Braden Really Changes It Up
Last night, Dallas Braden pitched his first game of the season, and first start after missing half of last year. It was more than encouraging as last year’s Opening Day starter struck out ten Mariners while walking just one and giving up just four hits over seven innings. Those ten strikeouts are a career best (his previous high was seven). Last year Braden was successful based on an unsustainable sub-5% HR/FB, so to be successful this year he is going to need to strike guys out like he did yesterday and in his 117 innings at Triple-A in 2007 and 2008.
Braden is an interesting pitcher; the lefty throws an 87 MPH fastball and along with it the slowest changeup in baseball. Last year, it averaged 72 mph. By comparison, Barry Zito‘s was 73.6, and no other starting pitcher had one slower than 78 mph. That separation of 15 mph is also one of the greatest between a fastball and changeup. And it works, as the pitch was worth almost two runs above average per 100 in 2009 and 2008, and he throws it often — 21% of the time last year.
Last night he threw it even more often, throwing 34 changeups out of his 91 pitches, and 32 changeups out of his 71 pitches to RHBs. It did not disappoint, inducing 12 of his 16 swinging strikes. It will be interesting to see whether he continues to use the change more often this season, as it is his best pitch.
To get a feeling for just how much slower his change is I plotted the pitches in Braden’s three-pitch strikeout of Milton Bradley in the first. He started off with a two-seam fastball (blue) for a swinging strike, then a four-seam fastball (green) up-and-in fouled off, and finally the changeup (yellow) down-and-away for a swinging strike three. I put a little dot every 0.075 seconds. The horizontal and vertical axis are not to scale, with the height exaggerated relative to the length.
By the time the change reaches the plate it is almost tenth of a second behind his two fastballs.
1
Dave, I love your 1337 (presumably R) skills.
Thanks.
“By the time the change reaches the plate it is almost tenth of a second behind his two fastballs.”
And it seems initially to be right in between the first two pitches from hitter’s perspective…falling below the strike zone by the time it reaches home. Although I don’t know how much that is a product of scaling.
Great visual!!!
I agree, that dot visualization method is excellent Dave. Fantastic way to present it.
Love the visual. Do this more often!
I know that so far this is an extremely small sample size, but look at the difference in pitch selection between 2009 and 2010. He’s now got four distinctly different pitches all of which he threw greater than 10% of the time. While the changeup itself is good, I think just as important to his success in his last start was the increased use of that two-seamer. Notice how closely the movement of the two-seamer matched that of the changeup. When you’ve got two pitches that differ by 13 mph but move the same, that’s hard to hit. Add on to that the fact that his slider and changeup are thrown at around the same speed but move totally different AND the fact that his fastball and two-seamer are thrown the same speed but move differently, and this guy has a lot of potential.
Granted, very small sample size of one start, but if he can keep up this sort of Pf/x in his next start and do moderately well too I’ll be bidding high on him for FAAB.
I just watched the highlights of Braden’s K’s from his start, and I think it might be the most fun to watch pitchers strike batters out on those bugs bunny change ups. It seems like the batter could go ahead and attempt another swing at the pitch after missing it the first time.
With a changeup, I think seeing him is as important as seeing the data.
Separation in MPH, path of the ball, all great info. If the batter doesn’t *believe* a fastball is coming, however, then he won’t stay as effective.
Sounds like his changeup fooled one team for one night at least. Such a tough call this early in the season. Some of the best pickups are early ones, but when to pull the trigger???
The thing is that even most changeup artists don’t pull off that kind of differentials, Chuck James had a very affective first year partially because he had nearly that kind of seperation. The next his fastball lost some zip and the next thing you kow he’s AAAA filler. A tenth of a second is long enough that even when they know its coming they won’t be able to affectively wait it out.
I like the visualization (as always) but think that something’s missing. While the dots are easy to get if you start from the left, I think most people will look at the graph starting from the right.
All the information in the graph save the dots are coded around the strike zone — height, color, comparison with other pitches. There’s no real reason to look at the left side of the graph except to pick up the dot spacing.
Instead, could you just make a line for, say, 100 mph (or whatever) and then use that as a baseline, and just mark each pitch at the time where a 100 mph pitch fastball would cross the plate?
(the reason for the 100 mph is that if you set it at 90, or a league average fastball, then an above average fastball would end up crossing the plate before the marker — ideally each pitcher would be gauged against their best fastball from that game to show the actual gap that the batter is seeing)
That would also allow you to shorten the “lead up” and further exaggerate the vertical path to let people get a better idea. It also has the added benefit that it’s logical in the sense that batters have to get around on the fastball, so the gap between the fastball and the other pitches is what counts.
Just a suggestion — don’t know how feasible it is.
FWIW, it didn’t occur to me to look at that graph in any way except from left to right — the same way I read text. Seemed clear enough to me. I don’t entirely get what you’re suggesting; I’d have to see an example to decide if it would be any better.
By marking each 0.075 seconds, you can only see the difference by visually “counting” the number of dots in each line from left to right.
My idea is to stick a single dot on the graph, say at 0.45 or 0.5 seconds for each line so that there is a single reference point for how fast the ball is going.
Then you only have to look at a single point, rather than following each dot consecutively.
Naw, I did the same thing: look at the strike zone and see the final dot, but then saw the incredible distancing when I read the graph from left to right.
Where did you get that graph? Or did you just make it?
Yeah I made it.
moyer doesn’t throw a slower change that zito? i know that its harder then 72, but it was down in the low 70s last few times he pitched.
Moyer’s changeup average 74 mph last year, and for many years before that. You can check the average pitch speed for any pitcher their page.
The leader board for slowest changeups in 2009 is here. Moyer, and Braden, do not appear because they did not get enough innings.
This is a reply to JDP: totally agree. Jamie Moyer’s changeup isn’t slower? Heck, his fastball is in the low 80s, so his changeup averages out faster than Braden’s? I find that hard to believe. Can the author cite where his stats are coming from? I’m not disagreeing with you, per se, just asking you to direct us to the website you got this information.
Braden added a cutter last season which he attributed to his success. If anyone remembers, he also in the minors threw a screwball, but A’s were concenred it led to his durability issues and he dumped iy.. His pure stuff in terms of velocity isnt great, but he has a wide array of pitches and battles out there. Reminds A”s fans of tim hudson with that bulldog mentality
I’m sure out in Fantasyland a million punters are looking at stats today, seeing 10K and “A’s pitcher” and adding Braden to their teams — afterall, Beane found Harden and Haren and Hudson (and Bailey and Mulder and Zito, oh my!) so it’s a lock, right?
They’re going to be sorely disappointed when Braden starts against one of the 26 or so teams with a better offense than the Mariners.