For a few years, now, Felix's fastball has been the center of debate. One group of Mariner fans wants to see him throw it less often and less predictably, as it has statistically been his worst pitch, while the other group wants to see him keep throwing it a lot in an effort to improve his command, as the fastball, they believe, is the pitch off of which all his other weapons are built. It's been a fascinating and novel discussion from the beginning, and 2008 only served to elevate its significance, as Felix started throwing his fastball more than ever before, with arguably disappointing results.

One of the main bullet points of the pro-fastball crowd is the argument that, were Felix to throw his fastball less often, that would in turn make his offspeed stuff less effective, as the success of his offspeed stuff comes at least in part out of its unpredictability. It makes intuitive sense. If batters aren't sitting fastball so much, then in theory they'll be better prepared to adjust to pitches that dive and bite. If true, this would support the position that Felix needs to keep attacking primarily with his heater.

With the help of PITCHf/x and the Fangraphs game logs, we can legitimately investigate this idea. I downloaded all of Felix's .xml files and went into each one individually to figure out how many fastballs Felix threw in each start. I then split them up into the following three groups:

Not the greatest set of sample sizes in the world, but it's not horrible, and the distribution is fairly even. Felix had eight starts in which he threw fewer than 60% fastballs, ten starts in which he threw more than 70% fastballs, and 13 starts in which he threw between 60-70% fastballs. For the curious, his overall range was 54% (April 6, vs. Baltimore) to 76% (July 18, vs. Indians).

Armed with these groups, we can then proceed to the data:

There's a trend in there, but not in the direction many people would expect. Instead of making his offspeed pitches better, Felix throwing a lot of fastballs actually seemed to make them worse. In fact, it just made Felix worse, period. When he threw a bunch of fastballs, his fastball was less effective, his offspeed stuff was less effective, and his K/BB dropped close to the league average. When he mixed his pitches better and threw more junk, he got more swinging strikes across the board, and his K/BB climbed higher. In other words, not only did throwing fewer fastballs improve Felix's offspeed stuff - it also improved his fastball. That was unexpected.

This isn't intended to be definitive. We're only dealing with one season of numbers, and there could be any number of reasons why the data looks the way it looks. But the fact that Felix's offspeed swinging strike rate was its lowest when he threw the most fastballs pretty strongly suggests that the two aren't correlated in the way that the pro-fastball people would think. It seems, rather, that the best way for Felix to throw off the timing of the hitters is to introduce a pitch mix closer to 55/45 or 60/40. It's probably not a coincidence that, in a season in which he threw the most fastballs, he put up the worst (meaningful) statistics of his young career. He should think about getting back to what made him effective in years past.

Short of getting some measurement of what's going on in a hitter's brain when he steps in against Felix, we'll never be able to prove the case either way. But while there's no evidence that throwing more fastballs improves Felix's offspeed stuff, there is evidence of the opposite, and with that in mind, I think it would behoove Felix and the Mariners to change away from his 2008 plan of attack. When a guy can throw four pitches for strikes, it isn't necessary to build off of any single one of them. It's best to use them all in an unpredictable mix, and it is by doing that that, in my opinion, we would be most likely to see Felix hit his peak.