I don't think the quantity of people picking Felix over Lee (or Sabathia or Price, for that matter) matters. If 80% of people have made a conclusion that they hold certain metrics over others, so be it. If a case can be made for it must be either Lee or Hernandez, you will have a solid split. Lee had lower walks (18! to 70), while Hernandez had much higher strikeouts (232 to 185). Hernandez had a lower ERA, Lee had a lower FIP, but their xFIP was extremely close. Lee had 1 more CG, but Hernandez had 37 more innings of work.

To me, is there a case for both? Of course. But while I agree that if either Felix or Lee win the Cy Young, its a triumph for sabermetrics, I don't think the margin of victory for whoever wins should matter. If its Hernandez, we will see that strikeouts and innings pitched (and probably ERA) were weighted more heavily than walks. What we should be ecstatic about is that win/loss record no longer is relevant.

Saying that the margin of victory is too wide, to me, sounds like when Bill Conlin (and 5 other idiots) didn't vote for Nolan Ryan for the HOF because they didn't want it to be unanimous. Margin shouldn't matter if its a good choice.

Or it could just be that readers are a bit like sheep. We read the good stuff about certain players and they stick with us. But I hope readers are smart enough to read B-R, BP, and FG and still make our own decisions. But you never know.

'be theirs confident with theirs'
is probably supposed to be 'that confident with theirs.'

you dont have to post this, but if given a choice id have gone with Felix.

Baseball-reference does not use ERA as a component in WAR. They use Runs Allowed. That means they do NOT use defense-independent stats. Essentially, if you believe the pitchers have some control over batted balls (and I do), baseball-reference WAR is a better indicator than any other. That was my biggest criterion in voting for the IBAs. I didn't think that any other pitching metrics proved that initial assessment wrong, so that's how I voted.

David: They do adjust for defense, though.

And to point out some problem for this, use Francisco Liriano. He posted a 3.62 ERA (I don't know his RA), which gets penalized because the Twins had an above-average defense. However, he also posted a BABIP of .340. Did he really benefit from the good defense? Is he really just that much worse than the rest of the Twins staff on controlling batted balls (noting that he had a GB% of 53.6)?

Loved this piece, and it summed up my feelings on the issue exactly. I wouldn't have worded the last sentence so strongly, but nonetheless, if you asked 1,000 aliens (really analytical, unbiased aliens) to study the numbers we have available and choose, there's no way they get even close to 80% for Felix, and that's what your driving at.

Like with anything (and everything) in life, people go with the crowd, and much of the talk around the Cy Young was framed "Who gets it: CC or the stathead pick - Felix?"

It was nearly the same behavioral mechanism that lead managers to vote Jeter the Gold Glove... no one is going to blame me for doing it.

In the end, Lee won't even get a sniff of the Cy Young, (unless you count a mostly distant 3rd or 4th sniffing), and that is the real travesty.

This is an excellent food-for-thought piece, and I agree with you in one way and disagree in another way.

First, I don't think it's totally fair to weight defense-independent metrics in a Cy Young race. These metrics are suggestive, not gospel...as you touched upon, we can't be sure how much control a pitcher has in all of this (and I agree with the earlier poster about pitchers having some control over BABIP, and some more than others). FIP says we should investigate how much Felix benefitted from his defense, but it proves nothing about that.

On the other hand...I'm very happy you brought this up in the first place. As the season came to the close, the Cy Young discussion really became "CC vs Felix", and it became a debate about which stats are most important and about how we can't penalize Felix for his low win total because that's affected by the team around him. What people overlooked is, just as CC benefitted in terms of wins due to his team, Felix benefitted in terms of the quality of his pitching. Having baseball's best defensive OFer is a good start, the defense was great overall, and the ballpark helps too. This was overlooked with CC as well...bad-hip ARod, statue Jeter, overrated Cano, Yankee Stadium...surely his quality was better than the end numbers?

I still think it's clear that Felix was head and shoulders better than CC this past season, but I find it strange that the sabermetric analysis of the Cy Young race always seemed to ignore the benefits Felix was given.

As for the margin of victory between Felix and Lee, are you sure it's so unreasonable? I think in the end you have to admit that any major difference in pitching quality between the two comes from personal preference of analysis, and that the only sure conclusion is that they both had great seasons in that regard. So shouldn't the vast innings pitched difference between the two definitively give Felix the win? I think it's a fair result.

Tyler, I agree. If 80% of voters independently arrived at Felix, that's great. I just fear that he's become the accepted sabermetric choice, and people aren't doing their due diligence.

Eric, thanks for the correction.

David, I also appreciate your distinction between ERA and RA. I disagree with your statement that if pitchers have some control over batted balls (which I do believe) baseball-reference WAR is a better indicator than any other. I think StatCorner/tRA do a better job that B-Ref and FG.

Will and Lee, I agree with your points.

Peter and Tyler, I tried to explore whether there was something to Felix and Lee leveraging their BABIPs. That was the substance of my analysis. I didn't find anything there. Seemed to me like a lot of ball in play luck.

Peter, if it comes down to personal preference as you say, then yes I think 80% is too wide a margin.

In the big Lee vs Hernandez sabermetric debate for the Cy Young, what about Justin Verlander? He gets essentially zero mention in this race, but is that right?

(all stats per fangraphs)

Lee - 7.1 WAR, 2.58 FIP, 3.23 xFIP
Verlander - 6.2 WAR, 2.97 FIP, 3.67 xFIP
Hernandez - 6.1 WAR, 3.04 FIP, 3.26 xFIP

Verlander had the highest K/9 rate of all 3 and the lowest HR/9 rate, but also the highest BB/9 rate. He also went 18-9 which is a far better W/L record than either for those that like more traditional stats.

Does Verlander deserve to win it? Maybe not. But his case is right in line with the other top candidates and it's sort of surprising how little mention he gets considering he finished with the 3rd highest WAR of any pitcher in the majors this season (last season he was 2nd to Greinke).

Luck counts.

WAR is a meaningless stat when trying to decide a CY YOUNG winnner. BABIP, FIP, none of that crap counts when awarding on what has already been decided.

Those things all matter, greatly, when predicting FUTURE outcomes, but what is done, is done.

Jeremy - You are right on! Cliff Lee should have won. Maybe next year when he is a "Yankee" he'll get his due.

WAR is not meaningless. It attempts to take out the things outside a players control and quantify what they have actually contributed towards their team's chances of winning games.

The Cy Young isn't an award for the luckiest pitcher of the season. It's for the best. And traditional stats like W/L have a lot more to do with things the picher can't control than things they can control.

Nice name Paul,

Agreed that WAR in its various guises has its uses, but no-one here at least, and less so in the wider baseball community is basing their CY pick wholly on WL records anymore - so you are making your point in a vacuum, either people here agree with your, or they are trolls trying to bait you.

The article is a good one, and plenty of kudos for at least the writer picks the pitcher who come 1st in fWAR (his preferred measure of pitching excellence) by a long way for his CY pick.

I'd pick Felix, but just because I think a pitcher has more influence than solely HR/BB/K, so fWAR for pitcher overestimates the C.Lee types (and also the Nolasco, Baker etc.. types) and penalises the low H/9 types because they cannot distinguish between where it is somewhat of a skill (i think Felix qualifies) and where it is mostly luck (say Cahill) - so its not included.

e.g. Lee in G1 WS
4 2/3 INN; 1BB, 7K = great FIP

however he had a bad night and gave up 5 doubles, FIP = good performance; actual performance = bad.

Obviously the situation is difficult for some in the SABR community as if the measure of choice...say fWAR says C.Lee laps the field, but your pick is Felix, then either your metric is incorrect (or shall we say not perfect), or you have some explaining to do to justify your pick (and then you turn into one of the various writers whose votes gets abused on the internet)

I do hope C.Lee gets some CY love, having 5 guys on the ballot helps, but there were some great perfomances by pitchers this year and anyone of 10ish guys deserve recognition.

I didn't actually say "it comes down to personal preference"...deciding who had the lead in terms of pure pitching quality is what comes down to personal preference. When this quality difference is this close, it's essentially negligeable...the quality difference between the two might as well be a tie.

Because of this, logically, if they are roughly the same quality pitcher, didn't the pitcher who threw MANY more innings have the better year? I say 20% for Lee is reasonable, for this reason: if 20% of people think Lee had clearly better quality (and so all choose him), 20% think Hernandez had clearly better quality (and choose him), the remaining 60% who see no difference in inning quality should ALL choose Hernandez on the basis of the innings difference.

"e.g. Lee in G1 WS
4 2/3 INN; 1BB, 7K = great FIP

however he had a bad night and gave up 5 doubles, FIP = good performance; actual performance = bad."

(1) FIP, just like virtually every other stat isn't meant to be used for obscenely small samples.

(2) Even if that game was representative of a players full season, I think there would be plenty of red flags [ie, like averaging 4.7 IP per start with a .363 BABIP?].

Thanks Eric R, you make my point for me better than I did

The red flags show that you shouldn't expect the FIP to continue, (i.e. it can be useful for predictive purposes going forward taking into account other factors)

But CY preference if based on fWAR which is based on FIP which rewards C.Lee for his WS G 1 performance? (if it was during the regular season)

FIP as a tool for making useful forward predictions - Yes
FIP to reward past performance - Not so much

Simply my humble opinion, but its why I don't buy the Lee over Felix for CY based on fWAR

Why do we think he pitched poorly in that game?

Chances are that if a random pitcher gives up 5 doubles, he was not good. Chances are that if Cliff Lee gives up 5 doubles, he was good, because he's Cliff Lee. What did he do wrong on those five pitches? How did he pitch poorly in that game other than by getting poor results?

General impression I had from the analysis at Fangraphs after the game was that he found a healthy part of the plate with numerous pitches in comparision to his more usual edge of the plate distribution

http://www.fangraphs.com/blogs/index.php/cliff-lee-was-not-living-on-the-edge/

Whether he is just unlucky or not, I'll leave up to brighter people than me to decide, but if you throw plenty over the heart of the plate, then you can't be too upset to give up a few well hit balls.

Lee got a lot less love in the CY voting than he deserved, but I honestly think that FIP overestimates the value of the Lees of the world (guys who lessen BB by giving up more hits), and underestimates the less hittable pitchers like Felix and Timmy

anyway peace out