Perception and Reality
If you’ve been following the reports about the Mariners starting pitching candidates this spring, you’ve no doubt read about how Hisashi Iwakuma’s “struggles” have been concerning to the coaching staff, which is part of the reason why there’s a chance he begins the year in relief, rather than as the #3 or #4 starter as expected. The other reason has been the “strong spring” performance of Blake Beavan, who began March on the outside looking in but has pitched well enough to give himself a real chance at starting the year in the rotation.
At least, that’s the story. Now, here’s the reality.
Hisashi Iwakuma: 57 batters faced, 2 walks, 9 strikeouts, 1 home run
Blake Beavan: 59 batters faced, 2 walks, 8 strikeouts, 1 home run
In the three things that a pitcher has some reasonable amount of control over, they’ve performed identically. Yes, Iwakuma’s given up 19 hits and Beavan just 12, but hits allowed aren’t really something that have any predictive value during the regular season, and that goes double for spring training. Evaluating a pitcher based on how many balls are falling in during Cactus League games in March is nothing short of absurd. It’s a bad use of results based analysis during the regular season, when teams are trying to win and rolling out Major League defenders on big league fields – during spring training, it’s insane.
In reality, Beavan hasn’t “performed” any better than Iwakuma this spring. If the Mariners actually do start the season with Beavan in the rotation and Iwakuma in the bullpen, it better be because they saw something in Iwakuma’s arsenal that they didn’t like, or because he showed worse stuff over here than he was throwing in Japan. There could be legitimate reasons for bumping Iwakuma from the rotation and giving Beavan the job instead – those reasons just don’t have anything to do with the results either pitcher actually got when they took the mound.
I guess it depends at least to some extent whether the hits are bloopers “falling in” or hard hit line drives. Why would Iwakuma allow so many more balls to “fall in” than Beavan? The batting average against Iwakuma is .358 vs. only .214 for Beavan. Iwakuma has allowed 21 baserunners in just 12 innings and Beavan has allowed only 14 baserunners in 15.1 innings. Why is Iwakuma so unlucky with the same defense behind him as Beavan?
Read this.
http://ussmariner.com/2006/08/29/evaluating-pitcher-talent/
Or this.
http://www.fangraphs.com/library/index.php/pitching/fip/
Hit rate over 60 batters faced is completely and utterly worthless. It’s rarely even useful over 600 batters faced.
Interesting post. Makes me wonder if some of the hype on Beavan is an attempt to shop him. I didn’t think he had been pitching bad, but I didn’t realize Iwakuma’s numbers were so similar. I’ve tried to pull the stats from the Mariners website, but they only list the position players. Anyway, thanks for the post. Has me thinkin’.
I think the early struggles of Iwakuma and the fact that he is having to make quite a few major adjustments is something the Mariners probably expected. The M’s may just want to make sure he is healthy.
On another note, completely agree the hits are just too small of a sample. Remember, in Spring Training balls are lost in the sun, different players are learning new positions, the difference in temp plays a major role, and the million other variables, including in 1996 or 1997 when Mariners minor leaguer Scott Podsednik was attacked by bees in LF. I was there in Peoria as a young buck and got to see Pod running all over left field to avoid the bees, which led him to misplace a ball, leading to a hit. Just sayin….
Stevemotivateir – at the M’s site, just toggle the “pitchers” tab at the top, to the right of the “Sortable player” header.
On top of that, we don’t even know what the pitcher is trying to accomplish. Maybe he’s concentrating on getting certain pitches right. Just get in a rhythm. Working the velocity up. Iwakuma is coming off injury in 2011, and that may be a factor as well in terms of what he’s trying to do.
With Iwakuma there are likely some special adjusments to be made, such as getting used to the American game, the language barrier, and seeing every batter for the first time. I would think it will be sometime into the regular season before the Mariners really know what they have.
I think what happens with Iwakuma is going to have more to do with evaluation rather than results so the media hype need not be a concern.
I was at the game today, sat a section over from GmZ and a squad of scouts. They were pretty engaged while Iwakuma was out there. I have no idea what they were talking about just that they were exchanging radar gun readings and delivery times on what seemed to be every pitch. Compared to Millwood’s innings, the Reds as a whole and the rule five kid…they seemed much more engaged with Iwakuma.
I’d guess they’re monitoring his velocity and trying to figure out why his sinker isn’t inducing groundballs as advertised. The homeplate umpire was seriously erratic, however ‘Kuma seemed to be around the zone (but I was a section away from a scouts eye view where the braintrust was) Also, he takes too much time with runners on. But who knows.
I have faith that they are doing their homework and not listening to the media. Dude is hands on.
“I’d guess they’re monitoring his velocity and trying to figure out why his sinker isn’t inducing groundballs as advertised.”
I’d say if it’s true his sinker isn’t working, it’s likely because it’s Arizona.
Iwakuma displaying below-average groundball rates would actually be a story to follow. I would think that GB number’s aren’t available for spring training, but does anyone have them?
Dave,
I read the two posts you linked… and I’m sure I’ll get some backlash for this, but I agree with dantheman.
I think things are oversimplified a bit when you categorize contact into GB’s, LD’s, and FB’s. How hard the ball is hit is what really matters.
I understand that it is a small sample size, but your comparison to Beavan fails under the same criticism, as there are just as many explanations for their similarities as for their differences.
I am a stat geek, but stats should be intended as a complementary, not primary, evaluation tool. In my opinion, stats should always take a back seat to good visual evaluation and intuition.
@marc
That was the place I had checked, but only the names were listed. So of course I just checked it again everything’s there 😛 I must have just caught it at a weird time or had a server error. That’ll teach me to check back!
“Now, here’s the reality.
Hisashi Iwakuma: 57 batters faced, 2 walks, 9 strikeouts, 1 home run
Blake Beavan: 59 batters faced, 2 walks, 8 strikeouts, 1 home run
In the three things that a pitcher has some reasonable amount of control over, they’ve performed identically.
[…]
In reality, Beavan hasn’t “performed” any better than Iwakuma this spring. ”
That’s only a small portion of reality. Even without Pitch f/x stats, we can ask whether they were throwing strikes or balls, missing bats, etc. And with a combination of radar gun readings and visual reports (not that those are available to most of us but the scouts and I’d guess STATS and other data companies have them) we can ask was his sinker sinking, was his curve curving, was his fastball fast, etc. etc.
While it’s correct to say that we can’t talk about a “strong spring” by simply looking at hits and runs given up, we can’t talk about a strong spring by simply looking at Ks, BBs, and HRs either. The stats that you put up do not show that “Beavan hasn’t ‘performed’ any better than Iwakuma this spring”. They show that he hasn’t performed any better on those three limited stats.
You guys can disagree with it all you want. It doesn’t really matter. The reality is that if you want to predict future pitching performance, you should completely ignore hit rate in almost every situation, and you should ignore hit rate in spring training in absolutely every situation.
If you’re deciding who should get a job based on how many hits they allow in Spring Training, you’re doing it wrong.
I think we all agree hit rate is useless. What we’re arguing is that the stats provided, in this case, are not providing enough context, small sample size or not, as to the performance of these two pitchers.
If I understand your post, you’re saying that either, or both, Beavan and Iwakuma’s performances are misunderstood because of the focus on hits allowed by each during Spring Training. As a way of proving your point, you listed stats that are just as open to context as hit rate. That’s where I took issue.
Isn’t it possible that the coaching staff saw that Iwakuma was getting batted around pretty good in comparison to Beavan… as opposed to them just taking a look at the BA against each? Your follow-up comment to Dan then told us to read to posts on statistical evaluation of pitchers, ignoring that what we’re saying is that “well hit” balls are not properly accounted for in the various sabermetrics generally available to us.
By the stats you provided in the post above, it is just as likely that Iwakuma is hitting too much of the plate and serving up BP as it is that he’s being screwed by a small sample size. In my opinion, this is a classic example of where statistical analysis highlighted a possible problem… and now it’s time to take a look at the tape.