ... Blackburn last year.

The contracts for Span and Blackburn were ill-advised from the start because neither was even in arbitration yet. Liriano is free agent eligible in two years and the team would also be buying low on him whereas they bought high on both Span and Blackburn (at least relative to how they performed this year).

Matt said...

Gotta do it, take the risk.
Think of it like the NFL. There aren't many good starting QBs out there, so teams have to sign the David Garrards, Jason Campbells, and Matt Hasselbecks and hope for the best.
Liriano is as close to an ace the Twins can afford right now; in a couple years when some bad contracts come off the books, maybe they can take a run at another one.

Anonymous said...

The span and blackburn deals are the result of the problematic way the twins have been doing business recently. They keep buying arb years from unspectacular players they already have team control of. When you buy up a players arb years, or pre arb years + arb years, the twins are forfeiting the ability to cut ties with the player to gain cost certainty. That cost certainty eliminates the effects of both overperforming and underperforming their contracts. I think the twins absorb a ton of risk when they sign players to these deals since the player is under team control either way. The player could get hurt, become replaceable, or simply not play well and the twins will have to pay them either way. My problem is that since the twins take on a lot of risk when they sign these deals they should be a fair amount back in the way of potential probable saving at the backend of the contract when the player is outperforming his deal. A guy like blackburn never had any chance of out performing his deal in a significant way even if everything went great because hes terrible.

I think liriano is a very different situation that blackburn because i think liriano could cost the twins a lot more if they try to sign him after this year or wait till he become a free agent.

Ed Bast said...

Liriano would be a great #2. For the right price, sure, sign him. But don't use it as an excuse not to pusue an "ace" via trade that will help in the playoffs.

I don't know why you insist on calling him an ace, either. You seem to think his 5 good innings in the ALDS overshadow his meltdown in the 6th, which apparently is unavoidable with good offenses (how Colby Lewis and CJ Wilson and Tommy Hunter and Cliff Lee managed to do this very thing, then, is an absolute miracle from heaven, I suppose?).

I completely disagree; his outing showed precisely why he is not an ace. When the going gets tough he folds - like every other Twin, I suppose - and he throws so many pitches in a game that often he can't get out of the 6th. Sorry, but playoff games are still regulation 9 inning affairs - 5 stellar innings doesn't mean much.

None of this matters, though. Due to some terrible contracts the Twins won't be able to afford an extension with him anyway. Nor are they really interested in the playoffs as an organization, so they won't go get an ace either. I suppose we're stuck with Frankie as our "ace".

Ryan said...

Ed,

What ace do you think we could pursue that is better than Liriano? And don't say Lee, we have no shot of signing him.

Nick N. said...

So, Ed, in your view, who's the ace with whom the Twins would have been better off in the playoffs? As I mentioned, Sabathia and Price (who you'd consider aces, I assume) both performed poorly in the playoffs. People talk about wanting to trade for Greinke, who is older and more expensive than Liriano and had a significantly worse season this year.

As Ryan asked, who exactly are you suggesting the Twins target as an ace that will be far better than Liriano in the areas you mention? Keep in mind that the likes of Lee, Halladay and Lincecum are not realistic.

Brad said...

Ed makes me LOL when he suggests that a pitcher with one of the best strikeout rates, an absolutely FILTHY slider, and extremely low home run rate can not be considered an ace on a team's staff.

I bet he thinks any closer who doesn't go 1-2-3 every outing is a poor closer, too.

Ed Bast said...

The ace the Twins would have been better off with in the playoffs is Cliff Lee. I said it when his name came up back in June. I said the Twins pitching staff couldn't beat the Yankees. The Twins, and Nick, didn't want to give up our 4th outfielder in 2014 for him. How did that work out for everyone?

We aren't going to be able to sign a FA ace, I understand that. No reason we can't make a trade, either in the offseason or midseason. Greinke is a real good option. And who knows who will be available come next trading deadline.

You're right, aces don't grow on trees. That's why when you have the opportunity, you've got to go for it once in a while.

Brad, Liriano is the "staff ace" if every team must be assigned 1 ace. He's a good pitcher. He's just not in the same league as Lee, CC, Price, Lester, et al. If you have to choose one of those 5 guys to start 1 game for you, are you guys honestly picking Liriano? If you are you're insane.

Also, I LOL when I read between the lines of some of these comments. Basically, folks are suggesting that the Twins roll with the same rotation next year, since we don't need "another" ace. A rotation, by the way, which has a cumulative 0 playoff wins. Yeah that's the solution: throw out the same guys and just cross your fingers that they'll miraculously figure it out.

Nick N. said...

Greinke is coming off a very pedestrian season, quit baseball at one point due to mental issues and has never pitched a particularly meaningful game in his career. I would love to know under what criterion you classify him an "ace" over Liriano.

Ryan said...

Ed,

You keep making comments about Lee but we would have had to give up a lot more than our "4th outfielder in 2014". We really couldn't match what Texas and NY were talking about offering for him. Also with Lee we were talking about renting him for half a season, not signing him to a couple year deal, which again we have no shot at.

I'd love if we strengthened our rotation. My point was that there just aren't any guys available that are better than Liriano.

Anonymous said...

An ace doesn't fold under pressure. How many times did Liriano take a great performance into the sixth inning, give up a run or two, and then follow it with a complete implosion if Gardy didn't pull him? The answer is more times than not this season. The general consensus is that Liriano folds faster than a lawn chair in a tornado. Aces get themselves out of an inning with minimal damage more times than not. If Liriano gives up a two run shot to the middle of an order, the 6,7,8,9 hitters proceed to abuse him when shutting those guys down is an ace's bread and butter. He gets frustrated and tries to pitch perfect. His numbers say he deserves the label, but whats between his ears does not.

Ed Bast said...

Nick, Roy Halladay never pitched in a particularly meaningful game prior to this year. How did he do in his first playoff start?

According to folks like you, Liriano followed up a far-worse-than-pedestrian '09 with what you seem to consider a Hall of Fame type '10, so there's precedence for that. As I recall Greinke was pretty good in '09, though, would you agree?

If you read my post, I never said to get rid of Liriano. Are you suggesting that a #1/#2 of Liriano and...Baker? is better than Greinke and Liriano? I guess I don't understand why anyone would be opposed to improving the rotation.

cy1time said...

Don’t you have to judge performance against a contract over the life of the contract? I’m not going to try to defend the Blackburn deal, but I expect that the Span deal is going to look good when this contract is up and we’re trying to figure out how to keep him in Minnesota. I was frustrated at times, watching him get picked off, but he’s a nice center fielder and lead off hitter. I expect the numbers will be back next year and for the next several years.

Anonymous said...

People perception that liriano folds on a regular basis is a ways from reality. No twins pitcher left with the lead more and had the bullpen lose the lead. In a likely offshoot of that statistic, no twins pitcher allowed 0-2 runs and failed to get the win than liriano did. He had a 3.62 era and only gave up 77 runs in 192 innings. If you exclude the 7 he gave up to the tigers in 1.2 innings his final era would have been 3.31, and that includes 3 not so good starts to end the season after the division had been clinched. He simply didnt give up that many runs. So how many times did he put the twins in a big hole early or implode late? Pretty rarely. With guys like liriano and baker, the sayings "familiarity breeds contempt", and "the grass is always greener on the other side" come to mind with a lot of twins fans.

Ed Bast said...

There was a 5 game stretch in June/July where he put the club in 1st inning holes of at least 3 runs in 3 starts. He imploded late in his biggest start of the year. Facts breed contempt, too.

Since you somehow think it's appropriate to remove numbers in evaluating him, I'll mention that after April, his ERA was 4.09.

Anonymous said...

There's certainly more upside in spans deal than in blackburns because span is a better player. The problem with spans deal is it is paying span what his rookie contract would have paid him if hed produced roughly what he did his first 2 years. It would be a good deal if span were to improve, but there wasnt a ton of reason to believe hed improve a lot over how hed played his first 2 years. They could have waited until his arbitration started to see if there was a shrewd deal to be made.

Anonymous said...

Ed, every pitcher in baseball had bad starts. Roy halladay pitch poorly against the giants etc, etc, etc. I removed the tigers game not to say the game didnt happen, but rather to illustrate that liriano didnt give up many runs to you people that like era, certainly not enough to consistently be imploding. If ERA was completely done away with and never used again there wouldnt be many people more excited about that than me. If lirianos era was 10 and he had the same peripherals he had this year id still think he was easily a top 10 pitcher in baseball.

Jay Daniel said...

I don't know the answer to this, but it seems plausible to me that the Twins blow leads after Liriano leaves the game because he is unable to go more than 6 innings most of the time. It's a lot more likely that the bullpen is going to blow a lead if it has to pitch 3 innings than if it has to pitch only 1 (or none).

A guy with less good stuff than Liriano (which is almost all pitchers) but who can pitch 230-250 innings a year with steam left over for the postseason seems to me to be more valuable as an "ace."

The question of whether that is in fact true seems to be the main field of debate here.

Dave said...

Brandon Webb people. The man. The ace. The answer.

J. Fetzer said...

"However, many people seem to have a vague and mystical conception of what an "ace" is."

I about shit myself reading this. Nick, that's for straighting it all out.

Anonymous said...

Cliff lee was given an early lead in the biggest game played this season, gave up the lead quickly, allowed a huge inning, unraveled to the point he couldnt complete his inning and didnt pitch deep into the game. Pretty cut and dry that hes not an ace.

Tim lincecum was clearly over come by the pressure and put his team into an early hole. Then when his team gave him a lead, he could put a single 0 up and failed to pitch deep into the game.

We've seen it too much from these two. Not "true aces". The search continues.

Dr. Truth said...

Anon, You're a complete idiot. You're the type of complete idiot who probably said after Game 1 of the ALCS, See, the Yanks ARE unbeatable. I wondered which idiot would make the first "Cliff Lee isn't the answer" comment today. Congrats on being a total douche.

Anonymous said...

Oh Dr truth, i tried to make it clear that my anti lee sentiment was a sarcastic poke at the stupid criteria laid out by a few dumb people on this comment board. Lee did just about everything that somehow excludes liriano from being an ace last night, and in a big game too. I still think lee is an elite pitcher just like i think liriano is an elite pitcher. My point was that if you cherry pick a tiny sample of any pitchers games and evaluated them against unreasonable criteria required by some of an ace, no pitcher in the league is going to measure up.

chris said...

Anon is right. You could see lee unravel in the 4th. An ace would have bared down and gotten out of that inning. And really if lee was an ace he would never have scored upon because thats how it works. You allow some baserunners, then that mental toughness sets in and you dont give up hits. Makes a lot of sense to me.

Ed Bast said...

Anon, except that my criticisms of Liriano are based on his entire breadth of work. Yes, his ALDS performance left something to be desired, but I saw that coming based on what I've seen out of the guy the last 2 years. It was one of the reasons I wanted him to start Game 2 instead of Game 1.

Lee has proved himself to be a postseason ace. Liriano hasn't. Please tell me what Liriano has accomplished post-surgery. Two awful seasons - did you see him pitch in 09? every time he got in to trouble he absolutely unraveled - and a very good year in 10, in which he pitched terribly down the stretch, including a poor postseason start.

Being an "ace" in my book - and we're getting way too bogged down on definitions - but being in ace doesn't mean having a dominant start every now and then and racking up a bunch of strikeouts. You need to be consistent, and you need to step it up in Sept./Oct. He simply hasn't done this. It's not debatable. Now, if your definition of an ace differs from mine - high K totals, a "nasty" pitch, good advanced stats, whatever - fine. Call him an ace.

All I'm saying is, based on what we've seen from him the last two years suggests we can't count on him in Sept/Oct. I want to see this team make a run at the WS. I think we need to bring in a top-of-the-rotation guy to help us win in the playoffs.

And I'm tired of the "small-sample-size" excuse applied to the Twins. Look: your sample size isn't going to get much bigger if you keep playing crappy and losing playoff games!

Anonymous said...

Judging lirano based on what he did the 2 years after he came back from injury isnt particularly fair since hes a very different pitcher this year than he was last. And if youre looking for a track record of late season games how is grienke at all superior to liriano. Never pitched a meaningful late season game and hes had one great season surrounded by lesser seasons.

Matt said...

A top 10 pitcher with an ERA of 10? You've just gotta be kidding here, right?

And, to those who use insults (like calling people dumb and/or stupid) and cheap shots as your way of debating others, knock it off.

Ed Bast said...

Read my f-ing posts! I want Greinke AND Liriano on this team! Or Liriano and another top-of-the-rotation guy! Read, THEN comment!

Anonymous said...

Yeah people who call other people dumb are stupid and i just assume everything they say is idiotic

Ed Bast said...

Matt, I missed that one. But that comment - "even if his ERA was 10 I'd think he was a top 10 guy with those peripherals" - seems to sum up the blogosphere's love affair with Liriano. He's a textbook example of a guy whose stats do not tell the whole story. Statheads pound home how elite his advanced numbers are; they latch on to him as a guy who doesn't get enough respect, and this confirms what sabermetrics is the superior way to evaluate baseball players, and everyone else just doesn't get it.

What all the stat-love doesn't seem to take into account is his actual performance. To say he struggles with the mental aspects of the game on occasion isn't even debatable; yet the stat guys yammer on about small sample sizes, etc., because "mental toughness" isn't quantifiable. To say he doesn't go deep into ballgames is a proveable, simple fact; yet stat guys don't find this important because well just look at his xFIP and WAR, man!

Let me say again: you cannot discount what actually happened. Liriano actually imploded in his ALDS start. He actually pitched poorly in stretches this year. He actually was very dominant in other stretches too. It's no slam on him to say that he's a very good pitcher, perhaps not quite in the upper echelon of starters in this league quite yet, and boy would it be great if we got another, perhaps better pitcher we can plug in at the top of the rotation.

For all my admitted bluster, my opinion simple enough: Liriano alone isn't enough to advance in the playoffs. Look at the history of this staff in the playoffs. Look at our potential opponents next year. I don't think it's a crazy opinion.

Anonymous said...

"sabermetrics is the superior way to evaluate baseball players, and everyone else just doesn't get it.
" Finally youre starting to get it."

"To say he doesn't go deep into ballgames is a proveable, simple fact" A simple fact but not necessarily true. Of the 44 qualified starting pitchers in the AL the average ip per start is 6.26, liriano was at 6.17 ip/s. And thats a very tight distribution. 8 pitchers averaged less than 6 innings, 2 averaged more than 7, only 9 averaged more than 6 and 2/3 including the 2 that averaged more than 7.

Now if you take the 5 pitchers from every AL team that pitched the most innings the average per start drops to 5.92 which liriano is solidly above. If you take every pitcher that made a start this season the average per start drops to 5.49. Liriano is certainly not elite as far pitching deep into games but hes not bad at it either. For a guy on a strict pitch count that strikes out a lot of batters shorter start should be expected. I dont think its a weakness of lirianos, not a strength either.

Nick N. said...

Ed, people aren't going after you because you said the team should replace Liriano with Greinke, they're going after you because you indicated that somehow Greinke qualifies as an ace while Liriano doesn't. Even under the critera you've laid out, that take makes no sense at this point in time.

I'm about done with this subject because the word "ace" has been so overused and mangled in this thread that I'm starting to wish it doesn't exist. But let me clarify my point...

I think it's fair to say that Liriano was not a true ace this year. He didn't hold up late in the season, and his inability to last deeper into games was a liability at times.

However, this was Liriano's first time throwing more than 136 innings in a major-league season (he exceeded that number by almost 60 innings) and he was only 26. The key takeaway from the article was that "it's hardly fair to assume we've seen the best of him."

Liriano now has a full strong season behind him to build confidence, he's pitched down the stretch in a pennant race, he's pitched in the playoffs. He might not be the "ace" that some of you would like yet, but there's a damn good chance he'll reach that level over the next three years. That is why locking him up is imperative, and that was why this post was written.

Ed Bast said...

Nick, agreed, much of the consternation here, I think, is centered around the word "ace". I'm as much to blame as anyone, except that I've made it clear that my definition of an ace may not be the same as the next guy's.

But, okay. Let's ban the word ace. It means nothing.

My point: The Twins need another frontline guy.

My opinion: Greinke is a better overall pitcher than Liriano.

Lost in the shuffle: Yes, by all means, resign Liriano if it's the right type of deal.

Really lost in the shuffle: Yes, Liriano had a very good year, and there's a chance he's going to yet develop into a truly elite pitcher.

Anonymous said...

the giants have 2 aces, the phillies have 3, the rangers have 1, the yankees have 1, the tigers have 1, blah, blah, blah. they aren't that rare. we don't necessarily need an ace, we just need someone that can pitch well in the postseason against the yankees. someone that doesn't back down gardy-style and has good stuff. liriano would have won game 1, but he let the yankees get comfortable in the box. they just got a feel for his stuff in the first 4 or 5 innings and then were able to jump all over it after that. if he had backed a couple off the plate or even hit someone as retaliation for the thome beaning, i think things would have been different. but we have to play nice and allow the yanks to walk all over us and the same shit happens every year. we need a pitcher to show the yankees that we are not afraid and even if we lose, we are going to play with fire and go after them. the yankees never felt threatened in that series. it was a sweep before it started.

Anonymous said...

You consider matt cain an ace? And this assessment is based on his long track record and not just he last 2 starts? Matt cain is a good pitcher but liriano is a lot better.

Anonymous said...

have you looked at cain's numbers over the past 4 years? he has been great and consistent. he has now proven that pitching in the postseason has no impact on him. he hasn't had major elbow surgery. i never said liriano wasn't an ace, i said that he can't back down in the playoffs. it's not about his stuff or his stats, more about his guts.

Anonymous said...

I say a few things im certain youll quibble with. Matt cain is quite a bit worse than his era. He has 3 post season starts, liriano has 1. Neither has proven anything about their guts in the post season.

Ed Bast said...

Nope see again, and I don't know how many times you have to keep telling stat heads this, but Matt Cain is not worse than his ERA is. ERA measures performance. He's been exactly as good as his ERA. Just because his xFIP is higher than his ERA doesn't mean Cain's not that good. It means the xFIP stat isn't that good at prediciting his performance.

To reiterate: xFIP is based on things that did not happen. ERA is based on precisely what did happen.

Anonymous said...

I said youd quibble with it. Im very confident in my analysis of era and how it doesnt really measure pitcher performance well. If you choose to disagree thats fine. You're wrong and hopefully someday you think about it some. ERA was arbitrarily chosen as the benchmark of pitcher performance, that doesnt make it reality. And i know you think your reasoning is sound but earned run average doesnt even precisely detail what happened in the irrelevant macro outcome way you believe it does. A straight run average would be the real world stat you are looking for.

I fully understand why people think era is a good stat, i thought the same thing at one time. But its not. Im not missing anything about the value of era, there are much better stats.

Ed Bast said...

ERA measures how many earned runs a pitcher gave up, right? xFIP measures how many runs a pitcher would have given up if his defenders were robots, the games were played in perfect weather, there was no such thing as a pressure situation, and hitters were computers.

Please explain to me how xFIP better measures how many earned runs a pitcher actually gave up (i.e performance).

I get xFIP. It's a hypothetical number. ERA is not hypothetical. I tend to go with the real over the abstract. Unfortunately for you, there are simply some things that cannot be accounted for in statistics.

Anonymous said...

Ed you are so wrong about xfip. Not even close. It intrinsically say nothing about how a pitcher would have pitched, it correlates 3 numbers that actually happened. And it actually assumes nothing about defense, weather and actually takes into account hitting environment. All these things are big advantages over era because these things are things either pitchers dont control or destandardize the statistic. You dislike statistics like xfip because you dont understand them and seem to have a lot of misconceptions about them. There is nothing mystical about xfip, i think youre hung up because its called expected, it is made up of real numbers. You prefer era because its habit and because you havent thought enough about what goes into pitching.

Anonymous said...

Expected Fielding Independent Pitching. This is an experimental stat that adjusts FIP and "normalizes" the home run component. Research has shown that home runs allowed are pretty much a function of flyballs allowed and home park, so xFIP is based on the average number of home runs allowed per outfield fly. Theoretically, this should be a better predicter of a pitcher's future ERA.

experimental stat. theoretically predicts future ERA.

xFIP = ((13*(.106*# of fly balls))+(3*BB+HBP-IBB)-(2*K))/IP+constant

takes the pitchers number of fly balls and multiplies it by the league average or homers/fly ball, then by 13. nevermind the fact that some pitchers are better or worse at keeping the ball in the park. Xfip is for everyone that believes in luck.

Ed Bast said...

Dude, you have literally no idea what actually goes into xFIP, do you?

You realize how hard it is to take seriously a hardcore stathead who doesn't know what the stat he's vehemently promoting actually means, don't you?

Look at the last anon's post. It's not hard to find that information if you want to.

I have no problem with xFIP and the like as predictive statistics - it would be great if more front offices used these sort of starts to evaluate players. But it's when folks like you throw out all other stats and start to use it to measure performance, clearly not knowing what the statistic even means, that I get wary.

Piece of advice: next time you tell someone they "don't understand" something, do the 30 seconds of research to make sure you understand it yourself.

Anonymous said...

haha do you really think ive never looked at the the xfip formula? And just to be clear ive never advocated xfip, i believe you threw xfip out as the stat head go to stat. When i evaluate a pitcher i look at k, bb, gb and typically shy away from hr rate mainly because i dont completely trust the park adjustments.

As for anons blurb about xfip, its a fine description of the statistic, but the part about it being a better predictor of future era is simply and editorial on the best use of xfip, not a statement about whether the statistic is based on a present reality. xfip is derived from a k rate, bb rate and hr rate, all those things actually happened. Every bit as much as an earned run was allowed. If i were to say theoretically era predicts future era that would be very true. If ERA wasnt meant to be used predictively theyd just give pitchers a run average since thats what really happened. Instead theyve removed runs (arbitrarily) deemed not the pitchers fault. That way a person could look at the stat and say "if the defense does not make mistakes we can expect the pitcher to allow this many runs". Most stats, and all worthwhile stats, are meant to be predictive.

Ed Bast said...

"xfip is derived from a k rate, bb rate and hr rate, all those things actually happened."

Well, not hr rate, because the formula assumes every pitcher should give up home runs at the exact same fly ball rate, so it basically eliminates hr rate. This is ridiculous, since it assumes crappy pitchers like Scott Baker give up homers at the same rate as good pitchers like Josh Johnson.

So, of the 3 things you mention, one of them isn't included in the formula whatsoever, and the others are modified by seemingly random coefficients (3,2). Why do you have to multiply by these numbers, or 13, or some apparently changing "constant"? How do these random numbers reflect what "actually happened"?

Anonymous said...

from now on there will be no fences in baseball. one out of every 10 fly balls will be considered a homerun. strikeouts are the only way to get outs. scott baker will be the best pitcher in the world.

Anonymous said...

xfip assumes that hr/fb is a constant and instead focuses on flyballs, the flyball rate is still a real number. Your problem is with the assumption, not that somehow the number is fictitious. And for every 1 questionable assumption xfip makes era makes 10. ERA assumes that all runs are created equally, all unearned runs are created equally, etc. which are far far far more questionable than the relative importance of fb% and fb/hr.

As for the coefficients they are formed from regressions around situation independent scoring data to weight the relative importance of each factor and then multiply the number so it is on the same scale as ERA. The numbers are ultimately just correlated constants and the dynamic inputs of the statistic are real measurable rate statistics. If you want to quibble with the weighting thats fine, and fair, one of the reasons xfip isnt one of my gotos. My assertion is that ground balls, strikeouts and walks are better indicator of how a pitcher pitched, and much more important, than simply did or didnt a pitcher allow an "earned" run.

Anonymous said...

As for all the senseless scott baker bashing, he didnt grade out well this year according to xfip and his gb/fb wasnt particularly high. He gave up as many homeruns as he did because he gave up so many flyballs.

Anonymous said...

BS! baker's xfip was better than cain's. and he gave up as many homeruns as he did because he threw meatballs. it's not senseless bashing. baker did not have a good year.

Anonymous said...

Difference in opinion i guess. Just like pavano was better than his 5 era in 09 and liriano was better than his 6 era in 09, baker was better than his era in 2010. And his xfip is lower than cains because hes strikes out more than cain, walks fewer than cain, and they have similar gb and hr/fb. The only thing cain does better is get batter to hit the ball at his guys significantly more often.

Anonymous said...

so you are saying that cain has been luckier than baker over the whole course of their careers?

Ed Bast said...

The only thing Cain does better than Baker is pitch.

When the advanced stats aren't able to magically explain every single thing that happens over the course of a season in a manner that proves/disproves your opinion, do the old sabermetric fallback: blame everything on luck.

Anonymous said...

i have had a very long string of bad luck that has caused me to not be able to make a major league team or even a minor league team for that matter. it's just bad luck. can't be explained by anything else.

Anonymous said...

I just did a quick search of this comment section. Luck has been brought up in 5 different comments, never by me. That being said, matt cain has without a doubt been luckier than baker during their careers. Now certainly matt cain does something that results in fewer runs that a statistic like fip wont account for, things like weak contact generation and such, but theres too large of gap for there to not be a fair bit of fortune.

And i dont really care that so many people dont believe in the predictive capabilities of advanced statistics, they certainly dont work all the time. But ive seen them work pretty effectively in the past, and most importantly the pitcher skill inputs make sense. Even if you dont believe groundballs, strikeouts and walks matter at all in effective pitching, putting stock in the capacity a bad stat like ERA, and trying to explain small sample size outcomes with nonsenses like pitcher mental toughness, and heart, etc, youre not going to do any better. Your just grasping at straws. Youre always going to be a prisoner of what just happened.

And to the anon whos minor league career was derailed by luck. My guess is you arent able to strike out professional players, probably have bad control, and give up a lot of fly balls. Both era and xfip would reflect that you are a bad pitcher. It seems like you think advanced statistics assume that all pitchers are the same and that the only difference between them is luck. Thats not the case.

Anonymous said...

that was the whole point, i was making fun of people that say that one pitcher was more lucky over a whole career instead of admitting that he is a better pitcher.

Ed Bast said...

"It seems like you think advanced statistics assume that all pitchers are the same and that the only difference between them is luck."

But this is precisely what xFIP does! To a T! It says, look, home runs should be hit on 10% of fly balls, period. Pitchers who give up HRs at a higher rate are simply unlucky. So, let's look at what would their ERA be if they gave up HRs at the same rate as the average pitcher.

xFIP is fundamentally based on the notion that all pitchers should give up HRs on 10% of fly balls, and anything more or less is completely luck.

Talk about grasping at straws.

Anonymous said...

i am with ed on this one. why not normalize the strikeout rate too, because inducing a weak fly ball vs a hard hit fly ball can have as much to do with a pitcher's performance as a strikeout. umpires can affect strikeouts and walks also. the strikezone is bigger for some pitchers, just like some pitchers have to pitch in smaller ballparks. let's just normalize everything and come up with a number and anything above or below that can be attributed to luck.

xfip may be a fun stat for some of you and may be a decent predictor of future era, but it doesn't show that a player pitched better or worse than his era. there is more to pitching than strikeouts, walks and giving up homeruns at a 10% rate.

Anonymous said...

Youre both very wrong about xfip. Ed you are particularly wrong this time. The formula for xfip considers the effect of flyballs on run scoring. Assumed in that, and supported by a lot of data, is that home run rates are consistent enough to be considered constant for the league. One big advantage to this is that it does not penalize or reward players for their pitching environments, thus standardizing the stat some and allowing for a more comparable stat team to team. xfip also assumes constant ld rate, constant 1b per fb, constant 2b per fb, constant 3b per fb, and all of it situation independent. All that information is wrapped up in the coefficient that is correlated, and consistent with a lot of data for the entire league (once again to not penalize pitchers for bad pitching environments). This is how the statistic attempts to remove things a pitcher doesnt not control. It cares about stuff a pitcher has a good deal of control over, strike outs, walks, and bated ball type. It doesnt care if the fly ball was caught at the warning track or just cleared the fence for a hr, just that a fly ball was hit, and flyballs arent good because a certain percentage of flyballs are going to fall in, and are more likely to be xbh. FIP uses the same type of weighting only it uses a less robust hr/fb ratio in place of a batted ball statistic.

To say that xfip assume all pitchers are the same is absurd and patently false. If pitcher A gets more k's, walks fewer and gets more groundballs than pitcher B, xfip will not tell you that these two pitcher have the same skill level, pitcher A is just more lucky that pitcher B. It will tell you pitcher A is better. The stuff you believe they are irresponsibly holding constant is just stuff the pitcher has little control over, choosing to look at the result of the pitch (k, bb, fb, gb, ld) and not the result of the play (1b, 2b, 3b, hr, out etc). You cant actually believe that pitcher have a huge amount of control on where a ball is hit. If they did why wouldnt they always have the ball hit to the short stop or some equivalent.

And to the anon who wants to standardize k rate, id be all for a long study to see if theres a difference in K rate between leagues and divisions. But for the most part, striking guys out is a skill. Its true that things like strike zone can have an effect on a strike out rate within a game, with a big enough sample size this noise is pretty minimal.

Ed Bast said...

I'm done with this mind-numbing conversation. It's like telling someone the sky is blue, and having them insist without a trace of logic that no, the sky is actually green and you're an idiot who just doesn't get it.

Anonymous said...

And to the anon who wants to standardize k rate,

it was a joke to show how stupid standardizing anything is.

If pitcher A gets more k's, walks fewer and gets more groundballs than pitcher B, xfip will not tell you that these two pitcher have the same skill level, pitcher A is just more lucky that pitcher B. It will tell you pitcher A is better.

fine, you take scott baker, i'll take matt cain. my team is already better than yours. who do you want next?

Anonymous said...

xfip assumes that all strikeouts are equal. a strikeout of the #9 hitter with no one on base isn't the same as striking out the #3 hitter with the bases loaded. sometimes pitchers are content with getting groundballs and only go after after the strikeout when they need it.

like i said, xfip is a good tool when you don't use it as the only stat.

Anonymous said...

Man Ed, lets keep it civil. I dont think youve ever in your life supported a claim with relevant numbers. Your arguments go like this, "era is a true measure of what happened. Stat head tell me thats not so. Well it is, now im going to attack advanced statistics without a strong understanding of the concept." I get that youve looked up xfip, you understand the formula, and youve read the stock explanation. But youve never thought about it, you dont get it. Your consistently wrong about everything baseball. I thought everyone had moved past the value of RBI and i find you are tired of people ripping on it. You misguidedly and inaccurately compare rbi totals to war. Its unbelievable. Its hard to believe someone could be wrong so often and have complete belief hes got the situation totally figured out.

Anonymous said...

"xfip assumes that all strikeouts are equal. a strikeout of the #9 hitter with no one on base isn't the same as striking out the #3 hitter with the bases loaded. sometimes pitchers are content with getting groundballs and only go after after the strikeout when they need it.

like i said, xfip is a good tool when you don't use it as the only stat."

Hopefully this was intended to be a sample size argument. This is the sort of thing sample size will level out.

Anonymous said...

"Hopefully this was intended to be a sample size argument. This is the sort of thing sample size will level out."

not necessarily. some pitchers try to pitch to contact and some of them are good at it. they get penalized because they don't go for the strikeout all the time. sample size will not level it out. pitchers that strike out more batters can look better in xfip even if they are getting crushed.

Anonymous said...

Balls in play are always worse than strikeouts. Even if you are trying to do it, at some point cheap hits will fall in and letting bat hitters put balls in play will hurt the pitcher. A statistic like xfip, fip, sierra does, and should, penalize a pitcher for this because it will lead to more runs than all strikeouts would. Certainly this could lead to longer outings, and that has value. When ever i hear dick and bert talk about someone needing to trust their defense and pitch to contact more I take it as slang for the pitcher is having problems with walks. If you can not pitch to contact and still not walk a lot of guys, its without a doubt a better option than letting guys put balls in play simple to keep a pitch count down, to an extent anyways.

Anonymous said...

what about a strikeout in the dirt where the wild pitch gets by the catcher and the batter gets first base. that is better than getting double play ground out or a foul out? i realize that this doesn't happen all that often, but you can't say that a ball in play is always worse. almost always it will be, but the goal of the pitcher is to get outs and prevent runs. you can take scott baker because he strikes out an extra half a batter per 9 innings and i'll take matt cain. cain has never had a hr/fb rate of over 8.4%, therefore he gets punished in xfip every year because the league average is around 10.

obviously xfip is a better predictor of future ERA for some pitchers, but at some point, if a guy outperforms xfip every single year, i think you just have to say that xfip isn't a good way to evaluate this certain pitcher and predict his future ERA. how does that sound?

Ed Bast said...

To recap:

I'm stating the fact that xFIP is theoretical. There isn't any gray area. It's theoretical. This is a fact.

You are stating an opinion that Scott Baker is better than Matt Cain.

I don't need to say any more.

Dr. Truth said...

Double plays are better than strikeouts.

Anonymous said...

To recap:

I'm stating the fact that era is no more reality based than xfip. There isn't any gray area. It's a made up, arbitrary, ambiguous construct. This is a fact.

And just to be clear, I never said matt cain was a worse pitcher than scott baker, ever. One of you losers said that because baker had a lower xfip, a statistic i have made clear i dont think is the best we can do for pitcher evaluation. Now i dont think theres probably much of a difference between baker and cain from a pitcher skill standpoint. Cains low era is a result of a 260 babip, and unless he generates an unheard of amount of weak contact, or has unbelievable defense, or some combination of the 2, a 260 babip is lucky. If that babip crept up to league average youd see his era balloon.

"cain has never had a hr/fb rate of over 8.4%, therefore he gets punished in xfip every year " What if matt cain has pitched in a pitchers environment that is 2 % more difficult hit a homerun in than an average environment? Then instead of punishing him because he has an incredibly minor ability to keep fly balls in the ball park at a lower than normal rate, he is simple not being reward for pitching in a pitcher friendly environment.

"Double plays are better than strikeouts." The idea isnt to look at an individual result of a batted ball but rather look at the results probabilisticly. A strike out will turn into an out 99% of the time. A ball in play with with a runner on 1b and less than 2 outs is going to be a double play ball roughly 10% of the time, itll turn into a hit 30% of the time, itll turn into a hr 5% of the time, etc. So with a guy on first and less than 2 outs i would take a k over a ball in play because over a large sample size the ball in play is going to lead to a lot more runs.

Ed Bast said...

"It's a made up, arbitrary, ambiguous construct. This is a fact."

ERA is a measure how many earned runs a pitcher actually gave up. What is arbitrary or ambiguous about it?

"And just to be clear, I never said matt cain was a worse pitcher than scott baker, ever."

Here's you: "The only thing cain does better is get batter to hit the ball at his guys significantly more often."

Anonymous said...

"Here's you: "The only thing cain does better is get batter to hit the ball at his guys significantly more often."" My statement is true. Still not an endorsement of baker over cain.

"ERA is a measure how many earned runs a pitcher actually gave up. What is arbitrary or ambiguous about it?" Are you kidding me. The designation of and earned run is exceptionally ambiguous and at times very arbitrary. Errors arent standardized from ballpark to ballpark. When an error does occur the inning is then recreated by the score keeper as if that error never happened. Pitchers on different teams play in very different ball parks, with very different defenses, in different scoring environments.

Ed Bast said...

So your problem with ERA, then, is that there's a huge conspiracy in the way scorers assign errors, it's so wild and out of control that ERAS are completely out of whack and meaningless?

Again, so your problem with ERA goes back to official scorers.

Anonymous said...

No, my problem with era is i dont think it effectively measures things pitchers can control. Im all for statistics attempting to remove bias', and while era actually tries to do this it does a bad job of it. Some of that has to do with how arbitrary the stat is, more of it has to do with it not measure pitcher skill independent of things like defense, ballpark, etc very well. You're the one who seems to have a problem with statistics that arent explicitly understandable in the present reality. Your main beef with xfip was that it was theoretical and predictive and therefore unsuited to be used to evaluate how a pitcher performed. My point was that ERA is no different. A guy on third and 2 outs, if a SS boots a routine GB, the scorekeeper will say "theoretically if an error had not of been made no run would have scored there. Its not fair that we count this run against his statistics because it theoretically shouldnt have happened." This is essentially the same thing a statistic like xfip tries to do.

Ed Bast said...

Liriano gave up 0 unearned runs this year. Baker only gave up 2 unearned runs, etc. Don't you think your "large sample size" takes care of these rare instances?

My problem is when xFIP is misused. It's potentially useful in projecting future performance, as I've said. Perfectly good statistic. But I don't believe it's at all appropriate to use xFIP to judge past performance. I.e., X had a better year than Y, look at his xFIP. Z should win the Cy Young, look at his xFIP. X was had an ace-like last year. Etc. My problem with this, again, is that xFIP is theoretical - you can't judge a player based on what could have happened, only what did happen, "fair" or not.

While you are correct that different environments, circumstances, etc. affect a pitcher's "basic" numbers, these happen to be part of baseball. You can't remove pitchers from their context because baseball is fundamentally rooted in context. Part of what makes a pitcher great is getting out of jams that may or may not have been their fault. Part of what makes a pitcher great is being able to get guys out any which way - not every out has to be by strikeout (a lesson Liriano would do good by).

One of the reasons I don't hate ERA like so many do is because even with so many variable like you mention, baseball games come down to which team scores more runs than the other. So minimizing runs is really the #1 task of a pitcher - not to rack up strikeouts, ground balls, etc. These things might help minimize runs, but I don't think it's particularly important how pitchers get outs, just that they get them.

Okay, I really mean it this time, I think we've spent enough time debating xFIP and ERA.

Anonymous said...

Then what you are advocating is a total run average not an era.

Nick N. said...

Wow. You guys got epic in here.

Anonymous said...

Im not playing games.

Ed Bast said...

It's not one of my prouder moments.