Life Is Easy, When You’re Up On The Mountain
And You’ve Got Peace Of Mind
Like You’ve Never Known
But Then Things Change And You’re Down In The Valley
Don’t Lose Hope For You’re Never Alone
(The McKameys, God On The Mountain)
I really don’t care about Carlos Gonzalez one way or the other. In another thread I likened him to an ex girlfriend, the one I broke up with a long time ago, only to see she looks stunning in her new dress now. I have no interest in discussing whether it was a good choice to let her go back then or not, and none in finding out how she compares to other girlfriends I have had since. Not even in knowing whether her true level of beauty is closer to her image in a dress or to a not so attractive one in jeans.
What I want to understand is whether that specific dress would make any woman appear equally more attractive, or if it is possible that it favors some more than the others. To understand that, I have to understand what effect such a dress has on appearance. To understand that, I have to understand how Coors really works.
And that’s what I did for the better part of the weekend. As a self proclaimed chief AN researcher and a self appointed head of AN Labs, I looked at every single pitch that was thrown in the Majors this year and tried to make sense of it all. It will not be brief, but if you engage on this journey with me, I promise to take good care of you, to show you data you will not find anywhere else and to give you more food for your thoughts. This is not your typical stats post, and I would love to see some non-stat crowd read it. If not, just jump all the way down to the summary.
The only thing I ask you is to leave your CG prejudice and/or agenda right here. This is not a story about him, he is just a perfect outlier to analyze. This is a long story made predominantly out of thin air.
What if I told you that Carlos Gonzalez’ poor road performance is due to his struggles to hit a certain pitch outside of Coors? And what if I told you that that certain pitch is not a curveball and that it actually all makes sense?
Until and including August 6th, there were 474,587 pitches thrown in the MLB, with just above fifteen thousand of them pitched in Coors Field. Looking at those pitches we can tell if there is any truth to the claim that batters in general have an easier time hitting in Coors than on the sea level.
Why do the pitches move?
This is pure physics. Curveballs curve, risers rise, sinkers sink; all as the effect of the spin on the ball. In case you missed it, I recently wrote an article explaining how the spin and velocity affect the movement of the ball. If you don’t want to understand the why of it, it’s just as well. Take my word for it, then.
How do the pitches move?
If you are more into empirical than theoretical, than this is a better way to look at things. What you will see in the following chart is the average movement on seven most often used pitches this year – I only used the ones where Pitch F/X pitch recognition confidence was more than 90%.
FF= Four-seam Fastball; FT = Two-seam Fastball; FC = Cutter; SI = Sinker; SL = Slider; CH = Changeup; CU = Curveball or “See you, Bobby Crosby”
- This is from a catcher’s point of view when a righthander is throwing*.
- The thick lines cross at the point where a pitch would land if it had no spin at all.
- The thin lines are two and a half inches per line grid.
- The colors represent the speed (the redder, the faster).
* I used both righthanders’ and lefthanders’ pitches, I just show them all as if they were thrown by righthanders.
This representation is somewhat misleading, because it centers around a point where a ball without spin would land. While it is understandable that scientists think this way, it is not how baseball players work – none of us has actually seen a 90mph fastball without a spin. Ever. Since the childhood, our eyes are conditioned towards the movement of the fastball, mostly that of a four-seam one, not only when hitting, but when throwing and catching.
So I added an eye, somewhat arbitrarily, to the point where “one would expect the ball to land”. So, if you were playing catch with someone who throws the ball fairly hard, the ball would land more or less there. Now, this looks much more natural – the riser really rises, the sinker really sinks and the curveball ends there where Crosby can not molest it.
What pitches are thrown in Colorado?
Now, if there is any truth to the claim that curveball doesn’t curve in Colorado, you would think that pitchers know about it, right? I mean, sabermetrics and all – if you are constantly getting killed on a non-curving curve, you will stop throwing it, I guess.
Let’s look at what pitches are thrown in Majors altogether and what in Colorado.
The right table excludes the Colorado pitchers* as their personal pitch selection would account for 50% of all the pitches in Coors and could seriously skew the picture. Here is what they do:
Colorado pitchers actually throw less fastballs at home than away. Also, somewhat more sinkers at home than on the road, but it could be just that Aaron Cook or Jeff Francis had a few more starts at home. Generally, neither comparison shows anything out of the ordinary. It is interesting that Colorado pitchers throw that many sinkers – it could be that they try to sign guys who throw sinkers or it could be that Bob Apodaca can teach a good one.
But, we did not find the smoking gun we were looking for – if anything, the curveball is thrown slightly more often at Colorado. Are all these people really that stupid? Would they willfully get hammered on a lousy curveball?
*One of the fun things about projects like this is that I run across the players I don’t know of. My new favorite name is Esmil “It’s A Thousand” Rogers. I am eagerly awaiting the appearance of his younger brothers, Escien, Esdiez and Esuno.
What happens when a certain pitch is thrown?
First, we need a baseline. Below is an overview what happens, percentage-wise, when a Major League pitch is thrown for a strike.
Don’t take these numbers as a gospel, but they do tell an interesting story. Just to make sure that nobody’s confused by this representation, let’s look at an example in the first column. When a four seam fastball is thrown for a strike, following happens:
On 28% of the occasions, it will be put in play. On 30% it will be taken for a strike. On another 30% it will be fouled off. And finally 12% of them will completely elude a swinging bat. On average one could say that fastballs are missed on 10% of the occasion, breaking pitches on 20% and that altogether that baseline is somewhere around 15%.
And now, let’s look at what happens when those pitches are thrown in in Colorado:
One thing we can see is that somehow the font got smaller on the second graph. I love Excel.
Slightly higher miss rate on the breaking pitches and lower on the fast ones.
So, by now we know that:
- Curveballs in Colorado are not thrown any less often than elsewhere
- Curveballs in Colorado are in general not hit with significantly more success than elsewhere
Before we have a look at the whole spinning/curving effects of Colorado, let’s come back to our Carlos
What does Carlos do differently on the road?
For The God On The Mountain Is Still God In The Valley
When Things Go Wrong He’ll Make Them Right
And The God Of The Good Times, Is Still God Of The Bad Times
And The God Of The Day Is Still God Of The Night
Well, we all know that Carlos Gonzalez was anything but Godlike in the valley, his recent punishment of Pirates‘ pitchers surrogates notwithstanding.
Is he being pitched differently on the road? Does the book on him say that he can not hit a sea-level curve and is this weakness being exploited? The following overview shows which pitches CG saw at home and which away, along with the rest of Colorado hitters as a baseline.
Wow, wait a second! Carlos Gonzalez actually sees more fastballs and less curveballs on the road than at home?! Are these people completely crazy?
Apparently not, because this is what Carlos Gonzalez does when he sees a certain pitch:
Now, this is staggering and completely unexpected. Carlos does struggle away and he does that in an epic fashion, but he struggles on fastballs?! He swings and misses on 24% of fastballs on the road, which is between double and triple what the league does. Or put in different words, CG swung at more than 100 fastballs while playing away games this year. He managed to put less than a quarter of the ones he swung at in play, while the league succeeds at it more than the half of the times.
But how does this make any sense? Fastball is a fastball, regardless where, it’s the curveball that performs differently in high altitude. Well, continue reading, you are in for another surprise.
How do the pitches move in Colorado?
I first analyzed all the pitches thrown in Colorado this year. Then I compared only the ones thrown by Rockies staff in Denver as opposed to those that they threw in the away games. Finally, to make sure I am not making any mistakes, I compared single Rockies’ pitchers for their home/away pitch movement. The graph you will see is not correct to the decimal point, but I am very confident that it is a good representation of what happens in thin Colorado air.
The washed out ones are the movement locations of the same pitches on the sea level. The bigger the distance between the foreground and a washout mark, the bigger is the “Coors effect” on that specific pitch.
This is in line with what one could expect based on calculated models of movement. Four points can be said about this:
- Pitches move differently in Colorado
- Pitches move less in Colorado and overall dispersion is much smaller (the visible rectangle versus the washed out one). Therefor, the element of surprise due to the pitch movement is smaller.
- Fastballs lose much more movement than curveballs and sliders
- Pitches are faster in Colorado (not visible from this graph, but have a look at Ubaldo Jimenez table below)
Here you can see that there is no significant difference in how his curveball and slider behave in thin air. The fastball, on the other hand, is a completely different pitch in the mountains. It loses more than a mile per hour less on the way to the plate, but it does so at the expense of extremely reduced movement. More than three and a half inches of fastball movement, its life, are lost. It is a much flatter and a little bit faster fastball.
So, it is surprisingly the fastball that suffers the most when thrown mile high.
Why should Carlos Gonzalez be affected by it more than someone else?
If we were to categorize Carlos Gonzalez, I think that a term free-swinger with power would be the one few would oppose. Let’s just briefly compare him with someone who is, well, something completely opposite:
Carlos loves to swing and when he runs into one, oh boy! Personally, I find his swing one of the nicest things there is to watch in all of Major Leagues. And make no mistake – his power is for real, even though Coors does have a 15% – 25% inflationary rate on home runs over last 4 years. He can hit them out of any park – if he can get a hold of one.
Would Daric Barton profit as much from playing in Colorado? I doubt it. He has a great eye, he reads the spin perfectly and rarely gets fooled by the pitches. Unlike with CG, Daric’s problem is not to know where the pitch will end, he already knows that on a sea level – it’s what to do with it once he gets a hold of it.
Or, think of it this way – you have a video game and you have two main characters in it. Barton is a 3/10 in power and 9/10 in pitch recognition. Gonzalez is 8/10 and 2/10 respectively. Coors is an upgrade of +2 in power and +5 in pitch recognition. You see what I’m getting at?
I have no time left to run correlation tests on the examples of players fitting one of the two molds and playing in Colorado and comparing their home/road split. It could be made form publicly available data, if anyone has energy to give it a go, it would be great supplement to this study and I would surely recommend it.
Summary and some thoughts
This is meant as both recapitulation and as a quick entry point for people who had neither the time nor patience to go through the whole thing. It’s also that I see those things in every SI article nowadays, and I want to be cool, too.
- Different pitches move differently due to their respective spins
- Visiting pitchers throw similar pitch selection in Colorado as they do elsewehere
- Colorado pitchers throw less fastballs at home
- Fastballs are the pitches most affected by thin air, as they lose a big part of their movement, unlike breaking pitches
- Carlos Gonzalez is being pitched more fastballs on the road than at home
- He struggles with fastballs on the road, yet is successful against them at home
- His performance on breaking pitches is pretty similar home and away
This is where the results of my research end, the rest is pure speculation.
- I can see it as plausible that different hitter types are affected differently by “Coors effect”.
- Coors seem to benefit more on a pitch recognition level than on a “added power” level, thus favoring free swingers with power.
- And, one last time, I don’t care about quantifying it and I am not running a conclusion that Carlos is a bad player, nor that his away splits are the true level of his skill. All players have better home splits. I am open to the possibility that Coors might improve CG’s home stats somewhat more than it would for a different type of a player, and if it does I could see why. Nothing more, so, please don’t misquote me.