The Case Against K/9 and BB/9

A few visitors to this site have asked why we no longer include strikeouts per nine innings (K/9) and walks p nine innings (BB/9) in pitchers' statistics tables. The simple answer is that there is no need to; K% and BB%, or the number of strikeouts or walks per batters faced, are better descriptors of pitchers' skill in striking out batters and avoiding walks.

You might be wondering why it matters at all. It's true that both sets of numbers will give you similar information about most pitchers. For some pitchers, however, the differences can be meaningful. Let's take a look at how Kevin Slowey, who is scheduled to make his major league debut tomorrow, compares to other Triple-A pitchers in strikeouts per nine innings:

    Name            K/9
1. Yovani Gallardo 12.5
2. Virgil Vasquez 10.6
3. Jeff Niemann 9.6
4. J.A. Happ 9.4
5. J.P. Howell 9.0
6. Andy Sonnanstine 9.0
7. D.J. Houlton 8.5
8. Matt Garza 8.3
9. Colby Lewis 8.3
10. Ron Chiavacci 8.3
11. Jason Hammel 8.3
12. Kevin Slowey 8.0

Slowey is ranked #12 in strikeouts per nine innings. That puts him among the top 25% of qualifying pitchers, but he's still within one standard deviation from the average.
Here are the Triple-A pitchers ranked by strikeouts per batters faced:

Name K%
1. Yovani Gallardo 34.9%
2. Virgil Vasquez 27.9%
3. Andy Sonnanstine 25.4%
4. Jeff Niemann 24.5%
5. J.A. Happ 24.4%
6. Kevin Slowey 23.4%
7. J.P. Howell 23.4%
8. Colby Lewis 23.0%
9. Jason Hammel 22.8%
10. D.J. Houlton 22.8%
11. Matt Garza 21.9%
12. Ron Chiavacci 21.5%

This ranking paints a different picture. Both Slowey and Andy Sonnanstine have impeccable control, and they both move up a few spots when we rank pitchers by K% rather than K/9. Pitchers who do surrender many hits or walk a lot of batters have inflated K/9 rates because they face more batters per inning and therefore have more opportunities to record strikeouts. Six of the eleven pitchers ranked ahead of Slowey according to K/9 are not really striking out batters more frequently than Slowey; they are simply facing more batters per nine innings of work.

K% and BB% do a better job of isolating the skills they measure, so it should come as no surprise that K% and BB% are more strongly correlated from season-to-season than K/9 and BB/9.

If you are more comfortable understanding K/9 and BB/9 than K% and BB%, you can quickly get a feel for the range of these numbers by browsing the league leaders tables. As a general rule of thumb an average professional pitcher will strike out about 15% of opposing batters and walk about 8% of opposing batters.

K/9 and BB/9 will probably persist if only because fans are comfortable using these numbers and 'innings pitched' is a more widely available denominator than 'total batters faced'. But the numbers are available here and many other places as well. There is just no reason to use K/9 or BB/9 when analyzing a pitcher’s abilities.

BallHype: hype it up!

Comments

1
I'm convinced.
Posted by smittybanton on May 31, 2007
2
I didn't ever think there was a debate?
Posted by siddfynch on May 31, 2007
3
There is no debate that I know of.
A few readers asked why we were no longer publishing K/9 and BB/9 numbers, so this is one explanation for anyone else who was wondering.
Posted by Chris Constancio on May 31, 2007
4
I'm still baffled why you don't include anything about steals for hitters.
Posted by galt on May 31, 2007
5
If K/9 has a better correlation from year to year should we completly disregard it? And I was wondering how K% coreelates from year to year?
Posted by uncwredsox on May 31, 2007
6
I'm relatively new to this site, but is there a reason there's no RBIs listed for batters?
Posted by parkerhodges on June 01, 2007
7
Because RsBI is perhaps the most overrated and irrelivant stat in the history of sports. We have statistics that measure an offensive player's skill set, why use such a flawed method of statistical reasoning?
Posted by ryandoumit on June 01, 2007
8
More importantly...who the heck is Virgil Vasquez?
Posted by siddfynch on June 01, 2007
9
@#7 - I agree that it doesn't measure anything but for those of us that come here to use the site/stats to write blogs, etc. it would be helpful rather than having to go to multiple sites. It is obviously up to the site owner and how he wants to present the info so he can take this comment for what it is worth.
Posted by bpelowski on June 01, 2007
10
#9 is accurate. While the "triple crown" stats aren't the best at telling true skills, it's a standard and quick way to show info about the player.
Because this site doesn't include info like RBI and steals, I find myself not coming here often as I would normally - I have to go to BA or Milb.com.
Posted by galt on June 01, 2007
11
Virgil Vasquez has come out of nowhere this season.

This is helpful feedback, by the way.
I don't find knowledge of players' RBI helpful at all in understanding player development. On the other hand, if the information is in our database (and it is), then there should be some way to present this kind of information for those of you who find it helpful.

Regarding SB, I think in other threads I have mentioned we're going to have a new 'baserunning' table with stolen bases among other stats this summer. There are still some bugs to be worked out, but it's coming this month. Same with the fielding stats table, by the way.
Posted by Chris Constancio on June 01, 2007
12
what happened to wOBA and SEC?
Posted by revredbird on June 02, 2007
13
You know, there is an actual benefit to K/9 over K% (at least in Fantasy Baseball), since pitchers with higher K/9 totals will generally be more helpful in the strikeout category than pitchers with lower K/9 totals. But the K% stat is more useful for real baseball.
Posted by locke000 on June 04, 2007

Please note that only registered users can contribute comments.

Current Reports

Sponsor


Login




Create Your Account
 

Organization Resources

RSS | PLAYER VIDEO | PLAYER LINKER
ABOUT FIRSTINNING | CONTACT US