Reyes' Troubles

As I've mentioned elsewhere, I'm a big fan of Anthony Reyes. I really think he can be a top-notch starter in the major leagues. Sadly, he hasn't managed to do this consistently, and he's currently stuck down in the minors, trying to get better control of his two-seam fastball and curveball (supposedly). I thought the demotion marked a good time to look at Anthony's career up to this point, and see if there were any trends that could be picked up on that might be guides to success, or hints as to his struggles.

So what I did was first remove the three relief appearences Reyes made in 2005 from his career stats. Then I put each of his 29 career major league starts (including postseason) into two categories - "good" or "not good". I based the decision on what ESPN's Game Score was for each start, assigning any start with a 55 or over to the "good", and all starts below 55 to the "not good" 55 was more or less an arbitrary decision, based on the fact that all the starts with scores higher than that seemed like ones where Reyes' pitching could be cited as a major reason the Cardinals won (or had a chance to win if they didn't). There are still some decent starts at lower scores, but more often the starts are ones where the best you could say is Reyes didn't hurt the team's chances of winning. I wanted to derive keys to success, so I'm focusing on his best work. It ultimately broke down that 11 of his starts were in the "good", 18 in the "not good". I tallied up the statistics, and then used rate stats to compensate for the different sample sizes. The various stats are listed in the table below (well, in theory, in practice it's kind of a mess), summaries and theories after that:

Stats Good Starts (11) Bad Starts (18)
Innings Pitched/Start 6.1 4.2
K/BB 4.06 1.35
K/9 Innings 8.47 6.14
BB/9 Innings 2.09 4.56
FB/GB 1.48 1.08
FB/9 Innings 11.21 12.61
GB/9 Innings 7.56 11.65
H/9 Innings 4.95 11.12
WHIP 0.78 1.74
HR/9 Innings 0.65 2.33
ERA 2.09 7.63
Game Score 64.55 39.33

Not surprisingly, all the stats are worse in the "not good" starts, which is, in its own way, a good thing. If he had better stats in his worst starts, I'd be confused. Walks rise (2.09 to 4.56), strikeouts fall (8.47 to 6.14). Flyballs increase a little (Note it's "flyballs", not "flyouts". That's how it was listed on ESPN, so I assume it means all flyballs, including those that drop for hits), but groundballs increase significantly, bringing Reyes - historically a flyball pitcher - close to an even flyball/groundball ratio. Hits per nine innings rise, and most unusually, so do home runs. You would think with more groundballs, homers would decrease, but that isn't the case.

Before I progress to my theory, I need to state my bias. It has long been my belief, since sometime last summer at least, that Duncan and LaRussa are screwing up Reyes by trying to force him to change his approach from "strike guys out, give up mostly flyballs", to match their "pitch to contact, groundball" philosophy. So I'm going to view the data through that lens. If you interpret it differently, let me know.

So, the theory. The bad starts represent times when Reyes attempts to approximate the pitch to contact style (and mostly fails), the good starts are when he sticks to what worked for him in the minors. Here's what I mean.

Duncan preaches staying down in the zone, on the corners, because it's harder to take a low pitch over the fence, and pitches on the corners are harder to drive, because you either overextend, or can't extend enough (if the pitch is inside). So Reyes tries to keep the ball down. he succeeds a little, thus the increase in grounders. But to do that, Reyes relies on his weaker pitches, the 2-seamer and the curveball. He has less control over these, and so misses the zone more often. Thus, the more than doubled walk rate. Walks also mean he's falling behind in the count, which is going to force him to throw strikes when the hitters know they're coming, like in a 3-1 count. The pitch Reyes can most consistently throw for a strike is his 4-seamer, which as nice as it may be, isn't going to work as well when the hitter can just sit on it. Nobody's fastball is fast enough when the hitter knows it's coming (see 2006 NLCS, Billy Wagner, So Taguchi). That's why you see the increase in hits, and especially home runs. Reyes falls behind because he misses low, and has to come back well inside the strike zone with a fastball the hitter is waiting for, and it it gets teed off on.

Conversely, in his good starts, Reyes works up in the zone more, relying on his 4-seamer and changeup (his two best pitches). Because he's up in the zone more, fewer groundballs (as you're more likely to get under a high pitch than you are to get on top of it). He throws strikes more consistently, because he has better control of what he throws, so fewer walks, fewer hitters' counts. That means that even though the hitters are pretty sure he'll throw the fastball up in the zone, they can't be certain, so they can't just sit on the 4-seamer, which makes it more effective when he does use it. Instead of being his primary pitches, the 2-seamer and curve become suggestive threats, something to remember he might throw.

What this suggests to me is that Iron Bill has been in the past, and should be in the future, allowed to pitch the way that worked for him in the minors, and has had his greatest success that way. Of the 11 "good" starts, 8 came in games where Reyes had more flyballs than groundballs. One came when the ratio was even (against K.C. in '06), and twice with more groundballs than flyballs (Sept. 3 and 27 of 2006, both with 7 groundballs and 5 flyballs). Granted, Reyes also has 9"not good" starts with more flyballs than groundballs, but it's still a better ratio than for the other two outcomes (see table below).

FB/GB Ratio Good Starts (11) Bad Starts (18)
FB < GB 2 6
FB = GB 8 9
FB > GB 1 3

25% good start rate with more groundballs and with equal numbers, but 47% good start rate when he has more flyballs. If not for that disastorous start he didn't expect the last day of the '06 season, it'd be fifty-fifty (his flyball to groundball ratio that day was 1/0).

So I think Anthony is best served by attacking the hitters high, getting the Ks and the flyouts. Do his career numbers reflect it? I did one final table, breaking his starts down into four categories: starts where he had more groundballs than flyballs (GB starts, 8 total), starts where the numbers were even (Balanced starts, 4), starts where he had more flyballs than groundballs, but not twice as many (Moderate Flyball Tendency MFT, 8 starts), and starts with twice as many or more flyballs than groundballs ( Heavy FB, 9). That table is below, with x representing the FB/GB ratio, and with discussion following:

GB Balanced MFT Heavy FB
IP/S 5.2 5.1 5.2 4.2
K/BB 2.18 0.92 3.6 1.75
K/9 Innings 7.46 4.56 7.14 7.45
BB/9 Innings 3.43 4.98 1.98 4.26
FB/GB 0.54 1.00 1.39 3.21
FB/9 Innings 7.66 12.03 12.1 16.38
GB/9 Innings 14.31 12.03 8.73 5.11
H/9 Innings 10.28 8.71 6.94 7.66
WHIP 1.52 1.52 1.00 1.32
HR/9 Innings 1.01 1.25 1.39 2.77
ERA 5.44 4.56 5.36 5.11
Game Score 46.25 47 51.5 49.78

Huh. Not really a trend here. Of course, the obvious caveat is the small samples, as none of these groups have more than 9 starts. It would seem Heavy FBs (2x as many FBs as GBs) should be avoided, though his World Series start did fits there, but so did that ugly start on short rest the last day of the '06 regular season, as well as his one bad start in September of that year. But still, high walk and home run rates make this a dangerous option, because it increases the odds someone is on base when the ball leaves the yard. His K/BB ratio in Balanced starts is ugly, not something to be encouraged, and it's his worst WHIP, though the ERA is the best he's got.

But I really think it comes down to his GB starts, versus his MFT starts. Those are the ones where he works deepest into games, and the starts he has the best K/BB ratios in, though the walk rate in the MFTs is considerably lower. The MFTs also have slightly fewer balls put into play per nine innings (20.83 to 21.97), which reduces the chances of errors causing problems, or balls falling in for hits, which is reflected in both fewer hits per nine innings, and a lower WHIP (1 to 1.52) compared to the GBs. On the other hand, the GB starts have a lower HR rate, at just about one per nine innings, versus 1.39. That's a comes out 20 HRs over a 180 inning season, versus about 28 HRs. Of course, the MFTs still have a lower ERA, probably because there are fewer baserunners when those homers are given up.

So, what's the solution? The first step would seem to be to let Reyes pitch his way. After all, that's how he put up the stats that got him so highly regarded in the first place, and they're the reason why he was getting a chance in the rotation in the first place (well, that and lack of reliable alternatives). However, for all the crap I give him, Duncan's track record of helping (mostly veteran) pitchers is pretty impressive. I don't think it's a bad idea for Kid Reyes to work on being able to get grounders when he needs them. I just don't think that approach should be getting emphasized at the cost of what's worked for him in the the past. His stats say the Cardinals' braintrust has let him pitch how he prefers in the past (his career FB/GB is 1.31, but it's 0.86 this season), and so that seems like the best course to take for now. Quit trying to force Reyes to pitch a different way, when you haven't managed to prove his way won't work.

If Duncan and LaRussa wind up being right, well then, I guess Reyes will have to change if he wants to stick in the bigs, and seeing as how he hasn't gotten some ludicrous contract yet, he might want to work towards that. Sorry again about the tables being messed up, if something isn't clear, ask, and hopefully I can clarify.