Steven Wright's ranking

I'll have to admit that I have the opposite feelings about knuckleballers as you do and maybe I'm just looking for reasons to not like him but there are some things that bother me about Wright.

1st. He pitched 35 innings less than he did in 2013. He put up decent season numbers but he was far more effective in the first 50 innings than the last 50 innings. I know it's easy to say that he was mostly used as a reliever but the other view of that situation is that he pretty much couldn't crack the PawSox rotation for most of the year.

2nd. Something I noticed last year that was also true this year. There really aren't many passed balls in the games he pitches no matter who is catching him. I'd blind bet that Wakefield, Wilhelm, Hough, Wood and both Neikro's had more passed balls per game than Wright. Something seems odd with that to me.

In the words of Mr. Baseball, "Catching a knuckleball is easy. You just wait until it stops rolling and pick it up."

Just checked, your right. What was he doing when the PawSox rotation was Ranaudo, Barnes, RDLR, Webster and Workman and why only 15 starts? LOL, I go into knuckleballer avoidance mode, I don't think I saw any of Wake's games after about 2004.

When I first noticed the passed ball thing was last year when we were talking about Vazquez' passed balls and that only 6 were from catching Wright.

Ah OK, I do remember when he rehabbed with Portland but I was thinking that was early. I just noticed that Wright is 30. Not a lot in knuckledog years but still 30.

I understand that I have an irrational love for the knuckleball which may partially influence my ranking of Wright, but the opposition some of you have towards knuckleballs also influences Wright's ranking negatively. Wright has statistically been one of the best AAA pitchers in the past two seasons, while showing improvements in certain areas. For example his walk rate has steadily decreased the past couple of seasons, and shot down to 2.1 BB/9 this season (4.6 > 4.3 > 4.2 > 2.1). I also don't think projections (whether zips, Oliver, etc.) can be accurately used for knuckleballers, there simply isn't great enough of a sample size to do so. By that I mean that his 3.13 ERA in AAA wouldn't necessarily translate to a significantly higher ERA in the majors because of the uniqueness of the knuckleball.

Steven Wright will most likely never turn into a star. He may not even become a successful major league pitcher. But the upside of Wright simply cannot be overstated due to the uncertainty around knuckeball pitchers. The upside of Steven Wright, combined with his success in the upper minors (and majors in a small sample), should make him a top 20 prospect in this system AT LEAST. I personally have him ranked at #10, and considered ranking him higher.

My issue w/ people's ranking of Wright isn't necessarily with where he's ranked, but who he's behind. Philsbosoxfan, for example with your rankings you placed the following players ahead of Wright:

Justin Haley
Noe Ramirez
Aaron Kurcz
Miguel Celestino
Heath Hembree

I'd argue that the greatest of ceilings among those players is Justin Haley who, maybe, has a chance to be a #5 starter. One of the other 4 guys could end up as a set-up man (my money is on Hembree), but there simply isn't much value to be had there. Meanwhile, in Steven Wright, you have a ceiling of a middle of the rotation starter (or Cy young candidate if you ask Eric), and the floor of a swing-man who can eat up innings out of the pen, both of which exceed the ceiling/floor of the five players above. I just don't think it's rational to rank Steven Wright beyond the top 20 in this system, no matter how much you hate knuckleballs. Just my opinion.

I don't doubt that I am biased against knuckleballers. On the other hand as you say pretty much everyone agrees that standard projection systems don't work so, I have no reason to believe that he's a starter in the majors until I see him go around the league and start seeing teams for a second time. You yourself pointed to improving walk rates but I question if that's a good thing for a knuckleball pitcher. Just because he has a knuckleball doesn't necessarily mean he has a good one, lot's of pitchers have them (Rusney Castillo has one). What if his best knuckleball is equal to Wake's worst knuckleball ? If I have no reason to believe that he's a starter, I'd take any one of those relievers over him. I only have him ranked as high as I do because there's a chance but I view it as a slim chance, you and eric view it as a strong chance. I also view his floor as AAAA, if his knuckler is hittable by MLB hitters, he isn't going to make it as a swing man either and it's not like as if he can dial up his knuckler for a 1 or 2 inning stint.

With Haley, he's a 23 year old pitcher with a very similar package to Couch except that his sinker is 2-3 MPH faster. He's 23 and evolving, Wright is 30, how much upside is left ?

I am against ranking players purely on statistics as Eric does. Such a method is a method is a proven failure.

Given that you can't rank a knuckleballer all that high. There is no.way to scout such a pitch. No way to tell if the knuckleballs that AAA pichers are flailing at will be successful against major league competition in the long run.

I agree on trying to find out while he has an option. The Sox won't experiment however.

Just for the sake of correctness, Haley is 23.

It's a bit wearying to repeat the same explanations about knuckleball pitching that I've offered before, but here's another shot with some extra ways of looking at it.

There is no such thing as a specific knuckle-ball hitting skill. A given well-thrown knuckleball is going to be as effective against MLB hitters as it as against Little Leaguers. The pitch moves at random, and that movement comes too late for any human being to react to. Nobody knows where it's going until it goes. If it moves an ordinary amount, you get weak contact, if it moves more than that, a swing and miss, and if an extraordinary amount, a bizarre passed ball. But you cannot learn to hit it. A good knuckleball pitcher can go around the league an infinite number of times, and get the same results. The only reason knuckleball careers end is that the pitchers get too old to consistently throw good knucklers.

Look at it another way. We could in theory build a hitting machine, one with much better "eyesight" than any human, that picked up the spin off a thrown pitch immediately and calculated its trajectory. That machine would hit unerringly well. But it couldn't hit a good knuckler. The knuckleball is the only pitch that actually breaks on its way to the plate, as opposed to following a smooth arc that gives the optical illusion of breaking. Our perfect hitting machine, if fed nothing but well-thrown knuckleballs, would swing and miss frequently, would hit weak grounders and the occasional popup, but only very rarely make hard contact, and then only by sheer luck, because the ball happened to break in the same plane as the swing.

When a knuckleball pitcher moves up a level, his bad knuckleballs, the ones that don't really break at all and hence are still in the plane of the swing that the hitter started before the ball broke at random, get hit harder. But that happen to every pitcher -- hanging curveballs and mislocated fastballs get hit harder as you move up a level, too. But for ordinary pitchers, their good pitches get hit harder, too, because hitting good pitches is the pre-eminent skill of MLB hitters. But there is no hitting a good knuckleball except by sheer luck. As a result, as knuckleball pitchers move from AAA to MLB, they lose less value than an ordinary pitcher.

The success that a knuckleball pitcher is having in AAA is a terrific measure of how frequently he's throwing good knucklers. Folks are inventing reasons for a AAA pitcher's success being less projectible than a conventional pitcher's, when not only is the precise opposite true, it's profoundly true. All the reasons that a pitcher might succeed in AAA disproportionately don't apply to knuckleballers. You can't get away with throwing high heat at 93 mph to MLB hitters, because they either won't chase or can hit it much harder; your FB command isn't good enough, so suddenly your killer changeup is much less effective, and a score of other stylistic scenarios that might give us pause about a conventional pitcher (and not just ones named Ranaudo and Webster respectively) -- none of them apply to a knuckleballer.

So: let's take a conventional pitcher and a knuckleball pitcher, having identical AAA success. They're going to throw the same exact pitches to MLB hitters.

1) It is the knuckleball pitcher who will be more successful, even if there are no stylistic weaknesses to the conventional one. The conventional pitcher will give up some well-hit balls on good pitches. That doesn't happen to the knuckleballer.

2) There is essentially no chance of a stylistic weakness for the knucleballer, which widens further the average gap between them once promoted.

Now, there are some people who feel that the capacity to throw a good knuckler has more variation and less predictability than the capacity to pitch well conventionally. That's based on a few anecdotes, like Wakefield completely losing it for more than a year early in his career. But I believe the evidence will show that knuckleball careers are no more inconsistent than conventional pitchers.

Steven Wright is not the best SP prospect in this system because his odds of being an ace are lower than Owens, Barnes, and probably Rodriguez. He has a higher floor than anyone, though. #2 starter ceiling, #4 starter floor, #3 starter projection. And that he's in his age 29 season is meaningless. As I have pointed out earlier, he has matched or outpitched Tim Wakefield when you align their careers at the point each started throwing the pitch.

This blew my mind.  My thought was that any pitch could break late based on the spin.  If knuckle balls can do it based on the spin (or lack thereof) then other sorts of spin ought to be able to create similar, although maybe less random or drastic, effects.  

I looked up curve ball trajectories and found what you said, Eric.  They follow a smooth trajectory and only appear to break at the last moment.  (The reasons they appear to break late are more interesting than I expected and in part involve the difference in the way our peripheral vision interprets spinning objects!)

The best explanation I found of knuckle balls was this one.  The conclusion by this researcher was that knucklers also follow a smooth path.  The difference with other pitches is the trajectory is random (though still smooth) and difficult to project until it's too late.  On the other hand things stated later in this video seem to undercut that assertion.  I'm guessing that's less because of Professor Nathan's being confused than because of poor reporting by the journalist, but who knows.

What's your take on this, Eric? 

Edit: Shortened the quoted material.

i don't disagree

Eric's argument can be summarized in four words "everyone's a Luddite but me.". To be honest I am finding this attitude quite tiresome. There are a lot of smart people working in baseball and for the team and you can't just declare them all to be stupid. In my experience whenever an arguer proclaims to have found a "profound truth.", they are instead profoundly wrong, and I think that's the case here.

In fact there are reasons for a knuckleball pitcher being less projectable going forward and hence far more risky.



Even if this is true, it is only true for a "well-thrown" knuckleball. Unlike a fastball or a curveball there is no way to grade a knuckleball and no way to tell the difference between a 60 knuckleball, which might be more effective against minor leaguers and an 80 knuckleball which hitters would be basically unhittable. The only way you can tell if a knuckleball is spinning enough to be as effective is by the actual outcome of the pitch. You could not watch a pitcher throwing a knuckleball on the side and tell from that if the knuckleball was good enough to get major league hitters out.

Secondly, a knuckleball is very difficult to master and throw consistently well. It's said that a knuckleballer could regularly load the bases with walks and then strike out the next three batters on nine pitches. There have been seven knuckleballers to debut in the last 20 years, and every single one of them had at least one strong season in the high minors. But only one of them, R.A. Dickey, became a regular starting pitcher.

Charlie Zink appeared to have achieved some sort of mastery over the knuckleball in 2008 when he achieved a very decent strikeout to walk ratio of 2.16 over 174 innings. For the rest of his career, the ratio was .45, a near 80 percent collapse. That sort of thing usually dosen't happen to pitchers unless they are injured, but it does happen to knuckleballers.

By saying that he had a lower floor than anyone else in the system and the upside of a #2 starter you are putting a lot of faith into Steven Wright's mastery of the knuckleball based upon roughly half a season of play. You are saying that he will not befall the same fate of just about every other knuckleball pitcher in the past 20 years. It's not an argument that anyone should take seriously especially with all the bluster you use in making it.

The physics of the knuckleball are pretty well understood. The pitch optimally rotates less than once on its way to the plate. The airflow over the seams, at some random point, changes in a discrete, nonlinear way, actually producing a force on the ball that changes the trajectory. In other words, as the seams rotate, there's suddenly a moment when the continuous airflow over the ball is disrupted and altered to a different sort of airflow. This is easy to visualize if you think of the ball in two different configurations in terms of where the seams are, not rotating at all -- each configuration will have a distinct airflow pattern based on where the seams are. Now, if you rotate the ball gradually from the first configuration to the second, the change in airflow pattern is not gradual, but at some point makes a jump from being closer to the first pattern to closer to the second. There's a tipping point where the seams have rotated into or out of play and the airflow over the ball alters.