Evaluating the BBWAA Cy Young voting through the years

Awards season has come and gone, and as usual, some of the winners were rather surprising. Some good surprises, and some bad.

Let’s start with Mike Trout, the winner of the AL MVP. Before the award was announced, I thought it was a done deal that Mookie Betts was going to win. He played on a playoff contender, had an incredible season, and was a new name in the debate. He seemed like the ideal candidate to once again take the crown away from Trout. But, Trout won, and it was a pleasant surprised, because while Betts had an incredible season and is a joy to watch, Mike Trout was once again the best player in the American League, and again, it wasn’t all that close.

Now, to Rick Porcello, who won the Cy Young award for the American League. This caused a bit of an outrage on twitter.

Ken Rosenthal also wrote a piece at Fox Sports suggesting that the BBWAA needed to be better, but that, in general, the BBWAA was doing better in awards voting than in years past:

I was therefore curious to see just how well or poorly the BBWAA had done in Cy Young voting over the years. In the sabermetric community, we often have a bleak view of awards. Ben Lindbergh and Sam Miller, on the Effectively Wild Podcast, even suggested that the MVP vote had been devalued. Since the award wasn’t going to the best player, then the award didn’t hold a lot of meaning to them.

I think as a community we could give the BBWAA more of a chance. We should examine if they are getting better, because we might someday reach a stage where awards such as the Cy Young or the MVP might be awarded to the best players.

So, with that in mind, I decided to quantify this query of mine by looking at the PWARP of the Cy Young winning pitcher, and comparing that to that season’s leader in PWARP. Sometimes, they were the same (or pretty close), meaning that the voters basically got it “right”.

The reason I am using PWARP from Baseball Prospectus is because I believe it is currently the best metric that covers overall performance for a pitcher. That said, it is not perfect. Therefore, if there is a 0.5 PWARP difference between the Cy Young winner and the PWARP leader for a specific season, I would not describe it as a miss, or as the BBWAA being “wrong”. When there’s that small of a difference between both pitchers, there are usually reasonable arguments to be made on either side. Now, when we are getting in the 2 wins or more difference territory, then I am more comfortable suggesting that the voters got it wrong.

I went all the way back to 1967 because before then the Cy Young award was awarded to the best pitcher in both leagues. After 1967, the rule changed and there were two Cy Young awards, respectively given to the best pitcher in the National League and the American League.

You’ll also notice that two pitchers won the award in 1969, in the American League. That’s because the two pitchers got the same number of votes. After the 1969 season, the voting was adjusted so that this wouldn’t happen again.

Let’s start with the Junior Circuit.

Since 1967, there have been 51 Cy Young awards given out. The leader in PWARP and the Cy Young winner have only been the same pitcher 17 times.

The biggest miss during that time happened in 1977 when the BBWAA voted for Sparky Lyle. Lyle that season finished with 1.23 PWARP, a good 2.17 ERA, and a pedestrian 4.10 DRA. Lyle’s biggest flaw, though, and the reason his PWARP was so low, is because he only pitched 137 innings. Lyle was the closer or relief ace for his team, and while that has value, it usually isn’t as valuable as a good starting pitcher.

Nolan Ryan, on the other hand, had an incredible season, finishing with an 8.3 PWARP. He struck out 341 batters that season, which is unheard of in today’s game, and pitched 299 innings. Put that together, and you have the biggest miss in American League Cy Young voting at 7.07 PWARP.

The second-worst was in 1982, which I’ve highlighted in the graph. Pete Vuckovich had a 5.01 DRA, he didn’t have the best ERA, and he didn’t have the most wins, although he did finish with 18. The only plausible explanation could be the unprecedented success of his team — the Brewers won the pennant in 1982 for the first time in their history. This honestly was a perplexing vote. I’m not sure why he was chosen, and it was obviously a pretty egregious decision considering that he was barely a replacement level pitcher.

But, those examples were from an era when sabermetrics hadn’t yet hit the mainstream. In recent years, since 2008, the biggest difference between the leader in PWARP and the PWARP of the Cy Young pitcher was 2.06...which was the difference between Chris Sale’s 2016 PWARP and Porcello’s PWARP. And as mentioned earlier, I think a number of writers would recognize that that was a bad vote.

Now let’s look at the National League.

The voters have strangely enough been better at voting for the National League Cy Young than the American League Cy Young. Out of 50 awards given out, the Cy Young winner and PWARP leader have been the same pitcher 23 times. And, the average difference in PWARP in the National League is at 1.4 compared to 1.8 in the American League.

The biggest and most encouraging sign in this graph is that since 2005, the voters have been getting much much better at voting. In fact, I would argue that the only real miss since 2005 has been the Jake Arrieta Cy Young win over Clayton Kershaw last year, where Kershaw was 2.94 PWARP better than Arrieta — a pretty big margin.

Arrieta’s win total — he had 22 victories to Kershaw’s 16 — might have helped his cause, but he probably won because of his 1.77 ERA compared to Kershaw’s 2.13 ERA. The problem is that ERA isn’t the best stat for evaluating overall performance. It simply is too noisy, and when one pitcher has the benefit of the Cubs defense, well, that needs to be accounted for somehow. DRA does that, ERA doesn’t.

There are also a myriad of other things DRA does better than ERA, but if you want to know why DRA preferred Kershaw last year, look at its NIP runs, which had Kershaw at -33.8 (where minus is good and plus is bad). The fact that Kershaw struck out 11.6 batters per nine innings and walked only 1.6 per nine were the reasons he was better. It’s not that Arrieta was better than Kershaw last year. It’s simply that we didn’t appreciate just how great Kershaw was. By PWARP, it was his best season to date.

It’s also the biggest miss in Cy Young voting by the BBWAA in that past 10 years, but not as many people complained because it wasn’t as obvious.

That said, the encouraging thing is that the BBWAA is getting better at the voting. Yes, there still are a few missteps here, and yes, the Porcello vote wasn’t great, but we shouldn’t diminish these awards. The more new statistics come out, and the more inclusive the BBWAA gets, the better the voting will be.

In the past, voting was largely a disaster. Pitchers who didn’t even belong in the convo were getting the award, and this happened regularly; now, it happens every once in a while. The days of the Pete Vuckovichs of the world netting this honor is long gone. The mistakes are still sometimes frustrating, but they’re not downright unconscionable.

Yes, it still sucks when the best pitchers don’t get their due. I wish they did. I wish that Jose Fernandez and Chris Sale were nominated, and I wish that one of them would have won. But, maybe we can take solace in the fact that the Cy Young award is getting better, and the BBWAA deserves credit for that. It’s not perfect but we can see change happening and it’s positive change.

Julien Assouline is a contributor at Beyond the Box Score. He’s written for Baseball Prospectus, The Hardball Times, and BP Milwaukee. You can follow him on twitter @JulienAssouline